information technology r&d: critical trends and issues

Information Technology R&D: CriticalTrends and Issues

February 1985

NTIS order #PB85-245660

—

Recommended Citation:Information Technology and R&D: Critical Trends and Issues (Washington, DC:U.S. Congress, Office of Technology Assessment, OTA-CIT-268, February 1985).

Library of Congress Catalog Card Number 84-601150

For sale by the Superintendent of DocumentsU.S. Government Printing Office, Washington, DC 20402

Foreword

New computer and communications technologies are obviously transformingAmerican life. They are the basis of many of the changes in our telecommunica-tions system and also a new wave of automation on the farm, in manufacturingand transportation, and in the office. They are changing the form and deliveryof government services such as education and the judicial system. Informationproducts and services have become a major and still rapidly growing componentof our economy.

A strong U.S. research and development effort has, in the past, been the sourceof much of this new technology. However, recent events, such as the restructur-ing of the U.S. telecommunications industry and the emergence of strong foreigncompetition for some technologies, have changed the environment for R&D. Con-sequently, the House Committee on Science and Technology, the House Commit-tee on Energy and Commerce, and its Subcommittee on Telecommunications, Con-sumer Protection, and Finance asked OTA to conduct an assessment of the currentstate of R&D in these critical areas.

In this report, OTA examines four specific areas of research as case studies:computer architecture, artificial intelligence, fiber optics, and software engineer-ing. It discusses the structure and orientation of some selected foreign programs.Finally, it examines a set of issues that have been raised in the course of the study:manpower, institutional change, the new research organizations that grew outof Bell Laboratories, and implications of trends in overall science and technologypolicy.

Information technology research and development in the United States is aremarkably adaptable system— important changes may already be taking placein funding patterns, institutional structures, manpower development, and gov-ernment policies. Hence, the policy issues for Congress are not so much to stimu-late change as to remove barriers to productive change, and to monitor and main-tain the health of the R&D enterprise in a time of rapid change in an industryso central to our national economy and security.

OTA gratefully acknowledges the contributions of the many experts, withinand outside the Government, who served as panelists, consultants, contractors,and reviewers of this document. As with all OTA reports, however, the contentis the responsibility of OTA and does not necessarily constitute the consensusor endorsement of the advisory panel or the Technology Assessment Board.

J O H N H . G I B B O N S

Director

OTA Information Technology Research and Development Project Staff

John Andelin, Assistant Director, OTAScience, Information and Natural Resources Division

Fred W. Weingarten, Communication and Information Technologies Program Manager

Project Staff

Donna L. Valtri, Project Director’Fred W. Weingarten, Project Director’

Chuck Wilk, Senior Analyst

Linda G. Roberts, Senior AnalystLinda Garcia, Analyst

M. Karen Gamble, Analyst

Prudence S. Adler, Analyst

Jim Dray, Research AnalystEarl Dowdy, Research Analyst

Lauren Ackerman, Research AnalystJohn Williams3

Peg Kay4

Administrative Staff

Elizabeth A. Emanuel, Administrative AssistantShirley Gayheart, SecretaryJennifer Nelson, Secretary

Marsha Williams, Secretary

Renee S. Lloyd, Secretary

Contractors

Myles Boylan J. F. Coates, Inc. Barbara Davies

Russell Drew Sidney Fernbach Michael Goetz

Stan Liebowitz Abbe Mowshowitz Larry Olson

Bruce Owen John Young

‘Project Director from November 1982 to December 1983.‘Acting Project Director from December 1983.‘Detailee from the Federal Communications Commission from October 1982 to October 1983.4Detailee from the National Bureau of Standards, May 1984.

iv

Information Technology Research and Development Advisory Panel

Professor

Geneva BelfordDepartment of Computer ScienceUniversity of Illinois

Steven BissetPresidentMegatest, Inc.

John E. BrysonMorrison& Foerster

NandiKishore M. ChitreDirectorSystems Planning DivisionINTELSAT

Ralph E. GomoryVice President andIBM Corp.Thomas J. Watson

JohnV. BarringtonDirector

Roger G. Noll Chairmanof Economics, Stanford University

Donald McCoyVice President and General ManagerCBS Technology Center

1thiel de sola Pool*Professor of Political ScienceMassachusetts Institute of Technology

Paul E. Ritt, Jr.Vice President and Director of ResearchGTE Laboratories, Inc.

Larry W. SumneyExecutive DirectorSemiconductor Research Corp.

Victor VyssotskyDirector of Research Executive Director

Research, Information SciencesResearch Center AT&T Bell Laboratories

Robert E. WesslundVice President for Technology Exchange

COMSAT Laboratories Control Data Corp.

William C. Hittinger George R. WhiteExecutive Vice President Senior Research FellowRCA/David Sarnoff Research Center The Harvard Business School

Bruce LusignanDirectorCommunication Satellite Planning CenterStanford University

Advanced Computer Architecture Workshop

Duane AdamsConsultantDefense Advanced Research Project Agency

James C. BrowneDepartment of Computer ScienceUniversity of Texas

Alan CharlesworthSenior Staff EngineerFloating Point Systems, Inc.

Kent K. CurtisSection Head, Computer ScienceDivision of Mathematical and Computer

SciencesNational Science Foundation

Sidney FernbachConsultantControl Data Corp.

Dennis GannonAssistant ProfessorDepartment of Computer SciencePurdue University

Robert G. GillespieVice Provost for ComputingUniversity of Washington

David KuckDepartment of Computer SciencesUniversity of Illinois

*Deceased Mar. 11, 1984

Neil LincolnVice PresidentETA Systems,

Roger G. Nell

and Chief ArchitectInc.

Professor of EconomicsStanford University

Paul Ritt, Jr.Vice President and Director of ResearchGTE Laboratories, Inc.

Paul SchneckLeader, Information Sciences DivisionOffice of Naval Research

Jacob SchwartzDirectorDivision of Computer SciencesCourant Institute of Mathematical Sciences

Fiber Optics Workshop

Thomas Giallorenzi, ChairmanSuperintendent, Optical Sciences DivisionNaval Research Laboratory

Michael BarnoskiConsultantPacific Palisades, CA

Melvin CohenDirector, Interconnection Technology

LaboratoryAT&T Bell Laboratories

Anthony J. DeMariaAssistant Director of Research for Electrical

and Electrooptics TechnologyUnited Technology Research Center

Charles W. DenekaDirector, Optical Waveguide TechnologyCorning Glass Works

Charles KaoITT Executive ScientistITT Corp.

Software Engineering Workshop

Richard A. DeMilloProfessor of Information and Computer

ScienceGeorgia Institute of Technology

George DoddDepartment HeadComputer Science DepartmentGeneral Motors Research Laboratories

Burton J. SmithVice PresidentResearch and Development DepartmentDenelcor, Inc.

Harold StoneProfessorDepartment of Electrical and Computer

EngineeringUniversity of Massachusetts

Jim ThorntonChief Executive OfficerNetwork System Corp.

Kenneth G. WilsonJames A. Weeks Professor of Physical

ScienceNewman LaboratoryCornell University

Herwig KogelnickDirectorElectronics Research LaboratoryAT&T Bell Laboratories

Henry KresselSenior Vice PresidentE. M. Warburg, Pincus & Co.

William StreiferSenior Research FellowXerox Corp.

John WhinneryDepartment of Electrical Engineering and

Computer ScienceUniversity of California, Berkeley

Amnon YarivProfessor of Electrical Engineering and

Applied PhysicsCalifornia Institute of Technology

Capers JonesManager of Programming Technology

AnalysisITT Corp.

Brian KernighanHead, Computing Structures ResearchAT&T Bell Laboratories

vi

Ann Marmor-SquiresChief TechnologistTRW Defense Systems Group

Robert MathisDirectorAda Joint Program OfficeDepartment of Defense

Harlan MillsIBM FellowIBM Corp.

Leon J. OsterwellChairmanComputer Science DepartmentUniversity of Colorado

C. V. RamamoorthyProfessor of Electrical Engineering and

Computer ScienceUniversity of California, Berkeley

Artificial Intelligence Workshop

Bruce BullockVice PresidentTeknowledge Federal Systems

Marvin DenicoffExecutiveThinking Machines Corp.

Ewald HeerProgram ManagerAutonomous Systems and Space MechanicsJet Propulsion Laboratory

Thomas F. KnightAssistant Professor of Electrical Engineering

and Computer ScienceArtificial Intelligence LaboratoryMassachusetts Institute of Technology

Mitchell MarcusMember of the Technical StaffLinguistics and Artificial Intelligence

Research DepartmentAT&T Bell Laboratories

Ronald OhlanderProgram ManagerDefense Advanced Research Projects Agency

Paul SchneckLeader, Information Sciences DivisionOffice of Naval Research

Ben ShneidermanProfessor of Computer ScienceUniversity of Maryland

Steve SquiresSoftware Technology Programming ManagerDefense Advanced Research Projects Agency

Terry A. StraeterVice President and Program DirectorGeneral Dynamics Electronics

Raymond YehProfessor of Computer ScienceUniversity of Maryland

Kamron ParsayePresidentSilogic, Inc.

Stephen PolitManager of Advanced DevelopmentArtificial Intelligence GroupDigital Equipment Co.

Azriel RosenfeldDirectorCenter for Automation ResearchComputer Science CenterUniversity of Maryland

David WaltzProfessor of Electrical Engineering and

Research ProfessorCoordinated Science LaboratoryUniversity of Illinois

Patrick H. WinstonProfessor of Computer Science and Electrical

Engineering and DirectorArtificial Intelligence LaboratoryMassachusetts Institute of Technology

vii

Information Technology Research and Development Federal

Charles BrownsteinHeadInformation Science DivisionNational Science Foundation

Douglas CrombieChief ScientistNational Telecommunications

Information Administration

John A. DalySpecial AssistantAgency for International Development

Craig FieldsAssistant Director for Systems ServicesDefense Advanced Research Projects Agency

Clinton W. Kelly, IIIAssistant Director, Systems Science DivisionDefense Advanced Research Projects Agency

Robert LovellDirector of CommunicationsNational Aeronautics and Space

Administration

Reviewers and Other Contributors

Robert AyresCarnegie-Mellon University

Richard BaxterBritish Embassy

Donald S. BeilmanMicroelectronics Center of North

CarolinaJohn Biddle

Computer and CommunicationsIndustry Association

Marlan BlissettUniversity of Texas

Erich BlochSemiconductor Research Corp.IBM Corp.

Justin BloomTechnology International, Inc.

Bill BoesmanCongressional Research Service

Jane BortnickCongressional Research Service

Paul BortzBrown, Bortz & Covington

Martha BranstadNational Bureau of Standards

Robert PowersChief Scientist

Workshop

Office of Science and TechnologyFederal Communications Commission

John RiganatiDivision ChiefSystem Components DivisionNational Bureau of Standards

Paul SchneckLeader, Information Science DivisionOffice of Naval Research

Elias SchutzmanProgram DirectorElectrical Optical Communications ProgramNational Science Foundation

Kenneth BrownAmerican Enterprise Institute for

Public Policy ResearchGlen Chancy

AT&T Bell LaboratoriesGary Coleman

Department of EducationArthur Cordell

Science Council of CanadaMatthew Crugnale

Gnostic Concepts, Inc.Beverly Daniel

General Accounting OfficeGeorge Dummer

Massachusetts Institute ofTechnology

Lauren ErlingADAPSO

Frank FisherUrban Institute

Penny FosterNational Science Foundation

Donald R. FowlerCalifornia Institute of Technology

Peter FuchsHarvard University

Osmund FundingslandGeneral Accounting Office

Robert GallawaNational Bureau of Standards

Howard GobsteinGeneral Accounting Office

Robert A. GrossColumbia University

Walter HahnGeorge Washington University

Glen HansenDepartment of Commerce

Richard HershUniversity of Oregon

Christopher HillCongressional Research Service

Joseph HullDepartment of Commerce

B. R. InmanMicroelectronics & Computer

Technology Corp.

VIII

— —

Charles L. JacksonShooshan & Jackson, Inc.

Don KashUniversity of Oklahoma

Todd LaPorteUniversity of California-Berkeley

Christopher LeMaistreRensselaer Polytechnic Institute

Jean-Pierre LetouzeyFrench Embassy

Thomas LinneyCouncil of Graduate Schools

Harold LinstonePortland State University

John G. LinvillStanford University

John LogsdonGeorge Washington University

Albert LumbrosoFrench Embassy

Enrique MarcatiliAT&T Bell Laboratories

Michael MarcusFederal Communications

CommissionNick Metropolis

Los Alamos National LaboratoryCharles H. Miller

AT&TRobert Miller

University of HoustonRobert Morgan

Washington University

In-House Reviewers

Audrey Buyrn, Program ManagerMartha Harris, Project DirectorPaul Phelps, Project DirectorHenry Kelly, Project DirectorJohn Alic, Project DirectorBarry Holt,AnalystEric Bazques, Analyst

Joel MosesMassachusetts Instituteof

TechnologyJane Myers

Patent and Trademark OfficeTom Nardone

Bureau of Labor StatisticsDorothy Nelkin

Cornell UniversityWilliamF.Nelson

GTE Laboratories, Inc.Arnie Packer

Manpower EconomistPaul Penfield

Massachusetts InstituteofTechnology

Robert PepperDepartment ofCommerce

Paul PolishukInformation Gatekeepers, Inc.

JohnM. RichardsonNational Research Council

Michel RobinFrench Embassy

Martha G. RussellUniversity of Minnesota

Grant SaveersDigital Equipment Corp.

Wendy SchachtCongressional Research Service

Jurgen SchmandtUniversity ofTexas

Dan SlotnickUniversity of Illinois

Ollie SmootComputer and Business

Equipment ManufacturersAssociation

Bill StotesberyMicroelectronics &Computer

Technology Corp.Gregory Tassey

National BureauofStandardsHenry Taylor

Naval Research LaboratoryAl Teich

American Association forAdvancementof Science

William WellsGeorge Washington University

Irwin WhiteNew York State Energy R&D

AgencyRobert Willard

Information IndustriesAssociation

John G. WilliamsCarruthers, Deutsch, Garrison&

WilliamsPaul Wilson

AT&TPaul Zurkowski

Information IndustriesAssociation

ix

—

Contents

Chapter Pagel. Introduction and Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

Z. The Environment for R&D in Information Technology in theUnited States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

3. Selected Case Studies in Information Technology Research andDevelopment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

4. Effects of Deregulation and Divestiture on Research . . . . . . . . . . . . . . . . . . . 111

5. Education and Human Resources for Research and Development . . . . . . . . . 139

6. New Roles for Universities in Information Technology R&D. . . . . . . . . . . . . 169

7. Foreign Information Technology Research and Development. . . . . . . . . . . . . 201

8. Information Technology R&D in the Context of U.S. Science andTechnology Policy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279

9. Technology and Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339

x

Chapter 1

Introduction and Summary

Contents

Goals for Federal R&D Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Principal Findings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

The Nature of Information Technology R&D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Software as Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Multidisciplinary Nature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Close Boundary Between Theory and Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Complexity .. .. . . $ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .. .. .. ... ... .

Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Advanced Computer Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Software Engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........Fiber Optics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Artificial Intelligence. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Issues and Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Impacts of Telecommunications Deregulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Scientific and Technological Manpower . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Changing Roles of Universities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Foreign Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Science Policy . ... ... . . . . . . ....... ..

Summary . . . . . . . . . . . . . . . . . . . . . . .. .

3

5

67788

99

111213

141416171819

21

Chapter 1

Introduction and Summary

American society is becoming increasinglyreliant on the use of electronic informationtechnology, principally computer and commu-nication systems. Economists and social scien-tists point out that information is a basic re-source for society and an important factor ofproduction for the economy. Information tech-nology provides business, governments, andcitizens with the basic tools necessary to com-municate and use information. Furthermore,the information industry, itself, including boththose that produce electronic hardware andthose that offer information services, is an in-creasingly important part of U.S. industrialstrength.

Because of this reliance on information tech-nology, public attention has been drawn to theprocess of innovation in that field, developing

new technological products and bringing themto the marketplace. Since research and devel-opment (R&D) is generally considered to be akey element driving innovation (although notthe only one), Congress is concerned with thegeneral health of R&D in information tech-nologies. Concerns include the effects thatchanges in the structure of the U.S. telecom-munications industry are having on industrialR&D, and the implications of new foreign pro-grams intended to challenge traditional U.S.market leadership in some areas of computersand communications. OTA examined theseand related questions at the request of theHouse Committee on Science and Technology,the House Committee on Energy and Com-merce, and its Subcommittee on Telecommu-nications, Consumer Protection, and Finance.

Goals for Federal R&D Policy

Currently, public concern with informationtechnology is focused on its role in promotingeconomic competitiveness. However, Federalinterest in promoting R&D and technologicalinnovation, dating back to the writing of theConstitution, has always been motivated by

●

several distinct goals, only one of which hasbeen economic growth. These goals, whichmay for any specific technology or programeither complement or compete with one another,are applicable to information technology.

● Support National Defense: A modem mil-itary force is dependent on communica-tions and computer technologies for manypurposes, including intelligence, missileguidance, command and control, andlogistics. Projected future applications in-clude battlefield management, fully auto-mated weapons, and computer-based arti-ficial intelligence advisors for pilots. Themilitary importance of information tech-

nology is demonstrated by the predomi-nant role the Department of Defenseplays in support of R&D in that field(nearly 80 percent of direct governmentfunding in this area is supplied by DOD).

Provide for Social Needs: Governmenthas always treated the telecommunica-tions system as a basic social infrastruc-ture. One major goal of communicationpolicymakers is “universal service. ” Newservices such as cable television and in-formation services delivered over telecom-munication channels can potentially im-prove the quality of life for Americancitizens. For example, OTA, in a recentreport, pointed out the role informationtechnology could play in improving thequality of and access to education andalso stressed the need for additional R&Dfocused on educational technology. Anotherrecent OTA report has described the po-

3

—

4 ● Information Technology R&D: Critical Trends and Issues

tential benefits of information technologyto handicapped individuals.

● Promote Economic Growth: The informa-tion technology industry (those who makeand sell or provide access to communica-tions media) and the information indus-try (those who use the new technologiesto produce and sell new information serv-ices and products) are a growing part ofthe U.S. economy. Their economic impor-tance is felt both domestically and inter-nationally. These industries also have animportant indirect effect, in that the tech-nologies and services they produce con-tribute materially to the economy in suchforms as productivity growth, betterquality of products, and improved mana-gerial decisionmaking. The health of theinformation industries depends in part ontheir ability to bring forth new productsand develop new applications; this ability,in turn, depends on R&D.

● Advance Basic Understanding of theWorld: Research in computer and commu-nication science has contributed to basicunderstanding in fields that transcendpure information technology. For exam-ple, research in artificial intelligence hasled to insights into brain functions andhuman thinking and problem solving.Communications researchers have madea Nobel Prize-winning contribution to ourunderstanding of the origins of the uni-verse. The study of computer languageshas both benefited from and contributedto theories of the evolution and structureof natural human languages. Similarly, in-formation technology researchers havecontributed to knowledge in such otherfields as fundamental mathematics, logic,and the theory of genetic coding.

Moreover, computers have become in-dispensable research tools in many areasof science ranging from physics and as-tronomy to biology and social science.Some scientists now speak of “computa-tional research” as a new, wholly differentform as important to the progress ofknowledge as is experimental and theoret-ical research.

●

●

Enhance National Prestige: The UnitedStates has always been a world leader incomputer and communication technol-ogies, both in the discovery of new basicknowledge and in its development intocommercial products. In certain areas ofmarketable technology, such as super-computers (very powerful computers, thelargest and fastest available on the mar-ket) or communication satellites, U.S.firms have in the past held an undisputedtechnological lead. Although such techno-logical dominance has nearly disappearedin most cases and could not have been ex-pected to continue as other nations devel-oped research capability and industrialstrength in information technology, pre-serving U.S. technological competitive-ness and leadership is still an importantpublic policy goal with implications that ex-tend beyond purely economic advantage.Support Civilian Agency Missions: Manycivilian Federal agencies need to use ad-vanced information systems to performtheir missions more efficiently or effec-tively. For example, computers and com-munication systems underlie the spaceexploration programs of the NationalAeronautics and Space Administration(NASA). Weather researchers working forthe National Oceanic and AtmosphericAdministration (NOAA) and fusion re-searchers for the Department of Energyneed access to the most advanced super-computers. Although the support of mis-sion agencies for basic research is limited,they have funded development and, often,applied research directed to meeting theirspecific needs.

Over the years, these Federal interests haveled to a wide array of policies, programs, andagencies designed to promote the conduct ofresearch and development in the United Statesand to appropriate its benefits to Americansociety. Many of them have helped stimulatethe development of computers and communi-cation technologies.

The principal issue examined in this study iswhether those policies, programs, and agencies

Ch. I—Introduction and Summary ● 5

are now adequate and appropriate for advanc- their particular characteristics, and the rapidlying information technologies in light of the escalating world competition in producing andemerging needs of society for these technologies, selling products and services based on them.

Principal Findings

In summary, OTA’s study of informationtechnology research and development reachedthe

●

●

following conclusions:

Information technology is an area inwhich the United States has historicallyshown great strength, both in basic sci-ence and marketed technology; and thisstrength has benefited the Nation in sev-eral ways. It has contributed to the growthof a strong economic sector, the informa-tion industry. It has contributed to na-tional security. It has stimulated theevolution of basic science.Most areas of information technology ex-amined in this study, including microelec-tronics, fiber optics, artificial intelligence,computer design, and software engineer-ing, are still in the early stages as tech-nologies. Much improvement in techno-logical capability remains to be developed,and that improvement will depend onboth fundamental research and techno-logical development. Hence, R&D will bean important factor in stimulating contin-uing innovation.By most measures, U.S. research and de-velopment in information technology isstrong and viable; however, those tradi-tional measures may not be realistic guidesto the future needs of the United Statesfor R&D in these areas. In particular, in-creasing competition from foreign nations(in particular, Japan) as well as the grow-ing intrinsic importance of informationtechnology to the United States, suggestthat more attention should be paid by pol-icymakers to the needs for R&D in infor-mation technology. Many of these na-tions, motivated both by economic concernsand broader social goals, have developed

●

●

●

national programs designed to foster in-formation technology.In response to these new pressures, indus-try support is growing rapidly for short-term applied research and developmentalwork, both within industrial labs andthrough support of university work. In-dustry has generally looked, and stilllooks, to the academic community to per-form and the Federal Government to fundthe long-term basic research that willunderlie future technological advances.Universities, traditionally viewed ascenters for basic research, are reexamin-ing their roles with respect to applied re-search and are forming new types of rela-tionships with industry and government.State governments, in particular, lookingfor stimuli to economic development seenew high-technology industry, includingthe information industry, as interested inbeing located near strong university basicresearch programs. Many states haveformed new organizations within theiruniversities in order to strengthen research,focus it toward problems of interest to in-dustry, and facilitate the transfer of tech-nology from the research laboratory to in-dustrial application. With most of theseexperiments, it is too early to tell whetherthey will be successful in stimulating in-dustrial development and whether theywill have beneficial or negative effects onthe universities.The Department of Defense is the pre-dominant source of Federal support of in-formation technology research and devel-opment, providing nearly 80 percent ofthe funding. Although some spillover tothe civilian sector from DOD research can

— —


be expected, its work is predominatelyfocused on military requirements and can-not be relied on to provide for all civilianneeds. An important issue for considera-tion by Congress is whether increasedfunding by nondefense agencies of long-term research in information technologyis needed to focus more work on civilianneeds.

● There is substantial concern that techni-cal and scientific information flows be-tween the United States and other coun-tries are unbalanced outward. Proposalsto redress that balance focus both on in-creasing access to foreign information andon restricting outflows of U.S. informa-tion. Policies that would enhance the ac-cess of U.S. information scientists to thework of foreign researchers (through suchvehicles as translation services, travelgrants, and foreign science exchange pro-grams) would help their own work. Suchprograms would also help science policyexperts to evaluate the state of foreigntechnology more accurately.

Steps have also been taken to tightencontrols over the export of technical in-formation on national security grounds.In considering such controls, Congressneeds to balance national security consid-

erations against both first amendmentrights and the possible negative impactson domestic R&D.Instruments for scientific research aregrowing more sophisticated and are be-coming obsolete at an increasingly rapidpace. As a result, they are more expensiveand yet must be replaced more frequentlyif researchers are to keep at the forefrontsof knowledge. In particular, computersoftware researchers need access to so-phisticated computational facilities anddata communication networks, and thoseworking on microelectronics need accessto chip processing facilities and computer-aided-design tools.Policies designed to stimulate informationtechnology R&D need to be evaluated forpossibly significant tradeoffs and exter-nal costs in other areas. Some costs areclear. For example, R&D tax incentivesare a short-term loss of Federal revenues.Policies designed to draw more highlytrained people into information technol-ogy may create manpower shortages inother areas of research. Devoting educa-tional resources to information technol-ogy education and training draws themaway from other areas of need,when overall resources are not

The Nature of Information Technology R&D

For purposes of this study, OTA defines the price. This high rate of technological

at a timegrowing.

improveterm “information technology to refer to elec-tronic hardware that is used to create, com-municate, store, modify, or display informa-tion and to programming or “software” thatis developed to control the operation of thathardware. Modern information technology isbased on the microelectronic chip, the large-scale integrated circuit that over the last dec-ade has transformed the computer and com-munications industries through steady andrapid increase of performance and decrease in

‘merit in microelectronics will continue into the1990s and probably beyond. Changes in theinformation technology products and servicesavailable in the marketplace will likely be asrevolutionary in the next decade as they havebeen over the last one that has seen the ap-pearance and growth of such technologies asthe personal computer, fiber optics, satellitecommunications, the video cassette recorder,and two-way cable television. Industry willcompete worldwide to bring these new prod-

Ch. l—introduction and Summary ● 7

Photo credit: Arthur A. Merin, National Magnetic Fusion Energy Computer Center, Lawrence Livermore National Laboratory

The use of computer modeling and graphics to study fusion energy. Outer magnetic surface of four-period helicalaxis stellarator (HELIAC)

ucts and services to use, and an important ele-ment of that competition will be research anddevelopment.

Information technology has characteristicsthat distinguish it from many other areas ofR&D and that affect policies designed to stim-ulate and direct it.

Software as Technology

“Software,” or programs, are sets of instruc-tions that direct a computer in its tasks. Theterm can also refer to information; for exam-ple, computer data bases, the information con-tent transmitted over communication lines, orthat contained on a video disk. Because soft-ware often can be the most complex part ofan information system, research and develop-ment on software is considered to be of equalimportance to research on hardware technol-ogy. Yet, to some people, software does not“look” like technology. The concepts and tech-niques that underlie programs are intangible,and their embodiment is usually in a form suchas a list of instructions on paper or a sequence

38-802 u - 85 - 2

of images on a video screen. Research in infor-mation technology includes research on bothhardware and software technology, and on theboundaries between them, for many computa-tional techniques may appear in either hard-ware or software.

Software as an area of research presentsFederal R&D support with some unusual prob-lems. For example, intellectual property as ex-pressed in software is difficult to protect. Itcan also be hard to distinguish between R&Din computer programming, work which ad-vances the state of software technology, andusing existing techniques to write new pro-grams and maintain old ones—a distinctionthat drew particular attention in the debateover R&D tax credits, for example.

Multidisciplinary Nature

R&D in information technology spans awide variety of disparate fields of work. Forexample:

• Solid-state physicists work to improve theperformance of the semiconductor devices


●

●

●

●

●

●

that form the heart of modern computerand communication systems.Mathematicians study abstract theoriesof computation, looking for methods offaster calculations, for ways to writeerror-free software, and to transmit themaximum amount of information reliablyover communication lines.Psychologists and brain physiologistsfind clues for writing “smarter” programsthat can solve more complex problems.Linguistic theorists contribute to the de-sign of new languages for programmingcomputers and build the foundations forcomputer programs that can understandand automatically translate human lan-guages.Users of computer systems, ranging fromphysical scientists to graphic artists, de-velop new types of software for their ap-plications. In this development process,they make fundamental contributions tosoftware technology.Sociologists, psychologists, and the man-agement scientists search for the most ef-fective ways to design office automationand decision support systems. Their re-search ranges from human factors, whichexamines direct human-machine interac-tion, to investigations of organizationaldecisionmaking.A few computer scientists, sociologists,historians of science, and experts in otherfields, examine the broader questions ofthe interactions of information technol-ogy with society.

As a result of this breadth of subject area,relevant Federal programs of R&D supportare distributed among many Governmentagencies and among many organizationalparts of those agencies. The National ScienceFoundation (NSF), for example, has at leastfour divisions, each in a different directorate,that support research directly related to com-puters and communications systems. Anotherimportant implication is that manpower pro-grams must be concerned not just with thesupply of “core” computer scientists and engi-

neers, but with a much broader base of re-searchers and users in many disciplines.

Close Boundary BetweenTheory and Application

Both the rapid pace of commercializationand the intrinsic nature of the technology re-sult in a short time lag between basic researchresults and commercialization. Because infor-mation systems are so flexible and becausemanufacturers are using powerful new designtools to create new types of computers, lesstime is required to implement a new concept.It may take only a few months for the resultsof a basic research project in pure mathematicsto be realized in a commercial software package.

This close relationship between basic re-search and commercial application, possiblymost closely paralleled by biotechnology, cre-ates unusual and difficult issues for institu-tions, such as universities, that are tradi-tionally concerned only with basic researchpublished in the public domain. These issuesinclude publication policies, intellectual prop-erty rights, faculty conflicts of interest, com-petition with the private sector, and researchpriorities. Federal agencies that fund basic re-search, such as NSF, also find themselves deal-ing with these issues more frequently.

Complexity

Modem computer/communications systemsare among the most complex technologies everassembled by human beings. Computers con-sist of millions of logical subunits. Computerprograms can contain millions of instructions.These complex machines, running complexprograms, are then interlinked through com-munications networks. To be able to under-stand, predict, and control the behavior ofthese technologies requires a powerful theoryof complex processes. No such theory yet ex-ists, although it remains a major goal of com-puter science.

It is because of this complexity that, eventhough computers and communication sys-

Ch. I—Introduction and Summary ● 9

terns are artificial devices—technologies cre- puters. Writing large, complex programs to doated and built by human beings-there is basic unusual tasks is a common research method-science devoted to understanding their behav- ology aimed at understanding basic principalsior. This science has a strong experimental underlying the structure of problems andcomponent to it. Many computer scientists software.and engineers need access to large-scale com-

Case Studies

Information technology is a broad subjectarea. To better understand and describe R&Dissues, four specific areas of technology wereexamined in more detail. These case studiescovered both software and hardware and bothcomputer and communications technologies.In addition, all four technologies have raisedpolicy issues in the last few years.

Advanced Computer Architecture

Although computers have increased in so-phistication over the years, and individualmodels differ from one another in detail, thehigher level logical design, called the “architec-ture, ” of most commercially available com-puters is based on concepts that date back afew decades.

New concepts of computer architecture nowbeing explored in the laboratory will underlietwo types of innovation: 1) the developmentof highly specialized, low-cost computer mod-ules that do specific types of tasks at extreme-ly high speeds; and 2) the future generationsof supercomputers.

Low-cost specialized computers are designedto perform specific types of computationaltasks very efficiently. When hardware is ex-pensive, as computer hardware has been in thepast, users usually try to allocate its cost overa variety of applications. Hence, the so-calledgeneral-purpose computer, which performs awide variety of computations relatively well,has been and continues to be the mainstay ofthe computer industry. However, most typesof computational tasks have unique character-istics. Since hardware costs have dropped

significantly, new market possibilities have ap-peared for inexpensive special purpose hard-ware that takes advantage of these character-istics. The general purpose computer systemof the future may serve mainly as a routingswitch, sending a computing task to the appro-priate one of several different specialized proc-essors attached to it.

Supercomputers, the label given to the mostpowerful machines on the market at any time,have received significant press attention late-ly. Until recently, U.S. firms have been theonly significant marketers of supercomputers.However, Japanese firms have now broughtout supercomputers that, by some measures,seem to perform comparably to U.S. machines.In addition, Japan has embarked on two ma-jor supercomputer projects. One is designedto develop the next generation of high-speednumerical calculating machines; the other, theFifth-Generation Computer Project based onartificial intelligence theories, is intended toproduce a machine that performs reasoningfunctions similar to human thinking processes.

The Federal Government is considering sev-eral responses to maintain leadership in thistechnology. For example, the Defense Ad-vanced Research Projects Agency (DARPA)is beginning a “Strategic Computing Pro-gram” to develop artificial intelligence capa-bilities for the needs of the Department of De-fense. The National Science Foundation hasembarked on a program intended both to in-crease their support of computer architectureresearch and to put more supercomputer ca-pability in the hands of scientific users. Both

IO ● Information Technology R&D: Critical Trends and Issues

A microprocessor on a chip. This chip contains a

NASA and the Department of Energy havedeveloped similar plans to increase the ad-vanced supercomputing capability availableto their researchers.

If they were fully implemented, these Fed-eral plans would constitute an important re-sponse to the growing need for access to ad-vanced computing resources. Over the lastdecade, computing resources available to basicresearchers, particularly in universities, havesteadily fallen behind the state of the art. Al-though the U.S. computer industry has tradi-tionally led the market in the supercomputerfield, basic researchers in other countries suchas West Germany and Japan may have more

Photo credit: AT&T

entire computer, as powerful as some minicomputers

and easier access to supercomputers from theUnited States than their counterparts atAmerican universities.

In addition to providing access to supercom-puters for agency supported researchers, suchFederal procurements would also stimulatethe market for the next generation machine,a market that has traditionally been small and,hence, unattractive to computer manufactur-ers. As the principal user of them, the FederalGovernment has always had a major influenceon the supercomputer market. This Federalstimulus may be particularly important for ex-perimental machines that are based on new ar-chitectural concepts. Markets and applications

Ch. 1-Introduction and Summary ● 1 1

for these machines are particularly unpre-dictable.

Should Congress decide that additional Fed-eral emphasis on computer design is in the na-tional interest, support of basic research shouldalso be expanded to provide the scientific basefor succeeding generations of advanced ar-chitecture computers, large and small. Al-though R&D on the next generation supercom-puters and other specialized architectures isunderway in a few industrial laboratories,firms participating in this specialized markettend to be small. Longer term basic researchis usually left to be performed in academic lab-oratories and receives little industrial support.Hence, Federal support of basic research canhave significant effect.

In the past, responding to funding limita-tions as well as the complexity of implementingcomputer hardware in a laboratory, agencieshave usually kept basic research in machinearchitecture at the theoretical, or “paper andpencil” stage. This restriction has not allowednew concepts to be tested as hardware. Toallow such testing, some of these conceptuallevel basic research projects need to be fundedat a level that allows them to be carried to ex-perimental and even prototype stages. Micro-electronics allows such hardware experimenta-tion to be done more easily and cheaply thanin the past, because prototypes can now beassembled from chips that form large logicalmodules containing thousands of functions.

Software has also been and continues to bea major problem area. Taking full advantageof a supercomputer’s speed and power requiresthe design of specialized programs. Signifi-cantly more research needs to be done onnumerical computational techniques and pro-gramming theory for new computer archi-tectures.

Software Engineering

The computer industry has transformedover the few decades of its existence from be-ing hardware dominant to being software dom-inant. In the early years, hardware was themost expensive component of a computer sys-

tem. Software was often given away free bycomputer manufacturers, and the costs of de-veloping custom programs were much lessthan the costs of purchasing and maintainingthe machines. The situation is now reversed.In most cases, costs of developing and main-taining software are now the dominant costsof creating and operating large computer ap-plications. Although developments in micro-electronics have steadily and rapidly reducedthe cost and increased the performance of com-puter hardware, improvements in the produc-tivity of programmers has come much moreslowly.

Since the limits of cost, reliability, and timerequired to create new applications are increas-ingly dictated by software instead of hardwareconsiderations, research directed toward im-proving the productivity of programmers anddesigners will have high leverage. “Softwareengineering” is the term used to describe thisarea of research.

Research in software engineering rangeswidely in content and approach. At one ex-treme is highly theoretical work directed at developing a fundamental understanding of thenature of programs and “proving” in somemathematical sense that they act as intended.At the other end of the spectrum are behavior-al and management scientists concerned withunderstanding how people interact with com-puter systems and how to manage program-ming projects. Not all research related tounderstanding programming and programbehavior is called “software engineering”;many fields of computer science and engineer-ing contribute.

The Department of Defense, faced with anannual software expense estimated at $4-$8billion, supports a large effort in software engi-neering. DOD has funded the design of a newprogramming language, ADA, and a “pro-gramming environment” called STARS thatprovides a set of automated tools for use bythe program designer. These tools are intendedto be standards, required to be used in the de-velopment of DOD software.

Because the economic stakes are so high, re-searchers at industry centers such as the Bell


Laboratories and IBM’s T. J. Watson Re-search Laboratories have also been active insoftware engineering research. University re-searchers in software engineering, supportedby Federal agencies, have concentrated on un-derstanding better the theoretical underpin-nings of computer programs.

Other countries have also recognized the im-portance of software engineering. It has beenmade an explicit part of the British informa-tion technology research program. The Japa-nese, realizing that software has been a majorweakness in their attempts to be competitivein the worldwide computer market, are experi-menting with what they call “software fac-tories. ” The French have always had strengthin the area of programming, particularly thetheory of programming languages. The De-partment of Defense language, ADA, was de-veloped in a French laboratory.

Although the potential benefits to the Na-tion of advances in software engineering arehigh, the implications for Federal support areless clear. Some experts arguing that lack offundamental research is not the key bottle-neck, point to the failure of designers to applyexisting basic theories and research method-ologies to the development of large-scale sys-tems. To do so would require closer industry-university researcher interaction and might behelped by the establishment of some large lab-oratory facilities to study design problems ona realistic scale.

Other experts, pointing out that we still lackthe fundamental theoretical breakthrough nec-essary to develop a true discipline of softwareengineering, argue that the problems will per-sist until such a theory is established. Ex-panded long-term Federal support of a broadrange of fundamental research on software isnecessary to develop a sound theoretical basisto programming.

Fiber Optics

The last decade has seen the rapid develop-ment of a new communications medium, infor-mation transmitted by pulses of laser lightthrough thin glass fibers. These optical fibers

can transmit far more information than cancopper wires or coaxial cables of the same size.Since information is transmitted as light, fi-ber optics are resistant to both electrical in-terference and eavesdropping. The potentialcapacity and economy of optical fiber is suchthat it not only will help keep communicationscosts down, but, over the long term, it will re-lieve the growing congestion of the radio spec-trum and may ease some of the demands forsatellite communications.

Optical fiber is already used for communi-cation lines between some major cities in theUnited States, and a transatlantic cable usingoptical transmission will be laid in the late1980s. By the end of the 1980s optical fiberwill be used heavily in nearly every stage ofthe telecommunication systems, from local in-traoffice networks to long-distance lines.

Although European firms are also develop-ing capability in fiber optics, the principalforeign competition has been from the Japa-nese, whose technology is roughly at a parwith that available from U.S. firms. Since alarge world market for fiber optical hardwareis growing, innovation in this technology iscritical to U.S. competitiveness in telecom-munications.

The technology of the fibers, themselves, isstill advancing rapidly along three generallines: 1) techniques to increase the informationcapacity of the fibers; 2) improvements to thetransparency of the glass, which allow longerdistances between “repeaters,” devices thatdetect the signals and retransmit them ondown the line; and 3) improved process tech-nologies for manufacturing fibers. Hence, eventhough fiber optics is now in the commercialmarketplace, the need for basic research con-tinues and the current state-of-the-art is a longway from any fundamental limits to this tech-nology.

Applied research and development on fiberoptics, because it is expensive and capital-intensive, is concentrated in a few large lab-oratories; and, since the economic reward fortechnological advance is clear, it is done pri-marily by industry. However, long-term devel-opment of fiber optics will depend on funda-

Ch. I—Introduction and Summary ● 1 3

-

“

l ,.

Photo credit: AT& T

Telephone lines circa 1887. Fiber optics is only themost recent of many technological developments thathave increased the capacity and shrunk the physical

size of communication lines

mental research in materials and solid-stateelectronics. Increased Federal funding of re-search in universities would achieve the dualobjectives of stimulating fundamental researchand training people at the Ph.D. level in fiberoptics technology. A current lack of trainedresearch personnel has been pointed out to bea handicap to progress in this field.

Artificial Intelligence

“Artificial Intelligence (AI)” is a term thathas historically been applied to a wide vari-ety of research areas that, roughly speaking,are concerned with extending the ability of thecomputer to do tasks that resemble those per-formed by human beings. These capabilities

include the ability to recognize and translatehuman speech, to prove the truth of mathe-matical statements, and to win at chess.

For two reasons public attention has re-cently focused on an area of AI called expertsystems. First, expert systems are seen to bethe first significant commercialization of 25years of research in artificial intelligence; and,secondly, they form the basis of the JapaneseFifth Generation Computer Project.

An expert system is an “intelligent” com-puter information retrieval system designedfor use in a specific decisionmaking task, saymedical diagnosis. It stores not only data,itself, but the rules of inference that describehow an expert uses the data to make decisions.By asking questions and suggesting coursesof action, the expert system interacts with adecisionmaker to help solve a problem. AI re-searchers now understand this process wellenough to be able to build profitable commer-cial systems for applications in very narrowareas of specialization, for example, advisingon repair of telephone cable or diagnosingpulmonary disease.

Although few computer scientists doubtthat sometime in the next century, people willinteract with computers differently than theydo now and that computers will show behaviorthat we would now call “intelligent,” they dif-fer over the nearer term significance of expertsystems. Some maintain that expert systemsrepresent an interesting and profitable, yetsmall, application area. In their view, AI con-tinues to move slowly and is a long way fromrealizing the potential predicted for it. Othersargue that expert systems are a model of howmost computing will be done within a decade;and, hence, that future U.S. leadership in com-puter science will depend on an aggressive pro-gram of research in this field.

The Department of Defense has always beena dominant supporter of AI research, includ-ing work on expert systems. It has funded thiswork by establishing and funding large-scalelaboratories at a few academic and nonprofitresearch centers. Civilian support is limited.

——

74 ● Iformation Technology R&D: Critical Trends and IssueS

NSF has steadily supported basic research inAI, but has done so at a relatively low fund-ing level, and NASA has funded work support-ing applications to autonomous spacecraft.DOD experience suggests that AI research isbest performed in large laboratories with ac-cess to largc-scale computing facilities and col-lections of specialized software. Some areas ofresearch will begin to require access to special-ized hardware and the facilities to design andbuild new systems.

This field in particular is experiencing anacute faculty shortage. The new commercialexcitement in A/I comes after a long periodof relative disinterest by funding agencies. Asa result, there are few researchers and research

centers working in the field, and competitionfor limited human resources is sharp. In par-ticular, experienced faculty are being drawnoff campus just as demand for newly trainedgraduates is growing rapidly.

The increased public attention and excite-ment about the promise of AI holds somedangers to the field. If Federal and privatesupport is contingent on high expectations forsignificant short-term advances, failure tomeet those expectations could result in anabrupt withdrawal of funding. Yet AI is a verydifficult field of research, just in its infancy.Much basic science remains to be learned inAI, and steady long-term support is requiredfor that learning.

Issues and Strategies

The level and nature of R&Din all fields, in-cluding information technology, is influencedby many social, economic, and political factorsbeyond those specifically related to scienceand technology. For example, the overall costof capital and the general health of the na-tional economy will affect management’s abil-ity and incentives to invest in R&D. Trade andexport policies that affect the access of domes-tic firms to foreign markets or vice versa willinfluence the willingness of U.S. firms to in-vest in innovation.

Some observers have suggested that seniorcorporate managers in many firms have down-played investment in R&D, which promisesrisky and long-term payoffs, in favor of short-term gains. According to this view, althoughthese managers may have been reacting ra-tionally in terms of immediate economic incen-tives and the desires of stockholders, the re-sult of their decisions has been a low rate ofinnovation and a long-term loss of competi-tiveness.

Although these broad factors set an over-all environment for R&D decisions, Congressalso deals with several specific policy issuesthat directly relate to science and technology.In this study, OTA concentrated on those di-

rect issues, particularly as they relate to in-formation technology.

Impacts of TelecommunicationsDeregulation

During the last few years, the fundamentalstructure of the telecommunications industryhas changed. AT&T has been subdivided intoseveral smaller (although still very large) in-dependent firms. Many telecommunicationproducts and services traditionally providedby regulated monopolies are now providedthrough a competitive marketplace. Improvedinnovation and international competitivenessin the telecommunications industry is one ofthe benefits intended by proponents of deregu-lation, and OTA expects such increased inno-vation will occur over the next decade. Theopen question is whether that burst of innova-tion will be at the expense of basic researchthat would lay the foundation for future inno-vation.

However, although technological develop-ment should increase, many science policy ex-perts have worried about the impact of deregu-lation on basic research. In particular, theyview Bell Laboratories as a national scientificresource and have expressed concern that anew, competitive AT&T will have decreased


Photo credit: AT&T

Alexander Graham Bell opening the first telephone line to Chicago


commitment to the basic research that hastraditionally marked at least part of the Lab-oratories’ activities.

OTA found that, although laboratory man-agement will be admittedly more interested inthe contribution of research programs toAT&T business, short-term effects on basic re-search will likely be small. In the first place,most large firms in historically competitiveinformation technology areas, such as comput-ers, have a history of supporting long-termresearch. Furthermore, AT&T senior manage-ment has a strong appreciation for the contri-bution that research at the Laboratories hasmade to the growth of the telephone system.They are unlikely to make a sudden shift inpolicy in the foreseeable future. Finally, mostof the areas in which AT&T Bell Laboratorieshas particular scientific strength are thosethat, although fundamental, are clearly relatedto information technology-e. g., solid-statephysics and computer science. Since the linebetween basic and applied research in thesefields tends to be particularly hazy, fundamen-tal work is likely to continue to be supportedeven were management to adopt a short timehorizon for results.

Long-term prospects for basic research at.AT&T Bell Laboratories will depend, in part,on AT&T’s success as a competitor in the com-puter and telecommunications industry. Thislong-term success will depend on many factorsincluding the regulatory environment, the eco-nomic climate, the skill of AT&T managementin operating under new rules, and the con-tributions made by the Laboratories to in-novation.

The divestiture has also introduced a newand potentially important organization, BellCommunication Research (Bellcore), to the in-formation technology R&D arena. Bellcoreserves the seven Bell operating companies,which hold it jointly. It commands the re-sources of a significant number of researchersfrom the pre-divestiture Bell Laboratories.Bellcore’s relationships with the operatingcompanies and its research activities are in thegenesis stage at present; thus, some time willbe required to determine its overall contribu-tion to the national R&D effort.

At the same time, the former Bell Labora-tories was a major performer of informationtechnology research in the United States, andthe divestiture in the telecommunications in-dustry is a structural and regulatory changeunprecedented in scale. The impact of thischange on industrywide research includingthat done at AT&T Bell Laboratories and Bell-core should be closely monitored by Congress.

Scientific and Technological Manpower

The Federal Government has historicallyassumed a role in stimulating the productionof technical manpower in areas of national in-terest. Policy has focused on increasing boththe quantity and quality of technical gradu-ates, but it has also played an important rolein equity. Federal scholarship and fellowshipprograms are vital mechanisms to eliminatefinancial barriers to entry into technologicalcareers.

Information technology is a fast-growingfield of national importance, and Federal pol-icies for developing R&D manpower in infor-mation technology may be deemed warrantedby Congress. However, those policies shouldbe broadly focused rather than aimed at nar-row fields of specialization, and they shouldbe consistent and steady over the long term.

Any policy designed to increase technicalmanpower will not take full effect for severalyears. The delays are inherent in the educa-tion system. They represent both the time ittakes educational institutions to respond bydeveloping new instructional programs andhiring staff, and the time it takes for a stu-dent to pass through the system. Hence, thesepolicies should be directed, not to short-term,but to long-term needs. If addressed to cur-rent shortages, Federal programs run the riskof overproduction by the time they take effect,for needs may have changed or labor marketforces such as high wages may have alreadyencouraged adequate new entrants in the field.

Detailed long-term needs for technical man-power are hard to assess in a field changingas rapidly as information technology. Hence,projections of future requirements differ sig-nificantly from one to another. They do not

Ch. I—introduction and Summary ● 1 7

account well for the appearance of new fieldsof specialization, such as the recent emergingarea of fiber optics, nor the sudden growth inimportance of older fields of research such asartificial intelligence. Thus, policies that dealbroadly with science and engineering manpow-er in all fields are most likely to be successful.

OTA found that, although some shortagesexist in specific fields, the long-term outlookis promising. Enrollments at university pro-grams, both at the undergraduate and gradu-ate level have been increasing rapidly, andeducators from some institutions have ex-pressed concern that in a few years there maybe an overabundance of computer scientistsand engineers, at least at the bachelor’s degreelevel. Similarly, as competition for admissionto undergraduate and graduate programs grows,so should the quality of the graduates.

Science education policy also addressesequity issues. Undergraduate and graduatetechnological education is increasingly expen-sive. Access barriers to high-paying careers ininformation technology could become formi-dable for some parts of society, due both tocost and poor preparation at the precollegelevel. To help offset these barriers, Federal pol-icy may need to address scholarships, fellow-ships, traineeships, and other forms of directassistance to students, as well as the improve-ment of precollege science education.

Changing Roles of Universities

The United States has traditionally lookedto its university system whenever nationalneeds called for improved technological inno-vation and the development of new expertise.This dependence on the education system,probably unparalleled in the world, has oftenresulted in experiments in institutional struc-tures associated with higher education. TheMorrow Act established the land grant col-leges to develop a scientific basis for agricul-ture and to train farmers in modern tech-niques. During World War II and after, majorfederally funded research laboratories were es-tablished in association with universities. Ex-amples include the Lincoln laboratories at theMassachusetts Institute of Technology, the

Lawrence Radiation Laboratory at the Univer-sity of California, Berkeley, and the Jet Pro-pulsion Laboratory at the California Instituteof Technology.

Once again, in response to some of the con-cerns raised earlier in this chapter, manyuniversities are experimenting with new struc-tures designed with the dual objectives of im-proving the quality of technical education oncampus and improving the flow of researchresults into industrial innovation. In contrastto many past experiments, these new struc-tures are, by and large, not in response to Fed-eral programs. Rather, many of them respondto initiatives both from industry looking forcloser ties with academic programs and fromStates who see strong academic technical pro-grams as attracting high-technology industryand, thus, serving as stimuli for economic de-velopment.

Some potential issues have arisen, but todate problems appear to have been or are be-ing resolved to mutual satisfaction by nego-tiation between the academic institutions andthe industrial sponsors. They include the fol-lowing:

● How intellectual property rights are dis-tributed among researchers, their institu-tions, and the industrial sponsors.

● When and under what circumstances re-search results may be published.

● Who establishes research priorities andstandards of scientific quality.

Over the longer term, other issues will alsobecome important, such as the overall influ-ence on the directions of academic basic re-search. Another long term concern is the po-tential imbalance of campus resources andattention between science and technology onthe one hand and other important scholarlyfields such as foreign studies, social science,the arts, or the humanities.

Another question of equity concerns the bal-ance between institutions, themselves. In thefirst place, are these programs “zero-sumgames” in which the gains of a few institutionsare mainly at the expense of all the others?Secondly, although the top research institu-tions have been able to negotiate arrange-


ments that, in their view, do not threaten theirother academic roles, less influential schoolswill have far less bargaining power.

Since many of these institutional experi-ments are new, it is too early to tell whetherthey will be successful in meeting their spon-sor’s expectations or whether the negotiatedsolutions to the major issues will work. How-ever, it is important that Federal science agen-ciestwo

●

●

watch these developments carefully forreasons:

Many of these new research institutionswill be performers of federally funded re-search or will be cosponsors of federallyfunded research.Many of the issues under negotiation be-tween universities and industry, such asownership of intellectual property rightsand restrictions on publication are cur-rently echoed as problems in government/university relationships.

Foreign Programs

Programs in other countries designed to pro-mote innovation in information technologyhave received attention in the United States.These programs have raised both concernabout increased competition in technologiesin which the United States has traditionallybeen a world leader, and they have providedmodels to those who suggest that the UnitedStates also needs to develop more coherentprograms to foster innovation by domestic in-dustries. Some examples of such programs arethe following:

●

●

The Japanese Fifth Generation ComputerProject is an attempt to move beyond cur-rent concepts of computer design to devel-op an entirely different type of machinebased on artificial intelligence concepts.The French La Filiere Electronique Pro-gram is a five-year national informationtechnology R&D program with long-termgoals of strengthening the French elec-

●

●

tronics industry and developing technol-ogy for social applications.ESPRIT, is a pan-European program in-tended to draw together the technicalresources of Europe to focus on R&D ininformation technology.The British Alvey program, constitutedin part as a British response to the Japa-nese Fifth Generation Computer Project,is a program of government support forresearch and development in semiconduc-tors and computer software.

Regardless of assessments of the specificprospects for any one program or its implica-tions for U.S. policy, they demonstrate in totalthe new competitive world that is developingin information technology. No longer can theUnited States expect to maintain unques-tioned technological leadership and unchal-lenged domination of world markets in infor-mation technology. Other nations now see itin their own interests to build scientific andindustrial strength in these areas and are tak-ing steps to do so. U.S. science and technol-ogy policy will need to adapt to take this newreality into account.

However, the foreign programs being estab-lished do not necessarily constitute usefulmodels for U.S. policies, for several reasons.In the first place, many of them are too newto determine success. Secondly, each is tai-lored to the unique patterns of government/in-dustry/academic relationships as they exist incountry of origin and may not be workable inthe country. Finally, many of these programsare designed to address particular bottlenecksto innovation that may exist in the specificcountry. U.S. science and technology has a dif-ferent balance of strengths and weaknesses.For example, some countries have a lack ofventure capital, others a shortage of scientificmanpower, and still others have strong cultur-al barriers between business and academic in-stitutions that impede technology transfer.

Information technology R&D is, in many re-spects, an international activity. For example,some domestic firms are owned or partiallyowned by foreign firms (and vice versa). Many


large computer and communications compa-nies are multinationals that operate labora-tories in several nations; and many companieshave agreements to exchange and cross-licensetechnology with foreign counterparts. Scien-tists have always viewed basic research as aninternational activity. Thus, international con-ferences, scientific journals, and exchange ofresearchers are common.

Programs designed to strengthen and pro-tect technology as a purely domestic resourceneed to balance this goal against the naturallimitations posed by this international char-acter. For example, the managers of ESPRIT,a program designed to stimulate Europeantechnology, must decide whether and on whatbasis to admit or deny the participation ofmultinationals based, at least partly, outsideof Europe. Companies concerned about whatthey may view as stringent controls over tech-nology in one country can simply move theirR&D efforts to their laboratories in othercountries.

Science Policy

Federal programs designed to stimulate in-formation technology R&D exist in the muchbroader context of Federal science and tech-nology policy.

Congress needs to ask whether existingscience and technology policy serves the needsof information technology R&D both in termsof its unique characteristics and the potentialimportance of the technology to the nation.Similarly, it is important to ask whether pol-icies designed specifically to support informa-tion technology R&D are consonant with thebroader scope of science and technology policy.

OTA identified three major sets of policyissues that, although applicable to a broadrange of science and technology, seem particu-larly important in the context of informationtechnology.

Institutional Focus. —The Federal Govern-ment has traditionally and purposefully sup-ported research and development through

several programs administered by differentagencies. Despite occasional calls for centraliz-ing R&D support within a single science andtechnology agency, this historical approachwas accepted for two basic reasons. In the firstplace, agencies with specific technologicalneeds were considered to be best suited to sup-port R&D focused on their needs. Secondly,multiple sources have been considered healthyfor the support of science, since diversity andredundancy are both important attributes ofscientific research.

The issue of government organization is be-ing reexamined, in the case of informationtechnology. Arguments in favor of a more cen-tralized and/or coordinated approach are basedboth on the changing nature of R&D and onthe challenges posed by new foreign programsstressing R&D in fields such as microelectron-ics, computer systems, and communicationstechnology.

Research is becoming increasingly expen-sive. Research equipment such as a microelec-tronic fabrication facility or a supercomputercan cost several million dollars in capital costs,plus several hundred thousands of dollars ayear to maintain and operate. Salaries ofresearch-level technologists are escalating, andmuch research in information technology nowrequires large teams of different specialistsand technicians. This cost and complexity ofR&D not only may make redundancy and du-plication of effort a luxury, but may even makesome types of research impossible to initiatewithout coordination of Federal programs.

Some foreign research programs targeted atspecific technologies have claimed successesin the sense of capturing a world market-themost notable being the Japanese program forthe so-called “64k” computer memory chip forcomputers that stores about 64,000 charactersof information on a microchip. While Japanesesuccess in this market may be attributed tomany factors, the existence of a specific gov-ernment program may have been a major in-fluence. Seeing such programs targeted atselected technologies in Japan and Europe hasled some experts to call for the United Statesto respond with a similar approach to domes-tic R&D.

20 ● /formation Technology R&D: Critical Trends and Issues

Photo credit: NASA

A communication satellite being launched from aspace shuttle. Satellite communication technologywas stimulated by both Federal military and civilian

R&D programs

Such programs, if initiated, would not likelybe restricted to information technology alone,but would be focused on a wide range of tech-nologies deemed critical. It would be a signif-icant departure from the past approach thatfeatured multiple government programs ofsupport for basic research, agency-by-agencysupport of applied research based on missionneeds, and, with a few exceptions, private sec-tor support for innovation and product devel-opment.

Military and Domestic R&D.–Militaryfunding tends to dominate the Federal R&Dbudget, particularly in the case of informationtechnology. Such a high level of DOD involve-ment is not surprising considering the impor-tance of information technology to its mission.The policy issue is whether the high level ofdefense support provides adequate underpin-nings for the broad development of civiliantechnology or whether a case exists for strongcivilian agency support of R&D.

There is no doubt that past DOD and De-partment of Energy support of R&D in suchareas as microelectronics and computer designhas been a spur to the entire information tech-nology industry. However, there is evidencethat, as the technology becomes more mature,military requirements have begun to divergefrom civilian interests. For example, in thearea of supercomputers and very high speedintegrated circuits, there is a significant doubtthat the DOD Strategic Computing and veryhigh speed integrated circuit programs, al-though responsive to DOD needs, will alsoserve civilian needs. Hence, there may be aneed to boost civilian support in informationR&D.

Influence of International Competitive-ness.—The growing concern over economicissues, in particular, international competitiveness has led to issues in which attempts totreat technology as a national resource thatcan be held and protected conflict with tradi-tional approaches to science policy.

For example, concern about the interna-tional flow of scientific and technical informa-tion conflict with the historical science policygoal of establishing international cooperationin science and technology. Open publicationof some scientific material has been challengedon the grounds that the information should berestricted to the United States. Some havesuggested that the practice of admittingforeign science students, who comprise a sig-nificant percentage of degree candidates ininformation technology, conflicts with U.S.economic interests. Restrictions have been im-posed on foreign scholars attempting to attendU.S. conferences and seminars.

Other policies assume that the United Statesdoes not need to draw on foreign science.Hence, research project support typicallygives little attention to the potential benefitsof travel by U.S. scientists to foreign researchlaboratories and conferences. Unlike the prac-tices of other countries, little support is pro-vided to provide translations and analyses offoreign research papers and monographs. YetU.S. scientists, in general, do not have facil-


ity in some of the significant scientific lan- information scientists do not even have facil-guages of today, particularly Japanese. Due ity with traditional scientific languages liketo the weakening of foreign language require- French and German.ments in some graduate programs, many U.S.

Summary

During the past decade, information tech- to monitor and manage this change. In par-nology has grown to major economic and so- ticular, Congress will need to:cial importance in the- United States andabroad. The desire to maintain a competitive

●

posture in worldwide markets as well as theurge to realize potential social benefits of newcomputer and communications systems has fo-cused public attention on the process of inno-

●

vation, in particular, on R&D. OTA has foundthat, in response to these pressures, the sys-tem has been remarkably adaptable. Federal

●

programs are changing rapidly in response tonew perceived needs; science and engineeringenrollments are increasing in response to ananticipated need for more technological man-

●

power; and universities and industry arechanging their traditional patterns of researchand relationships. The policy problem will be

—

determine and maintain an appropriatelevel and balance of R&D support for in-formation technology (including researchon the social impacts of these tech-nologies),remove unintended barriers to govern-ment and private sector R&D efforts,assure that these changes in the R&D sys-tem do not have unintentional side effectson other sectors or goals of the U.S. sys-tem of science and technology, andassure access by researchers in all disci-plines to powerful new computational anddata communication technologies.

Chapter 2

The Environment for R&DInformation Technology

the United States

●m●m

Contents

PageIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Concepts for R&D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

The Roles of the Participants and the R&D Environment . . . . . . . . . . . . . . . . . . . . . . . . 28Federal Government Role in R&D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28Industrial R&D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34Universities’ Role in R&D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37Conflicts in Perspectives, Goals, and Policies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

Measures of the Health of U.S. R&D in Information Technology, . . . . . . . . . . . . . . . . . 43Information Industry Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43U.S. Patent Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

A Synthesis: The Changing U.S. R&D Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

TablesTable No. Page

l. Federal Obligations for Total Research and Development: Fiscal Year 1984 . . . . . . 282, Federal 0bligations for Basic Researching Information

Technology-Related Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293. Federa1 0bligations for Applied Research in Information

Technology-Related Fields. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294. The Top 15 in R&D Spending . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445. Federal and Department of Defense Obligations for Basic Research,

Applied Research, and Development, 1983 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 456. Federal and Department of Defense Obligations for Basic

and Applied Research by Field, 1983. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467. Technology Distribution of the Top 50 U.S. Patent

Electrical Categories 1978-80 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468. Percentages of U.S. Patents Granted in Information

Technology and in All Technologies, 1981 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499. Foreign-Origin U.S. Patents in Some Components of Information Technology. . . 49

IO. Foreign-Owned U.S. Patents unselected Telecommunications Classes, 1982 . . . . . . 49

FiguresFigure No. Page

l. Interrelationships in the R&D Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262. Science, Engineering, and Technician Employment

Within High-Technology Industries, 1980. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353. Federal 0bligations to Universities and Colleges for R&D Plant . . . . . . . . . . . . . . . 394. Estimated Ratio of Civilian R&D Expenditures to Gross

National Product for Selected Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455. U.S. Patents in Information Technology, SIC Codes 357, 365,366,367 . . . . . . . . . 486. Share of Foreign Patenting in the United States for the Three

Most Active Countries in Selected Product Fields: 1981 . . . . . . . . . . . . . . . . . . . . . . 507. U.S. Patent Activity of 10 Foreign Multinational Corporations, 1969-80 . . . . . . . . . 50

Chapter 2

The Environment for R&D in InformationTechnology in the United States

Introduction

“Information technology” is a generic termfor a cluster of technologies (discussed in detailin ch. 9) that provide automated capabilitiesfor

● data collection;● data input;● information storage and retrieval;● information processing;• communication; and● information presentation.

The information technology industry has be-come an integral component of U.S. industrialstrength. In common with other high-tech-nologyl industries, the robustness of informa-tion technology depends in part on a base ofresearch and development (R&D). However,several interacting factors are straining long-established U.S. policies vis-a-vis research anddevelopment:

●

●

●

●

rising costs and complexity of R&D;intensive competition for both domesticand foreign markets;limited resources; andaccelerating technological advances.

This report describes those factors and ex-amines their effects on the information tech-nology industry and its R&D base, raisingquestions both about current policies andabout proposals for improving the competitiveposition of the United States in internationalinformation technology markets.

‘The term high technology is used throughout this chapterto refer to those industries characterized by a high proportionof R&D expenditures per employee, or a significantly larger pro-portion of skilled workers than the industry average, and arapidly evolving underlying technological base. Thus computersand electronics are within the scope of this meaning; auto andsteel manufacturing are not, in spite of the recent trend to mod-ernize and automate.

The world’s major countries are coming toview the development of high technology—and particularly information technology-asa key to economic gains, important social ob-jectives, and national defense and prestige.These countries have adopted national indus-trial policies in information technology, in-vesting hundreds of millions of dollars in thehope of achieving preeminence, both in R&Dand in commercial markets. This growth offoreign competition2-especially from Japan–has stimulated a concomitant growth of R&Din the United States.

The trend toward internationalizing R&D,manufacturing, and distribution is increasing.American companies are deciding that tech-nological strength can also be improved bylicensing technology from other domestic andforeign businesses, by acquiring equity posi-tions in firms with needed technology, and byestablishing R&D and manufacturing opera-tions in foreign countries in order to obtainaccess to rapidly changing commercial ap-plications.

Figure 1 shows some important componentsof the R&D process and diagrams their inter-relationships. The remainder of this report willfocus on those components.

●

●

This chapter describes some of the keyplayers in the process and discusses somemeasures of health of the informationtechnology industry.Chapter 3 Presents four case studies, eachdealing with an important element in thecluster of technologies that comprise in-formation technology.

‘See also ht.ermtiond Competitiveness in EZectrom”cs (Wash-ington, DC: U.S. Congress, Office of Technology Assessment,OTA-ISC-200, November 1983).

25


SOURCE: Office of Technology Assessment.

●

●

●

●

Chapter 4 discusses the recent divestitureof AT&T in the context of its potentialeffect on the R&D activities of Bell Lab-oratories.Chapter 5 considers the availability of ●

trained personnel to the R&D process.Chapter 6 examines new university-indus-try institutional relationships and theirchanging roles in the R&D environment.Chapters 7 and 8 focus on some of the ex-

ogenous elements: the science and tech-nology policies of foreign governments inchapter 7; the science and technology pol-icies of the United States in chapter 8.Chapter 9 describes the technologicalunderpinnings of information technology,the directions of key research areas, andthe characteristics of the informationtechnology industry.

Ch. 2—The Environment for R&D in Information Technology in the United States ● 27

Concepts for R&D

R&D includes a wide variety of activities–ranging from investigations in pure science toproduct development. Because segments ofinformation technology draw on so manyscience, engineering and other disciplines—computer science, manufacturing, electrical,mechanical and industrial engineering, physics,chemistry, mathematics, psychology, linguis-tics—it is difficult to assign particular effortsto the general category of information tech-nology.

As defined by the National Science Foun-dation (NSF), R&D is categorized as follows:3

●

●

●

●

Research is systematic study directedtoward fuller scientific knowledge or un-derstanding of the subject studied.In basic research the objective is to gainfuller knowledge or understanding of thefundamental aspects of phenomena andof observable facts without specific ap-plications toward processes or productsin mind.In applied research the objective is to gainknowledge or understanding necessaryfor determiningg the means by which a rec-ognized and specific need may be met.Development is the systematic use of theknowledge or understanding gained fromresearch directed toward the productionof useful materials, devices, systems, ormethods, including design and develop-ment of prototypes and processes.

The definitions of “applied” and “basic” re-search are especially troublesome in a field asdynamic as information technology, in whichlaboratory concepts evolve into marketableproducts very rapidly. In the area of artificialintelligence, for example, the work is oftenbasic in the sense that it seeks new ways ofunderstanding complex symbolic processes,and applied in the sense that much of the workis directed at prototype applications. This

‘National Science Foundation, FederaJ Funds for Researchand Development Fiscal Yws 1981, 1982, and 1983, VolumeXXXI Detailed Statistical Tables, NSF 82-326 (Washington,DC, U.S. Government Printing Office, 1982), p. 1.

fuzziness has led to differing judgments as towhich projects are applied and which are basic,and to confusion in data collection, because theterms are not applied uniformly by Federalagencies, industry, or academia.

Apart from the difficulty in drawing a cleanline between basic and applied research, thereis an additional problem in identifying the setof industries that collectively comprise the in-formation industry. For purposes of this re-port, we use the term to include electronics,computer, and telecommunications equipmentmanufacturers, providers of computer-basedservices, and commercial software developers.

There are, then, two major areas of ambi-guity in any discussion of information tech-nology R&D: ambiguities inherent in designat-ing an effort as “information technology” andambiguities arising from the overlap of basicand applied research. Because of this, quan-tification of R&D efforts in information tech-nology is necessarily approximate and thenumbers cited in this report should be re-garded as estimates, not as “hard” data.

One further term, “innovation,” should alsobe clarified. As used in this report, innovationis a process that includes research and devel-opment, manufacturing or production, and dis-tribution. 4 The Nation’s innovative capacitydepends on the effective functioning of allparts of the process. Success in the market-place requires proficiency in some—not wellunderstood—combination of those parts. Thereare other factors that influence marketplacesuccess, such as the timing of the introductionof commercial products, the influence of entrepreneurs, and a variety of government policies.Thus, while excellence in research and devel-opment provides no assurance of leadershipin the commercial marketplace, it may verywell be a necessary (if not sufficient) ingredientof success.

‘International Competition in Advanced Technology: Deci-sions for Amen”ca, Office of International Affairs, National Rt+search Council, 1983, pp. 21-22.


The Roles of the Participants and the R&D Environment

Industry, universities, and Government(State, local, and Federal) are the three keycontributors to R&Din the United States. Theeffectiveness of the Nation’s R&D is depen-dent on the vitality of each of the participantsand on their interrelationships. Federal sup-port of R&D and Federal policies that main-tain a healthy economy encourage industrialinvestment in R&D.5 A vigorous industry, inturn, provides a large Federal and State taxbase, making possible added support for aca-demic institutions. For their part, well-financed academic institutions generate well-trained personnel as well as the dynamic knowl-edge base necessary to fruitful R&D efforts.

In addition to describing the roles of eachof the participants and the environment forR&D, this section identifies some of the di-verse changes taking place in the R&D en-vironment-those already in place, and othersin transition-that may profoundly modifysome longstanding institutional patterns.

Federal Government Role in R&D

The Federal Government plays several keyroles in information technology R&D. As amajor user of information technology products(about 6 percent of the total automated dataprocessing market and most of the market forsupercomputers), its requirements are of con-siderable interest to the industry. As a spon-

‘David M. Levy and Nestor E. Terleckyj, Effects of Govern-ment R&D on Pn”vate Investment and %xiucti”w”ty: A Macrosconomi”c Analym”s, National Planning Association, revised Jan.5, 1983, pp. 17-19.

sor of research, the Federal Government fundsroughly half of the total R&D carried out inthe United States6 and about two-fifths of theresearch by the electrical machinery/commu-nications industry (a key component of the in-formation technology complex). In addition,the Federal Government itself performs about$11 billion of R&D7 (table 1) in its own and con-tract laboratories.

Beyond that, the Government helps to shapethe environment in which private firms maketheir R&D decisions. In some cases, this is aresult of deliberate Government policy in-tended to stimulate (or suppress) industry in-vestment. At other times, the environment isaffected by uncoordinated actions—intendedto serve other purposes-taken by a varietyof Federal entities including the Federal Re-serve Board, the courts, regulatory bodies anda plethora of executive branch agencies includ-ing the Departments of Justice, Commerce, State, and Defense, the National SecurityCouncil, and the Environmental ProtectionAgency.

Government Funding of R&D inInformation Technology

The Federal Government provided about 65percent of the funding in 1982 for R&D inscience and engineering fields in the Nation’s

———— ..—‘Federal Support for R&D and Innovation, Congressional

Budget Office, April 1984, p. iii.‘Pmbalde Levels of R&D Exped”tums in 1984: Fomcas t and

Analysis, Columbus Division of Battelle Memorial Institute,December 1983, p. 1.

Table 1 .–Federal Obligations for Total Research and Development: Fiscal Year 1984 (Estimated)(millions of dollars)

Extramurala

United States and Territories

FFRDCSC FFRDCS

C FFRDCSC

administered administered Other administered Us.Industrial by industrial Universities by universities nonprofit by nonprofit State and local supported

Total Intramuralb firms firms and colleges and colleges institutions restitutions governments Foreignd

Total all agencies $45,497.0 $10,969.9 $22,957.4 $1,614.4 $5,270.7 $2,291.9 $1,335.5 $6832 $189.1 $1848

NOTES: aAll organizations outside the Federal Government that perform with Federal funds,bAgencies of the Federal Government.cFederally funded R&D centers.dForeign Citizens, organizations, governments, or international organizations, such as NATO, UNESCO, WHO, performing work abroad financed by the Federal

Government.SOURCE: “Federal Funds for Research and Development, Fiscal Years 1982, 1983, 1984, ” vol. XXXII, Detailed Statistical Tables NSF 83-319, p. 30

Ch. 2—The Environment for R&D in Information Technology in the United States ● 2 9

universities and colleges.8 Data collected bythe National Science Foundation (see tables2 and 3) indicate that Federal obligations forbasic research in fields related to informationtechnology included $103.7 million for com-puter science and $115.4 million for electricalengineering in fiscal year 1984. For applied re-search, funding levels are $145.8 million forcomputer science and $568.3 million for elec-— . . —

8Nationa.1 Science Foundation, Early Release of Summary Sta-tistics on Academic Science/Engineering Resources, Divisionof Science Resources Studies, December 1983. Also see Fed-eral R&D Funding The 197585 Decade, National Science Foun-dation, March 1984, p. 11.

Table 2.—Federal Obligations for Basic Research inInformation Technology-Related Fields

(millions of dollars)

Computer ElectricalYear science engineering Total

1974 . . . . . .1975 . . . . . .1976 . . . . . .1977 . . . . . .1978 . . . . . .1979 . . . . . .1980 . . . . . .1981 . . . . . .1982 . . . . . .1983a , . . . . .1984a . . . . . .

NA

$26.5931.0240.2842.9646.2252.2167.4580.25

$103.66

$38.4547.7653.0855.1457.4162.0370,5978.5193.6391,89

$115.38

NA

$79.6786.1697.70

104.98116.80130.71161.07172.14

$219.04aNatiOna} science Foundation estimatesNA—Not available.

SOURCE National Science Foundation, “Federal Funds for Research and De-velopment, Detailed Historical Tables” Fiscal Years 1955-S4, ” p 275

Table 3.—Federal Obligations for Applied Researchin Information Technology-Related Fields

(millions of dollars)

Computer ElectricalYear science engineering Total

1974 . . . . . . NA $230.79 NA1975 . . . . . . 239.201976 . . . . . . $46.99 244.61 $291.601977 . . . . . . 58.34 327.59 385.931978 . . . . . . 66.97 375.22 442.191979 . . . . . . 63.31 355.84 418.151980 . . . . . . 82.38 446.56 528.931981 . . . . . . 69.32 478.17 547.481982 . . . . . . 103.49 518.56 622.051983 . . . . . . 121.18 525.75 646.921984a , . . . . . $145.85 $568.33 $714.18aNatlonal science Foundation estlfnates.

NA—Not available

SOURCE National Science Foundation, “Federal Funds for Research and De.velopment, Detailed Historical Tables: Fiscal Years 1955-84, ” p 327

trical engineering. These categories aloneamount to $933 million in basic and appliedR&D funding. In addition, there are otherR&D areas related to information technology—e.g., mathematics, physics, and materialssciences.

The Department of Defense (DOD), whichhas the largest of the Federal agency R&Dbudgets, is becoming increasingly dependenton electronics and computer science. By 1985,those fields will absorb nearly 25 percent ofthe total DOD R&D spending.’ Within DOD,the Defense Advanced Research ProjectsAgency (DARPA) has heavily funded effortsin artificial intelligence, microelectronics, net-working and advanced computer architecture.

No other agency compares with DOD, whichaccounts for about 60 percent of the FederalR&D budget.l0 NSF, for example, accounts forabout 3 percent of the Federal total, with afiscal year 1983 appropriations of just over $1billion” and $1.32 billion and $1.5 billion infiscal years 1984 and 1985 respectively.12 NSFfunding, which is heavily weighted toward the“basic” end of the R&D scale, supports re-search through grants, scholarships, univer-sity laboratory modernization, the establish-ment of university-based “centers of excellence, ”and similar programs. Many of the informa-tion technology-related disciplines are fundedthrough NSF programs: communications, elec-trical engineering, optoelectronics, mathe-matics, physics, materials research, informationsciences, and so on. These information tech-nology-related fields accounted for over $90million in fiscal year 1984.

Since 1972, NSF has been the primary Gov-ernment force in creating university-indus-try cooperative research centers, providingsome of their startup funds, planning grants,and advice during their first years of opera-..__ ——.—.— -

Wited in a speech by Dr. Leo Young, Office of the Under Sec-retary of Defense for Research and Engineering, DOD, at theIEEE 1984 Conference on U.S. Technology Policy, Feb. 22,1984, Washington, DC.

Ioprobab]e Leve]s of R&D Experd”tures in 1984, op. cit.“Federal Funds for Research and Development, Fiscal Years,

1981, 1982, and 1983, op. cit., p. 25.lZFi9c~ yea 1985 Nat,ion~ Science Foundation Budget Wti-

mate to Congress.

30 . Information Technology R&D: Critical Trends and Issues

tion. The centers are expected to become self-supporting. 13 By the end of fiscal year 1984,NSF was involved with 20 of these centers andexpects to make awards to at least 10 newcenters in 1985.

A number of the centers are involved withinformation technology-related research. Forexample:

●

●

●

●

●

Rensselaer Polytechnic Institute’s Centerfor Interactive Computer Graphics is do-ing applied research in computer-aided de-sign (CAD).North Carolina State University’s Com-munications and Signal Processing Lab-oratory is primarily engaged in basic re-search.Both basic and applied research is beingperformed at the University of RhodeIsland’s Robotics Center.Ohio State University is doing basic andapplied research in robotic welding.Georgia Institute of Technology’s Centerfor Material Handling is engaged in ap-plied research.

The National Science Board has recommendedbroadened NSF support for engineering re-search, an area long neglected in favor of theagency’s traditional emphasis on basic scien-tific research. Congress approved about $150million for NSF grants for engineering re-search in fiscal 1985, about 10 percent of theagency’s total research budget.14 The Presi-dent’s budget request for fiscal year 1985 callsfor more funding for this purpose, as well asfor establishment of cross-disciplinary engi-neering centers at universities which would,among other effects, promote research on com-puters and manufacturing processes.15 Gov-ernment funding of engineering equipmentand facilities at universities decreased from$42 million to $17 million per year between1974 and 1981,16 but has been rising sincethen.

laDoD i9 aISO supporting this program through m $8 mill-ion grant, primarily for laboratory equipment.

“’’National Science Foundation Starts to Broaden Supportof Engineering Research, ” The Chronicle of Higher Education,Jan. 18, 1984, p. 17, and interviews with NSF officials, Janu-ary 1985.

“Ibid.16Probable Levels of R&D Expemh.tums in 1984, op. cit.

The Pattern of Government Funding of R&Din Information Technology

The Federal Government has had a long his-tory of funding R&D in information technol-ogy-related fields. It is currently the majorsponsor of those types of information technol-ogy R&D in which it has special interests.These include artificial intelligence, supercom-puters, software engineering, and very largescale integrated circuits (VLSI), all areas intheir technological infancy and with enormouspotential for military as well as commercial ap-plications. There is a long list of related tech-nologies that have been stimulated by Govern-ment-often defense or other mission agencies—sponsorship of R&D including radar, guid-ance systems, satellite communications, andmany others.

There are some historic examples of inten-sive Government sponsorship of technologi-cal development in areas where the potentialbenefit was expected to be great, but the risksand costs of research were high and thereforeunattractive to industry —e.g., computers,aviation and communications satellites. Oneof the classic illustrations of a successful, ma-jor Government contribution to informationtechnology R&D is in the field of satellite com-munications. The National Aeronautics andSpace Administration (NASA) (which current-ly accounts for about 7 percent of the FederalR&D budget) had the leading role in pioneer-ing technological progress toward commercialdevelopment, accelerating the time frame forthe introduction of this technology, influenc-ing the structure of the U.S. domestic and in-ternational telecommunications common carrierindustries, and effecting significant cost sav-ings over the long run.17

In these cases, the Government, through theundertaking of a number of risky and expen-sive R&D programs and with extensive pri-vate sector involvement, developed a largepool of baseline technology that served toprove the feasibility of geostationary satellite

171Mofis Teub~ ad Edwud Steinmuller, Government POl-icy, Innovation and Economic Growth: Lessons From a Studyof Satelh”te Comrnum”c.ations, Research Policy 11 (1982) 27-287,North Holland Publishing Co.


communications. These R&D programs werefor the purposes of proving the feasibility ofvarious technological advances such as geosta-tionary orbiting satellites, electromagneticpropagation of signals from outer space,traveling wave tubes, automatic station keep-ing, and aircraft communications. The NASAprograms initiated to undertake the extensiveR&D included the SCORE, ECHO, and RELAYprograms, the SYNCOM series of launchesthat paved the way for Intelsat I, the firstcommercial communications satellite, and theApplications Technology Satellites series. Thecosts for the RELAY, ECHO, and SYNCOMPrograms alone through 1965 were over $128million-an amount that few companies could–or would—commit, particularly consideringthat the feasibility of synchronous satelliteoperation was seriously questioned.18

It is also interesting to note that theseNASA programs likely had some importantside-effects on the structure of the U.S. inter-national satellite communications industry.Because AT&T was the only private companyto have heavily invested its own funds for sat-ellite communications R&D—with focus onthe nonsynchronous TELSTAR system—it islikely that AT&T would have dominated thenew international and domestic satellite com-munications services industry. Instead, theNASA programs, through continuous trans-fer of technology to, and close interaction with,commercial firms stimulated the competitionthat followed the 1972 Federal Communica-tion Commission’s decision allowing open en-try into the domestic satellite communicationsservices industry.

The market for the supply of satellite com-munications equipment was also open to com-petition due to the expertise gained by NASAcontractors. In addition, the international sat-ellite network that evolved is owned and oper-ated by INTELSAT, an international consor-tium, with the U.S. portion owned and oper-ated by COMSAT, a broadly based private/public corporation.

181 bid., p, 277.

Other Federal Government Policies

The Federal Government has many othermeans for promoting (or suppressing) privatesector R&D activities including antitrust pol-icy, patent policy, tax credits, technologytransfer from Federal laboratories and feder-ally funded R&D centers, and the promotionof Research and Development Limited Part-nerships (RDLP). Export controls, whether fornational security or political purposes, serveas a negative influence in promoting privatesector R&D. A major source of corporate fund-ing for R&D, international sales, is lessened,and the open exchange of technical data islimited. Six policies intended to promote pri-vate sector R&D are reviewed below.

PATENT POLICYPrevious policies assigning Federal owner-

ship of patents based on Government-fundedR&D have been modified in recent years withthe intent to stimulate patenting and commer-cialization of invention. Public Law 96-517(1980), which permits small businesses, not-for-profit institutions, and universities to ob-tain patents based on Government-sponsoredR&D, is intended to encourage university-industry collaboration and patenting. TheGovernment’s right to patent ownership wasfurther reduced by Presidential Memorandum(February 1983). This memorandum modifiedthe Federal Acquisition Regulations by ex-tending the concepts of the current law toallow all Government contractors to retainpatent rights.

There are obvious tensions in this situation,since it is sometimes argued that the publicshould own patents derived from research ithas funded. The counter-argument is thatGovernment-owned patents tend not to be-come commercialized and the public reaps noreal benefit from them. For example, Federalefforts to license its patents have resulted ina meager 4 percent being licensed, in contrastwith 33 percent for university-owned patents .19

19 Lm~ing Felker, Us. I)epartrnmt of Commerce, Office ofProductivity, Technology, and Innovation, during interviewswith OTA staff, January 1984.

32 • Information Technology R&D: Critical Trends and Issues

Allowing universities and businesses to retainownership stimulates commercialization, butmay also have the effect of distorting theuniversity’s traditional role as a developer offundamental knowledge.

TECHNOLOGY TRANSFERThe Stevenson-Wydler Technology Innova-

tion Act of 198020 was an attempt “to improvethe economic, environmental, and social well-being of the United States,” through suchmeans as: establishing organizations in the ex-ecutive branch to study and stimulate tech-nology; promoting technology developmentthrough the establishment of centers for in-dustrial technology; stimulating improved uti-lization of federally funded technology devel-opment by State and local governments andthe private sector; and by other activities.21

The act has been selectively implemented.Most of the Federal Laboratories have estab-lished Offices of Research and TechnologyApplications (ORTAs) which collect and dis-seminate the results of their respective Lab-oratory’s research. The Center for the Utili-zation of Federal Technology, in the NationalTechnical Information Service of the Depart-ment of Commerce, serves as a central clear-inghouse.

However, the heart of the act, the Coopera-tive Generic Technology Program, has neverbeen implemented. In February 1984, the Sec-retary of Commerce issued the first report tothe President and the Congress on the prog-ress of Federal activities conducted pursuantto the Act.22 It appears that much of the workcited in the Report as “Stevenson-Wydler”activities would have been performed even ifthe act had not existed. For instance, the newpatent policies discussed above and the R&DLimited Partnership (RDLP) discussed belowwere both cited as “Stevenson-Wydler’ initia-tives.23

— — —‘Public Law 96-480.“For more details see the Stevenson-Wydler Technology In-

novation Act of 1980, Report to the President and the Congressfrom the Secretary of Commerce. February 1984.

‘gIbid.“Ibid., p. 4.

The Act has probably had an effect on theactivities of the Federal Laboratory Consor-tium (FLC). In 1984, the FLC established anaward for excellence in technology transferand issued 26 such awards. The Federal Lab-oratories, however, are mission-oriented; andno Federal Laboratory has a mission empha-sizing the development of commercial technol-ogies.24 Thus, the concept of cooperative gen-

eric research laboratories envisioned by theAct has not been tested.

TAX CREDITSTax credits for businesses performing R&D

have been expanded through the EconomicRecovery Tax Act (ERTA) of 1981, which willexpire in 1986 unless extended.25 A key pro-vision of ERTA allows companies to claim 25percent tax credits for their qualified R&Dcosts above their average expenditures for theprior 3-year period. The law also allows for in-creased deductions for manufacturer’s dona-tions of new R&D equipment to universities,and provides a new capital cost recovery sys-tem for R&D equipment (modified later by theTax Equity and Fiscal Responsibility Act of1982–TEFRA).

Opinion is mixed as to whether this taxcredit is effective in stimulating R&D invest-ment. One study,26 based on the limited avail-able data, observes that the tax credit maywell be helpful in encouraging increased R&Dbudgets. However, a current study finds verylittle effect on increased R&D spending dueto the tax credit.27 Battelle Memorial Instituteattributes at least part of the increased indus-try investment in R&D to the tax credits.28

.—~ioTA MemormdD, “l)weloprnent and Diffusion of COlll-

mercial Technologies: Does the Federal Government Need toRedefine Its Role?” March 1984, p. 26.

“public Law 97-34, August 1981.‘Eileen L. Collins, An Early Assessment of Thins li&D In-

centives Prow”ded by the Econonu”c Recovwy Tax Act of 1981,National Science Foundation, PRA Report 8307, April 1983.

“preliminary findings of an ongoing study by Edwin Mans-field, financed by the National Science Foundation.

as~obab]e ~ve]s of R&D Expencli”tures in 1984, pp. 2, 11-12, op. cit.

Ch. 2—The Environment for R&D in /formation Technology in the United States • 3 3

R&D LIMITED PARTNERSHIPS

The Department of Commerce has been pro-meting wider use of R&D Limited Partner-ships (RDLPs), and offers advisory assistanceto businesses in their use. RDLPs are intendedto attract venture capital to commercial R&Dby limiting the potential losses to the venturecapitalists while still permitting them to re-tain patent rights, if any, and to have pros-pects for receiving royalties or a subsequentbuy-out by the company. RDLPs are some-times used for conducting R&D with relativelyshort-term payoffs—e.g., 3 to 4 years. The useof RDLPs primarily affects the segments ofthe innovation sequence from prototype throughproduct development.

During 1982, $275 million is estimated tohave been invested in RDLPs, mainly throughlarge brokerage houses. In 1983, the amountis estimated at $490 million; in 1984, it was$220 million.29 Although these investmentshave tended to go for biotechnology, somehave been allocated to information technologyprojects in fields such as computers, software,microelectronics, telecommunications, robot-ics, and artificial intelligence. The drop in fund-ing in 1984 is believed by investment bankersto be due to two trends. First, there is a gen-eral drop in investor interest in high tech-nology. Second, some investors appear to beconcerned about possible changes in tax lawsthat may give less favorable treatment ofR&D tax deductions and capital gains.

ANTITRUST POLICYThere have been administrative proposals

and congressional bills that would limit theuse of the treble damage penalty against com-panies found guilty of antitrust violations andestablish clearer guidelines for companies con-sidering cooperative research activities.

There have been arguments noting that theantitrust laws have not had a chilling effecton cooperative research since they are rarely

ZgData based on interviews by OTA staff with officials fromthe Office of Productivity, Technology, and Innovation, U.S.Department of Commerce, and key sources in the investmentbanking community, Jan. 29, 1985.

used. However, until recently businesses havebeen exceptionally cautious about such ven-tures because of concern over litigation.

Some of the questions that arise in consid-ering more liberalized interpretation of anti-trust legislation concerning joint research are:

●

●

●

Will U.S. companies, long accustomed toperforming much of their R&D individu-ally, be able to adapt swiftly to a differentmode of operation? What will be the realcommitment to shared research? How willintellectual property issues be resolved?Will there be new opportunities for collu-sion among joint R&D partners that re-create historical antitrust problems?Will joint R&D dilute the benefits of com-petition even in basic research? Will smallfirms be disadvantaged?

in the closing days of the 98th Congress, theNational Cooperative Research Act of 1984was passed. The Act eliminates the treble dam-age penalties in antitrust cases involving jointR&D ventures when those ventures meet theconditions of the Act, in particular, by provid-ing prior notification to the Federal TradeCommission and the Justice Department.

INDUSTRIAL POLICYAn important topic debated in Congress

concerns industrial policy and the appropriaterole of Government in strengthening industry.One approach would provide an environmentgenerally conducive to industry reinvestment,productivity improvements, and increasedcompetitiveness through tax, antitrust, pat-ent, and other policies. A different approachwould assist selected industries, create a high-level industry-labor-Government advisorycouncil, and provide loans and loan guar-antees.

Among the issues that surround the debatesare whether the Government could be effec-tive in selecting industries for support; whetherbusinesses, without further encouragement,would invest their resources in areas mostbeneficial to the Nation’s competitiveness; andwhether foreign national industrial policiespose insuperable problems for U.S. businesses.

—


There may also be important lessons from thevarious foreign experiences with targeted in-dustrial policies, many of which may not besuitable as models for the United States (seeCh. 7: Foreign Information Technology R&D).

Industrial R&D

The corporate motivation for performingR&D centers about the need to maintain or im-prove market share and profitability for boththe short and long term. For high-technologybusinesses in general, R&D plays a critical–although not singular-role in helping the firmto sustain or improve its competitive position.

Funding and Licensing of Research

The information technology industry in theUnited States spent about $10.8 billion forR&D in 1983. As is typical of a high-tech-nology industry, the firms in information tech-nology often spend a large proportion of theirsales revenue on R&D—for supercomputermanufacturers the ratio was nearly 20 percentin 1982, and it is believed to have been com-parable in 1983. In 1983, overall R&D spend-ing by computer manufacturers rose by 19.5percent; spending by computer software andservice vendors rose by 38.9 percent; andtelecommunications R&D rose by 31 percent.The importance of R&D to these firms can beseen from the fact that even during the recenteconomic recession they continued to makesubstantial R&D investments.

While information technology companiesperform much of their R&D in-house, they alsomake use of research originating in univer-sities, other companies, and the Federal Gov-ernment through licensing and other arrange-ments for technology transfer. Licensing andcross-licensing are often used as means of ac-quiring technology quickly and for recoveringR&D expenses.

Protection of Research ResultsLeadtimes in research and product develop-

ment are very important for capturing markets,recovering R&D expenses, and contributingto profitability and further R&D investment.A 6-month leadtime in getting products intocommercial markets can make the difference

between market dominance and substantiallosses. The information technology industry’ssignificant investments in R&D, the high mo-bility of technical personnel, and the increas-ing internationalization of R&D, encouragerapid diffusion of technical data and frequentintroduction of new products. These charac-teristics intensify the need for legal protectionof new ideas and products.

Certain areas of information technology areespecially vulnerable to “borrowing” and thedegree of legal protection available is uncer-tain. Software, for instance, can be copywriterbut cannot be patented except in certain in-stances. Policy is being made in the courts, vir-tually on a case-by-case basis, and the re-sultant ambiguities satisfy no one. The entireproblem of intellectual property rights has be-come a matter for national attention.

Industry-University Links

Chapter 6 of this report describes in detailthe relationship of the information technologyindustry and the universities. Internationalcompetition is causing U.S. industry to be-come increasingly sensitive to the importanceof academia both as a performer of informa-tion technology-relevant basic research and asa supplier of trained personnel.

The information technology industry is amajor “consumer” of technically trained per-sonnel. As shown in figure 2, the office/com-puting and communications industries are ri-valed only by the aircraft and parts industryin terms of overall employment of scientists,engineers, and technicians. According to sta-tistics compiled by the National Center forEducation Statistics,30 some 21,400 electrical,electronic, and communications engineeringand 25,500 computer and information sciencemajors graduated in 1982. Within those disci-plines, less than 2 percent are unemployed.31

There is some controversy surrounding theputative shortage of future manpower for in-formation technology research (discussed at

~ONation~ Center for Education Statistics, Survey of EmnedDegrees Conferred, reported to OTA by Dr. Vance Grant, Jan.3, 1984.

‘] Congressional Budget Office, Defense Spendeng and theEconomy, Table A-7, p. 59, February 1983.


Figure 2.—Science, Engineering, and Technician Employment WithinHigh-Technology Industries, 1980

(Thousands)

o 20 40 60 80 100 120 140

Machinery, except electrical

Office and computing

Construction and related

General industrial

Metalworking

Other

Electrical machinery

Communication

Electronic components

Electronic industrial appliances

Other

Transportation equipment

Aircraft and parts

Guided missiles/space vehiclesa

Motor vehicles

Other

Instruments

Measuring control devices

Engineering/scientific equipment

Photographic equipment

Other

Chemicals

Industrial organic

Plastic materials

Industrial inorganic

Drugs

Other

-

I

J

confidentiality rules and/or Statistical reliability data release

SOURCE: National Foundation, Changing Employment Patterns of Scientists, Engineers, and Industries, 1977-80


length inch. 7) but there is certainly no presentoversupply of well-trained R&D personnel noris there doubt that industry is dependent onthe university system to produce the man-power necessary to maintain a sufficient levelof R&D.

Industry is both influenced by, and influ-ences, university training programs. In addi-tion to its needs for traditionally trainedgraduates, there is a growing need for gradu-ates with multidisciplinary training. For ex-ample, companies involved in fiber optic com-munications require researchers trained inboth physics and electronics-a combinationwhich is not part of traditional curricula. Asimilar situation applies in the “expert” sys-tems field, where a wider range of skills areneeded and broad scientific training is espe-cially valued. These shifting industry needsplace demands on universities to alter theircurricula and to create multidisciplinary in-stitutional structures, and they increasinglyrequire frequent retraining of professionaltechnical personnel.

In some cases, information technology firmsrequiring special skills not normally producedby academia have compensated for the short-fall by providing additional cross-training foremployees, by helping selected universities todevelop new curricula, or by furnishing sup-plemental teaching staff. For example, in 1983IBM announced that it would make $10 mil-lion available to support university research-ers and another $40 million earmarked for thedevelopment of curricula in computer-aided de-sign and manufacturing.

Foreign Government PoliciesPolicies and practices of foreign govern-

ments and companies can influence the prof-itability of U.S. companies or deny them mar-kets, and thus effectively restrict their abilityto invest in R&D. These policies and practicesinclude pricing exports at below cost in orderto capture larger market share and the advan-tage of scale, targeting specific advanced tech-nology markets through government-spon-sored industrial strategies, creating nontariffbarriers (e.g., discriminatory certification prac-

tices), restrictions on foreign direct invest-ments, exclusion of U.S. subsidiaries fromR&D programs funded by the host govern-ment, preferential treatment of domestic pro-ducers in government procurement, and ex-port credits.32

Jointly Funded Research

Recently, the industry has made what maybe the beginning of a major shift from its tradi-tional pattern of conducting independent R&D,toward undertaking some joint or cooperativeefforts. There are a number of examples inwhich companies are jointly supporting basicand some applied research through newlyformed cooperative organizations, which relyheavily on university and corporate research-ers. Among these new organizations are theMicroelectronics Center of North Carolina, theSemiconductor Research Corp., and the Micro-electronics and Computer Technology Corp. Adetailed discussion of these arrangements andthe policy issues arising from them is con-tained in chapter 6.

These cooperative research efforts werespurred by escalating R&D costs, by a per-ceived limited supply of science and engineer-ing talent, and by the apparent erosion of in-formation technology industry’s internationalcompetitive position. Some leaders in the in-dustry argue that it has neither the resourcesnor the time to continue its established pat-tern of across-the-board duplicative R&D. Thisdoes not mean that information technologycompanies intend to slacken their competitiveR&D work vis-a-vis proprietary technologies.If anything, the cooperative projects are ex-pected to lead to more innovation and morecompetition at the level of the participatingcompanies.

Cooperative research programs require acareful distinction between proprietory andnonproprietory technology. Nonproprietarytechnology is made up of:

● generic technology, consisting of scien-tific and engineering principles that form

32 For mom ~t~g, s Internat.iorlal Competition ~ Adv~~Technology: Decisions for America, op. cit., pp. 28-37.


a competitively neutral technology basethat can be shared by all firms withoutreducing the potential benefits for any onefirm; and

● infratechnology, consisting of the knowl-edge base necessary to implement productand process design concepts. It includessuch things as basic data characterizingmaterials, test methods, and standards.Like generic technology, the infratechnol-ogy is competitively neutral.

The various cooperative arrangements are con-cerned with the nonproprietary technologies.

Cooperative Government-industry develop-ment of nonproprietary technical standards isa related area important to industry. A recentexample is the cooperative effort of the Na-tional Bureau of Standard’s Institute for Com-puter Sciences and Technology (ICST) and 12information technology firms in developingand demonstrating networking technology foroffice systems. A similar cooperative projectis aimed at developing networking technologyfor the factory floor. These efforts are basedon the development of nonproprietary stand-ards. The programs, which began joint dem-onstrations in 1984, permit the products of dif-ferent manufacturers to work together com-patibly and therefore expand the market forthem.

In the long term, continued expansion ofU.S. cooperative research activities could havepolicy implications for the appropriate amountsand focus of Federal funds for R&D, for uni-versities’ needs for outside support, and forinvigorating segments of the university re-search environment—and the potential foraltering the status of U.S. R&D relative toother nations.

Universities’ Role in R&D

The exceptionally broad nature of the under-pinnings of information technology, and theescalating complexity associated with con-tinued advances based on research, indicatean increasing role for universities. Major ad-vances in fundamental knowledge are often

the result of decades of dedicated and expen-sive research-efforts which few commercialfirms would be willing to undertake. In eachof the four areas of information technologiesselected for the chapter 3 case studies (ad-vanced computer architecture, fiber optics,software engineering, and artificial intelli-gence), universities have made and are con-tinuing to make valuable research contribu-tions in technologies that the private sectorcommercializes.

The intensity and breadth of university re-search is dependent on a wide variety of fac-tors, ranging from the prestige of the institu-tion, graduate enrollment, ability to retainqualified faculty and researchers, adequacy offunding for researchers, adequacy of facilitiesand laboratory equipment, affiliations withmajor companies, and increasingly on interac-tions with other researchers domestically andinternationally. It is also dependent on a largeproportion of foreign graduate students– asmany as 50 percent in some universities-par-ticularly in disciplines such as engineering.

The intensity of university R&D is also de-pendent on the level of funding for researchprovided by the Federal and State govern-ments, as well as by industry. About 85 per-cent of the funding for university and collegeR&D came from external sources.33 Federaland State governments as well as industryprovide funding for research, for scholarships,for laboratory equipment, and for real estate.The universities accounted for about one-halfof all basic research expenditures in 1984, with70 percent of their funding provided by theFederal Government.

Laboratory Research Instrumentationand Facilities

During the past few years, problems con-cerning the obsolescence of university labora-tory research instrumentation and facilities,and a lack of access to supercomputer equip-ment, have been recognized in many academicdisciplines. This problem is not specific to in-

$~Nation~ &ience Foundation, Early Release Of summ~’Statistics, etc., table 1, December 1983.


formation technology, but this field is amongthose affected. A decline in university researchcapabilities could result in a significant declinein the overall rate of the Nation’s scientificadvance.

One study notes that when the appropriateinstrumentation needed to conduct specific re-search is unavailable, then research objectivesare altered to match that which is available.34

This source also reports that leaders withinthe scientific community have estimated thecost of updating university research equip-ment to lie between $1 billion and $4 billion.In particular, instrumentation with costs be-tween $100,000 and $1 million at U.S. researchuniversities is reported as becoming obsolete.35

One estimate stated that selected instrumentcosts have increased fourfold since 1970.36

Compounding the problem is the fact thatthe most up-to-date research equipment hasa short lifetime-only 3 to 8 years.37 Anotherstudy which compared university laboratoryequipment with that of industrial laboratoriesfound that the median age of university lab-oratory equipment was twice that of the equip-ment found in the laboratories of companiesperforming high-quality research.38

The same study also noted that until re-cently, not a single top-line supercomputerwas installed in service at a U.S. university.A number of foreign universities, however, are

‘—~4TeStim~y of Charles A. Bowsher, Comptroller General,GAO, before the Senate Committee on Commerce, Science, andTransportation, Subcommittee on Science, Technology, andSpace, Research Instrumentation Needs of Universities, May27, 1982.

‘sObsolescence of Scientific Instrumentation in ResearchUniversJ”ties, Emerging Issues in Science and Technology, 1981,A Compendium of Working Papers for the National ScienceFoundation, National Science Board, p. 49.

‘Su”ence, vol. 204 (1979), p. 1365, as reported in Obsolescenceof Sci”entifi”c Instrumentation in Research Um”versities.

“International Competition in Advanced Technology: Deci-sions for Ameri”ca, op, cit., pp. 47-48.

‘L. Berlowitz, R. A. Zdanis, J. C. Crowley, and J. C. Vaughn,“Instrumentation Needs of U.S. Universities,” ~“ence, vol. 211,Mar. 6, 1981, p. 1017, as reported in Obsolescence of ScientificInstrumentation in Research Um”versities.

equipped with supercomputers.3g In part as aresponse, the Federal Government is increas-ingly sharing its supercomputers with its con-tractors, many of which are universities. Dur-ing fiscal year 1984, $6 million was authorizedfor NSF to buy access to the equivalent of onesupercomputer for scientific use. This amountwas increased to $40 million in fiscal year 1985and will contribute to the establishment of sixor seven supercomputing centers nationwideover the next 5 years. In addition, four U.S.universities have recently acquired supercom-puters.

Further acquisitions of supercomputers byuniversities may be curtailed both by finan-cial limitations and by the difficulty of assem-bling the expert staff needed to maintain thefacilities. Consequences of obsolescence arelikely to include foregone opportunities to per-form frontier university research, less-than-optimal training for graduate students, a con-tinuation of the migration of faculty and newgraduates to industrial laboratories, and a de-terioration in the quality of U.S. instrumen-tation, because university researchers tradi-tionally provide valuable feedback andinnovative improvements to the instrumentmanufacturing community.

A number of factors have contributed to theobsolescence of university laboratory equip-ment. Among these are the long-term decreasein Federal funding for R&D plant in univer-sities and colleges since 196540 (fig. 3); an ap-proximately four- to six-fold increase between1970-80 in the costs of state-of-the-art instru-

— ———~gArnOng these me: West Germany’s Max Pl~* Institute

and the Universities of Karlsruhe, Stuttgart, Berlin, and KFA;Japan’s Universities of Tokyo and Nagoya; England’s Univer-sities of London and Manchester France’s Ecole Polytech-nique; and in Sweden, onehalf time access by universities toa major auto manufacturer’s supercomputer.

40’’The Nation’s Deteriorating University Research Facilities,A Survey of Recent Expenditures and Projected Needs in Fif-teen Universities,” prepared for the Committee on Science andResearch of the Association of American Universities, July1981, p. 4. This survey covered 15 leading universities and sixacademic disciplines.


120

100

20

01963 1965 1967 1969 1971 1973 1975 1977 1979 1980 19811982

Fiscal yearSOURCE: National Science Foundation.

mentation41; and the short lifetime of state-of-the-art equipment and high maintenancecosts. In addition, during periods of decreasedfunding, university laboratory administratorstend to forego instrumentation purchasesrather than reduce project staffs.42 Further,until recently, Federal funding for researchprojects did not allow for purchase of instru-mentation if it was to be shared with otherprojects.

Quantification of the ProblemUntil recently, much of the information con-

cerning the extent of the obsolescence prob-lem has been anecdotal. However, there arenow a number of initiatives to provide statis-tical data quantifying its scope and the effecton specific disciplines. In addition to the needfor suitable data collection, there is a need to—. ——.—

‘l’’ Revitalizing Laboratory Instrumentation, The Report ofa Workshop of the Ad Hoc Working Group on Scientific In-strumentation, ” National Research Council, March 1982, p. 1.This source estimates a sixfold increase in costs, while othersestimate a fourfold increase.

‘Testimony of Charles A. Bowsher, Comptroller General,GAO, op. cit.

identify critical areas affected, the trends orrate of change, and likely influence of new ini-tiatives to relieve the problems.

A GAO investigation into the instrumenta-tion of obsolescence issue found “a tremen-dous lack of information”: an absence of trenddata on nationwide research equipment ex-penditures by universities, a lack of consen-sus on university laboratory needs, and nocomprehensive indexes that would measurechanges in the price of equipment43 44 or thecosts to maintain it. GAO also found that therapidly increasing costs for instrumentationin conjunction with relatively level funding ofbasic research (in constant 1972 dollars) atuniversities and colleges for the period 1968-81 combined to have a “large effect on re-

..— [GAO testimony].

illustration of cost escalation is the $100,000 electron microscope of the 1960s, which could distinguish ob-jects smaller than of a meter. By the scan-

ning transmission electron microscopes had improved the reso-lution by a factor of 1,000, and cost more than $1 million.(Testimony of Dr. Edward A. Knapp, Director, NSF before theHouse Subcommittee on Science, Research, and Technology,Feb., 1984.)

38-802 0 - 85 . 4


searchers’ acquisition and maintenance of re-search equipment. ”

A National Science Foundation study45 sur-veyed the status of university laboratory re-search instrumentation in 1982 in three ma-jor disciplines. The study, when completed in1985, will provide some useful insights into theextent of the problem nationwide. The studypolled 43 academic institutions concerning thecondition of their instrumentation in computersciences, physical sciences, and engineeringdisciplines in order to develop national esti-mates of the findings. Preliminary data fromthe study, which covers equipment with pur-chase prices ranging from $10,000 to $1,000,000in use in 1982, may not either confirm or refutethe notion of a serious problem with instru-mentation obsolescence in these three fields,but does seem to demonstrate that the prob-lem may not be uniform. For example:

●

●

●

University officials classified 26 percentof the research equipment listed in thesefields in 1982 as “not in current use. ”Some portion of this undoubtedly is obso-lete. Seventeen percent of the laboratoryequipment associated with computer sci-ence research was obsolete; 24 percent ofthe physical sciences and engineeringequipment was obsolete.One-half of all of the academic researchinstruments in the fields surveyed thatwere still in use in 1982 were purchasedduring the 1978-82 period. Only 12 per-cent of the computer science instrumenta-tion was purchased prior to 1972, and 78percent was purchased during 1978-82.Concerning state of-the-art equipment, 98percent of the computer science equip-ment in this category had been acquiredsince 1978, compared with 80 percentof the engineering research equipment.Eighty-four percent of all of the state-of-the-art equipment surveyed was listed asin excellent condition, as compared with42 percent of all equipment covered by thesurvey.

“’’One Fourth of Academic Research Equipment ClassifiedObsolete, ” Science Resources Studies Highlights, NSF, 1984.

●

●

The replacement value of all instrumenta-tion in use was estimated at 42 percentabove the original purchase price (almostmatching the inflation rate).Two-thirds of all research instrument sys-tems in use in 1982 were acquired partlyor entirely with Federal funds.

These preliminary findings indicate the needto develop data providing a comprehensivepicture of university research instrumentation.The 43 universities surveyed account for 94percent of the R&D expenditures in each ofthe three disciplines (computer sciences, phys-ical sciences, engineering) covered and had in-strumentation inventories that cost nearly $1billion-a significant portion of which wasfunded by the Federal Government. Exactlyhow much total funding is needed to equip thelaboratories adequately is not known. How-ever, it is possible to make some very approx-imate, inferential estimates based on the avail-able data. For example: given an instrumentinventory of $1 billion and assuming that theequipment has a 4-year lifespan, one-quarterof the equipment ($250 million current dollars)would be needed annually to upgrade theequipment assigned to those three disciplinesin the 43 universities.

Remedial ActivitiesFederal agencies, State governments, and

industry have begun to address the instrumen-tation problem. For example, the Departmentof Defense initiated a $150 million 5-year pro-gram in fiscal year 1983 to fund instrumenta-tion in areas of research in support of its mis-sion. DOD’s University Research Instrumen-tation Program is based in part on a 1980study46 of the instrumentation needs of U.S.university laboratories to conduct defense-related research. The pervasiveness of theproblem is illustrated by the estimated 2,500responses from the academic research commu-nity to an initial DOD invitation for proposalsfor funding.

‘American kmciation of Universities Report to the NationalScience Foundation, Scientific Instrumentation Needs of Research Universities, June 1980. See also 13erlowitz, et. al., op. cit.


In addition, NSF’s appropriation increasedfrom $195 million in fiscal year 1984 to about$234 million in fiscal year 1985 for support ofadvanced instrumentation. Some $122 millionin fiscal year 1985 (up from $104 million infiscal year 1984) will be allocated to researchinstrumentation for individual project grants,and the remainder for instrumentation formulti-user regional instrumentation centersand major equipment in national centers.

The universities are also concerned aboutobsolete or inadequate research facilities, orbuildings–which Federal agencies have notfunded since the 1960s. One preliminary esti-mate of the funds needed to fully upgrade fa-cilities at the Nation’s major research univer-sities is between $990 million and $1.3 billionper year.47 NSF is leading the interagencySteering Committee on Academic ResearchFacilities (which includes representation fromDOD, DOE, the National Institutes of Health,and U.S. Department of Agriculture) to ad-dress this issue. The committee is expected torecommend that a study be initiated to clarif yrequirements for additional buildings or mod-ernization programs for the Nation’s academicresearch institutions. In addition, the NationalScience Board addressed this issue during itsJune 1984 session and recommended that fund-ing for facilities become a component of thefiscal year 1986 NSF budget. It recommendedthat NSF conduct pilot programs for R&D fa-cilities construction in three areas of priorityresearch (large-scale computing, engineeringresearch centers, and biotechnology), and thatNSF support the Committee by obtaining im-proved data on the condition of university fa-cilities. The House Authorization bill for thefiscal year 1984 budget of the Department ofDefense directed that agency to determine theneed to modernize university science and engi-neering laboratories for national security pur-poses. Congress has requested NIH to makea similar determination with respect to itsmission.

47 Adequacy of Academic Research Facilities, A Brief Reportof a Survey of Recent Expenditures and Projected Needs inTwenty-Five Academic Institutions, National Science Founda-tion, April 1984.

State Government and Industry InitiativesAmong the various State government ini-

tiatives to improve university research capa-bilities are those of North Carolina, Massachu-setts, New York, California, and Minnesota(see ch. 6).

Industry is also contributing at a significantlevel to academic information technology re-search and education. For example, seven com-puter vendors alone have made recent commit-ments to contribute some $180 million in cashand equipment to universities. One source“conservatively” estimates the level of dona-tions of computer equipment to higher insti-tutions of education to exceed $100 million in1982. Among the major contributors wereIBM, Digital Equipment Corp., Apple Com-puter, Inc., Hewlett-Packard Co., Wang Lab-oratories, Inc., NCR Corp., and Honeywell,Inc.

Two other recent examples further illustratethe trend: Brown University built a $1.5 mil-lion computer science facility based on contri-butions from IBM, Xerox Corp., Gould, Inc.,and others; the University of California atBerkeley has commitments of $18 million incash and equipment from firms such as Fair-child Camera & Instrument Corp., AdvancedMicro Devices, Bell Laboratories, DigitalEquipment Corp., GE, Harris, Hewlett-Pack-ard, Hughes Aircraft, IBM, Intel, NationalSemiconductor, Semiconductor Research Corp.,Tektronix, Texas Instruments, and Xerox.48

Changing University Role

The role of university research maybe at thethreshold of significant change. Faced with theincreasing expense and risks associated withresearch, a limited supply of trained personnel(especially in needed multidisciplinary skills),and intensifying competition, U.S. industry istaking steps to bolster the universities’ rolein the performance of research in information

48These donations are seen as motivated by business strate-gies, and to some extent, by the 1981 changes in the Federaltax regulations which provide tax advantages for donations ofnew equipment to schools.

42 . /n formation Technology R&D: Critical Trends and Issues

technology. State governments are promotingtheir universities’ research capabilities to at-tract technology-intensive industry.

The National Science Foundation’s supportof engineering research at universities is alsocontributing to the change. With a $150 mil-lion appropriation for fiscal year 1985, NSFwill increase the range of engineering projectssupported and will help to establish more uni-versity engineering centers to promote re-search on computers, manufacturing proc-esses, 49 and other nationally important tech-nologies. Many of these joint activities are em-phasizing strengthened linkages among theties, promoting entrepreneurship, and improv-ing the overall scientific and technologicalbase of State and local communities.60 Theyare also serving to add to both the supply andthe quality of degreed professionals, and tomodernize the tools available in participatinguniversity laboratories.

The universities may find themselves in aposition in which they are looked to as criti-cal to U.S. competitiveness in domestic andworld markets, to our ability to maintain tech-nological prominence and to remain reason-ably self-sufficient in critical areas for nationalsecurity purposes. Undoubtedly, for these andother reasons—such as the growing need forlife-long education for many professionals–there will be forces for change in the role ofuniversities.

Conflicts in Perspectives, Goals,and Policies

These various participants-academia, in-dustry, and Federal and State governments–work together in a sort of dynamic balance,

49Nation~ &ience Foundation starts to broaden suPPort ofengineering research, The Chrom”cZe of I+@her Education, Jan.18, 1984, p. 17, and updated by NSF officials, January 1985.

‘For more detailed information, see Technology, Innovation,and Regional Economic Development, Background Paper No.2, Encouraging High Technology Development, Office of Tech-nology Assessment, February 1984.

despite some differing perspectives amongparticipants, and some discords in goals andpolicies. For example:

1.

2.

3.

4.

National economic goals for improvingproductivity and competitiveness in inter-national markets are supportive of ahealthy information technology industry.However, productivity improvements areviewed by some as possibly resulting infewer employment opportunities, andlower job skill requirements and paylevels.National science endeavors are fosteredby measures that increase fundamentalknowledge, consistent with universityand scientists’ objectives-including thesharing of research results international-ly—but may be contrary to national secu-rity objectives of controlling technologytransfer and industry concerns for pro-tecting research data.Universities’ and scientists’ interests inconducting undirected (basic) researchmay be in conflict with industry’s andmission agencies’ need for achieving spe-cific results. The current trend toward in-creased university-industry collaborationand toward university patenting may re-sult in more directed university researchand less independence of universities.U.S. policies toward opening universityadmission to foreign students have beenenormously successful, but other regula-tions encourage emigration of aliens aftergraduation, thus depriving U.S. firms andacademic institutions of needed talent.

The most striking observation concerningthe roles of the various participants is thatthey are in a state of flux. To date, the direc-tions of the changes appear to be: 1) modifiedinterrelations among the participants in theR&D process, 2) a significantly larger role inresearch for participating universities, and 3)a potential strengthening of national capabil-ities to conduct R&D in information tech-nology.


Measures of the Health of U.S. R&D inInformation Technology

There is no single indicator of the health ofR&D in information technology. However, acombination of indirect indicators can providean impression of its overall vigor in the UnitedStates. Indicators of industry growth includethe level of funding for R&D, the availabilityof trained personnel, trade balances for infor-mation technology exports and imports, andpatent trends. Although the indicators usedare not comprehensive, taken together theyportray a robust industry with significantgrowth in sales and investment in R&D. Theyalso show that these industries account for theemployment of a high proportion of the Na-tion’s technically trained work force as wellas a substantial proportion of its industry- andgovernment-funded R&D. Paradoxically, theyprovide a varying contribution to the U.S.trade balance, and a declining proportion ofthe total number of information technologypatents granted in the United States.

These indicators, while generally promisingin themselves, provide far less than a completepicture of the state of health of U.S. R&D inthe information technology industry. Interna-ational competition in both information tech-nology markets and in R&D is intensifyingand U.S. leadership in many of these areas isbeing seriously challenged. Also, as describedpreviously, aspects of the R&D process are influx and it is too early to tell whether thechanges are for better or for worse. In addi-tion, while this report focuses on informationtechnology R&D, several other factors playcritical roles in U.S. competitiveness. Theseinclude marketing strategies, manufacturingcapabilities, and global macroeconomic andtrade conditions.

Beyond that, as noted earlier in this chapter,the “information technology industry” is anill-defined entity and the available statisticsare often noncomparable. Much of our infor-mation is based on statistics pertaining to asmall subset of the Standard Industrial Code

(SIC) for manufacturing companies withoutaccounting for the significant revenues fromservices. One statistical comparison will serveto illustrate the problem of noncomparability:In 1982, shipments of all electronic computingequipment establishments totaled $34.1 bil-lion; in the same year, data processing reve-nues of the top 100 information technologycompanies totaled $79.4 billion.sl

Because of these data inconsistencies the in-formation presented is skewed by the noncom-parable databases; and, for that reason, quan-tifications can only be regarded as approximate.

Information Industry Profile

The review presented in chapter 9 of busi-ness statistics for the U.S. information tech-nology industry indicates that this industryis generally robust as measured in a varietyof ways, and in comparison with U.S. indus-try as a whole. For example, for the 1978-82period: sales revenue grew by 66 percent com-pared with 40 percent for the composite U.S.industry; profits grew by 36 percent comparedwith 6 percent for the composite. Profits-to-sales ratios were about 9 percent v. 5 per-cent; and the growth in the number of employ-ees averaged 12 percent v. a negative 8 per-cent.52

Concerning R&D, the information technol-ogy industry is also vigorous. R&D expendi-tures compare very favorably to the compos-ite industry when measured as a percentageincrease over the period, as a percentage ofsales, or in terms of R&D expenditures per em-ployee. This industry accounted for 28 percentof the total R&D spending by all industries.

51Based on a draft report, ‘l’he Computer Industry ~d ~n-termtiomd Trade: A Summary of the U.S. Role, by RobertG. Atkins, Information Processes Group, Institute for Com-puter Sciences and Technology, 1984.

‘zThese data primarily represent large firms. See table 52 (ch.9) for limitations.

—


In fact, a listing of the 15 U.S. companies thatspend the most on R&D—as a percentage ofsales, and in dollars spent per employee-is al-most completely populated with informationtechnology companies such as Cray Research,Telesciences, Advanced Micro Devices, LTX,Amdahl, Computer Consoles, and ConvergentTechnologies53 (table 4).

The electrical and communications industryincreased R&D budgets by approximately 13percent in 1984 and 1985. These increasesresulted in large part from R&D on semicon-ductors and telecommunications. During 1984and 1985, electrical and electronics companiesplan to accelerate their investment in commu-nications R&D, including work on integratedpower semiconductors and cellular radio.”Thus, the industry viewed broadly is com-mitted to well-funded R&D.

Comparison of R&D Funding bySelected Countries

Funding levels for R&D are generally rec-ognized as important to the innovation proc-ess, as noted earlier. The United States hasfallen behind two of its major competitors, asmeasured in terms of total outlays for R&D

63~ugjne9s week, R&I) Scoreboard 1982, June 20, 1983,pp. 122-153.

“Science Resource Studies, Highlights, National ScienceFoundation, NSF 83-327, Dec. 15, 1983, p. 2, and NSF 84-329,Oct. 15. 1984.

as a percentage of Gross National Product (ex-cluding expenditures for defense and space) .66Figure 4 shows that both Japan and West Ger-many have been outpacing the United States(as well as France and the United Kingdom)by this measure for more than a decade. Bothof these countries have relatively small R&Dexpenditures for defense and space purposesas a percentage of GNP—e.g., 2.5 and 5.6 per-cent in 1981 for Japan and West Germany, re-spectively, in contrast with 31 percent for theUnited States, 29 percent for France, and 30percent (in 1975) for the United Kingdom.

The Influence of DOD Funding of R&D inInformation Technology

DOD funding for R&D in information tech-nology reflects its growing dependence on thistechnology and its reluctance to be dependenton foreign sources for technology critical tonational security. Defense spending for R&Dgenerally has ranged from a high of 90 percentof Federal R&D spending in 1953 to a low of50 percent during 1976-80, and is expected torise to 70 percent in 1985.56

Table 5 shows the distribution of DOD fund-ing for 1983 among basic and applied research,

.-—66% willi~ c. B~Srn~ U.S. Civilian and Defense R&D

Funding: Some Trends and Comparisons with Selected In-dustrialized Nations, Congressional Research Service, Libraryof Congress, Aug. 26, 1983.

‘Ibid.

Table 4.—The Top 15 in R&D Spending

In percent of sales In dollars per employee

1. TeleSciences. . . . . . . . . . . . . . . . . . . . . . . . . . .......31 .6°/02. Policy Management Systems . .... . . . . . . . . . . . . . .26.63. Fortune Systems . . . . . . . . . . . . . . . . . . . . . ........22.34. Management Science America . ................20.85. King Radio. . . . . . . . . . . . . . . . . . . . . . .............20.06. Dysan . . . . . . . . . . . . . . . . . . . . . ..................19.47. Advanced Micro Devices. . . . . . . . . . . . . . . . . . . . . .. 19.48. Modular Computer Systems. . ..................17.69. ISC Systems . . . . . . . . . . . . . . . . . . . . . ............16.6

10. Computer Consoles . . . . . . . . . . . . . . . . . . . . . ......16.611. LTX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16.412. Ramtek. . . . . . . . . . . . . . . . . . . . . . ................15.613. Applied Materials . . . . . . . . . . . . . . . . . . . . . ........15.614. Auto-trol Technology . . . . . . . . . . . . . . . . . . . . . .....15.415. Kulicke & Soffa Industries . . . . . . . . . . . . . . . . . . . . .15.3

1. Ultimate . . . . . . . . . . . . . . . . . . . . ... ... ... ... ..$37,0892. Fortune Systems . . . . . . . . . . . . . . . . . . . . . ......19,3903. TeleSciences. . . . . . . . . . . . . . . . . . . . . . .........18,7974. Convergent Technologies . . . . . . . . . . . . . . . . . . . . 18,7215. Activision. . . . . . . . . . . . . . . . . . . . . . ............16,6676. Cray Research. . . . . . . . . . . . . . . . . . . . . . ........16,4677. Management Science America . ..............15,5638. Amdahl . . . . . . . . . . . . . . . . . . . . . ...............15,4139. Digital Switch . . . . . . . . . . . . . . . . . . . . . .........15,017

10. Policy Management Systems . ................14,67711. Applied Materials . . . . . . . . . . . . . . . . . . . . . ......14,54512. Auto-trol Technology . . . . . . . . . . . . . . . . . . . . . ...14,41313. Computer Consoles . . . . . . . . . . . . . . . . . . . . . ....13,81614. Network Systems . . . . . . . . . . . . . . . . . . . . . ......13,29215. LTX . . . . . . . . . . . . . . . . . . . . . ..................13,229

SOURCE: Standard & Poor’s Compustat, Inc., as cited in Business Week, “A Deepening Commitment to R& D,” July 9, 1984, p. 64; and “The U.S. Still Leads the Worldin R&D Spending, ” June 20, 1983, p 122


Figure 4.— Estimated Ratio of Civilian R&DExpenditures to Gross National Product

for Selected Countries

2.6 ‘2.5

2.4 West

1.7

1 France I1.1

1.0 ,

. 9 -1

1961 63 65 67 69 71 73 75 77 78 81 83

Years

and development. DOD spending for basic re-search accounts for some 13.6 percent of thetotal Federal obligations, while applied re-search accounts for 33.9 percent, and devel-opment accounts for 71.3 percent.

Table 6 shows that the DOD R&D budgetdominates some fields of Government R&Dspending for basic and applied research. For

example, in funding for basic research in elec-trical engineering, DOD accounts for some 69percent; in computer sciences, 55 percent; andin mathematics 42 percent. In applied re-search, the DOD is a major Federal funder forelectrical engineering (90 percent), computersciences (87 percent), and mathematics (29 per-cent). These, as well as others, are disciplinessupported primarily by the DOD R&D bud-get and that have a central influence on ad-vances in information technology for the Na-tion.67

There have been many commercially appli-cable advances in information technology thathave their origin in, or had strong early sup-port from, DOD funded R&D. These includevery high speed integrated circuits (VHSIC),digital telecommunications, and new high-per-formance materials. However, there are somemajor disadvantages for the commercial sec-tor to DOD funded R&D. Among these are:security classifications which tend to slow ad-vancements in technology; rigid technicalspecifications for military procurements whichhave limited utility for commercial applica-tions; and the “consumption” of limited, val-uable scientific and engineering resources formilitary purposes, which may inhibit commer-cial developments. This issue is discussed inmore detail in chapter 8.

U.S. Patent Activity

It is generally accepted that patenting is ameasure, even if imperfect, of the effectivenessof R&D activities. A key observation is thatpatenting in information technology is among

571bid.

Table 5.—Federal and Department of Defense Obligations for Basic Research,Applied Research, and Development, 1983 (Estimated) (millions of dollars)

Total R&D Basic research Applied research Development

Total FederalGovernment . . . . . . . . $42,973.8 100 ”/0 $5,765.2 100.0 ”/0 $7,499.7 100.0% $29,708.9 1OO.OO/o

Department ofDefense . . . . . . . . . . . . $24,519.6 57.1 % $ 782.1 13.60/o $2,543.9 33.9 ”/0 $21,193.6 71.30/0

SOURCE: National Science Foundation, “Federal Funds for Research and Development Fiscal Years 1981, 1962, and 1983,” vol. XXXI, Detailed Statistical Tables (NSF82-326) (Washington, DC: U.S. Government Printing Office, 1982), pp. 174, 179, 181, 183, Percentages calculated from data in the table. As cited in Boesman,“U.S. Civilian and Defense R&D Funding; Some Trends and Comparisons With Selected Industrialized Nations,” Congressional Research Service, Libraryof Congress, Aug. 26, 1983.


Table 6.—Federal and Department of Defense Obligations for Basic andApplied Research by Field, 1983 (Estimated) (millions of dollars)

DOD as apercentage of

Total Federal total Federalfunds DOD funds funds

Basic research:Electrical engineering . . . . . . . . . . $103.8 $71.8 69.1 ‘/0

Computer sciences . . . . . . . . . . . . 73.0 40.0 54.8Mathematics . . . . . . . . . . . . . . . . . . 100.7 42.6 42.3

Applied research:Electrical engineering . . . . . . . . . . $520.5 $471.2 90.5 ”/0Computer sciences . . . . . . . . . . . . 91.5 79.4 86.7Mathematics . . . . . . . . . . . . . . . . . . 48.5 13.9 28.6

SOURCE: National Science Foundation, “Federal Funds for Research and Development, Fiscal Years 1981, 1982, and 1983,”vol. XXXI Detailed Statistical Tables (NSF 82-326) (Washington, DC: U.S. Government Printing Office, 1982), pp. 73,75, 79, 82, 98, 101, 104, 109. Percentages calculated from data in the table. As cited in Boesman.

the most intensive of all technologies. U.S. pa-tenting of foreign origin68 in all technologieshas doubled in the past two decades to 41percent—indicating escalating world competi-tion for U.S. patents in general. A small num-ber of foreign multinational corporations havea dominant (but perhaps somewhat diminish-ing) role in the proportion of foreign-originU.S. patents. These “multinationals” empha-size information technology patents. The over-all picture derived from this review of patentdata confirms the finding reported in chapter3 that foreign competition in information tech-nology is increasing.

U.S. Patent Data

The top 50 electrical patent categories(ranked by actual numeric growth in the num-ber of patents) received 8,139 patents during1978-80 time period (table 7). Within thesecategories, semiconductors and circuits ac-counted for 48 percent and computers 15 per-cent, respectively. In the computer category,General Purpose Programmable Digital Com-puter Systems was the most active, as in pre-vious years, receiving 632 patent documents.Miscellaneous Digital Data Processing Sys-tems received the second largest number of

5aThe coUtV origin of a patent is determined by the coun-try of residence of the first named inventor.

Table 7.—Technology Distribution of theTop 50 U.S. Patent Electrical Categories 1978.80

Percent ofRanked by actual file growth categoriesSemiconductors and circuits . . . . . . . . . . . . 480/oComputers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 ”/0Other . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 ”/0Total number of patents in the

50 patent categories . . . . . . . . . . . . . . . . . 8,139SOURCE: Tenth Report, Technology Assessment and Forecast, U.S. Patent and

Trademark Office, U.S. Department of Commerce, November 1981,pp. 16, 24

patents in the electrical category, with 548patent documents.59

Solid-state devices, integrated circuits, andtransistor categories together account for 24of the 50 categories in the total ranked by ac-tual growth from 1978 to 1980.60 The percentgrowth in the number of these patents gener-ally ranges from about 40 percent to 59 per-cent.61 Solid-state devices account for 8 of the11 highest growth entries. Lasers, laminag-raphy, and fiber optics are also among theinformation technology segments included inthe high patent growth entries. The two sub-classes of fiber optics inventions show patent

5Tenth Report, Technology Assessment and Forecast, Pat-ent and Trademark Office, U.S. Department of (knmerce, November 1981, pp. 14-18.

‘Actual growth is the numeric increase resulting from addi-tions to the patent copies (including cross-reference copies) tothe file in the 3-year period 1978-80, Ibid p. 11.

“Percent growth, as used by the U.S. Patent and TrademarkOffice, is computed by dividing the actual growth for the 3-yearperiod exarnined (1978-80) by actual growth for the 8-year period(1975-80), and multiplying by 100. Ibid., p. 11.


growth rates of over 70 percent. The listing,in fact, is composed almost exclusively of in-formation technology inventions. Computertechnology patents showed growth rates ofover 70 percent for the 1978-80 period, wellabove the average of 46 percent for all tech-nologies during the same time period.62

There are reasons for caution against gen-eralizations concerning the use of patent sta-tistics—e.g., variations in the importance andthe degree of “invention” of different patents;the propensity (or absence of it) of some com-panies, and perhaps countries, to patent as op-posed to using other alternatives-e. g., tradesecrets, or lead times in the market place; thecost factor as a disincentive to patenting, aswell as concern for antitrust allegations basedon patent dominance; rapid technologicalchange (making patents of limited value); dif-ferences in the scope of patent categories thatmay give a misleading impression of substan-tial amount of patenting activity in a broadlyscoped subcategory or vice versa.

Nevertheless, the evidence shown aboveclearly seems to support the observation thatthe level of patenting for information technol-ogy in the United States is vigorous and maybe indicative of extensive R&D in this field.

Foreign-Origin U.S. Patents

In 1973 the U.S. Patent and Trademark Of-fice (USPTO), began detailed documentationof foreign-origin patent activity through itsOffice of Technology Assessment and Fore-cast (OTAF). One of OTAF’s reports,63 whichis cited extensively in this section, providesuseful background and many important find-ings on foreign patenting in the United States.Among the findings are:

1. Because patents obtained in the UnitedStates convey no protection in other coun-tries and vice versa, inventors tend to pat-ent in more than one country, and espe-

621 bid., p. 22-26.etIbid., SW for ex~ple, Section I, Part IV— “Most Foreign

Active Patent Technologies, ” pp. 27-32, and Section II, Pat-ent Trends: Foreign Multinational Corporations PatentingTrends in the United States, pp. 33-46.

2.

cially in countries that represent largepotential markets. As a consequence, U.S.patent statistics tend to mirror trends intechnological activity worldwide.Although foreign-origin patenting in alltechnologies averaged only 20 percent ofthe total U.S. patenting for the years1963-66, the percentage share has contin-ued to increase, reaching 40 percent of thetotal for the year 1980, and 41 percent forthe 1981 to mid-1983 period.

Figure 5 illustrates the long-term decline inthe number of U.S.-origin information technol-ogy patents granted in the United States be-tween 1968 and 1981, and the relative leader-ship position of Japan compared to France,West Germany, and the United Kingdom. Itis not clear as to why the total number of U.S.patents has declined steeply between 1971 and1980, but the U.S. Patent and Trademark Of-fice advises that except for Japan, the trendwas worldwide during that period. (Note thatin 1979 a shortage of funds at the Patent Of-fice limited the number of U.S. patentsgranted, artificially lowering the total for thatyear).

As shown by table 8, the share of U.S.-originpatents decreased from 79 to 58 percent of thetotal during 1968-81, while the share of Japa-nese-origin information technology patentsgranted in the United States increased from3 to 19 percent.

The two “top 50” electrical category listsnoted earlier reveal a significant proportion offoreign-origin U.S. patents in the high patent-growth categories. Fifty-four of the entries inboth lists show the percentage of foreign ori-gin to exceed the average of 38.5 percent forall technologies for 1978-80. This is not sur-prising, since the high-growth patent subclassesare pursued by companies in all of the devel-oped countries.

Table 9 shows the percentage of foreign-ori-gin U.S. patenting in patent category group-ings dealing with some components of infor-mation technology. Three of the five categorygroupings examined (two in fiber optics andone in television) show foreign-origin patenting

48 . /nformation Technology R&D: Critica/ Trends and Issues

Figure 5.–U.S. Patents in Information Technology, SIC Codes 357, 365, 366, 367

14,000

12,000

11,000

10,000

9,000

8,000

7,000

6,000

5,000

4,000

3,000

2,000

1,000

0

1968 69 70 71 72 73 74 75 76 77 78 79 80 81

Years

SOURCE: U.S. Patent and Trademark Office, Office of Technology Assessment and Forecast.


Table 8.—Percentages of U.S. Patents Granted inInformation Technology (IT) and in All Technologies (ALL), 1981

United States Japan West Germany United Kingdom France

Year IT ALL IT ALL IT ALL IT ALL IT ALL

1968 . . . . . . 79 77 3 2 5 6 4 4 3 21981 . . . . . . 58 60 19 13 8 10 4 4 4 31982 . . . . . . NA 59 NA 14 NA 9 NA 4 NA 3NA—Not availablealnformation Technology (IT) here in~l”de~ Slc Codes 357, Office Computing and Accounting Machines; and 365-367, com-

munication Equipment and Electronic Components.

SOURCES The Office of Technology Assessment and Forecast, US Patent and Trademark Office, All Technologies Report,1963-June 1983, and Indicators of the Patent Output of US. Industries (1963-81). IT numbers were calculated fromdata developed with assistance from the National Science Foundation, Science Indicators Unit.

Table 9.— Foreign-Origin U.S. Patents inSome Components of Information Technology

Percentforeign origin

Title 1/81-6/83

Light transmitting fiber, waveguide, or rod . . . 48.2°/0Laser light sources and detectors . . . . . . . . . . 50.6°/0Color and pseudo color television . . . . . . . . . . 52.8°/0Active solid-state devices, e.g., transistors,

solid-state diodes . . . . . . . . . . . . . . . . . . . . . . 40.80/oGeneral-purpose programmable digital

computer systems and miscellaneousdigital data processing systems . . . . . . . . . . 34.9°/0

NOTE The percent foreign origin IS determined by dlwding the total number ofU S patents granted between January 1981 and June 1983 to foreign.resident Inventors by the total patents granted in the same time period,and multlply!ng by 100

SOURCE Off Ice of Technology Assessment and Forecast, Patent and TrademarkOff Ice, U S Department of Commerce

to be significantly higher than the average of41 percent.

Table 10 shows that in selected telecom-munications categories, the Japanese share offoreign-owned U.S. patents ranges from 38 to56 percent. For all categories of telecommu-nications, Japanese residents received 45 per-

cent of the foreign origin U.S. patents from1980-83.64 Figure 6 depicts the shares of U.S.patents for communications equipment andelectronic components among Japan, WestGermany, and the United Kingdom.

These findings are consistent with commentsfrom OTA workshop participants concerningthe growing intensity of foreign competitionin R&D. The statistics no doubt understatethe level of foreign ownership of U.S. patents,since they do not take into account patents ofU.S. origin that are controlled by foreign in-terests, e.g., patents issued to U.S. residentsor companies that are foreign-owned or for-eign-controlled, or the inclination of foreignmultinational corporations to patent in othercountries (see section below on Foreign Multi-national Companies). Even understated, how-ever, the intensity of foreign influence overU.S.-patented technology” is clearly signifi-cant. By way of providing perspective, it

‘Patent Profiles: Tekcomnnmications, Patent and TrademarkOffice, Office of Technology Assessment and Forecast, U.S. De-partment of Commerce, August 1984, p. 15.

b51bid., p. 4.

Table 10.—Foreign-Owned U.S. Patents in Selected Telecommunications Classes, 1982

Percent foreign Percent Japanese Percent Japaneseof total of total of foreign

Telephony. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43% 17 ”/0 390/0Light wave communication . . . . . . . . . . . . . . . . . . . 53 ”/0 20 ”/0 380/oAnalog carrier wave communications . . . . . . . . . . 43 ”/0 240/o 560/oDigital and pulse communications . . . . . . . . . . . . 40 ”/0 160/0 38%Telephony–Class 17911 R.l AA: IAT.1 FS, 1 H-1 MF, 1 MN.l SS, 1SW.106, 108 R-I9OLight wave communications–372/43-59 & 75, 357/17 & 19, 455/600-619, 370/1.4, 350/96.1-96 34Analog carrier wave communlcat!ons—455/ l.355Dlgltal and pulse communication (excludes I!ght wave, Includes error detection and A/D & D/A conversion) —375/all subclasses; 371/1-6 & 30-71, 178/all subclasses,

340/347, AD347, AD.347 SY, 332/9 R-15, 329/104.109

SOURCE Reports prepared by Off Ice of Technology Assessment and Forecast for publication In late 1964


Figure 6.–Share of Foreign Patenting in the UnitedStates for the Three Most Active Countries in

Selected Product Fields: 1981

Communication equipmentand electronic components

SOURCE: National Science Board, National Science Foundation, Scienceindicators—19S2, 1983.

should be noted that some other countrieshave an even higher percentage of foreign-origin patents, e.g., Canada, 93.4 percent; theUnited Kingdom, 84.2 percent; France, 67.6percent. Japan has 16.6 Percent.66

Patent Statistics,

4,000

3,500

3,000

2,500

2,000

Foreign Multinational Corporations

Another important OTAF finding concernsthe role of foreign multinational corporations(FMNCs) in patenting. Taking into accountthe relative annual sales of the 10 major FMNCsand their ranking among the “Fortune 500”companies, OTAF has found a strong correla-tion between ranking by patents and sales lev-el.67 A comparison of the 10 FMNC's patent-ing with total U.S. patenting for 1969-80 isshown in figure 7.

In addition to noting that the FMNC's own-ership of U.S. patents has recently (1980) lev-eled off to about 5.5 percent, the OTAF studyobserves that:

●

●

The 10 FMNCs own or control, on the av-erage, 4.7 percent of all U.S. patentsgranted each year.The extent of the 10 FMNC's ownershipof U.S. patents doubled from 1969 to1976–although the rate of increase haddiminished to near zero by 1980.

Op.

Figure Patent Activity of 10 Foreign Multinational Corporations, 1969-80

7.0%

6.00/0

5.0%

4.0%

3.0%

0

1969 70 71 72 73 74 75 76 77 78 79 80

YearsSOURCE: Tenth Report, Technology Assessment and Forecast, Patent and Trademark Office, U.S. Department of Commerce, November p.


● More recently, (1979-80) the percentage an indication of the concentration of foreign-of U.S. patents granted to the 10 FMNCshas begun to decline in proportion to totalforeign origin U.S. patents suggesting adiminished role in ownership of U.S. pat-ents for these particular FMNCs.

The trend, while changing, shows that aboutone in every eight U.S. patents of foreign originis owned or controlled by only 10 FMNCs68—

owned U.S. patents by a few multinationalfirms.

These statistics further confirm the testi-mony of OTA workshop participants concer-ning growing foreign competition in informa-tion technology R&D, and the observationthat other countries have developed nationalpolicies and programs that target information

—.—. ——6aIbid. ((_J’I’AF Study), P. 40.

A Synthesis:

technology.

The Changing U.S. R&D Environment

Some measures, such as investment in R&Dand growth in profits, indicate that R&D inU.S. information technology is vigorous. Othermeasures, such as competition in advanced-technology products and foreign ownership ofU.S. patents, indicate a less robust situation.Thus, although information technology re-search and development is making marked ad-vances, it is—at the same time-undergoingpressure from foreign competition. In responseto the pressure, the participants in the R&Dprocess are initiating a variety of changes.These changes are discussed in later chaptersof this report.

Industry continues to invest heavily in in-formation technology R&D–an indication ofits belief in R&D’s importance to competitiveness. The increasing costs of R&D are mak-ing new institutional arrangements such asjoint research ventures and closer ties withuniversities more attractive. However, indus-try experts recognize that although R&D isnecessary to competitiveness it is not suffi-cient to ensure it; other components of the in-novation process are also important to main-taining competitiveness in international trade.

Universities are encouraging new institu-tional arrangements with industry, and the im-portance of their role in the R&D process (par-ticularly in performing basic research) maybegrowing. There are widespread problems re-lating to both the quantity and quality of uni-versity equipment and facilities for conduct-ing information technology R&D, althoughthese conditions may be improving. SomeState Governments have become active inhelping their universities to improve researchcapabilities and in encouraging university-industry pairings (see ch. 6).

Finally, the Federal Government has adoptedpolicies intended to encourage private sectorinvestment in R&D and to facilitate the trans-fer of technology from Government to indus-try. The Federal Government’s (especiallyDOD’s) expenditures for information technol-ogy R&D are growing rapidly and continue tohave a strong influence on the direction oftechnological development in some informa-tion technology areas.

52 ● information Technology R&D: Critical Trends and Issues

Observations

Some useful observations can be made basedon the changes taking place in the U.S. R&Denvironment. First, the growth in foreigncompetition-whatever its effect on U.S. jobsand trade balances for the long term—is stim-ulating R&D investment, as shown in chapter9. There has probably not been a time sincethe turn of the century when new products andproduct improvements have been marketed insuch rapid succession as has been the casewith information technology, nor a peacetimeera when R&D had such a central role in theaffairs of nations.

Second, the U.S. information technology in-dustry is facing a “new world” of foreign com-petition. The intensive level of targeted andwell-funded foreign competition is not likelyto decline in the foreseeable future. Each ofour major competitors’ governments believein the central importance of information tech-nology as an essential ingredient for achiev-ing economic, social, or national security goals,as well as the penalties-in terms of worsen-

ing trade balances and job losses—associatedwith falling by the wayside in the competitiverace. As a consequence, they have establishednational policies and programs to enhancetheir domestic industrial position.

Third as foreign competition has inexorablystrengthened in the post-World War II era, thebroad margin of error that the United Statesonce enjoyed has essentially vanished.

The long-term effects of several factors—na-tional industrial policies, nontariff trade bar-riers (e.g., prohibiting the import of certainproducts or services, incentives for industrialinnovation, interest rates, export controls—will determine the winners and losers, as na-tions maneuver to remain competitive. TheUnited States will need to find ways to moni-tor its position relative to international com-petitors and to refine its policies as needed tokeep in step with the changing global R&D en-vironment.

Chapter 3

Selected Case Studies inInformation Technology Research

and Development

Contents

PageIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . 55

Case Study 1: Advanced ComputerArchitecture . . . . . . . . . . . . . . . . . . . . . . 55

Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . 55Changing Computer Architecture . . . . . . 56Computer Architecture R&D . . . . . . . . . . 57Critical Areas of Research . . . . . . . . . . . . 63Manpower . . . . . . . . . . . . . . . . . . . . . . . . . . 65International Efforts.. . . . . . . . . . . . . . . . 65

Case Study 2: Fiberoptic Communications 67Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . 67Advantages . . . . . . . . . . . . . . . . . . . . . . . . . 67Commercialization Trends . . . . . . . . . . . . . 69United States R&D..... . . . . . . . . . . . . . 71Directions of U.S. Research . . . . . . . . . . . 73Research in Japan . . . . . . . . . . . . . . . . . . . 74

Case Study 3: Software Engineering . . . . . 75Findings . . . . . . . . . . . . . . . . . . . . . . . . . . 75Introduction . . . . . . . . . . . . . . . . . . . . . . . . 75Software R&D Environments . . . . . . . . . . 76Content and Conductor Software

Engineering R&D...... . . . . . . . . . . . . 79International Efforts in Software

Engineering R&D...... . . . . . . . . . . . . 85

Case Study 4: Artificial Intelligence . . . . . . 87Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . 87Introduction . . . . . . . . . . . . . . . . . . . . . . . . 87Artificial Intelligence R&D

Environments . . . . . . . . . . . . . . . . . . . . . 89Content and Conduct of Artificial

Intelligence R&D . . . . . . . . . . . . . . . . . . 96Foreign National Efforts in Artificial

Intelligence R&D . . . . . . . . . . . . . . . . . . 105

TablesTable No. Page11. Major Federal Advanced Computer

Architecture R&D Projects . . . . . . . . . . 57

Table No. Page12. Milestones in the History of Computer -

13.

14.

15.

16.17.

18.

19.

20.

21.22.

Architecture . . . . . . . . . . . . . . . . . . . . . ..5 71982 Federal Spending for ComputerArchitecture R&D..... . . . . . . . . . . . . . 58Federal Open Access SupercomputerFacilities . . . . . . . . . . . . . . . . . . . . . . . . . . 58Summary occurrent and PlannedGovernment Open Access AdvancedComputer Architecture SoftwareDevelopment Facilities. . . . . . . . . . . . . . 59NSF Plans for Computer Research .,.. 60Summary of New CommercialSupercomputer Systems UnderDevelopment . . . . . . . . . . . . . . . . . . . . . . 611983 DARPA, DOE, and NSF UniversityFunding of Advanced ComputerArchitecture . . . . . . . . . . . . . . . . . . . . . . . 62Major Government Sources of ArtificialIntelligence R&D Funding. . . . . . . . . . . 94NSF Grant Awards in ArtificialIntelligence . . . . . . . . . . . . . . . . . . . . . . . . 94Strategic Computing Cost Summary . . 94Representative Expert Systems . . . . . . 104


8.

9.

10.110

12.

13.

14.15.

Computer Architecture FunctionalElements . . . . . . . . . . . . . . . . . . . . . . . . . . 56Optical Fibers Are Small, Lightweightand Versatile . . . . . . . . . . . . . . . . . . . . . . 68The Software Life Cycle. . . . . . . . . . . . . 80Artificial Intelligence . . . . . . . . . . . . . . . 90Artificial Intelligence Science andEngineering . . . . . . . . . . . . . . . . . . . . . . . 92Strategic Computing Program Structureand Goals . . . . . . . . . . . . . . . . . . . . . . . . . 95A Semantic Network.. . . . . . . . . . . . . . . 98Markets for Artificial Intelligence . . . . 101

Chapter 3

Selected Case Studies in InformationTechnology Research and Development

Introduction

These case studies present a microcosm ofthe R&D process in information technology.The diversity of research and development ef-forts in information technology make it vir-tually impossible to examine in detail all of themany fields and disciplines. Therefore, fourfields have been selected for detailed analysis.They are: Advanced Computer Architecture(ACA), Fiber Optics (FO), Software Engineer-ing (SE), and Artificial Intelligence (AI).

These fields were selected for several rea-sons. They depict the wide range, diversityand inter-relatedness of the fields and applica-tions of information technology. An analysisof them provides a broad overview of the scien-tific, technical and institutional issues in in-

formation technology R&D. These fields werechosen to include both hardware and softwareand both computer and communications tech-nologies. They also illustrate the mix of long-term goals and near-term capabilities, thus re-flecting the importance of these differentperspectives in the R&D process. These fourareas, moreover, are among those consideredto be critical in determining the direction andpace of advance of information technology asa whole. The importance of advances in thefour fields is exemplified by the developmentof government funded national R&D pro-grams in Japan, Britain and the EuropeanEconomic Community.

Case Study 1: Advanced Computer Architecture

●

●

Findings

The technology of advanced computer de-sign is critically important for the expan-sion of information technology in manyfields. There is extensive R&D activityunderway in universities, in industry, andin the National Laboratories aimed at pro-ducing and exploiting new computer de-signs; but there is considerable uncertaintyover which new designs will be viable.Since their invention, electronic computershave been based on one architectural model,the von Neumann sequential processing ar-chitecture. The limits of computationalspeed achievable with this design are beingreached; significant further increases incomputer performance will require parallel

processing architectures, which are inherentlymore complex to design and to use.VLSI (Very-Large-Scale Integrated Circuit)design facilities, based on powerful com-puters, are now being used to develop andtest computer architectural designs, includ-ing parallel processors and special designsfor certain dedicated operations, such ascommunications signal processing, imageprocessing, and graphics.Software has been difficult to produce forcomputers of advanced, high-performancedesign. As the variety and complexity of ar-chitectural types increases, the difficulty ofdeveloping and integrating software will in-crease. Therefore, research in software de-velopment for novel computer designs willbe critical.

55

38-802 0 - 85 - 5


●

●

●

The Federal Government has been a majordriver of advanced computer architecturebecause of its scientific and national secu-rity applications. The Government has in-fluenced the evolution of computer designthrough the funding of R&D and the pro-curement of state-of-the-art systems. Thisleverage, though still important, is dimin-ishing as commercial applications for ad-vanced computers grow and as Federal re-quirements become a smaller fraction ofsales.National programs in Japan, Great Britain,France, Germany, and the European Com-munity have been established to pursue ad-vanced computer R&D. The Japanese haverecently demonstrated an ability to produceadvanced architecture computers of com-petitive performance to American products.American companies and universities pur-suing R&D incomputer design face-dif-ficulties:

Universities cannot afford design andtesting facilities for developing an architec-tural idea to the point where its performancecan be assessed.

Companies face large, risky investmentsin the design of new high-performance com-puter systems. Markets for novel machinesare initially small and expand only slowlyas new applications are exploited and soft-ware becomes available.

Changing Computer Architecture

Computer architecture is the internal struc-ture of a computer, the arrangement of thefunctional elements that carry out calculationsand information manipulations. (see fig. 8).

Since the early 1950s, all electronic comput-ers (with a few exceptions) have been designedaround one basic architectural model, the vonNeumann machine, invented by mathemati-cian John von Neumann. In this architecture,instructions and data are stored in memory,fetched one by one in sequential fashion, andacted on by the processor. Computer designis now changing, encouraged by two factors.

Figure 8.—Computer Architecture FunctionalElements


First, the limits imposed by physical lawson the computational speed attainable withtraditional computer design are being ap-proached. Science and engineering demandcontinual advances in computational speed toincrease the precision of calculations and toimprove accuracy in models and simulations.1

Sequential processing constitutes a severe re-striction on the precision and completenessthat these calculations and models can achievein a reasonable amount of time. Therefore,computer designers are studying architecturesthat can make possible decomposition of largecalculations into pieces for simultaneous proc-essing by a number of computational units inparallel.

Second, as information technology is appliedin more and more areas, special problems areencountered that impose unique demands oncomputer capabilities. Until recently, systemdesigners have relied on software to apply thecapability of von Neumann processors to prob-lems. Now, it is possible to create special in-tegrated circuits to address specific problems.

IThese are the applications normally associated with so-called“supercomputers.” Major current applications are in, for ex-ample, weather modeling and the simulation of nuclear weap-ons explosions. The reader is referred, for a discussion of theapplications of and policy issues surrounding supercomputers,to Supercomputers: Foreign Competition and FederaJ Fund”ngby Nancy Miller, Congressional Research Service, Issue Brief83102, latest update July 12, 1984.

Ch. 3—Selected Case Studies in Information Technology Research and Development ● 57

Computer architecture R&D is making possi-ble the economical design of custom computerarchitectures for specialized applications in-cluding telecommunications and data acqui-sition signal processors, image and graphicsprocessors, and symbol processors for themanipulation of nonnumerical information.

The impending changes in computer archi-tecture promise cost-effective solutions tomany problems, but they also challenge thedesigners, suppliers and buyers of computersystems. Designers will need to have moredetailed appreciation of applications; com-puter vendors will be faced with more com-plexly segmented markets; and buyers willneed to be more sophisticated in defining theirneeds and in choosing among a wider offeringof products.

Computer Architecture R&D

Federal Government Involvement

The Government has had considerable in-volvement in advanced computer architectureR&D (see table 11), both as a funder and a per-

former of work. Major elements of the soft-ware development work for each generation ofthese systems have been done by the NationalLaboratories, especially Los Alamos (LANL)and Lawrence Livermore (LLNL).2 Moreover,the impetus for the development of each suc-cessive generation of advanced architecturescientific computers has come predominantlyfrom government demand for faster, higher ca-pacity, and more sophisticated systems forweapons, intelligence, energy, and aerospaceapplications (see table 12). The National Labsstill constitute the greatest concentration ofusers of supercomputers (see table 15).

The Federal Government has provided be-tween $15 and $20 million in annual fundingfor advanced computer architecture R&D inrecent years (see table 13). In addition to spon-soring research in universities and in indus-try, the Government has performed computerresearch at the National Labs.

— — — —The first Cray-1 computer was placed in Los Alarnos National

Lab without any software.

Table 11 .—Major Federal Advanced Computer Architecture R&D Projects

Machine Year delivered Agency Contractor Major useENIAC . . . . . . . . . . . . . 1945NORC . . . . . . . . . . . . . 1950— 1950CDC 1604 . . . . . . . . . . 1959LARC . . . . . . . . . . . . . 1961STRETCH . . . . . . . . . . 1961CDC 6600 . . . . . . . . . . 1964ILLIAC IV . . . . . . . . . . 1972MPP . . . . . . . . . . . . . . 1983s-1 . . . . . . . . . . . . . . . . –

ArmyNavyNSANSAAECAECLLNLDARPA/NASANASALLN L/Navy

University of PennsylvaniaIBMSperryControl DataSperryIBMControl DataBurroughs/University of IllinoisGoodyear Aerospace—

Ballistics calculationsOrdinance researchClassifiedClassifiedNuclear weapons designNuclear weapons designNuclear weaponsAerodynamicsImage processingSignal processing

SOURCE Office of Technology Assessment

Table 12.—Milestones in the History of Computer Architecture

Class Date Typical machines Major innovationI . . . . . . . . . . . . 1953 IBM 701 Vaccuum tubesII . . . . . . . . . . . . 1960 CDC 1604 TransistorsIll . . . . . . . . . . . 1964 IBM 360 1/0 processing

CDC 6600 Freon coolingIv , , ... , ., . . . 1970 IBM 370 Integrated circuits

CDC 7600v . . . . . . . . . . . . 1972 Illiac IV Parallel processing

TI ASC Pipeline architectureVI . . . . . . . . . . . 1976 Cray-1 Vector processingSOURCE Office of Technology Assessment


Table 13.—1982 Federal Spending for ComputerArchitecture R&D (millions of dollars)

Department of Defense . . . . . . . . . . . . . . . . . . . . . . . . . $12.0a

Department of Energy , . . . . . . . . . . . . . . . . . . . . . . . . . 2.9a

National Aeronautics and Space Administration . . . 1 .5a

National Science Foundation . . . . . . . . . . . . . . . . . . . . 1.2Total . . . . . . . . . . . . . . . . . . . . ... ... ... ... ... ... .$17.6

aEstimates.

SOURCE: Department of Defense and Department of Energy numbers from theOffice of Technology Assessment Workshop on Advanced ComputerArchitecture; NASA number from personal communication with PaulSchneck; NSF number from Summary of A wards, FY 1982, NSF Direc-torate for Mathematical and Physical Sciences, Computer SciencesSection,

DOE spent $645,000 at Los Alamos and$500,000 at Lawrence Livermore NationalLabs in fiscal year 1983 on two experimentalresearch projects on advanced computer archi-tecture hardware.3 In addition, Los Alamos isleasing a Denelcor HEP-1 to experiment withparallel processing software concepts. TheNavy, in conjunction with Lawrence Liver-more, has been involved with the design andconstruction of an advanced architecture com-puter, termed the S-1 Project. The design isintended to handle signal processing tasks forNavy missions. Approximately $20 millionhas been spent over the last 4 years on the S-1 Project.4

Facilities and help in advanced computer ap-plications development are provided to re-

9Edward Oliver at the OTA workshop on Advanced Comput-er Architecture, July 14, 1983.

‘Personal communication from Sidney Fembach, Consultant,Control Data Corp.

searchers in science and engineering fields tosupport the missions of several Governmentdepartments (DOD, DOE, NASA), and to fur-ther basic research (NSF). Seven Federal fa-cilities provide limited open access to certaingroups of researchers (see table 14).

The National Laboratories plan to add moresupercomputers over the next few years, so itcan be expected that Government scientistsand engineers and contractors on missionagency work will have access to state-of-the-art large-scale computing facilities (see table15). Academic researchers will have limited ac-cess to these facilities for work in fields relatedto agency missions (e.g., fusion energy, atmos-pheric and ocean sciences, and aerodynamics).These facilities also provide support for soft-ware development.

The Department of Defense, through theDefense Advanced Research Projects Agency(DARPA), has formulated ambitious plans forresearch and development in advanced com-puter-based systems. Included will be effortsto develop high speed signal processor archi-tectures and to integrate numeric and symbol-ic processing in advanced computer architec-tures for use in intelligent weapons systems.5

In April 1983, the National Science Foun-dation organized a working group to study

————‘This program, called “Strategic Computing, ” is covered in

some detail in the Artificial Intelligence Case Study later inthis chapter.

Table 14.—Federal Open Access Supercomputer Facilities

Facility Major system Research users Charges

NCAR . . . . . . . . ........2 CRAY 1-As Atmospheric and ocean sciences No charge to NSF users. $2,200 perprime CPU hour for others

MFECC . . . . . . ........2 CRAY 1 Magnetic fusion energy No charge1 CDC 7600 community

LANL . . . . . . . . . . . . . . . .Open 1 CRAY Government agencies, labs, and $636 per prime CPU hour3 CDC 7600s nonprofit institutions1 CYBER 825

NASA-Ames . . . . . . . . . .CRAY 1-S Computational fluid and No charges to NASA grantees.CDC 7600 aerodynamics $2,000 per CPU hour for CRAY

NASA-Goddard . . . . . . . . CYBER 205 NASA-funded and NASA-project No charge to NASA grantees. $1,000related per CPU for others

NASA-Langley. . . . . . . . .CYBER 203 NASA and NASA-funded $1,300 per CPU hourscientists

NASA-Lewis . . . . . . . . . .CRAY 1-S Principally aerodynamics related No charge for NASA supportedSOURCE: A Nafioml cornwtiw ~rrtirortmerrt for Academic l?esearcl? National Science Foundation, October 1983


Photo credit: U.S. Department of Energy, Los Alamos National Laboratory

View of a part of the main computing facility at Los Alamos National Laboratory. CRAY 1 in foreground

Table 15.—Summary of Current and Planned Government Open AccessAdvanced Computer Architecture Software Development Facilities

Number Planned additionsAgency Current systems of users fiscal years 1984-88

National Science Foundation. ...2 Class VI 850 1 Class VllDepartment of Energy . . ........3 Class VI 6,400a 1 Class Vl, 5 Class VllNational Aeronautics and Space

Administration . . . . . . . . . . . . ..3 Class Vl, 1 special 3,000 1 Vl, 1 Vll, 1 specialTotals . . . . . . . . . . . . . . . . . . . . ..8 Class Vl, 1 special 10,250 2 Vl, 7 Vll, 1 special

performing classified work.SOURCE: A Computing Environment for Academic Research, National Science Foundation,


what was recognized as a critical scientific im-perative. The report of that group stated:6

Computing facilities have a decisive effecton the kind of research which is done by aca-demic scientists and engineers. During the1950s and 1960s the Government encouragedthe growth of computing in research, re-search methods were transformed in disci-pline after discipline, and the United Statesenjoyed a large, ever-widening lead in quan-titative research and modeling complex phe-nomena. In the 1970s Government supportslackened and academic computing facilitiesno longer kept pace with advancing technol-ogy . . . Science has passed a watershed inusing computers for research. Computers areno longer just tools for measurement andanalysis but have become the means for mak-ing new discoveries . . . Academic research incomputer architecture, computational math-ematics, algorithms, and software for paral-lel computers should be encouraged to in-crease computing capability.

In response to this imperative, the NSFworking group recommended an expansion inspending for academic research in advancedcomputers, and improved access to computingfacilities including 10 new supercomputer fa-cilities and special networks to make these sys-tems widely available (see table 16).

The House Committee on Science and Tech-nology considered R&D in advanced com-puters and access to powerful computer sys-tems by scientists and engineers in manyfields to be crucial elements in the advance-

6A Natiomd Computing Environment for Acadenu”c Comput-ing, prepared under the direction of Marcel Bardon by the Work-ing Group on Computers for Research, Kent K. Curtis, Chair-man, July 1983, pp. 1-2.

Table 16.—NSF Plans for Computer Research(million of dollars)

Fiscal years1984 1985 1986

Local facilities . . . . . . . . . . . . . . . . $45.5 $ 90.9 $106.7Supercomputers . . . . . . . . . . . . . . . 14.4 70.0 110.0Networks . . . . . . . . . . . . . . . . . . . . . 2.1 7.3 11.5Advanced computer systems and

computational mathematics. . . 8.0 20.0 33.0

Total . . . . . . . . . . . . . . . . . . . . . . . $69.9 $188.1 $261.2SOURCE: A National Computing Environment for Academic Research, National

Science Foundation.

ment of science and technology. Accordingly,they approved a budget of $40 million in fiscalyear 1985 for NSF’s Advanced Scientific Com-puting initiative, thus doubling the adminis-tration’s request for this program.7

Industry’s Role

Three U.S. companies are developing nextgeneration (Class VII) supercomputer sys-tems. In addition, two Japanese companies(Fujitsu and Hitachi) have introduced new sys-tems and a third (Nippon Electric-NEC) is de-veloping a new supercomputer, planned for de-livery in 1985 (see table 17).

Cray will introduce the Cray-2 in 1985. Thiswill be a four processor vector machine.8 TheCray-3 is scheduled for introduction in 1986.It will be an 8 to 16 processor vector machinewith Galium Arsenide (GaAs) (see ch. 9) cir-cuitry. Cray sees integrated circuit technologyas critical. The Japanese are the major sup-pliers of state-of-the-art fast bi-polar memorychips, and one-half of the integrated circuitsin current Cray machines are Japanese made.9

Control Data (CDC) spun off developmentwork for its next generation advanced archi-tecture machine to a new company, ETA Sys-tems, which CDC capitalizes with $40 millionfor 40 percent ownership. This approach is be-ing taken by CDC because small groups withdedication, entrepreneurial spirit, and a per-sonal stake in the success of the project areconsidered important.10 ETA Systems willspend $4 million to $6 million the first yearon direct R&D costs. Plans are for the firstdemonstration machines to be available late

7Authorizing Appropriations to the National Science Foun-dation, House Committee on Science and Technology, Report98-642, Mar. 30, 1984, pp. 8-9.

8Vector computers have specialized architectures that achievehigh speed calculation of mathematical formulas by treatingentire arrays of data (vectors) as processable by single instruc-tions, saving time on certain calculations that can be arrangedas a series of vectors.

‘L. T. Davis, “Advanced Computer Projects, ” presentationat the Frontiers of Supercomputing Conference, Los Alamos,Aug. 15, 1983.

‘OW. Norris, “A Conducive Environment for Supercomput-ers, ” banquet address at the Frontiers of Supercomputing Con-ference, Los Alamos, Aug. 18, 1983.


Table 17.—Summary of New Commercial Supercomputer SystemsUnder Development

Maximum speedCompanv Model (M FLOPS’) Available.Cray Research . . . . . . . . Cray-2 1,000 mid 1985

Cray-3 NA 1985-86ETA Systems . . . . . . . . .GF-10 10,000 1986-87b

GF-30 30,000 NADenelcor . . . . . . . . . . . . . HEP-2 4,000 1985-86NEC . . . . . . . . . . . . . . . . .SX-2 1,300 March 1985NA—Not announcedaM FLOpS (million floatlng ~olnt ~Peration~ per second) a measure of computer performance on high preci SiOn CalculationsbL M Thorndyke at the ~ro”tier~ of s~pe~~o~p”t~~g conference, LOS Alamos, August 1983.

SOURCE The IEEE Committee orI Super Scierrt/f/c Computers.

in 1986, and volume production of two machinesper month is planned for 1987. The design willemploy two to eight vector processors with themaximum eight processor version to sell in therange of $20 million.11

Denelcor, a former maker of analog com-puters, developed a parallel processing com-puter design (HEP-Heterogeneous ElementProcessor). Since 1982, this design has beenavailable for sale or lease. Considered to be anexperimental machine by users at facilitiessuch as Los Alamos, this design is a steptoward a new generation of computer archi-tectures. Work is currently underway on theHEP-2, which should be competitive with Crayand CDC machines if component and softwareproblems can be overcome. Moreover, the via-bility of Denelcor efforts will require highersales than have so far occurred with the HE P-1 .12

In addition, other U.S. firms including com-puter companies (Digital Equipment, Hewlett-Packard, Honeywell, IBM, NCR and Sperry),telecommunications companies (Harris), semi-conductor companies (Advanced Micro De-vices, Intel, Monolithic Memories, Mostek,Motorola and National Semiconductor), elec-tronics companies (Allied, Eaton, GeneralElectric, RCA and Westinghouse) and aero-space companies (Martin-Marietta) are in-volved to some extent in research on parallel

—. —“L. M. Thorndyke, “The Cyber 2XX Design Process, ” pres-

entation at the Frontiers of Supercomputing Conference, LosAlamos, Aug. 15, 1983.

“B. Smith, “Latency and HEP, ” presentation at the Fron-tiers of Supercomputing Conference, Los Alamos, Aug. 15,1983.

processing, data-flow or multiprocessor ar-chitectures. 13

Industry representatives characterize theadvanced computer architecture business asrisky. The market is small: approximately 100Class VI supercomputers have been installedworldwide as compared to tens of thousandsof less powerful computers. Developmentcosts are high: design tools include other ad-vanced architecture machines for hardwaresimulation and software development.

The Role of Universities

Although as many as 50 U.S. universitiesare involved in advance computer architecture(ACA) research,14 significant funding levels areavailable in only a few major schools. (Seetable 18.)

University research in advanced computerarchitecture is characterized by a series ofstages of elaboration of a concept including:1) theoretical paper and pencil work; 2) simu-lation of ideas on existing computer systems;3) “breadboard” wiring of designs with off-theshelf components; and 4) full-scale engineer-ing and construction of prototype machinesin which state-of-the-art components, softwareand peripheral devices can be integrated totest the design on full-scale problems. OTAfound that few if any projects currently under-

“’’Next-Generation Computing: Research in the UnitedStates,” IEEE Spectrum, November 1983, pp. 62-63.

“The OTA workshop on Advanced Computer Architectureconcluded that every major Computer Science and ElectricalEngineering department has some interest.


Table 18.—1983 DARPA, DOE, and NSF University Funding ofAdvanced Computer Architecture

Institution DARPA DOE NSF

California Institute of Technology . . . . . . . . . .Massachusetts Institute of Technology. . . . . .University of California, Berkeley . . . . . . . . . . .Stanford University . . . . . . . . . . . . . . . . . . . . . . .New York University . . . . . . . . . . . . . . . . . . . . . .University of Illinois . . . . . . . . . . . . . . . . . . . . . .University of Texas . . . . . . . . . . . . . . . . . . . . . . .University of Wisconsin . . . . . . . . . . . . . . . . . . .Duke University . . . . . . . . . . . . . . . . . . . . . . . . . .Other . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Totals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 ,000 a 5502,000 250 1022,000a

2,000a

600 150155 200107 120

95197570

7,000 a 1,757 1,339a Est imate

SOURCE DARPA numbers from a personal communication with Duane Adams; DOE numbers from Edward Oliver at the OTAWorkshop on Advanced Computer Architecture; NSF numbers from Summary of Awards: Fisca/ Year 1983, NationalScience Foundation, Division of Computer Research.

way in universities have funding to carry a de-sign concept through to the final, systemsengineering stage. Several projects will pro-duce prototypes, but the elaboration of an ideainto a system with software and supportingperipherals to demonstrate the performanceand utility of the concept on real problems re-quires funding on the order of twice what thecurrently best funded projects receive.

The major distinction between efforts pur-sued in industry and in universities on ACA,aside from the commercial and product devel-opment orientation of industry work, is thatmore radical and advanced designs are beingpursued in universities, whereas evolutionarydesigns are sought by industry. This is a re-sult of the stake that industry has in the ex-isting base of software and users and the needfor upward compatibility of systems. Univer-sity researchers have a greater ability to pur-sue revolutionary designs that could requirecompletely new programming approaches andtechniques.

Facilities Requirements

The increasing availability and capability ofVLSI circuitry and computer-aided designtools are expected to have significant impacton computer design.15 Prototype productiontime and cost will decrease. Both general pur-——- -- - —

‘sS. Trimberger, ‘‘Reaching for the Million-Transistor Chip,IEEE Spectrum, November 1983, p. 100.

pose and custom application architectures areimplementable in VLSI, opening opportunitiesfor the testing and evaluation of many morecomputer architecture ideas. However, the ini-tial investment required for VLSI design andfabrication equipment is very costly and willprobably remain so. It is unlikely that mostuniversities will be able to afford this equipment.

Other expenses associated with ACA re-search include computer-based simulation fa-cilities. There may be a need for current gen-eration supercomputers at universities tofacilitate and test the design of new architec-tures. Bell Laboratories currently devotesmost of its Cray computer to VLSI circuitdesign.

Supercomputers are also used by Cray andControl Data for software development, sothat software is available when a new hard-ware design is completed. Universities wouldbenefit from access to these software devel-opment tools, giving researchers the chanceto test ideas experimentally. But here again,the costs associated with the procurement andoperation of these design and computing re-sources are beyond the means of universityproject budgets. Some universities are form-ing consortia to spread the cost of microelec-tronics design and fabrication facilities acrossseveral institutions (see ch. 6). Shared super-computer facilities are a key element to NSFplans for Advanced Scientific Computing.


Annual operating costs for governmentsupercomputer facilities average more than$10 million.” Only three U.S. universities cur-rently operate such facilities and these are uti-lized at less than 50 percent of capacity. Thereason for this low usage is the high cost ofcomputer time on these systems, ranging from$2,000 to $3,000 per hour. University ACA re-search is therefore usually done on minicom-puters which lack the capability of generatingsophisticated, high-resolution graphics of im-portance to the design of integrated circuits.

Critical Areas of Research

Currently, the goal of most advanced com-puter architecture research and developmentis parallel computation.

The two major U.S. industrial developers ofsupercomputers, ETA Systems and Cray Re-search, and the Japanese manufacturers,Hitachi and Fujitsu, are pursuing parallel ar-chitectures in a conservative incremental fash-ion, contemplating the production of machineswith up to 16 parallel processors by the late1980s. Vector architecture will remain thedominant method for achieving fast numeri-cal processing in these systems.

Universities, by contrast, are pursuing anumber of methods of achieving “massivelyparallel” computation with upwards of 1,000processors working in concert. One of the basicproblems of computing in parallel is the re-quirement for communication and coordina-tion among the individual processing elementswhen they are working on pieces of a singleproblem. Often the results of one process arerequired for another process to go forward.Several architectural solutions to these dif-ficulties are under study, and extensive evalu-ation of different approaches must be donebefore their viability in real-world problemscan be assessed. Detailed simulation of con-cepts and testing of prototypes is required,and present university funding is inadequateto support such work.

‘6A National Computing Environment for Academ”c Re-search, op. cit., p. 22.

There are currently more than 50 conceptsfor parallel processing architectures under con-sideration in academic and industrial institu-tions. However, there are no standard metricsfor comparing the performance of different ar-chitectural designs or the software to be usedon parallel machines. Nor is it likely that anyone metric could fully measure differences inperformance, since different applications placedifferent demands on systems. The develop-ment of such metrics and the establishmentof suitable test facilities for implementingstandard design evaluations are critical issuesin advanced computer architecture. The Fed-eral Government may have a role in this areaby setting voluntary standards for computerperformance measurement, and by providingfacilities for testing.

Thus far little attention has been devotedto the problems of symbolic, as opposed tonumeric processing architectures. In the past,the von Neumann architecture has been usedfor both kinds of computations; the focus ofadvanced computer architecture R&D hasbeen on computers for “number crunching”applications, or high precision calculation,simulation and modeling for science and engi-neering. The increasing importance of artifi-cial intelligence is encouraging the design ofspecial machine architectures, both to ease the

programming of artificial intelligence applica-tions, and to speed the processing of symboliccomputations. Several companies are now pro-ducing machines for artificial intelligence, andthe list is expected to grow. ’7

Three other areas of technology are criticalto the development of advanced computer ar-chitecture systems: integrated circuits, circuitpackaging, and algorithm and software design.

Integrated Circuits (IC)

An order of magnitude (1 OX) increase in com-puter speed is expected from improvementsin IC materials and manufacturing techniques.Silicon will remain the dominant IC substratematerial through 1990 because the technology

“See the case study on Artificial Intelligence in this chapter.

64 Ž Information Technology R&D: Critical Trends and Issues

is well understood and the practical limits ofdevice density, speed and power consumptionhave yet to be reached. Silicon will be the basisof a growing set of special purpose VLSI ar-chitectures. Gallium Arsenide (GaAs) digitallogic and memory circuits are growing in im-portance, and will be used in the Cray-3.18

University based research in chemistry, phys-ics and microelectronics are expected to makesignificant contributions to increased inte-grated circuit capability.

Packaging

Interconnections among the logic elementson some complex chips occupy over half of theuseable chip area, and affect the performancespeed of chip functions. Currently, three sand-wiched layers of interconnection within a chipare typical, and it is expected that as manyas 12 layers will be common in a decade.19

As chips have become more complex, con-taining greater numbers of logic elements, thenumber of “pins,” or inputs and outputs, re-quired for communication with them hasgrown. The connection of sets of chips hasthus become more complex, and the difficultyof simultaneously housing and powering chips,and dissipating the waste heat from chip setsis forcing advanced computer designers to findmore sophisticated methods of packagingthem.20

Cryogenic liquid cooling equipment is re-quired for most existing and planned super-computers. Facilities must be provided for therefrigeration and storage of the coolant. Andthe size, weight, and reliability of the coolingequipment must be considered in the purchaseand use of these systems.

—.———“L. T. Da~is, op. cit.“J. A. Armstrong, “High Performance Technology: Direc-

tions and Issues, ” presentation at the Frontiers of Supercom-puting Conference, Los Alamos, Aug. 15, 1983.

20MCC is devoting some of their initial efforts to packagingtechnology; and ETA Systems sources estimate that 60 per-cent of the R&D effort for the GF-10 will be in packaging. ETAis planning to use liquid nitrogen cooling to obtain a doublingin speed from CMOS silicon integrated circuits (L.M. Thorn-dyke at the Los Alamos Conference).

Packaging is also a critical factor in super-computer manufacturing costs. Cray machinesare currently hand wired. In an effort to re-duce costs, the Japanese are developing de-signs that lend themselves to automated man-ufacturing procedures.

Software and Algorithms

The lack of applications software for super-computers has been a significant barrier totheir adoption and use.

One half of recent Cray Research R&D fundshave reportedly been devoted to software de-velopment, 21 and the vectorizing FORTRANcompiler, a software program that helps pre-pare standard FORTRAN code for executionon the Cray vector architecture, has been a ma-jor factor in the commercial success of theCray-1 line.” Users of vector computers, in-cluding the Cray, are quite pleased to obtain20 percent of the maximum rated speed ofthese machines on typical problems.23 Workis continuing in industry and universities todevelop software to make vector machinesmore effective and easier to use, and this workwill be of critical importance through thisdecade.

The introduction of parallel processing de-signs and the proliferation of special purposecomputer architectures will make softwareproduction and design more complicated.24 Thecreation of new high-level languages that aremore easily understood by users, and othert o o l s a n d programming support environmentsthat facilitate the expression of logical, sym-bolic, mathematical, scientific, and engineer-ing concepts in computable form, could greatly

—2’Rollwagen, op. cit.*’Nippon Telephone and Telegraph, the Japanese state tele-

communications monopoly, has chosen a Cray-XMP over re-cently introduced Japanese supercomputers of comparable orsuperior speed, reportedly because of the software, and the ex-perienced team of field representatives, available from Cray.“NTI’ Picks Cray Super CPU,” Ektronz”cAfews, Oct. 10, 1983,p. 87.

Z~David Ku*, at the OTA workshop on Advanced ComputerArchitecture, July 14, 1983.

*’Paul Schneck, at the OTA workshop on Software Engineer-ing, Nov. 17, 1983.


expand the utility and lower the costs of oper-ating advanced computer systems.25

In order for the speed potential of advancedcomputers to be used, problems must eitherbe programmed to take advantage of the com-puter design, or the architecture must be de-signed to handle the unique characteristics ofthe problem. The mediating factor between theproblem and the architecture is the algorithm,or the structured procedure for solving theproblem. In the future, computer designerswill have to be more cognizant of computa-tional algorithms and the effects of computerarchitecture on programming and problemsolution, and thus they will need greater knowl-edge of applications. Similarly, designers ofcomplex programs, especially scientists, mathe-maticians, and engineers, will need to have agreater appreciation of the inherent capabil-ities and limitations of particular computer ar-chitectures as they become more dependent onadvanced computers in their work. Collabora-tion between the users and designers of futurecomputer systems is critical to both the util-ity and the commercial success of advancedarchitecture computers.26

Manpower

There is a shortage of people capable of de-signing software for advanced architecturecomputer systems. In particular, people skilledin the design of software and software toolsfor use in sophisticated scientific and mathe-matical applications are scarce.27 There is aneed for people who understand scientificproblems in a range of disciplines, and who candesign and implement computer systems tosolve those problems.

Attracting talented faculty to train the nextgeneration of computer researchers is a prob-lem. The difficulty results, in large measure,

—- . —- — — .“M. B. Wells, “General Purpose Languages of the Nineties, ”

presentation at the Frontiers of Supercomputing Conference,Los Alamos, Aug. 17, 1983.

Z60TA Workshop on Advanced Computer Architecture, JUIY

14, 1983.Z7This point was emphasized in one applications area in Par-

ticular, telecommunications (Paul Ritt at the OTA workshopon Advanced Computer Architecture, July 14, 1983).

from the uncompetitive salaries and low jobmobility offered by universities. The problemis expected to become acute as demand forcomputer architectures employing symbolicprocessing and artificial intelligence capabil-ities increase. (See Artificial Intelligence casestudy.)

International EffortsJapan

Two Japanese firms, Fujitsu and Hitachi,have introduced advanced architecture com-puters whose performance is competitive withthe fastest available American supercom-puters. Early copies of these machines havebeen installed in three Japanese universities.A third company, Nippon Electric, has an-nounced plans to introduce a supercomputerin 1985.

The Japanese Government is funding twonational efforts in advanced architecture re-search and development: “High Speed Com-puting Systems for Science and Technology’and the “Fifth Generation Computer System”program.

The Electrotechnical Laboratory of theAgency of Industrial Science and Technology(AIST), an arm of the Ministry of Interna-tional Trade and Industry (MITI), is manag-ing a 10 year project (January 1981—March1990) called “High Speed Computing SystemsFor Science and Technology. ” It is focused onmicroelectronics research and development(GaAs, Josephson Junctions and High Elec-tron Mobility Transistor devices), parallelprocessing systems with 100 to 1,000 com-puting elements, and systems components, in-cluding mass storage and data transfer de-vices, to support high-speed scientific andengineering calculations. Total governmentfunding will be on the order of $100 million.A consortium of six major Japanese computercompanies (the Technology Research Associ-ation) has been formed to conduct much of thiswork in industrial laboratories on “consign-ment” from MITI.28

———. - — --—“’’Super Computer-High Speed Computing Systems for Sci-

ence and Technology, ” Science & Technology in Japan, October-November, 1982, p. 16.


The “Fifth Generation Computer System”program is managed by the Institute for NewGeneration Computer Technology (ICOT). Theprogram is described in more detail on page105, as part of the artificial intelligence casestudy.

Great Britain

Great Britain has reacted to the Japaneseefforts (in particular the Fifth Generation proj-ect) by establishing a national plan for re-search in advanced computer systems, knownas the Alvey Programme for Advanced Infor-mation Technology. It is discussed below inthe Software Engineering and in Artificial In-telligence Case Studies. Researchers in Brit-ish universities have made significant contri-butions to advanced computer architectureresearch. For example, the University of Man-chester has an advanced prototype of a “data-flow” architecture machine.29 Work is also be-ing pursued in industry .30 Inmos, Ltd. hasdeveloped the Transputer, a device which com-bines processing and communications func-tions on a single chip. This device has beenspecifically designed for the connection of anumber of units to achieve concurrent proc-essing.31

France

France has considerable interest in advancedcomputer architecture research and develop-ment. Two industrial companies, Cii-Bull andCGE, as well as seven government funded in-stitutions including five universities are pur-.

‘9A. L. Davis, “Computer Architecture, ” IEEE Spectrum,November 1983, pp. 98-99.

‘°Five have been identified, see “Next-Generation Comput-ing: Research in Europe, ” IEEE Spectrum, November 1983,pp. 65-66.

““lhnsputer Does Five or More MIPS Even When Not Usedin Parallel, ” Iann Barron, Peter Cavill, David May and PeteWilson, Electronics, Nov. 17, 1983, p. 109.

suing work in this area.32 Three “supercomput-er” projects are expected to produce machinesby 1985-88, but these systems are not likelyto be speed rivals of American and Japanesesystems of similar vintage.33

West Germany

West Germany has four universities and sixindustrial companies working on advancedcomputer architecture R&D.34 The govern-ment is providing about $4 million per yearfor research at the universities on parallel proc-essing. A project at the Technical Universityin West Berlin has received a grant from theMinistry of Research and Technology to de-velop a full-scale prototype.35

The European Community

The Commission of the European Commu-nities (EC) has initiated ESPRIT (EuropeanStrategic Program for Research and Develop-ment in Information Technology) to pursue in-formation technology R&D on a cooperativebasis with industry, universities, and thegovernments of the EC countries pooling theirefforts. Four institutions (three Belgian andone French) have thus far announced plans tostudy parallel processing with ESPRIT fund-ing.36 The proposed ESPRIT plan calls for ‘hedevelopment and use of computerized facilitiesto study new computer designs.37 (For moreinformation on ESPRIT, see ch. 7.)

‘z’’ Next–Generation Computing: Research in Europe, ” IEEESpectrum, November 1983, pp. 64-65.

39Report of the IEEE Super Scientific Computer Comm.f”ttee$Oct. 11, 1983.

“IEEE Spectrum, November 1983, op. cit., pp. 67-68.“’’Western Europe Looks to Parallel Processing for Future

Computers, ” Ekctrom”cs, June 16, 1983. p. 111.“A. L. Davis, op. cit.“Proposal for a Council Decision adopting the first European

Strategic Programme for Research and Development in Infor-mation Technologies (ESPRIT), Commission of the EuropeanCommunities, June 2, 1983, p. 24.


●

●

●

●

●

●

●

Case Study

Findings

Fiber optic communications

2: Fiber Optic Communications

technology isan important export (e.g., a $3 billion worldmarket projected for 1989), and is a key ele-ment in improved productivity and reducedcosts in telecommunications systems.Fiber optic communications technology isdeveloping rapidly and the potential bene-fits from continued research are extensive.The most significant research is concen-trated in only a few large firms and univer-sities, in part because of the expense andthe required long term commitment.Increased Government funding in this tech-nology-now at a low level—would enlargeuniversity participation and accelerate tech-nological advances.The scarcity of trained research and devel-opment personnel is a continuing handicap.A large proportion of university research-ers in fiber optics are foreign nationals andmany return to their native land after com-pleting graduate studies.Research conducted in Japan and Europeis among the world’s most advanced and ex-change of research information internation-ally among colleagues is critical to scientificadvancement.Japan is the world’s leader in several keyaspects of the technology, and is a strongcompetitor to American firms.

The first successful transmission of voicesignals using energy from the Sun was accom-plished in 1880 by Alexander Graham Bell andSumner Tainter using a device called the photo-phone, but was abandoned because weathermade the system unreliable.38 Since 1970, in-terest has resumed in using light energy fortelecommunications, in the form of fiber op-tic communications because of two technologi-cal advances: the laser and light transmissionthrough low loss silica glass fibers.

—. —..—38 Report on Research at the University of Arizona, vol. 1, No.

1, fall 1983, p. 11, published by the Research Office, Universi-ty of Arizona.

Advantages

Fiber optic communication is the transmis-sion of light signals through transparent glassor plastic fibers, where the signals are gener-ated by lasers or light emitting diodes (LEDs)and received by photodetectors, which convertlight signals to electrical signals. The majorcomponents of fiber optics technology are thefiber cables and connectors, transmitters andreceivers, and repeaters, or regenerators,39

which amplify and reconstitute the signals pe-riodically along the fiber.

There are several properties of fiber opticcommunications that make them attractivefor telecommunications applications:

●

●

●

●

large bandwidth, meaning that largeamounts of information (voice conversa-tions, computer data, graphics) can betransmitted rapidly. For example, a quar-ter-inch diameter optical cable with twofibers carries as much data as a 3-inchcopper cable with 20,000 wires.40 (See fig.9.) Conservatively, the capacity of a singlepair of fibers currently available commer-cially is about 4,000 voice grade circuitsin field applications and about 400,000voice grade circuits under controlled lab-oratory conditions;less susceptibility than copper wire toradio frequency interference, providingless cross-talk, higher quality signaltransmission, and immunity from electro-magnetic pulse (EMP) effects-character-istics of value for both civilian and mili-tary uses;lower loss of signal strength, meaningthat fewer repeaters are needed;resistance to “noninvasive” or covertwiretaps;

. . .—sg]n Stmdwd copper telephone wires, Signal regeneration is

required at about onemile intervals. Repeaters add significantlyto the installation and maintenance costs of transmissionsystems.

4~High Technology, “Fiber Optics: Light at The End of theTunnel, ” March 1983, p. 63.


Figure 9.—Optical Fibers Are Small, Lightweight, and Versatile

Photo credits: AT& T Sell Laboratories


small size and low weight, factors thatcontribute to ease of transport and lessneed for underground and building ductspace;declining cost compared to other terres-trial telecommunications technologies.41

Commercialization Trends

The technology has progressed rapidly intothe commercial marketplace since 1966, whenresearchers 42 first proposed the possibility ofpurifying the glass used in optical fibers to re-duce losses in signal strength. In 1970, thefirst low loss fiber was produced by CorningGlass Works, and by 1977 prototype systemswere being installed by AT&T and GeneralTelephone and Electronics Corp. in the UnitedStates. Today, many developed countries havefiber optic communication systems in opera-tion or plan to install them.

The United States, Canada, Western Eur-ope, and Japan accounted for an estimated 96percent of a $550 million world market in fi-ber optics communications equipment in 1983.The world market was approaching $1 billionfor 1984,43 and is projected to expand to $3 bil-lion by 1989,44 as countries satisfy their tele-communications system expansion and re-placement needs with the increasingly cost-competitive fiber optic communication sys-tems rather than with microwave radio, cop-per twisted wire pair, and coaxial cable.

Applications in the United States

Telecommunications, the major market (85percent)45 for fiber optics, can be described infour segments: long distance; interoffice trunksthat connect telephone central offices; localfeeder lines; and local area networks. Long dis-

——tlThiS Compuison is based on the relative cost per chmnel-

mile, which is the number of voice circuit equivalents (chan-nels) multiplied by the distance of the transmission link.

42Kao and Hockrnan, I(IT Standard Telecommunications Lab-oratories, England.

43D. G. Thomas, “Optical Communications, ” Research andDevelopment, June 1984, p. 203.

“Signal, September 1983.“The remaining 15 percent of the noncommon carrier applica-

tions are said to be in vehicular,, industrial control systems,and in CATV. High Technology, op. cit.

tance applications are currently the most costeffective for fiber optic communications, andcomprise the vast majority of current use inthe United States.

Interoffice trunking provides links betweenintracity telephone facilities. Feeder lines in-clude intracity transmission links between car-rier facilities and subscriber distribution points,while local feeder lines extend to subscriberlocations. Local area networks (LANs) servelimited communities, for example, within abuilding or a building complex. They are justemerging and will become an important mar-ket for fiber optics.

There are a growing number of installationsof fiber optic systems in long distance telecom-munications. Among these are AT&T’s North-east Corridor route. In 1983 a line connectingWashington, DC and New York City was in-augurated. In 1984 this line was extended toCambridge, MA, and Richmond, VA. The to-tal Northeast Corridor line will use 45,000miles of fiber over the 750 mile route. The firstphase of a west coast route, which will even-tually extend from Los Angeles to Oaklandand Sacramento, has been completed. A trans-atlantic cable (TAT 8), engineered to carry40,000 voice circuits at a cost of less than halfthat of its predecessor, is scheduled for in-stallation in 1988. By March 1983, AT&T hadalready installed over 100,000 miles of fiberand projections are for another 300,000 milesby 1990.4’

Other firms planning major systems includeUnited Telecommunications, Inc., with its23,000 mile, $2 billion lightwave network tobe completed by 1987.47 Southern New Eng-land Telephone Co., in a joint venture, willroute its system through 20 States along rail-road rights of way; MCI, with a 4,000 mile sys-tem intended to serve the east coast;48 andCable and Wireless, a British company, with

4Qptoelectronics Supplement to JZkctrom”c News, “Fiber Op-tics Market Still Baffles Suppliers, ” p. 5, Apr. 4, 1983.

47’’ 23,000 Mile Fiber Network Planned by United Telecom, ”The Journal of Fiber Optics, June 1984.

4aFiber Optics Industry Service: Competitive Environment,Gnostic Concepts, Inc., 1983, pp. 1-2.


a 560 mile network that will link major citiesin Texas.

Fiber optics is already an attractive replace-ment for copper cable in some local telecom-munications applications: 1) in trunk lines be-tween telephone central (switching) officeswhere the average 10-mile distance can bespanned by optical cables without repeaters,thus installation and maintenance costs arelower, and 2) in local feeder lines in large cit-ies where crowding in underground utilityducts is a growing problem.

Local area network applications are expectedto grow rapidly according to some industryprojections, particularly as user requirementsfor bandwidth increase to the 10 to 100 mil-lion bits per second (Mbs) range. The past slowgrowth of fiber optic applications in LAN sys-tems is due to the variety of technical needsamong different customers, and a lack of uni-form technical standards. In contrast withlong distance communications systems, whichhave traditionally paid extensive attention totechnical standards development, a Federaltelecommunications standards committee onlyrecently (July 1984) held its first meeting todevelop Federal guidelines for LANs. Oncestandards issues are settled, growth in the useof fiber optics within buildings and buildingcomplexes is expected to be rapid.49

Other factors limit the adoption of fiber op-tic communications technology. The large baseof installed copper wire in the AT&T plant(some 827 million miles) is likely to be replacedslowly. The large capacity (up to 100 televi-sion channels) of some CATV systems and thefact that this expensive investment in coax-ial cable has been made quite recently in manycities, suggests that fiber optics will not bewidely used for cable television for some time.Advances in nonoptical transmission tech-niques are enabling considerable increases inthe information carrying capacity of copperwire pairs.

‘gAviation Week and Space Technology, “Promising FutureSeen for Optical Fibers,” pt 1, Oct. 12, 1983, pp. 44-77,

In applications other than communicationscarrier uses, such as aerospace and militarysystems, fiber optic technology is already be-ing exploited. These are primarily in com-mand, control, and communications applica-tions including guidance and control systemsfor aircraft, spacecraft, and missiles; opticallymultiplexed data bus transmission systems;electronic warfare and sonar applications; andadvanced instrumentation systems. The im-perviousness of fiber optics to electromagneticinterference, along with light weight, smallsize, and high information rates, make it ofspecial value in aerospace and military appli-cations.

Foreign Applications

Installation of fiber optic communicationssystems have been completed, or are plannedin a number of countries. A small sample ofthese includes:

●

●

●

West Germany. By late 1984, 10 broad-band integrated fiber optic local area net-works will be built in Berlin, Hamburg,Hanover, Dusseldorf, Nuremburg, andMunich, called the BIGFON (BroadbandIntegrated Fiber Optic Local Network)network. In addition to having access tothe public switched telephone network,subscribers will also be able to access theintegrated telex and data network and re-ceive radio, television, telephone, and fullmotion picturephone. The long-term goalis to include all telephone subscribers inthe nation.France. Several fiber optic systems are be-ing installed by the French. The govern-ment has decided to upgrade the nation’santiquated telecommunications networkby leapfrogging toward the most advancedtechnology available-especially fiber op-tics. They also plan to install a submarinecable between France and Corsica in 1985,providing over 7,600 channels operatingat 280 Mbs.Japan. One of several projects beingundertaken is an 80 km fiber optic routein the suburbs of Tokyo. The system oper-ates at 400 Mbs, providing video confer-


encing and facsimile services. In addition,the Japanese have installed a broadbandnetwork for data communications inTsukuba to facilitate scientific and tech-nical communications.

An Intelligent Network System (INS)is being developed by Nippon Telephoneand Telegraph’s Yokosuka Laboratorynear Tokyo. Services include voice, data,CATV, still and motion pictures, facsim-ile, TV conferencing and high resolutionTV. The INS is expected to make exten-sive use of fiber optic technology. Anotherproject involves a 45 km undersea fiberoptic cable south of Tokyo, operating at400 Mbs. The system will be extendedover a 1,000 km route between islands.

● United Kingdom. British Telecom, the na-tion’s telecommunications authority, hascommitted itself to using only fiber opticsin the trunk network from 1984 on. By1990, half the trunk network will be fiberoptic systems.

Mercury Communications, a communi-cations carrier in limited competition withthe government telecommunications au-thority, has begun installing an intercityfiber optic network using the British railrights of way.

The British plan to construct a nation-wide cable telecommunications system forthe delivery of television programming,FM radio, pay television, and text. An ex-periment with a small number of homesis being conducted using fiber optics tech-nology. This is considered to be the pro-totype for the national system.

United States R&D

Much fiber optics R&D in the United Statesis being conducted by a few large companies:AT&T Bell Laboratories; Corning Glass Works;ITT; and, to a lesser extent, GTE; a few uni-versities; and some smaller companies. Manyof the commercial products now available area result of R&D performed by Bell Labs.AT&T began funding optic communications

research in 1960 and related laser research in1958.

There is concern that many U.S. companieshave been inclined to undertake research onlywhere the prospective payoff is likely to oc-cur within a very few years. This attitude hasthe effect of shifting investigation away frompromising areas such as research on infraredsystems, where another 5 to 10 years of workmay be required before commercial productsbecome available. Thus, the importance ofstable Federal funds for basic research isunderscored by short-term planning within in-dustry.

The expense of research in fiber opticsmakes it difficult for small firms to play a rolein R&D, except in some areas of product com-mercialization. In addition, there is a tendencyfor equipment purchasers to prefer vendors ofcomplete lines of components, which works tothe detriment of small firms. As a result, smallspecialty firms often must rely on DOD for re-search funds, on takeovers by larger firms, oron venture capital in order to remain competi-tive. Regardless, these firms play an impor-tant role in the technology by providing in-novative ideas and products, by serving asconduits for the commercialization of univer-sity-based research, and by filling niches thatmight not be attractive to larger firms.

Because of the considerable expense asso-ciated with research in fiber optics technology,and the low level of available funding, fewuniversities have major research programs.The importance of cost is illustrated by the$500,000 or higher cost of Molecular BeamEpitaxy equipment (needed for growing alter-nating epitaxial layers on semiconductor lightsources and detectors) and $200,000 for fiberdrawing equipment (needed to produce fiberand to experiment with different fiber designs)required to perform research. Very little uni-versity research is focused on glass fibers, butinstead is directed toward light sources anddetectors. The principal universities with ma-jor research programs are the California In-stitute of Technology, the University of 11-

38–802 O - 85 - 6

72 • Information Technology R&D: Critical Trends and lssues

linois, Stanford, Cornell,Massachusetts Institute

Government Funding

Princeton, andof Technology.

the

University research has been supported fordecades through NSF funding for the supportof research and, more recently, training andlaboratory equipment. NSF supports severaluniversities, such as the University of Arizonain Tucson, Northeastern University, and Cor-nell University’s Submicron Facility. AlthoughNSF funding levels in this technology are notlarge $1.75 million in 1983–they representa consistent source of funds, often supportingfundamental research with very long term po-tential payoff. Additional levels of fundingwould likely accelerate the rate of technologi-cal advance.

Between 18 and 20 projects are being fundedby NSF in about a dozen universities. Manyof these are concerned with advancing theo-retical knowledge in areas such as laser tech-nology, the development of pioneering opticand optoelectronic integrated systems andbistable optical switching devices, researchinto infrared lasers and detectors, and the ap-plication of integrated optical interface circuitsin local area networks at gigabit (billions ofbits) per second data rates. Another $300,000of NSF funding is available for upgradinguniversity laboratory equipment.

The DOD funds fiber optics research throughmission-oriented procurements. Approximately95 percent of the research is carried out by in-dustry. DOD has some $32.4 million commit-ted to the development of cables and connec-tors, light sources and detectors, radiationeffects exploration, and to sensor and commu-nications applications. An estimated $12.6million of this is committed to research, prin-cipally applied research. In addition, the mil-itary departments allocated about $22 millionfor fiscal year 1984 among seven procurementprograms for applications ranging from sur-veillance, shipboard and long distance com-munications, and helicopter flight controlsystems.

Cooperative ResearchNSF has several activities directed at en-

couraging cooperative research between uni-versities and industry, and transferring tech-nology into industry. One of these noted inchapter 2, the Industry/University Coopera-tive Research Centers program, provides plan-ning grants to aid universities in establishingindustry affiliations and support for specificscientific or engineering technologies.

NSF has awarded a $75,000 grant for plan-ning purposes to the University of Arizona atTucson, Optical Sciences Center. The Univer-sity held its first meeting with industry inearly 1984 to begin determining mutual inter-est in specific areas of cooperation and indus-try support. Most of this research is expectedto be directed toward long term projects inphysics and materials science with potentialapplications in optical logic circuitry and op-tical computers, with limited attention to fi-ber optics.

Another NSF activity funds specific proj-ects where research is undertaken cooperatively by universities and company investigators.Funding is at levels of about $100,000 per yearover a 2 to 3 year period, on the average. Oneof the funded projects is being undertakenjointly by Bell Laboratories and the Univer-sity of Arizona, Optical Sciences Center. Thisproject’s long-term research is in high speedoptical, bistable switching devices operatingat picosecond (1 trillionth of a second) rates.

The current level of cooperation in researchis not extensive, but holds promise for broad-ening the base of research. Among the prob-lems and issues to be worked out are findingways for competing firms to share researchdata and establishing a balance between uni-versity investigators’ interest in long-term re-search and companies’ desire for short-termpayoffs.

Manpower and Industry Support

Fiber optics research requires training inboth physics and in electrical engineering. Be-


cause universities, in general, do not providethis cross disciplinary training at the bache-lor and masters degree levels, companies hir-ing recent graduates provide supplementaryin-house training. Some companies also estab-lish an affiliation with universities. An exam-ple is the affiliation between Corning GlassWorks and the University of Rochester Insti-tute of Optics, in which the company providessome faculty and funding. Another exampleis the support from United Technology Re-search Center to the Rensselaer PolytechnicInstitute which is providing real estate for newfacilities and adjunct faculty to teach special-ized engineering courses.

The University of Rochester is the onlyschool in the United States to offer an under-graduate degree in optics. Only three schoolsin the nation offer graduate degrees in optics:The University of Arizona in Tucson, theUniversity of Rochester, and NortheasternUniversity.

A factor aggravating the availability oftrained Ph.D. graduates in this field is thata large proportion of the students are foreignnationals. Industry argues that immigrationlaws make it difficult for the student-graduateto remain in the United States after gradua-tion, although many would prefer to stay andperform research.

Directions of US. Research

Among the areas of R&D focus identifiedduring the course of this study are:

Fibers

The early fiber optic cables put in use wereof the multimode type, in which lightrays en-ter the fiber at a variety of angles and travelthrough the core of the fiber reflecting fromits inner refractive surfaces. However, singlemode fiber technology, where lightrays followa single direct path along the fiber core, hasimportant advantages. Single mode fibershave greater information carrying capacityand allow a tenfold increase in the distance be-tween repeaters for regenerating signals.

Today’s single mode fiber systems are able totransmit, without repeaters, up to 200 Mbs(million bits per second) over 80 to 100 km.(This information rate is sufficient to carrysimultaneously a video channel, high fidelityaudio, data, and many telephone calls.) In lab-oratory tests, this performance has been ex-ceeded by about 10 times, suggesting fargreater potential gains from research. Im-provements from research are expected to con-tinue in both multimode and single modefibers.

Improvements have been made in loweringattenuation (losses in signal strength) in bothtypes of fibers by a factor of 100 since 1970,primarily due to development of methods toreduce impurities in the fibers.

Activities are being directed toward furtherimprovements in optical fibers. These includeresearch into different cross sections for fibercores, such as circular, triangular, and oval,as investigation into new materials such asplastics, and improvements in fiber splicingtechniques. Research into plastic materials isabout at the stage of 1975 era research in glassfibers, and promises even lower cost, moredurable fiber materials for some applications.

Longer wavelength (1.7 and 4.0 micron) ma-terial for fibers is also receiving attention, asthese show promise of decreased attenuationby a factor of 10 to 100 over that of currentlyavailable fibers, with long-term prospects fortranscontinental or transoceanic transmissionwithout the need for repeaters.

Light Generators and Detectors

Light generator and detector technologiesare important areas of research. Research iscontinuing to improve the lifetimes of thesedevices, their spectral stability, the narrow-ness of spectral emissions, switching speeds,current threshholds, and receiver sensitivity.The most recent advance is the cleave-coupledcavity laser—a device notable for its wave-length stability and capability of changing fre-quencies rapidly, making it attractive as amultisource generator. It has been demon-

74 . /formation Technology R&D: Critical Trends and Issues

strated at 274 Mbs over 100 km of fiber opticcable by Bell Laboratories, and at 1.6 Gbs(billion bits per second) over 40 km by theJapanese.

Coherent detection, a technique to improvereceiver sensitivity and to increase the infor-mation carrying capacity of fibers, is beingpursued in many research laboratories, andmay become important if a variety of obstaclescan be overcome.

Research is also continuing into ways of in-tegrating light sources, detectors, and theassociated circuitry into single chips thuslowering cost and increasing reliability.

Optical Multiplexer

Optical multiplexer (and demultiplexers)are devices that combine (or separate) differentsignals so they can be sent through the sameoptical fiber. Wavelength multiplexing is al-ready being used in AT&T’s east coast andwest coast systems, and a few experimentalsystems in Japan, Canada, and Europe. Re-search in wavelength multiplexing techniquesshould lead to important cost savings for wide-band systems.

Connectors and Splicing Techniques

Research is continuing to simplify tech-niques for splicing together separate segmentsof optical fiber and to achieve lower losses dueto the splice.

Bell Labs recently announced the develop-ment of an ultraviolet splicing system thatcontributes only 0.03 decibels to signal loss,using an optical test signal to assist in thealignment of fibers.

Switches

Switching permits a signal to be routedthrough specific paths to subscribers. Switch-ing is a bottleneck in optical communicationssystems because the conversion of signalsfrom light to electronic (current switch tech-nology is electrical) causes delays in movingthe signals through the system. Improve-ments in switching capabilities hold promise

for reducing the number of conversions re-quired from optical to electronic and viceversa. Current switches are expensive, limitedin applications, and of unpredicted reliability.

Optical switching research is being con-ducted at Bell Laboratories and the Univer-sity of Arizona, where experimental, roomtemperature switching rates, for “turn on, ” of50 picosecond (50 trillionths of a second) havebeen measured.

Storage

Research is continuing into methods forstoring optical signals on fixed and volatilememory devices. Improvements in storage willmake possible store and forward and electronicmail features for optical networks.

Amplifiers and Repeaters

Regenerative repeaters detect a signal, thenamplify, reshape, and retime it into a replicaof the original signal. The regenerated signalthen modulates a laser or light emitting diodefor transmission along the next span of opticalfiber. Decreasing the number of repeaters re-quired along a line depends in part on amplifi-cation capabilities.

Integrated Circuits

Research is continuing to improve capabil-ities for putting optical and optoelectroniclight generators and detectors onto single in-tegrated circuits, and to increase the opera-tional bit rates. Recent breakthroughs holdpromise for reducing the number of discretecomponents required in fiber optic systemsand expanding bit rate capabilities.

Research in Japan

While research in fiber optics is being ac-tively pursued in the United Kingdom, France,and West Germany, the most advanced for-eign research has been undertaken in Japan.

Japan’s research is being conducted, at leastin part, to support the development of a newnationwide broadband telecommunications


network that will make extensive use of fiberoptic technology. Research is also being sup-ported for future commercialization and inter-national markets. Fiber optic research sup-porting this network is performed mainly inprivate companies, and is supported by theMinistry of International Trade and Industry(MITI) through the Optical Measurement andControl System (OE) project. OE plans to de-velop optoelectronic integrated circuits, tran-sistors, and GaAs/GaAIAs lasers and detec-

tors. The budget is approximately $100 millionfor the 1979-87 time period. Half of this amountis to be devoted to a coordinated research fa-cility, information exchange among research-ers, and the remainder on projects in six orseven companies, including Hitachi, NipponElectric Corp., Toshiba, Mitsubishi, Fujitsu,and Matsushita. Japanese investment in in-frared laser research is estimated at between$3 million and $4 million.

Case Study 3: Software Engineering

Findings

Software is an important factor in informa-tion technology, exceeding four times thecost of hardware in large systems. The rela-tive decline in hardware costs is shifting thefocus of R&D to software. The complexityof new applications and information sys-tems is also forcing the focus of informationtechnology R&D toward software issues.R&D in software engineering has producedprototypes of software design tools and pro-gramming environments (integrated sets oftools) that promise significant productivityincreases. But the cost of retooling, includ-ing retraining software manpower, and theuncertainty associated with innovationin software development are retarding adop-tion of innovative tools and techniques.In order to speed the acceptance of softwareengineering innovations, an applied researchbase needs to be established to scientificallytest and validate new software developmenttechniques, and to disseminate informationon their performance in specific applicationsenvironments. Such an applied researchbase would link basic research in univer-sities to applied research and developmentefforts in industry and government.Foreign efforts, particularly those in Japanbut also national targeted efforts in Europe,show signs of movement toward such an ap-plied software engineering research base.

●

●

The Federal Government, through the De-partment of Defense, is making some effortsto create a software engineering appliedresearch base for national defense purposes,but the applicability of this research baseto the general problem of software produc-tivity in the American economy is un-certain.It is difficult to differentiate software pro-duction activities from R&D, especially de-velopment. Much software production is acreative design effort. There are some as-pects and types of programming that areclearly not R&D, but the dividing line is dif-ficult to define.

Introduction

The term software refers both to the instruc-tions that direct the operation of computersystems, and the information content, or data,that computer systems manipulate. Softwareis thus a logical rather than a physical prod-uct. An adequate organizing formalism orcalculus for software creation has not yet beendiscovered. 50 Therefore, the development oflarge, complex software systems dependsheavily on the insight and creativity of sys-

50” We are in a business that is 35 years old . . . and I inviteyou to think where civil engineering was when it was 35 yearsold . . . they had not discovered the right angle yet. ” HarlanMills at the OTA workshop, Nov. 17, 1983.


terns designers and programmers. The meth-ods presently employed to develop and testsoftware are ad hoc, without a strong scien-tific basis. Thus, software systems are expen-sive to build and maintain, and can be unreli-able in operation. The problems associatedwith inadequate software methods are grow-ing as information technology uses spread andare relied on for larger, more complex, andmore critical applications.

Research and development efforts have pro-duced new methods that show promise for im-provement in the productivity and reliabilityof software creation, testing and maintenance.These include Advanced program editors anddebuggers, new languages, design methodol-ogies and integrated programming environ-ments. However, the introduction of innova-tive techniques into software production isproving difficult. The adoption of new meth-ods often requires the conversion of large ex-isting program inventories. This is expensiveand risky because evidence that one method-ology is better then another is not systemati-cally collected. As well, most software devel-opment is oriented toward the single projectat hand and often relies on antiquated pro-gramming habits and attitudes. The high jobmobility of programmers and software sys-tems designers perpetuates individualism andfragmentation in software methodology.

The establishment of software engineeringthen, involves not only the development ofsuperior methodologies, but a transformationof the programming process—from an art toa science and from a labor intensive to a capi-tal intensive effort. The pressures on softwareproduction for larger, more complex, more costeffective, and more reliable systems, make thepresent situation untenable.” A concerted ef-

——— ... —- —-——‘l’’ According to a projection made at an early 1981 data proc-

essing managers conference, Department of Defense softwarecosts would increase nearly three times as fast as the depart-ment’s budget, and nearly 20 percent faster than expendituresfor computers during the 1980s, ” B. M. Elson, “Software Up-date Aids Defense Program, ” Aviation Week and Space Tech-nology, Mar. 14, 1983, p. 209. Just as the explosion in num-bers of telephone operators in the 1930s and of bank clerks forcheck processing in the 1960s forced a move to automated sys-tems, the sheer demand for manpower in computer program-

fort among researchers, educators, data proc-essing managers, systems designers, and pro-grammers, and support from corporate andGovernment management is required for thistransformation to occur.

Software R&D EnvironmentsVarieties of Software and Characteristicsof Software Development

Software is classified as being of two gen-eral types: applications software that is de-signed to apply computer power to a specifictask or tasks, such as computer-aided designof automobiles, or payroll or inventory man-agement in a department store; and systemssoftware that is used to manage the compo-nents of an information system itself, such ascomputer operating systems that control in-put and output operations.

In general, systems software is an integralcomponent of the hardware because its job isto control the hardware, including the peri-pheral equipment–e.g., disk, printer, andmemory usage, and to schedule and accommo-date the creation and execution of applicationsprograms. A recent trend, encouraged by thespread of personal computers, has been towardthe use of standard systems software so thata large number of applications programs canbe made comparable with hardware from dif-ferent suppliers. The manufacturers of hard-ware or specialized software vendors writemost systems programs, and more of the sys-tems programs are being embedded in hard-ware-in programmable integrated circuitmemory called ROM (Read Only Memory).Thus end-users now generally do not create oralter systems programs.

Much applications programming is done byusers. For personal computers in homes andfor supercomputers at the National WeatherService, programs must be written to tell themachines how to solve problems and organize

ming seems to be increasing the pressure for the introductionof less labor intensive software development techniques; seeT. C. Jones, “Demographic and Technical Trends in the Com-puting Industry, ” DSSD User’s Conference.

information. Much of this work is done byhighly trained professionals knowledgeableabout computer systems and the problems tobe solved.

Not all applications programming is soft-ware R&D. For example, an economist writinga short program to calculate a unique set ofstatistics that may never be used again or byothers is not engaged in software R&D. Con-versely, a physicist developing a program fora supercomputer to calculate formulas used innuclear reactor design certainly may be in-volved in R&D. There is a large area in be-tween these examples that is ambiguous, andno hard data is available concerning the timespent creating different categories of pro-grams. The Internal Revenue Service hasfaced this difficulty in defining the types ofsoftware work that are eligible for the R&Dtax credit. Their proposed solution has beento consider the costs of developing computer

software as not eligible for tax credits, unlessthe software is “new or significantly improved, ”or “if the programming itself involves a sig-nificant risk that it cannot be written. ”52

Software R&D in Industry

The information processing industry isdevoting a large and growing amount of re-sources to software development. Purchasesalone, currently some 12 percent of totalspending for software, are expected to exceed$10 billion in 1984. Approximately 10,000companies of various sizes, from one man oper-ations to divisions of major corporations, aredeveloping software for sale. Estimates of thetotal number of programmers in industryrange from 500,000 to nearly a million,53 and——-——

‘*’’Credit for increasing Research Activity, ” Federal Regis-ter, Jan. 21, 1983, pp. 2799-2800. There has been considerablecontroversy over this issue. See W. Schatz “A Taxing Issue, ”

June 1983, pp. 58-60.53ADAPS0 estimates, and T. C. Jones, op. cit. p. 83.

78 ● Information Technology R&D: Critical Trends and IssueS

the distinction between hardware and soft-ware developers is becoming increasinglyblurred.54

Every major computer and telecommunica-tions equipment manufacturer and service pro-vider develops programs and software toolsto make their products suited to the needs ofusers. Some firms such as Hewlett-Packard,reportedly spend nearly tw0thirds of theirR&D budgets on software.66 More than 40 per-cent of the technical people at Bell Labs areinvolved with software development.56

Interest in the improvement of software pro-ductivity has led some large corporations toestablish formal programs for the develop-ment and evaluation of software practices—tointroduce scientific and engineering tech-niques into the evaluation of software devel-opment and use. AT&T currently employs ap-proximately 300 Ph.D.s in software research,while IBM has approximately 150 and ITThas about 20 of these researchers developingand using formal experimental methods ofevaluating software engineering techniques.Other large companies, including ControlData, Xerox, and Honeywell, are also begin-ning to use experimental R&D studies as thebasis for improvements in software produc-tion.57

The Federal Role

The National Science Foundation funds uni-versity and some corporate research in soft-ware engineering. The NSF Software Engineer-ing Program within the Division of ComputerResearch awarded $2.2 million in grants infiscal year 1983.56 Additional research relatedto software engineering is funded by other pro-grams in the Computer Research Division(e.g., the Software Systems Science, Computer— — . ——

54See, for example, S. B. Newell, A. J. De Geus, and R. A.Rohrer, “Design for Integrated Circuits, ” Scieme, Apr. 29,1983, pp. 465-471. See also “The Changing Face of Engineer-ing, ” Electrom”cs, May 31, 1983, pp. 125-148.

‘SW. P. Patterson, “Software Sparks a Gold Rush, ” indus-try Week, Oct. 17, 1983, pp. 67, 69-71.

56AT&T: 1982 Annual Report, p. 19.‘70TA workshop on Software Engineering, Nov. 17, 1983.58Summary of Awards, Fiscal Year 1983, National Science

Foundation, Division of Computer Research.

Systems Design, Theoretical Computer Scienceand Special Projects Programs) bringing thetotal funding to $5 million to $10 million peryear.59

The Federal Government is the world’s larg-est user of data processing resources. A recentGAO study found that 95 to 98 percent of theGovernment’s applications software is customdeveloped. 6o The cost of software for the De-partment of Defense is estimated to be $4 bil-lion to $8 billion per year.61 DOD operates apatchwork of incomparable systems and com-puter languages.62 The incomparability of soft-ware contributes to increased software devel-opment costs, through schedule slippages,lengthy testing programs, and problems incontracting for hardware and software services.

DOD has moved toward the developmentand use of a single standard computer lan-guage. The rationale for this is the potentialfor saving several hundred million dollars ayear through lower personnel training costs,increased programmer productivity, and sub-stantial reuse of standard code modules. Aftercompetitive development of four separate lan-guages and extensive design evaluation, thePentagon chose a language developed by theFrench company Cii Honeywell Bull. Thename of the language, Ada, is trademarked,and compilers using the Ada name are strictlycontrolled and validated to assure that the lan-guage remains standard. Ada is expected tobe the primary DOD computer language by1987.

There is some resistance to use of Ada. Sev-eral years ago, the Air Force developed its ownquasi-standard language, Jovial. Comparisontests between Ada and Jovial will continue for

59Estimates by OTA.‘“’’ Federal Agencies Could Save Time and Money With Bet-

ter Computer Software Alternatives, ” General Accounting Of-fice, GAO/AFMD-83-29, May 20, 1983, p. 1. Even standard ap-plications such as payroll are largely custom designed. ThisGAO report found that there are at least 78 different Federalcivilian payroll systems.

elElson, O p . Cit., p. 209.

’21. Peterson, “Superweapon Software Woes, ” Science News,May 14, 1983, pp. 312-313. Testing of software alone is esti-mated to account for nearly half of this cost.


the rest of the decade.63 The Army is more en-thusiastic about Ada. The Navy, which hasmore software already written than the AirForce and Army combined, is interested, butthe task of switching the Navy to Ada will beenormous. The Navy has over 450 differentsystems and subsystems with embedded com-puters, and the number of Navy computershas been doubling every 2 years.64

A limitation of Ada is that a new program-ming language only addresses about 20 per-cent of the total software problem. As detailedbelow, coding of computer programs is 20 per-cent or less of the effort of developing andmaintaining software. Government computersystems in particular require an enormousamount of documentation, and Ada has no fa-cilities for automated documentation.65

To deal with these problems of software de-sign, production and maintenance, DOD hasproposed a new initiative called AdvancedSoftware Technology.66 This is envisioned asa 10 year program costing $250 million.67 Theadministration requested a funding level of$19.3 million in fiscal year 1984, but the Sen-ate Armed Services Committee cut the pro-gram to $10.5 million because, “The Com-mittee is not convinced that the necessaryplanning has been done to justify a budget ofnearly $20 million in the first year. ”68 Includedin the plan, as it has thus far been developed,is a provision for a Software Engineering In-stitute to help formulate and standardize soft-ware engineering techniques and practices.69

— . —“J. Fawcette, “Ada Tackles Software Bottleneck, ” High

Technology, February 1983, pp. 49-54.64petergon, op. cit., p. 313. It has been estimated that there

are some 50 million unique lines of Navy software code in a va-riety of languages currently in use, It would cost, it has beenreported, some $85 billion and take several years to rewrite thismass of code.

e5The OTA workshop on Software Engineering, NOV. 17, 1983.“This initiative is known within DOD as STARS, Software

Technology for Adaptability, Reliability and Serviceability.eTPeterson, Op. cit.eaomm”bus Defense Authorization, 1984, Report to Accom-

pany S.675, Committee on Armed Services, U.S. Senate, p. 131.‘BOTA Workshop on Software Engineering, NOV. 17, 1983.

Content and Conduct of SoftwareEngineering R&D

Software engineering ideally is a set of con-cepts and tools for transforming descriptionsof tasks to be performed by computer systemsinto digital code that machines can under-stand. Software engineering research involvesthe study of methods to understand, improve,implement, and evaluate these concepts andtools, and to embody them in software devel-opment systems or programming environ-ments to facilitate the entire software lifecycle.

The Software Lifecycle

As can be seen in figure 10, there are sev-eral stages in the life-cycle of a piece of soft-ware. Each of these stages is characterized byits particular set of objectives and methodsthat influence each stage of research and ex-pected improvements. Testing occurs continu-ously throughout the life of a software and isan integral part of software production andmaintenance. Documentation, an activity notdepicted in figure 9, is of preeminent impor-tance in every phase of the software lifecycle.Comprehensive records of every activity andsoftware characteristic must be created to aiddesigners, programmers, maintainers, andusers in understanding the structure and oper-ation of the software system.

Requirements Specification

This initial phase of software developmentinvolves the description of the system objec-tives and the tasks that the end users wantperformed. This description requires a thor-ough knowledge of the application by the soft-ware design project team. It is accomplishedby rigorous and continuing communication be-tween the project team and the end users.

OTA advisors and published sources empha-size that requirements specification is themost critical phase of project development, be-cause all of the later stages must build on thefoundation laid in this activity. At the sametime, it is the most difficult activity in the soft-ware lifecycle to develop a rigorous method-


Figure 10.—The Software Life Cycle

SOURCE: Data Communications.

ology for. Requirements specification involvesthe understanding, explicit organization, andintegration of what are often idiosyncraticpractices of information usage. Often it is notpossible to completely specify requirements atthe beginning of a project; also, requirementscan be expected to change over the course ofa long software development project.

Computerized tools now available or underdevelopment may offer assistance in the speci-fication process. 70 They are designed to helpusers describe and specify system properties,functions and performance requirements.Some of these tools allow the specifications tobe checked by computer for consistency andcorrectness; some of them can generate simu-lations to help the user analyze the operationof a specified system. Thus far these tools haveworked well only on a limited range of ap-plications.

A difficult problem is specifying how theknowledge, experience and habits of workersin particular environments can be describedand how these abstract specifications can be

7WTA workshop on Software Engineering, Nov. 17,1983.

converted into machine code. Research is pro-ceeding in knowledge representation andknowledge engineering (see Artificial Intelli-gence Case Study). Near term prospects areuncertain for identifying core concepts fororganizing the different kinds of knowledgefound in the variety of environments in whichcomputers are applied, or designing broadlyapplicable very-high-level languages which canautomatically render abstract specificationsin machine-executable procedures.71

The experience now being gained in researchand development of “expert systems, ” and themore fundamental research efforts in knowl-edge representation, machine learning andhuman cognition should eventually contributeto the process of computer systems require-ments specification. But for the near term, thisphase of software development will remain la-bor intensive and will require special skills (in-cluding human relations) and specific knowl-edge of applications among its practitioners.

— — . —“R. Yeh, “Software Engineering, ” IEEE Spectrum, Nov.

1983, p. 92.


Prototyping

Once the project team has a working knowl-edge of the requirements of the software, a pre-liminary design of the final system is made sothat the team can obtain feedback from theusers. Thus, fine adjustments can be made atan early stage.

A software prototype can be as simple as apaper and pencil sketch that runs through theworkings of the system, or as complex as alarge and intricate computer simulation. Thechoice of prototype complexity is based on thenature of the application, the level of trainingand the sophistication of the end users, thesize of the final software system, and thecriticality of the application.72

Research and development of the prototyp-ing phase involves the design, developmentand testing of methods and tools to make pro-totyping more understandable to systems de-signers and users, and more automated so thatit is a logical and accepted part of software de-velopment. Research efforts in universitiesand industry are focusing on languages thatmake the quick production of prototypes pos-sible, and on the design of man-machine inter-faces to enhance and verify the effectivenessof communication between the system and itsusers.

The major issue in prototyping involves theacceptance of this activity and the awarenessamong system designers of its importance.Prototyping is commonly ignored in currentsoftware development practice because of itsexpense and difficulty and the lack of auto-mated support. But experts contend that, inthe long run, prototyping can save time andeffort because changes to requirements speci-fications become more costly as each phase ofsoftware development proceeds.

Design

The design stage of software developmentinvolves the reduction of the software speci-

72 For example, air traffic control software requires extensivesimulation, whereas an inventory system might need only a simp-le sketch and some text to check against user needs.

fications to a set of procedures that can be pro-grammed for the computer. The softwaredesign team analyzes the application and seg-ments the design problem into subproblemsor modules that can be programmed by in-dividuals or small groups.

The segmentation or modularization of thesoftware design breaks the problem into man-ageable pieces to maximize the ease of pro-gramming and testing the segments, and tofacilitate the interchange or replacement of themodules to simplify maintenance. Propersegmentation is of particular importance be-cause it has been demonstrated that as the sizeof the programming project team grows, moreand more time is spent communicating amongteam members and coordinating communica-tion among program segments and less in ac-tually writing code. Thus segmentation has acrucial impact on the overall productivity ofthe software development team.

The design phase of software productionrelys heavily on the experience and intuitionof the design team and is considered to be bestlearned by apprenticeship. Understanding offundamental mathematical principles is alsoimportant. 73

Research on the design phase of softwareengineering is embedded within the largerframework of R&Don “programming environ-merits. ” The objective of software engineer-ing is to provide as much support as possiblefor the creative designer. Facilities should bemade available to ease the rendering of ab-stract principles and problems into workable,testable, reliable, and cost-effective solutions.Software design can be made more productivein much the same way that design of complexintegrated circuits has improved: by the ap-plication of computer-aided-design (CAD) toolsbased on sophisticated graphics workstations.A number of companies are making significantproductivity gains using CAD tools for soft-ware design.

. —“Data Communications, op. cit. p. 81, and AZosa”c, op. cit.

pp. 4-5.


In the longer term, software design produc-tivity could improve significantly through theexpanded development and use of softwareLibraries, or systems to store and access stand-ard, reusable software modules that performfunctions required in many programs. As nowpracticed, software design generally segmentseach program problem in a unique fashion,both to accommodate the unique features ofnew applications and to make the most effi-cient use of hardware. Program modules areusually not designed for reuse, and there is nostrong incentive to do so because facilities arenot provided for storing separate modules orfor classifying and finding appropriate seg-ments for new programs, and because the proj-ect development team is generally unable tospread the cost of creating reusable modulesacross a large organization. The result is thatdesigners and programmers are constantly“reinventing the wheel” in software devel-opment.

Researchers in computer and informationscience have identified basic techniques andhave designed tools to provide generic pro-gramming modules, data structures and algo-rithms, and to classify and organize them forstorage and access. But the software commu-nity has been slow in adopting the facilitiesnecessary to make software libraries work be-cause of high initial cost, poor managementunderstanding of such systems, and reluctanceamong programmers to modify long-establisheddesign practices.

coding

Coding is the stage of software productionthat makes a procedural description of the sys-tem executable by computer. More emphasishas traditionally been placed on this activitythan on any other, even though it is estimatedto comprise only 15 to 20 percent of the totalsoftware development effort.74 The reason forthe traditional emphasis on coding is that itrepresents the basic mechanism by which pro-

— .————. .—“Data Communications, op. cit. p. 58, and E. B. Altman,

“Software Engineering, ” Mim”-Micro Systems, December 1982,p.184.

grammers control the operation of computersystems.

Many tools are already available to pro-grammers to help produce efficient and reli-able code. The most familiar of these are thehigh-level languages such as FORTRAN,COBOL, and Basic that were designed to aidprogrammers in coding certain broad types ofapplications. Compilers translate high-levellanguage into machine executable binary dig-ital code. Editors and debuggers facilitate theentry, testing, and correction of code lines.

Research and development in the improve-ment of computer code writing continues. In-novations over the last decade include struc-tured programming which greatly simplifiesthe tasks of testing and deciphering code dur-ing maintenance, and new high-level languagesdesigned to be easy to learn and use or tech-nically sophisticated (no languages yet intro-duced appear to be both simple and highlyflexible).

A recent trend, that has proceeded hand inhand with the introduction of personal com-puters, has been the development of codingsystems or languages that can give end usersof computer systems more control—and re-move the need for the help of a professionalprogrammer for each application. An exampleis Visicalc.

Large companies, including manufacturersof mainframe computers, are interested indecentralizing the control of computer pro-

gramming and use by providing software sys-tems that are responsive to the needs of non-computer professionals. A proposed solutionto the software bottleneck dilemma is to farmout some of the work now done by professionalprogrammers to end users. IBM, for example,asserts that up to one-half of computer ap-plications development can and should be han-dled by the end-users with personal com-puters.75

75AppZication Development in Practice, Tech Tran User Sur-vey, Xephon Technology Transfer, Ltd., 1983, pp. 53-54. Someexperts consider this solution to be dubious, and believe thatit may in fact merely spread software problems.


The proliferation of personal computers inthe business world has some other possible im-plications, both for the process of software de-velopment and for employment patterns ofprofessional software development personnel.End-users programming on personal com-puters could allow central data processing de-partments to concentrate their efforts on thelarger, more complex, and more critical ap-plications. A reduction in the backlog of smallprojects could also encourage the developmentof new large-scale applications that heretoforlanguished because of long waiting times. Itis conceivable that the demand for profes-sional programmers will slacken in such indus-tries as financial services which have tradi-tionally employed a large part of the dataprocessing workforce in coding and maintain-ing applications programs. This could resultfrom trends toward: 1) small-scale applicationsdevelopment by end-users with personal com-puters using commercially available softwarepackages, and 2) increased use in data proc-essing shops of standardized and automatedsoftware development tools, also commerciallyobtained, for large-scale projects. But increas-ing demand for innovative, off-the-shelf ap-plications programs and software develop-ment tools suggests that demand for the mosttalented software designers will increase ascompetition intensifies in these expandingmarkets.

There is a secondary impact of the introduc-tion of personal computers on software devel-opment. Upper management is becoming fa-miliar with the problems and possibilities ofsoftware and computer use through experiencewith personal computers. Management awareness of and commitment to change in the soft-ware development process may be a key fac-tor in introducing and accepting software engi-neering concepts and innovations.

Large and complex applications will con-tinue to depend on the efforts of professionalsoftware designers and programmers. Thus,the introduction of automated tools and tech-niques such as CAD are imperative. Pressurewill continue for more productive and reliable

software engineering as the complexity ofcomputer-based systems increases and asmore of society’s functions are entrusted tocomputerized systems.

Testing and Validation

An important activity in software produc-tion and maintenance is the testing of theproducts of the various phases of the softwarelifecycle. Until recently software developmentbudgets included sufficient funds for adequatetesting in only a few highly critical projects.

It is impossible, no matter how extensivethe testing activities, to guarantee large soft-ware systems to be error free. A risk assess-ment based on the cost associated with soft-ware defects or breakdown in a given applicationmust be made to determine the level of effortin testing that is justified. A rigorous testingregime can double the cost of software devel-opment.

There are many concepts, tools and tech-niques available and undergoing research ordevelopment for testing and validating soft-ware. They range from traditional methodssuch as manual “desk checking” or “walkthroughs, ” to statistical methods to determinethe probability of errors in a set of code, tosome as yet highly experimental automatedprogram provers that verify that properlystructured programs perform as specified. Thebroad applicability of program provers will depenal on some fundamental breakthroughs inprogram theory.76 Until the requirementsspecifications and design stages of softwaredevelopment are better understood and sup-ported, manual techniques will be the keytesting methods available. Although someautomated tools are in use, few exist as in-tegrated packages. They must generally bepieced together by individual projects; and— —.

“Some experts contend that the pursuit of formal verifica-tion methods such as program provers may prove fruitless be-cause of the increasing complexity of software and the rapidchanges in software development practices. See R. A. DeMillo,R. J. Lipton and A. J. Perlis, “Social Processes and Proofs ofTheorems and Programs, ” Communications of the ACM, May1979, pp. 271-280.


they themselves can become sources of highcost and errors.77

Maintenance

This is the phase of the software engineer-ing lifecycle that is concerned with the revi-sion of software that is in use. Software main-tenance consists of two kinds of activities:correction of errors that went undetected inthe course of software development and test-ing but that crop up in the use of the software;and changes in programs resulting from al-tered or additional requirements specifica-tions. It has been estimated that 50 to 90percent of current software costs involve main-tenance. Some software systems may cost 25times as much to maintain as to develop.78 Be-cause of poorly designed, badly written, andinadequately documented code, much of themaintenance programmers’ time is spent try-ing to understand programs rather than chang-ing them.

A solution to the maintenance problemadopted in many cases is to discard old code.This is because experience indicates that the“patches” made in programs when they arefixed or revised often increase the complexityand difficulty of maintaining code and makeit inefficient to run. Therefore an average of10 months worth of programming develop-ment effort is thrown away each year by main-frame based data processing organizations.Unfortunately, this old code is often replacedby poorly designed new code which again isdifficult and expensive to maintain.

Poor requirements specification and designpractices, idiosyncratic coding styles, andpoorly organized documentation increasemaintenance costs. Because requirementswere vaguely specified or have changed, thedesign proceeded on assumptions that werefalse or later rendered invalid. Bad designassumptions lead to poor segmentation andstructuring of programs which exacerbate the

77w. R. Adrion, hf. A. Branstad, and J. C. Cherniavsky,“Validation, Verification, and Testing for Computer Software,”ACM Computing Survey, June 1982, p. 183.

7sOlsen, op. cit. p. 58.

problems of making changes and testing. Achange in one part of a poorly designed pro-gram will require changes in other parts of theprogram, multiplying the cost and time formaintenance. Novel coding of standard func-tions increases the difficulty of predicting theeffects of changes. Large and cumbersome,poorly organized and incomplete documenta-tion make it difficult for maintenance pro-grammers to know what the system is sup-posed to do.79

Junior programmers with little experienceare generally assigned to maintenance tasks,thus increasing the delays in bringing softwareback into use, and also perpetuating theknowledge and acceptance of poor program-ming styles and habits.

The Scientific Basis for Software Engineering

There are some well understood scientificmethods that can contribute significantly tothe improvement of software engineering toolsand practices. The application of scientificallydesigned, experimentally based statisticalmethods for the evaluation of the effectivenessof software tools may produce the conceptsand methods for at least an order of magni-tude increase in software productivity.

In particular, in the critical area of humaninteraction with computers, a body of researchmethods have been developed in universitiesthat, if widely and systematically applied,could identify, quantitatively assess, and pro-duce improvements in programs for makingcomputers more responsive and easier for peo-ple to program and use. Research in behavioralpsychology and psychophysics is being ap-plied to the design of computer systems andsoftware engineering tools on a limited basisin a number of university programs.

The limited scale of commitment to appliedresearch in software engineering maybe a ma-jor reason for lack of dissemination and useof these well understood techniques. Acad-

7%3TA advisors related that program documentation captures,on average, only 60 percent of the information needed to fullyunderstand a program.

Ch. 3—Selected Case Studies in Information Technology Research and Development • 85

emics are generally involved in small-scale,fundamental research oriented projects thatcannot test large size programs or a large num-ber of possible software development aids.Some researchers suggest that applied re-search on a massive scale following the modelof medical clinical trials is necessary to dis-seminate and make use of the knowledge em-bodied in available techniques.

Industry participation is an important in-gredient in increasing the scope of experimen-tation because of its large and diverse softwareefforts. As a user of innovative tools and sys-tems, industry must also be involved in theresearch so that essential feedback is providedto tool designers and evaluators. Inadequateappreciation of the gains that can be achievedin software productivity and a concentrationon short-term results have made industrymanagement reluctant, until recently, to makea commitment to large scale or long-term ex-perimental software development studies.

As mentioned earlier, some large companiesare becoming interested in applied software re-search and are hiring computer and humanfactors scientists to develop research pro-grams. However, competitive and proprietaryconsiderations and the uncertainty of softwareprotection by copyright or patent appear tohave inhibited an integrated applied researcheffort among companies. Antitrust laws mayalso have a chilling effect on industry basedjoint applied research.80

An alternative approach is for the FederalGovernment to use its software developmentefforts as a test-bed for software engineeringresearch. Some moves in this direction are be-ing discussed within the defense community,and the National Bureau of Standards has aresearch effort in software engineering in itsInstitute for Computer Science and Tech-nology.

SOEleven major defense contractor companies recently an-

nounced plans to form a consortium for the pooling of effortsin software engineering R&D. The Justice Department hasreportedly given the go-ahead for the development of a formalbusiness plan patterned after the Microelectronics and Com-puter Technologies Corp. (MCC). M. Schrage, “Software Re-search Group Set, ” Washington Post, Oct. 10, 1984, pp. C1-C2.

International Efforts in SoftwareEngineering R&D

JapanThe Japanese have had a software engineer-

ing effort since 1970.81 The Information Tech-nology Promotion Association (IPA), a consor-tium of private companies and the govern-ment, together with long-term credit banks,provided $25 million in 1983 for software R&Dand technology transfer. Current projects in-clude: purchase or development of softwarepackages for rent to users; a Software Main-tenance Technology Project, to be completedin 1985, which is developing work stations forprogram analysis, documentation, testing, andproduction management; and a cooperative ef-fort among users, suppliers, and researchersto develop computer-aided design, computer-aided instruction, and software engineering in-novations.

A noteworthy aspect of Japanese softwareengineering efforts is the establishment ofsoftware factories. These are consortia, staffedby people from participating companies, whichare creating integrated environments for soft-ware production, testing, and maintenance.Impressive programmer productivity gainsare reported to have resulted from the aver-age reuse of 30 percent of computer code innew applications.82 Japanese software fac-tories appear to owe their success in raisingproductivity levels, in part, to the fact thatonly restricted kinds of applications with fixedand well understood requirements are at-tempted. It is uncertain how effective thesesoftware factories would be on more difficultapplications. Nevertheless, these concerted,cooperative efforts provide a test-bed for newsoftware engineering concepts and tools in awell capitalized development and productionenvironment. As well, they train people in theuse of advanced software development tech-niques, serve as a dissemination mechanism

81 Summary of Major Projects in Japan for R&D of Informa-tion Processing !fechnology, prepared by Arthur D. Little (Ja-pan), Inc., under contract to OTA, July 1983.

‘*Raymond Yeh at the OTA workshop on Software Engineer-ing, Nov. 17, 1983.


when those people return to their companies,and raise management awareness of the im-portance and efficacy of software engineering.Thus software factories have potentially sig-nificant implications for future Japanese soft-ware development efforts on a broad scale, re-gardless of their impact on individual softwaresystems.

Britain

As a part of the Alvey Programme for R&Din information technologies (see ch. 7), Brit-ain plans to spend approximately $100 millionover 5 years (1984-89) on software engineer-ing R&D, which is approximately 20 percentof the total Alvey Programme effort.83 Bri-tain’s traditional strength in software will bebuilt upon, expanded and modernized to min-imize dependence on software imports.84 Aninitial focus of the Alvey effort will be themeasurement of the cost effectiveness of soft-ware engineering through analysis of softwareimports and exports. Also, the program willtrack the capitalization of software develop-ment and analyze the relationship of capitalintensity to programmer productivity. Effortswill be made to establish formal output meas-ures for software cost and quality to helpevaluate new tools and methods and to estab-lish a system for software warranties.85

The software engineering effort will consistof three main thrusts:86

1. Exploitation of existing tools and meth-ods, the transfer of technology from uni-versities to industry, and increased in-vestment in software production capitaland in training.

—.. .. —-. .—83A fiO/@UIMIl e for Advanced Information Technology: The

Report of The Alvey Co-”ttee, Department of Industry, HerMajesty’s Stationery Office, p. 47.

84A]vey Softwwe En@”nMring—A Strategy Overview, Pre-pared by Alvey Software Engineering, Director D. E. Talbot,Department of Trade and Industry, p. 1.

as~ftwm &h”abih”ty and Metn”cs ~gramm e: Ovem”ew, prf+pared for the Software Engineering Directorate by the Centerfor Software Reliability, Alvey Directorate, Department ofTrade and Industry, April 1984, p. 4.

‘Alvey Software En@”neen”ng-A Strategy Overview, op. cit.,p. 3.

2.

3.

Integration of tools and methods for theimprovement of hardware and softwaredevelopment which will be focused in anInformation Systems Factory.Innovation software engineering throughincreased levels of R&D aimed at devel-oping new tools and evaluating their ef-fectiveness.

The Information Systems Factory, to be es-tablished by 1989, is based on the premise thatinformation systems’ functional requirementsmay be written independently of whether agiven function is to be implemented in hard-ware or software. Therefore, it is reasoned, thetrend should be toward the design of moduleswhose uses are known in relation to other mod-ules, such that the decision of whether to im-plement a function in hardware or softwaremay be made on the basis of economic, time-scale and other cost-benefit criteria. The In-formation Systems Factory concept will evolveinto the 1990s, by which time advances in fun-damental research are expected to make pos-sible a truly intelligent software engineeringproduction environment with fully automatedand integrated tools and a large library ofstandard function modules.87

The British consider the understanding ofsoftware engineering concepts to be too prim-itive at this point to make a final determina-tion of a single direction to pursue to producea fully integrated software engineering envi-ronment. Like those in the United States, cur-rent British research projects are too small inscale to adequately test software in diverse ap-plications to determine the efficacy of newtools and methods. The Alvey Programme willfund several approaches for full-scale develop-ment, both to select useful techniques and tobuild technology transfer bridges between re-search, development, and production. In thiscontext, the software engineering effort is ex-pected to contribute to, and to benefit from,the parallel Alvey Programme efforts in com-puter-aided design for VLSI circuits andknowledge-based systems.88

——..—————“Ibid., pp. 5-7.*aIbid., pp. 7-8.

Ch. 3—Selected Case Studies in Information Technology Research and Development Ž 87

France

Like the British, the French have been notedin the past for the quality of their software.89

The major French R&D centers for softwareengineering are run by L’Institute Nationalde Recherche en Informatique et en Automati-que (INRIA) and the Centre National d’Etudes

“yThree significant computer languages, Algol, Prolog, andAda, were originated in France.

Telecommunications (CNET). Together, theselaboratories employ approximately 100 re-searchers in software engineering and relatedstudies. 90In addition, 11 other French Gov-ernment and industry labs are pursuing R&Din software engineering.91

wRe~ewch and De\~e]opnlent i n h’lectrom”cs uSA-France1982/1983, French Telecommunications and Electronics Coun-cil, March 1984.

glI~~E Spectrum, November 1983, PP. 64-65.

Case Study 4: Artificial Intelligence

Findings

● Artificial intelligence (AI) research seeks tomake computers perform in ways that dem-onstrate human-like cognitive abilities: per-ception and action in complex situationsand environments; interaction with humansin natural language; and common sense andthe ability to learn from experience.

● The capabilities of artificial intelligence areoften subject to exaggeration. AI has gonethrough several waves of optimism and dis-appointment as new fundamental conceptshave been discovered by research, as newcomputational techniques and equipmenthave allowed these concepts to be tested,and as the limits of these concepts havebeen realized when applied to real-world orlarge-scale problems.

● In the last several years, the first real com-mercial products of about 25 years of AI re-search have become available. These sys-tems are of three types:— expert systems that aid human experts

in analyzing complex situations and inmaking decisions,

— natural language processing programsthat serve to make the interaction of hu-mans and computers more natural, and

— image or vision processing systems thatcan make robots and other automatedprocesses more flexible in operation.

● Presently available AI technology is ofvalue in only a limited range of applications.But the promise and potential have led to

several national efforts to push AI technol-ogy forward.In the United States, AI research has beensupported principally by the Department ofDefense. Plans for putting artificial intelli-gence concepts to work in defense applica-tions have recently been announced. Theseapplications are far beyond the capabilitiesof any present systems.Basic AI research is conducted at a rela-tively small number of the nation’s topuniversities by a small number of research-ers. These researchers are under conflictingpressures from companies wishing to cap-italize on the production of highly valuablesystems, from students demanding trainingin AI, and from personal desire to pursuetheir own research interests.Only a few universities have the facilitiesand equipment to compete with industrialAI labs. Therefore some highly motivatedresearchers are drawn out of academiawhere they would be available to performneeded fundamental research and to trainfuture generations of researchers.

Introduction

The goal of artificial intelligence (AI) re-search is to create systems that demonstratesome of the following characteristics: theability to assimilate unstructured informationand to act independently in complex situa-tions; the capability of natural language in-teraction (e.g., in English) with humans; com-

38-802 0 - 85 - 7


mon sense, and the ability to learn fromexperience. 92 The pos s ib i l i t y t h a t m a c h i n e smay be made capable of such activities hasburst into public consciousness as a result ofthe increasing power of microelectronics-basedcomputers and their application in many newenvironments, the establishment of nationalefforts with the aim of creating intelligentcomputer systems, and the sudden emergenceof some commercial products that embodycharacteristics of artificial intelligence.

Some experts contend that the ultimate aimof computer science is to produce intelligentmachines. Just as the industrial revolutionwas driven by the urge to amplify human andanimal muscle power with more efficient me-chanical devices, the computer revolution isdriven by the desire to amplify human intel-lectual power with electronic informationmachines.93

An expanding user population and the in-creasing diversity and complexity of the ap-plications of current and planned informationsystems, are compelling designers to find waysto build intelligence into computers. Com-puters are increasingly being required to ex-change information among people with diverseinterests, backgrounds, and levels of computersophistication, so they are being designed withsoftware that permits people to interact withthem in forms more closely resembling natu-ral language. New varieties of sensors and in-creasingly diverse sources of information areproviding input, thus computer systems mustbe more flexible to make use of informationwith varying levels of importance and con-fidence attached to it. Processing and output

9~D. w~tz, “Artifici~ Intelligence: An Assessment of TheState-of-theArt and Recommendation for Future Directions, ”AI Magazine, fall 1983, pp. 55-66.

93 For exmple, the Japanese 5th Generation project is moti-

vated by a belief that intelligent computer systems will be oneof the cornerstones of a healthy economy and society in the lateryears of the 20th century and beyond. See T. Moto-oka, “Chal-lenge of Knowledge Information Processing Systems (Prelimi-nary Report on Fifth Generation Computer Systems), ” pp. 3-89, and H. Karatsu, “What is Required of The 5th GenerationComputer–Social Needs and its Impact,” pp. 93-106 in FifthGeneration Computer Systems, T. Moto-oka (cd.), JIPDEC-North Holland, 1982, 287 pages.

facilities are needed that present informationin forms that are tailored to the unique needsof individuals, or that can control operationsin complex and perhaps unpredictable circum-stances.

The complexity of the environment in whichthe system is to operate is perhaps the mostimportant dimension to be considered in thedesign of AI computer systems. As the num-ber of environmental variables increases, andthe number of possible responses multiplies,automated systems must exhibit increasingdegrees of intelligence. Currently, computersystems are being designed for environmentsin which they will be required to resist confu-sion and error from ambiguous or contradic-tory input signals, automatically coordinatediverse activities under changing conditions,and provide logical and understandable ex-planations of their actions to operators.

The highest degree of machine intelligencemight be the ability to automatically adapt toconditions that were not specifically antici-pated by the designers. Such ability would re-quire a machine to have “common sense, ”broad world knowledge from which to inferreasonable courses of action, and also theability to assimilate new knowledge by learn-ing, through self-organization of experience.Machines with this degree of intelligence arespeculative, and attempts to push present AIconcepts toward such capabilities illustrateboth the difficulty of the problems, and the in-adequacy of present concepts as a foundationfor machine intelligence on a scale approachingthe cognitive abilities of humans.94

There have been some notable recent suc-cesses in applying AI concepts to real-worldproblems. The introduction of a number of AIsystems into commercial use has elicited a de-mand in industry for AI software that canmake computer systems more responsive andproductive. Three types of artificial intelli-gence products in particular are generating in-

“Some tasks requiring “intelligence” are already performedbetter by machines than humans, f~r example long division;other tasks may never be performed by machines. See Waltz,p. 55. See also M. Michael Waldrop, “The Necessity of Knowl-edge, ” Science, Mar. 23, 1984, pp. 1279-1282.


tense interest among industrial companies.These are: natural language data base inter-faces that allow users to access informationstored in a computer data-bank with Englishlanguage queries; expert systems that aid peo-ple in complex tasks that require experienceand detailed knowledge to perform; and robotvision systems that promise increased flex-ibility in manufacturing and other robot ap-plications. Industry sees these systems asenhancements or potential replacements forrare and/or expensive human skill.

Similarly, the Department of Defense is in-terested in artificial intelligence to enhance thepower, efficiency and reliability of increasinglyautomated computer-based weapons and com-mand and control systems. DOD’s new “Stra-tegic Computing” program seeks to push thefrontiers of AI science and technology for usein battlefield management systems, autono-mous (unmanned) tanks and submarines, andautomated expert assistants for airplanepilots.

There is concern, among both AI research-ers and observers of information technologyR&D, that enthusiasm about recent successesand fear of foreign competition are encourag-ing irrational expectations for artificial intel-ligence. These forces are drawing limited AIR&D resources toward risky, short-term de-velopment work, and away from some of thetough questions that AI research should andcould answer. As Arno Penzias, Nobel laureateand Vice President for Research at Bell Lab-oratories has written about artificial intelli-gence R&D:

A crash effort in any one area would inevi-tably pull talented people away from otherareas . . . whose payoffs to society might beeven greater in the long run. Our challenge isto improve computers and extend their exper-tise into all the areas where people will needthem and want them. We can best do thatwith a balanced program of research and de-velopment. . . using knowledge to help createa society that provides a meaningful life forall the people in it.95

‘6A. A. Penzias, “Let’s Not Outsmart Ourselves in Thinking-Computer Rush, ” Wall Street Journal, Sept. 13, 1983.

The sudden discovery of promise in artifi-cial intelligence by society at large raises sev-eral questions. Of immediate concern are thequestions of whether AI, as it is now under-stood, can meet industry and military expec-tations for performance, or whether there willbe expensive failures in applying this technol-ogy, and by implication a casting of doubt onthe entire AI enterprise, because present con-cepts are immature and limited; and whetherthe present educational structure, or some rea-sonable extension of it, can meet the growingdemand for trained AI R&D talent. For thelonger term, there is a question whether theprogress of AI research can be sustained inthe face of feverish commercial development,or whether the limited number of researcherswill be drawn away from the study of funda-mental and difficult problems and the teachingof new AI people. There are also profoundquestions concerning how artificial intelligencetechnology might be used, and how these usesmight affect society in general, especially interms of employment and the potential thetechnology offers to enhance or compete withhuman intellect.

Artificial Intelligence R&D Environments

The Roots of AI

The idea of automated intelligence has in-trigued mankind for more than a century.% Inthe 1950s, two important figures in the earlyhistory of electronic computers, John von Neu-mann and Alan Turing, both expressed confi-dence that within a short time computingmachines would equal or surpass human in-tellectual capabilities.97

‘E. Charniak, “Artificial Intelligence-An Introduction, ” presentation at the 1983 Conference of The American Associationfor Artificial Intelligence, Washington, DC. One can conceiveof artificial intelligence as the culmination of the mecha”sticparadigm of scientific thought that has been dominant sincethe time of Newton. See Science and Change, 1500 to 1700 byHugh Kearney, McGraw Hill, New York, 1971, for a discus-sion of the emergence of the mechanistic paradigm. Somehistorians suggest that Man has always sought to representand embody life and intellect in his artistic and useful creations.See J. David Boulton, Turing’s Man: Western Culture in theComputer Age, (Chapel Hill) University of North Carolina Press,1984,

“See J. von Neumann, “The General and Logical Theory ofAutomata, ” pp. 99 and 109, and A. M. Turing, “Computing


Improvements in the understanding of com-putation and knowledge organization coupledwith advances in computer speed and mem-ory capacity made possible through improve-ments in electronics, have encouraged inter-mittent waves of optimism that super electronic“brains” were just a few years in the future.Time and time again, these waves of optimismhave ebbed in the face of a common problem.As the systems that researchers contemplateand build are required to deal with more in-formation, and as the situations in which theyare required to operate become more realistic(more perceptually complex and unpredicta-ble), the fundamental computational principlesand methods they have to work with becomeless efficient and impossibly sl0W.98

The term “artificial intelligence” was coinedby John McCarthy in a grant application in1956 to describe the subject of a conferencethat he was organizing.99 This meeting, heldat Dartmouth College, brought together re-searchers in different fields whose commonconcern was the study of human and machinecognition. The conference established AI as adistinct discipline, and also served to definethe major AI research goals: 1) to design ma-chines that think, and 2) to understand andmodel the thought processes of humans. ThusAI research is grounded in computer scienceand electrical engineering and also has itsroots in cognitive psychology, linguistics, andphilosophy. There is an appreciation amongmany researchers that AI is an interdiscipli-nary study born of the interaction of these twogoals, and that the intersection of the sepa-rate traditions constitute what has come to bea core of concerns that distinguish artificialintelligence from other fields of science andengineering. loo (See fig. 11.)

Machines and Intelligence,” p. 245, in Perspectives on the Com-puter Revolution, Z. Pylyshyn, (cd.), Prentice-Hall, EnglewoodCliffs, NJ, 1970.

‘aJ. T. Schwartz, “Research in Computer Science: Influences,Accomplishments, Goals,” Report of The Information Z’echnol-ogy Workshop, R. Cotellessa, Chairman, National Science Foun-dation, Oct. 10, 1983, p. 18.

~gcharni~, op. Cit., p. 5-‘OONils Nilsson, “Artificial Intelligence Prepares for 2001, ”

Presidential Address given at the Amual Meeting of the Amer-ican Association for Artificial Intelligence, Aug. 11, 1983.

Figure 11 .—Artificial Intelligence

SOURCE: Randy Davis and Chuck Rich.

As an esoteric study melding some aspectsof computer science, linguistics, psychology,and philosophy, artificial intelligence researchhas grown up, for the most part, in an exclu-sive set of universities. Some industry-basedresearch efforts at AT&T Bell Laboratories,Xerox Palo Alto Research Center, Bolt Beranekand Newman, and SRI International spanmore than a decade, but most of the work hasbeen concentrated at Massachusetts Instituteof Technology, Stanford University, and Car-negie-Mellon University, the leading researchand academic centers offering a full range ofAI subdiscipline. The concentration of effortand expertise in these top tier universities hastended to restrict entry into the field, and tofoster a very tightly knit community of re-searchers.

Some years ago, AI research and trainingprograms began to spread to a wider set ofuniversities, and this expansion of academicprograms continues. These schools have pro-grams built around one or two graduates of

Ch. 3—Selected Case Studies in Information Technology Research and Development . 91

the established programs, and thus they of-fer only restricted AI curricula. Also, a num-ber of new industry-based R&D programs arebeing formed. The newer industrial programsin AI R&D have also tended to center on theideas of one or two researchers from the topschools. The lifespan of these industrial effortsmay be limited if they are unable to draw innew talent from the small available pool andare unable to obtain infusions of new ideas. 101

The difficulty of establishing and maintain-ing new AI R&D programs results from thefact that there is only a small base of existingresearchers. Second, the nature of AI work issuch that an interdisciplinary team is requiredin most cases to produce programs of usefulsize and complexity. This “critical mass” ofworkers knowledgeable in AI concepts existsonly in a few universities and companies andrequires a number of years to pull togetherbefore fruitful work can be expected.l02

OTA found that a major transformation isoccurring in AI, resulting in two major direc-tions for the R&D community. On one hand,fundamental research will advance theory. Theother direction is taking existing AI conceptsand developing their applications. As in otherfields, the two directions reflect the divergenceof interest between basic scientists and ap-plied scientists or engineers (see fig. 12). How-ever, AI is distinct from other fields that haveundergone this transformation in at least foursignificant ways: 1) AI is a very young (scien-tifically immature) field, 2) there are fewtrained AI practitioners, 3) the expectationsfor the technology are high, and 4) the paceof transformation to an applied study is fast.The divergence of interests is expected toeventually result in the concentration of thetraditional “top-tier” university centers of AIresearch on the fundamental scientific ques-tions, an expansion of industrial R&D effortsto deal with the development of applications,

‘“’ The OTA workshop on Artificial Intelligence, held Oct. 31,1983. There is some controversy on this point, but it is generally agreed that some “critical mass” of AX trained people, work-ing in close proximity where they can exchange ideas, is a crit-ical factor in performing advanced AI work.

“zWaltz, op. cit., pp. 64-65.

and the establishment of new academic pro-grams for the training of applied AI scientistsand engineers.

The growing commercial and military impor-tance of AI is forcing researchers to make dif-ficult choices in how they spend their time.They must balance a number of commitments:to the study of research issues that have long-term significance; to the teaching and super-vision of students;103 to consulting for indus-try and government; and to startup companiesin which they develop personal interests. Tobe sure, not all of these commitments aremutually exclusive. Long-term research andstudent supervision can reinforce one another.Consulting and entreprenuerialism can becomplimentary. But the wearing of many hatsby AI faculty may pose conflicts apart fromthe overcommitment of time.

First, there is a danger that students maybe judged, consciously or unconsciously, ontheir contribution to the business interests offaculty, rather than on their progress in learn-ing the science of AI.104 Second, the progressof AI, as well as that of other sciences, is de-pendent on the free exchange and public dis-semination of research results. Yet proprietaryinterests may inhibit that flow of informa-tion.105 Third, many researchers face a conflictbetween the objectives of the sponsors of re-search and their own conception of the bestdirection for the research. (See ch. 6 of this re-port on New Roles for Universities.)

This latter conflict may be particularly acuteamong those in AI research because of the factthat the social implications of artificial intel-

IOWne OTA advisor said that at his university, one quarterof the students now entering computer science and electricalengineering want to study AI. AI faculty across all schools areobliged to supervise, on average, some ten graduate students;in other fields the average ranges from less than one to aboutfour. As well as teaching and advising, faculty must spend timewriting proposals and seeking funding and equipment for grad-uate thesis work.

‘04 MIT has developed a policy that prohibits faculty fromsupervising students that work for companies in which thefaculty member has substantial interest. Patrick Winston atthe OTA Workshop on Artificial Intelligence, October 31, 1984.

105This point was made by advisors at the OTA workshop onArtificial Intelligence, and is discussed by Jordan J. Baruchin an editorial in Science, Apr. 6, 1984.


Figure 12.—Artificiai Intelligence Science and EngineeringResearch Topics

Motivations

I Requirements

The evolution of the field of artificial intelligence is producing new relationships and newdistinctions within the Al science and engineering community.

SOURCE. Office of Technology Assessment.

ligence are so profound. They include thereplacement of humans with machines inskilled work, the increased reliance on auto-mated systems and the risk of system failure,and affects on man’s perception of himself, andhis institutions. These conflicts raise ethicalquestions such as: how much responsibility dothe creators of powerful technologies have forthe social impacts of their creations? or, shouldscientists consider the public interest in de-ciding what research should be done?l06

reader is referred to Relations of Science, Governmentand Industry: The Case of Recombinant DNA, ” by CharlesWeiner, ch. 4 in Science, Technology, and the Issues of TheEighties: Outlook, A. H. and R. Thornton (eds.),Westview Press, Boulder CO, 1980.

Until recently, artificial intelligence researchwas supported by four Federal agencies, pri-marily the Defense Advanced Research Proj-ects Agency (DARPA) plus the National Sci-ence Foundation, the Air Force Office ofScientific Research, the Office of Naval Re-search, and a handful of companies. Now, sig-nificant levels of work are beginning to befunded in several corporations as the profit po-tential of AI grows. The Department of De-fense plans to expand AI research in univer-sities and defense contractor companies.

Industry Efforts

An increasing number of industrial firmshave begun AI R&D efforts in the past few


years. Market research sources report that in-dustry spent a total of $66 million to $75 mil-lion in 1983 on AI products.107 As many as 15or more of the largest U.S. corporations areon the threshold of expanding the develop-ment of expert systems.l08 The Microelectron-ics and Computer Technologies Corp. (MCC)is devoting a fair amount of its early R&D ef-forts to AI related questions. 109 Three U.S.companies, Xerox, Symbolics, and Lisp Ma-chine, Inc., currently offer computer systemsespecially designed for AI program develop-ment. 110 Four other computer companies, Sper-ry, Apollo Computer, Data General, Hewlett-Packard and Digital Equipment, are also de-veloping AI computers.111 Of most significanceis the number of startup companies develop-ing AI products, particularly expert sys-tems. 112 At least a dozen firms have beenfounded in the past few years and they are ob-taining people, techniques, and seed ideas fromthe top universities and from nonuniversitycenters such as SRI International.

““T. Manuel, and S. Evanczuk, “Commercial Products Beginto Emerge From Decades of Research, ” Electrom”cs, Nov. 3,1983, pp. 127-129. It should be noted that these numbers re-flect a very broad definition of AI with which many expertswould disagree.

‘“*J. Johnson, “Expert Systems: For You?, ” Datanation,February 1984, pp. 82, 84, 88.

‘“’See ch. 6, in this report.‘‘“These so-called “Lisp Machines’ (because they use the AI

programming language Lisp) represent an estimated marketvalue of $50 million.

“’These machines are useful in many types of complex pro-gram development; nearly one-half of the sales of these ma-chines, by at least one of the current vendors, are to users thatare not necessarily developing AI applications. J. M. Verity,“LISP Markets Grow, ” Datamdion, October 1983, pp. 92-94,98,100.

“*See Verity, op. cit. p. 94, and Kinnucan, P., “ComputersThat Think Like Experts, ” High Technology, January 1984,p. 42. One OTA advisor believes that venture capital will bea significant source of funding for AI R&D. At least $16 mill-ion has thus far been invested in startup companies develop-ing expert systems, J. W. Verity, “Endowing Computers WithExpertise, ” Venture, November 1983, p. 49.

Government Funding

The Department of Defense is the lead agen-cy funding AI research. The Defense AdvancedResearch Projects Agency (DARPA) estab-lished the three university “centers of ex-cellence” at MIT, Carnegie-Mellon, and Stan-ford, and has awarded contracts since the early1960s. The Office of Naval Research (ONR)and the Air Force Office of Scientific Research(AFOSR) have been funding AI work since the1970s. The Air Force in particular is interestedin AI techniques for image understanding.DARPA has proposed a significant expansionin its AI efforts (see “Strategic Computing”below). ONR and AFOSR funding is expectedto remain stable in the near term. 113 (See table19.)

The National Science Foundation has awardedgrants for university and some industrial AIprojects since the late 1950s. In recent yearsthe NSF AI research budget has remained sta-ble, between $5 million to $6 million annually.(See table 20.) Two NSF directorates, Comput-er Science and Information Science and Tech-nology, provide the bulk of these funds, anda number of the awards (13 of 80 in 1983) aremade jointly between these two divisions, orwith the NSF Office of Interdisciplinary Re-search, or with other programs within NSF.

“STRATEGIC COMPUTING”DARPA has embarked on a major effort to

push the frontiers of computer technology andartificial intelligence for application in futuremilitary systems. The “Strategic Computing”program plans to spend $600 million over thenext 5 years (see table 21), over and above pastlevels of funding, and spending will accelerateto unspecified levels as prototype developmentgets underway in the late 1980s. The projectcalls for the demonstration of significant mil-

I I ‘Personal commumcatmns from Dr. David Fox, Director of!tIathematics and Information Science, Air Force Office of Sci-entific Research, and Paul Schneck, Head, Information SciencesDivision, Office of Naval Research.


Table 19.—Major Government Sources of Artificial Intelligence R&D Fundinga

Estimates

Low High

National Science Foundation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.0 6.0National Institutes of Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5 4.2Office of Naval Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 1 .4b

Air Force Office of Scientific Research . . . . . . . . . . . . . . . . . . . 2.5 3.0Defense Advanced Research Projects Agencyc . . . . . . . . . . . . 15.0 20.0

Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27.2 34.6asources indicate that these levels have obtained for the past several years, and with the exception Of DARPA, are eXrx?C@dto remain stable or increase modestly in the near term.

bThiS does rro~ include $.goo,@o in robotics research, some of which iS in artificial irIt011i9encf3CTlliS represents the base level Of C)ARPA funding before the initiation of the “strategic computing” Pr09rWI (began in fiscal

year 1984).SOURCES: NSF numbers from Swnrnary of Awards publications of the Division of Computer Research, FY 1982-83, and the

Division of Information Science and Technology, FY 1981-82 and 1983, National Science Foundation. NIH numbersfrom Susan Stimler, Director, Biomedical Research Technology Program, Division of Research Resources, NationalInstitutes of Health. ONR numbers from Paul Schneck, Leader, Information Sciences Division, Office of NavalResearch. AFOSR numbers from David Fox, Director, Division of Mathematics and Information Science, Air ForceOffice of Scientific Research, DARPA numbers from Ronald Ohlander, Program Manager, Defense Advanced Re-search Projects Agency.

Table 20.—NSF Grant Awards in Artificial Intelligence

Fiscal year 1982 Fiscal year 1983–Number–of Institutions . . . . . . . . . . . . . . . . . . . . . . . . . 41 38Number of awards . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Total awards granted ($OOO)a . . . . . . . . . . . . . . . . . . . . $6,077 $5,150Top 12 institutions: (Dollars in thousands)

MIT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 946 240Stanford University. . . . . . . . . . . . . . . . . . . . . . 485 587Rutgers University . . . . . . . . . . . . . . . . . . . . . . . . . . . 245 442University of Maryland . . . . . . . . . . . . . . . . . . . . . . . . 351 317University of Illinois . . . . . . . . . . . . . . . . . . . . . 478 141University of Pennsylvania . . . . . . . . . . . . . . . . . . . . 302 288SRI International . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302 202New York University. . . . . . . . . . . . . . . . . . . . . . . . . . 360 130Yale University . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 301Carnegie-Mellon University . . . . . . . . . . . . . . . . . . . . 185 227University of Massachusetts . . . . . . . . . . . . . . . . . . 87 229University of Michigan. . . . . . . . . . . . . . . . . . . . . . . . 40 337

Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3,951 3,441aThese numbers reflect the awards In each fiscal year and not I he amount Of funding dur!ng those Years sixteen Of the 67awards made In 1982 were mult!-year grants (most were 2 year grants), 5 of the 80 awards made I n 1983 were multl.year qrantsTherefore, total funding probably Increased somewhat between 1982 and 1983

SOURCE Office of Technology Assessment, and the National Science Foundation, Divisions of Information Science ancl Tech.nology and of Computer Research

Table 21.—Strategic Computing Cost Summary

Fiscal year1984 1985 1986 1987a 1988a

(Dollars in millions)Total military applications . . . . . . . . . . . . . . . . . $ 6 $15 $27 TBD TBDTotal technology base . . . . . . . . . . . . . . . . . . . . 26 50 83 TBD TBDTotal infrastructure . . . . . . . . . . . . . . . . . . . . . . . 16 27 36 TBD TBDTotal program support . . . . . . . . . . . . . . . . . . . . 2 3 4 TBD TBD

Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $50 $95 $150 TBD TBD%ut-year funding levels to be determined by program progress.SOURCE: Defense Advanced Research Projects Agency.


itary capabilities within 10 years (see fig. 13).114

The major areas of focus are:

● autonomous (unmanned) vehicles,● pilot’s associate systems to aid combat

pilots in coping with the complexity ofcurrent and planned aircraft, and

● battle management systems to help com-manders make decisions under conditionsof uncertainty.

To lay the groundwork for the developmentof such systems, DARPA plans to fund thebuilding of Galium Arsenide integrated circuitpilot production lines for the manufacture oflow power and radiation resistant microelec-tronics. In addition, DARPA is focusing——.

‘14Strate@”c Computing, “New-Generation Computing Tech-nology: A Strategic Plan for its Development and Applicationto Critical Problems in Defense, ” Defense Advanced ResearchProjects Agency, Oct. 28, 1983, p. vii.

efforts on research and development of ad-vanced computer architectures to meet thecomputational speed, and physical size andweight requirements of mobile systems (seethe case study of Advanced Computer Archi-tecture in this chapter), as well as artificial in-telligence R&D in vision, speech, natural lan-guage, and expert systems.116

Some of the goals of the “Strategic Comput-ing” program, particularly the vision systemrequirements for autonomous vehicles, havebeen described as “extremely ambitious.’’’”Some describe this program as unprecedentedin the history of U.S. Government funding ofscience and technology.117 Unlike the Manhat-

‘]51bid., app. IV, p. 2.‘]’ Ronald Ohlander, at the OTA workshop, Oct. 31, 1983; see

also Strate&”c Computing, pp. 22-23.“’Mitch Marcus at the OTA workshop, Oct. 31, 1983.

Figure 13.—Strategic Computing Program Structure and Goals

MajorDevelop a broad base of

machine intelligence technologygoals to increase our national

security and economic strength

Military Autonomous Pilot’s Battleapplications systems associate management

&●

Inteligentfunctional Natural language Vision Expert systems

capabilitiesNavigation Speech Planning and reasoning

Hardware/softwareHigh-speed signal-processing General purpose systems

systemarchitecture Symbolic processors Multi-processor programming

and operating systems

Technologybase

Microelectronics Silicon and GaAs technologyVLSI systems

1

Networks Research machines Rapid machine

infrastructure I implementation systems and foundriesprototyping

Interoperability protocols Design tools

SOURCE’ DARPA.


tan Project or the Manned Moon Landing Mis-sion, which were principally engineering prob-lems, the success of the DARPA program re-quires basic scientific breakthroughs, neitherthe timing nor nature of which can be pre-dicted.

The kind of work that is being done now[in AI research] that would support, for in-stance, an autonomous vehicle system, isprimitive in relation to the problems that areto be addressed [by the DARPA project].Operation in a complex combat environmentthat may have multiple targets with camou-flage, different kinds of obstacles with vary-ing degrees of threat and impedance associatedwith them, and the integration of variouskinds of sensors, for example touch and vi-sion, and the appropriate knowledge repre-sentation to deal with them, is an enormousproblem that will be solved only by signif-icant strides in basic AI research, and notjust development in narrow vehicle applica-tions.118

It is expected that the Strategic ComputingProgram will approximately quadruple the an-nual Federal funding of R&Din AI and relatedhardware. ll9 This major increase in R&D fund-ing will have significant and far reaching ef-fects on the artificial intelligence communityand research environment. Undoubtedly, theincreased flow of money will have some posi-tive effects on AI research. In particular theavailability of funds for modern equipmentshould make university laboratory facilitiescomparable with increasingly well equipped in-dustry labs.120 The increased availability ofgraduate student financial aid in the form ofresearch assistantships should draw in andhelp hold more qualified post-graduate train-ees, thus expanding the potential faculty base,and relieving some of the pressure on currentfaculty to obtain funding for graduate re-search. However, as industry scales up to meet

“’Marvin Denicoff, at the OTA workshop, Oct. 31, 1983.1‘gFrom about $30 million to approximately $120 million (aver-

aging the $600 million program over its 5 year projection). OneDARPA source estimated that one-fourth of the program fundswould be spent in universities. (Duane Adams at the OTA Work-shop on Advanced Computer Architecture, held July 14, 1983.)

““This may be an inducement for researchers to accept re-search appointments.

military requirements, additional pressures onthe limited AI manpower resources are ex-pected. 121 Although the expansion of trainingprograms should eventually increase the sup-ply of manpower, the rate of that expansionis limited by the small existing faculty base.122

The current imbalance in the supply and de-mand of AI manpower will continue, and likelyincrease, and may intensify competition amongcommercial, military and academic AI R&Dagendas. Thus some applied research, such aswork toward “intelligent library systems” andcomputer assisted education, and fundamen-tal research that is unlikely to have high im-mediate commercial or military value, may beneglected. 123

Content and Conduct of ArtificialIntelligence R&D

Artificial intelligence as a field is a set ofsomewhat loosely related R&D activities thatrange from the study and implementation ofmachine sensing (e.g., vision and speech) andpattern recognition algorithms and systems,to theoretical and practical work on automaticproblem solving, inferencing and reasoningstrategies and programs. Of central concernis the concept of knowledge, a set of informa-tion (facts, procedures, patterns) that is in-tegrated and processable as a whole, and thusconstitutes a useful representation of a partor domain of the world. AI deals with howknowledge is built up and used in computer-based systems: how it is collected, stored, ac-cessed, manipulated, and transferred. AI R&Dseeks methods of formalizing and represent-ing knowledge in consistent and unambiguous,yet flexible ways so that these tasks can beperformed by machines.

“’Ronald Ohlander, at the OTA workshop, Oct. 31, 1983.‘*zIn the three AI subfields from which most commercial and

military applications are expected, expert systems, natural lan-guage understanding, and vision, there are approximately 60faculty at the top schools turning out some 30 Ph.D. graduatesper year, half of whom take industry jobs. D. Waltz, “Artifi-cial Intelligence: An Assessment of the State-of-the-Art andRecommendations for Future Directions,” prepared for the NSFInformation Technology workshop, Jan. 5-7, 1983; and a per-sonal communication from Azriel Rosenfeld, Apr. 25, 1984.

‘z’ David Waltz at the OTA workshop, Oct. 31, 1983. A pri-vate funding organization, the Systems Development Founda-tion, is sponsoring research directed at “intelligent libraries. ”

Ch. 3—Selected Case Studies in Information Technology Research and Development Ž 97

Although investigators in a number of fieldsperform work that contributes to and overlapswith the subject matter of AI, a number ofareas are considered the core of AI research.

Symbolic Computation

Since AI is concerned with the processingof knowledge, intelligent machines are re-quired to manipulate symbols that representobjects, concepts, and qualities, as well assymbols that represent numbers or quantities,which are the focus of traditional computation.These qualitative symbols may be representedwithin a computer as logical relationships oras lists of symbols with pointers linking vari-ous objects, concepts, and qualities. The ma-nipulation of symbols provides a mechanicalmeans of achieving inference, in particulardeductive inference, in which a problem solu-tion or a conclusion is tested against facts ina knowledge base to determine its truth orvalidity.

A branch of AI research and development,which forms a link with R&D in advancedcomputer architectures, is the design of com-puting machines that optimize efficient man-ipulation of symbols. Some in computer ar-chitecture research maintain that this R&Deffort has not received sufficient attention. 124

‘*’Sidney Fernbach at the OTA workshop on Advanced Com-puter Architecture, July 17, 1983.

m

Photo credit, AT&T Be// Laboratories

Specialized microprocessors promise increasing levelsof “intelligence” in computer systems

Pattern Recognition

Of particular importance to the branches ofAI concerned with sense perception, such asmachine vision and speech understanding, isthe idea of matching input from the environ-ment with symbolic representations of pat-terns (e.g., visual objects or speech utterances)stored in the system. Raw sensory input, inthe form of electrical signals from a televisioncamera or a microphone, must proceed throughseveral levels of processing to produce pat-terns that are comparable with stored symbol-ic representations and recognizable as havingunique meaning.

In general, these processing levels include:●

●

●

Formation: The sound or light signals aredigitized and stored as a set of simplephysical parameters such as frequencyand amplitude.Analysis: Patterns of variations in the pa-rameters, for example areas of light anddark in pictures or variations in pitch andintensity in an utterance, are detected andmeasured to produce a detailed physicaldescription of the input.Interpretation: The patterns, now repre-sented by sets of measurements, can beeither directly compared with templates,stored descriptions of entire patterns, orfurther analyzed to extract features,which are parts of patterns that are ofparticular value in defining and identify-ing the patterns.

Knowledge Representation

This area of research deals with ways of ex-pressing knowledge in computable form andmaking and exploiting connections among thefacts or propositions in a knowledge base. Abasic concept in knowledge representation isthe propositional formalism, which is simplythe structured way that facts are expressedso that ambiguities are avoided and process-ing operations are as efficient as possible.


As well as a formal structure or syntax,knowledge representation requires a methodof capturing meaning, or semantics, in aknowledge base. Meaning in a qualitativesense is expressed by the relationships thatare stated to exist among concepts. One meth-od of representing relationships is the Seman-tic Network (see fig. 14).

A knowledge representation scheme that issimilar to semantic networks has been devel-oped in which the attributes of concepts areexpressed and related. Frames provide slotsto fill to structure knowledge about things,thus they represent expectations about whatattributes members of a given frame will pos-sess. These “expectations” can be exploitedto imply procedures. For example, a computercan be programmed to inquire about the at-tributes of a newly introduced object.

Similarly, knowledge representations calledscripts can represent expectations about a

knowledge domain. But instead of represent-ing concepts, scripts describe situations andactions that are expected in those situations.A classic example is a restaurant script inwhich one expects people to order from awaiter (or waitress), to eat, and to pay the bill.Thus, invoking the restaraunt script would in-voke the entire set of expectations surround-ing restaurants that have been programmed.

Another knowledge representation scheme,one that was developed in the mid-1960s, isthe production system. This essentially con-sists of a set of if-then rules that specify a pat-tern, for example, “if the temperature is lessthan 65 degrees” and an action to be taken,“then turn the furnace on. ” The importanceof this scheme is its ability to express proce-dural knowledge and to initiate operationsdepending on prevailing conditions. Its majorweakness is that a fixed rule must be statedfor every condition that is to be encountered.

Figure 14.– A Semantic Network

I

1 J ImI r

Parakeets II

I “

Chickens ,

Cows

SOURCE: Randy Davis and Chuck Rich.


All of the knowledge representation schemesthus far adopted by AI researchers use consid-erable computer memory if they contain a siz-able or complex domain of knowledge. Tradi-tionally, researchers and builders of programshave had to “shoe-horn” knowledge into com-puter systems with limited and expensivememory capacity. The precipitous decline inthe cost of computer memory is alleviatingsome traditional AI problems, making possi-ble for the first time the production of cost-effective systems with enough knowledge totackle “real-world” problems. Advances incomputer architecture design, especially theability to make customized VLSI processorsdedicated to symbolic computation and capa-ble of parallel processing, should have a greatimpact on the kinds of problems that may beaddressed by knowledge-based systems.

Other difficulties inherent in the applicationof existing knowledge representation schemeswill not necessarily be alleviated by lowermemory costs and advances in hardware ca-pability because they involve the limitationsof the schemes themselves. Existing knowl-edge representations deal only with surfaceknowledge, which is either explicitly stated inrelationships or deductively inferred by thechaining of propositions. Deep knowledge,knowledge gained from inductive inference(reasoning from facts to general principles),and commonsense knowledge that is routinelylearned by children through experience, arenot expressable using current knowledge rep-resentation systems. Other problems involvethe inflexibility of procedural knowledge andthe difficulties of programming machines toreason about unforseen occurances.

Given the large collection of facts, relationsand patterns in a useful size knowledge base,even the fastest computers can bog down.Therefore, an AI program must concentratethe search of its knowledge base on those por-tions that are most likely to hold the facts, pat-terns, or solutions that are needed in a givensituation. The problems associated with thesearch of knowledge bases are ubiquitous inAI. They result from what is termed the “com-

binatorics explosion”: as the size of a knowl-edge base increases, the number of possiblepaths to a sought-after piece of informationincreases exponentially. AI interest was orig-inally in developing general knowledge AI sys-tems, but the combinatorics explosion sets alimit on how broad or useful such systems canbe. More recent work has focused on specificproblem domains where a detailed represen-tation of the world can be built and exploited,and where the search is controlled by knowl-edge about the problem domain, or heuristics.

Machine Learning

The study of heuristics (“rules of thumb”or knowledge acquired from experience), howthey may be used to process knowledge, andhow they may be generated automatically bycomputer systems are major pursuits of AIresearch. Ideally, a computer could modify itsrepertoire of procedures as new facts are addedto its knowledge base. Likewise, new proce-dures should suggest new relations amongfacts, but current knowledge representationand programming techniques are incapable ofsupporting changing knowledge and rulebases.

Commonsense Knowledge

Human knowledge is based on a wide rangeof experience acquired through sense organs.This experience is fed into a cognitive systemthat can integrate that knowledge in order tocope with a complex, dynamically changingworld. Much of what people know involvesrelationships that are obvious but that dependon explanations beyond the comprehension ofmost. Representing commonsense knowledgein computer systems is a major challenge toresearchers, and is essential for certain kindsof intelligent machine behavior, such as rea-soning about the chronology of events in thecourse of a disease or reasoning about themovement and characteristics of robots.

Pragmatic Knowledge

Current artificial intelligence systems em-ploy syntax (structure) and semantics (mean-


ing) to represent knowledge about the domainsin which they operate. Systems of higher in-telligence, ones capable of choosing among arange of responses in complex situations, ofadapting to changing (and unanticipated) con-ditions, and of interacting with humans andperhaps other intelligent systems on a highlevel of understanding, must employ ways ofrepresenting and reasoning about the needsof users and about knowledge itself. Truly in-telligent machines would respond to people ofdifferent backgrounds and needs in ways thatare appropriately tailored to individuals. Theywould also possess an understanding of theuses and limits of knowledge-a mechanismfor judging the appropriateness of informationto a given situation. For example, a robot tankwould need to be able to distinguish betweentrees it could run over and those that it wouldhave to avoid.

Current syntax- and semantics-based arti-ficial intelligence systems employ knowledgerepresentations that are inadequate to supportpragmatic knowledge. Frame- and script-based representations can, in a rudimentaryway, deal with expectations about situations,but those situations must be tightly circum-scribed and clearly defined at the time of pro-gram design. Research has yet to identifyknowledge representation schemes that cansupport the processing of pragmatic knowl-edge in complex and dynamically changingdomains.

Commercially Available Artificial Intelligence

Notwithstanding the fundamental limita-tions of present artificial intelligence concepts,programs are being developed that are prov-ing useful in a widening array of applications.Two market forecasts125 describe the commer-cial potential of AI systems as exploding overthe next decade. From the small 1983 salesbase of $66 million to $75 million, the marketfor AI products is expected to rise to $2.5 bil-lion to $8.5 billion by the early 1990s.126 Thechanging mix of products and sectors in which.—--—— —. ---

‘2sInternational Resource Development, Inc., and DM Data,Inc.

‘z’ Manuel and Evanczuk, op. cit., pp. 127 and 129.

they will be used is depicted in figure 15, whichshows AI emerging from the laboratory intothe home, the factory, and into schools. Threetypes of AI-based systems in particular are ex-pected to experience rapid growth-vision sys-tems, natural language understanding sys-tems, and expert systems.

COMPUTER VISION SYSTEMSVision systems are being designed as input

subsystems to enhance the utility and flexi-bility of computer-based automation. Visionsystems are increasingly used in industry forquality control inspection, for product iden-tification, and for robot guidance and con-trol.127 Machine vision will be a crucial func-tion in many future automated systems,including aspects of the DARPA “StrategicComputing” initiative. For example, an ad-vanced vision system will be an integral com-ponent of autonomous vehicles.

Some sources estimate a current annualmarket for machine vision systems of $35 mil-lion. This market is projected to double eachyear for the next 5 years, and should reachabout $1 billion by the end of the decade.128

More than 250 companies, most of them in theUnited States, are designing and selling ma-chine vision systems.129

Vision systems consist of cameras and com-puter processors to analyze and interpret theinformation collected by the cameras. Thecameras provide the processor with frames,typically 60 per second, collected by scanningthe camera’s visual field. A frame forms an im-age consisting of a matrix of dots, or pixels(picture elements), numbering in current sys-tems 64 X 64 or 256 X 256 pixels. Higherresolution cameras are being developed, butas the resolution (number of pixels per frame)increases, higher speed processors are neededto handle the increased information within rea-

IZIThe reader is referred to the OTA report, cornpu~er~z~Manufacturing Automation: Employment, Education, and theWorkplace, April 1984, pp. 89-92, for a discussion of currentindustrial applications of machine vision systems.

‘*s’ ’Machines That Can See: Here Comes a New Generation, ”Business Week, Jan. 9, 1984, pp. 118-120.

12gIbid., p. 118.


Figure 15.— Markets for Artificial Intelligence

1993 total $8,536 million •

SOURCE International Resource Development Inc.

sonable time limits.130 Each pixel representsan area of light or dark in the image. Most sys-—— —

Stanford University researcher has developed a robot cartthat can navigate in a simple environment at a speed of 3 to5 meters per hour while analyzing one image of its surroundingsfor each meter of progress using a computer capable of proc-essing a million instructions per second. For such a system to

terns allow a pixel to represent only black orwhite, though systems are being developedand are in use in which pixels may be one ofas many as 64 shades of gray, and color visionsystems will soon be available. Gray scale andcolor systems will be capable of higher acuity,but such systems require much higher proc-essor speed and more sophisticated algorithmsto interpret the increased information flow.

The processor analyses the image to extractpatterns that may represent edges or texturesthat can be used to characterize and recognizeobjects in the visual field. Simpler systemshave stored “templates,” or representationsof entire objects, in memory to which the ex-tracted patterns are compared. More sophis-ticated systems store “features,” or charac-teristic parts of objects. In these systems, setsof features are extracted from the image andare combined to represent constraints on thelist of possible objects that may be present;recognition involves the identification of anobject (or objects) that could satisfy those con-straints.

A major problem in machine recognition isinferring three-dimensional information fromtwo-dimensional information in an image ofthe scene.131 One method that has been adoptedto solve this problem is to project “lightstripes” on objects in the scene; depth cuesand contours can be inferred from distortionsin the stripes.

Current vision systems are primitive in com-parison to human visual recognition capabil-ity. The human retina can perform at least 10billion operations per second, and the brain isundoubtedly capable of much higher proc-essing speeds.132 As important, human experi-ence with the visual world provides a store ofknowledge with which the eye and brain maymake inferences and interpretations to resolve

naviagate at walking speed would require a computer capableof from 1010 instructions per second. T. and R.

“Computer Vision: The of Imperfect Inputs, ” Spectrum, November 1983, pp. 88-91. Computers of this

speed are being developed, but may require facilities as largeas a room just to provide cooling for the hardware.

‘3’ personal communication, May 25, 1984.‘32 and op. cit., p. 88.


ambiguities caused by shadows and variationsin lighting, the orientation of objects, and thepresence and often overlap of multiple objectsin a scene. Also, subtle variations and move-ments can be detected and used by the humaneye and brain to distinguish superficially simi-lar objects (e.g., the faces of identical twins).

In order to make vision systems based oncurrent processor architectures and algo-rithms work, the number of environmentalvariables that may cause ambiguities must bereduced. For example, objects on an assemblyline may be arranged so that only one face ispresented to the camera or only one object ata time is in the visual field. Lighting may becontrolled so that perceived edges can be in-terpreted as facets of an object and will notbe the result of shadows or variations in thereflectance of surfaces of an object; or struc-tured lighting may be used to infer distanceand contour.

If vision systems are to operate in less con-trollable circumstances, such as out-of-doorsor in the home, much higher speed processorsand more sophisticated algorithms for analyz-ing and interpreting visual scenes will beneeded. Custom designed, massively parallelVLSI-based processors dedicated to visualanalysis are being looked to to provide morepowerful hardware. Research in vision algor-ithms is proceeding to provide more powerfulconcepts and more sophisticated software. Theparallel advance of hardware and softwaresolutions, and fundamental advances in AI re-search, will be required to produce systems ca-pable of the interpretation of complex andunstructured visual scenes.

NATURAL LANGUAGE UNDERSTANDINGSYSTEMS

These systems are intended to allow com-munication with computers in English or otherhuman languages, freeing users from the needto learn a computer language. In the words ofone researcher:133

The ultimate goal of creating machinesthat can interact in a facile manner with

——. . .-—“’Waltz, Al Magazine, op. cit., p. 56.

humans remains far off, awaiting break-throughs in basic research, improved infor-mation processing algorithms, and perhapsalternative computer architectures. How-ever, the significant progress experienced inthe last decade demonstrates the feasibilityof dealing with natural language in restrictedcontexts, employing today’s computers.

Automatic natural language understandingwork began with the goal of producing sys-tems that could translate one natural languageinto another. Such systems are in use now, butare proving to be limited to rough translation,requiring a human translator to produce finalfluent text.

The goal of another class of natural lan-guage systems is to understand documents:to produce an abstract of a piece, perhaps alertpeople who might be interested, and answerquestions on its content. Such systems mightbe paired with document generators that couldproduce instruction manuals.134 Documentunderstanding and generation systems arestill largely experimental, but they have thepotential to augment human knowledge throughthe production of intelligent library systems,aiding people in searching and evaluating largebodies of information.

More immediate-term applications are inusing natural language systems as “front-ends” or interfaces to computer systems. Nat-ural language systems are in operation thatserve as interfaces to data bases, for example,to help store and find personnel or inventoryrecords. One can expect that model systemswill soon be available that demonstrate thefeasibility of controlling complex systems suchas power generation facilities and weaponssystems with natural language interfaces. Inconjunction with expert systems, natural lan-guage systems are being developed for com-puterized medical advisor, trouble-shootingand repair, mineral exploration, and invest-ment analysis applications. Future applica-tions, requiring advanced work to produce,might include use in the creation of graphicaldisplays and in computer-aided education.135

—“’Ibid., p.-57.“’Ibid.


There is also current work in producingspeech understanding systems, which wouldreplace keyboards with voice input devices.Primitive systems are available, but truespeaker independent systems with large vo-cabularies that can handle continuous speech,as opposed to isolated words separated bypauses, are as yet unattainable with presenttechnology. Advanced integrated circuit sig-nal processors should produce some progressover the next decade.136

In part, the capabilities of natural languagesystems have been limited by the cost of com-puter memory. Natural language systemsmust have large vocabularies to be “natural’to users. This problem is diminishing becauseof the fall in computer memory prices.

Two other problems are replacing memorycost as an upper limit to system size and func-tionality. First is the “combinatorics explo-sion” problem: as the size of the systemvocabulary increases, the time required forpresently available computer architectures toprocess a sentence increases exponentially.Novel machine architectures based on paral-lel processing promise some relief for this prob-lem. Second, as noted in the previous discus-sions of commonsense reasoning and pragmaticrepresentation, researchers are finding dif-ficulties in designing systems with broad anddeep knowledge of the world, and with prag-matic understanding of situations. In lan-guage, this kind of knowledge is crucial to thesolution of ambiguities.

EXPERT SYSTEMSThese AI-based programs are designed to

serve as consultants in decisionmaking andproblem solving tasks that require the applica-tion of experience and judgment. ’37 Expert—— —.. —————

“’For discussions of the state-of-the-art in voice input com-puter systems, the reader is referred to: A. Pollack, “ComputersMastering Speech Recognition,” New York ~“mes, Sept. 6, 1983,pp. Cl and C7; R. D. Preuss, and D. J. Jurenko, “Digital VoiceProcessing,” Astronautics and Aeronautics, January 1983, pp.44-46; and R. J. Godin, “Voice Input Output, ” Ehxtrom”cs, Apr.21, 1983, pp. 126-143.

“’Expert systems illustrate an interesting point about humanintelligence, “Paradoxically, it has proven much easier toemulate the problem-solving methods of some kinds of special-

systems consist of a set of “if-then” ruleswhich express the knowledge and experienceof an expert, and the actions one would takewhen faced with a set of conditions in the do-main of his expertise. They also generally havea separate knowledge base which states factsabout the domain, to which the program canrefer to make inferences and deductions aboutgiven situations and conditions.

Expert system programs are developed fromextensive interviews with recognized expertsin the field of application. Such interviewsoften reveal many unwritten (and even uncon-scious) rules and criteria for judgment that theexpert applies in solving problems.138 The in-terviewer, called a knowledge engineer, thencodes the rules and facts in a way that is effi-ciently processable by computer. Extensivetesting and validation must be conducted be-fore a system is considered complete enoughto use in actual practice.

This area of AI application is receiving con-siderable industry and military interest be-cause expert systems may be capable of reliev-ing some of the demand for high-priced expertsin the fields in which they are applied. Ironi-cally, because they require man-years of effortfrom scarce knowledge engineers to develop,expert systems are quite expensive. Thereforethey are currently being attempted commer-cially only in applications that promise a par-ticularly high payoff.139

Research in this field has demonstrated thatexpert systems can rival human expert judg-ment if the domain of application is properly

ists than to write programs that approach a child’s ability toperceive, to understand language, or to make ‘commonsense’deductions, ” R. O. Duda, and E. H. Shortliffe, “Expert Sys-tems Research, ” Science, Apr. 15, 1983, pp. 261-268.

138This fleshing out of unwritten rules is a possible side ben-efit of the development of an expert system; it can lead to pro-gress in the field of application itself. Conversely, some fearthat this codification of knowledge in automated systems could“ossify” and constrain progress in certain fields by foreclos-ing the possibility of the unexpected discovery of new solutionsthrough serendipity.

‘3gSome of the enthusiasm recently derponstrated by compa-nies for expert systems may be unfounded; OTA advisors werewary of overstated claims for the usefulness and applicabilityof current generation expert systems.


chosen.140 This means that the applicationmust have certain characteristics for an expertsystem to be successfully applied. These in-clude: 14

1

● There must be recognized experts in thefield;

● The task normally takes the expert a fewminutes to a few hours to accomplish;

● The task is primarily cognitive (as op-posed to manual);

● The skill is teachable to neophytes;● The task requires no commonsense.

More generally, the application domainmust be restricted enough to be expressablein a finite set of rules, numbering less than2,000 at present, so that available computerscan solve problems in a reasonable amount oftime. Even so, most current expert systemstake longer to perform a task than a compe-tent human expert.142

Table 22 shows a representative sample ofcurrent expert systems and their uses. Severalcompanies are forming or expanding effortsin developing expert systems, and venturecapital interest in startup companies is re-portedly brisk.143

———.‘401n fact, one OTA advisor said that the success of current

applications depends more on the choice of domain than it doeson how well the program is written. Patrick Winston at the OTAworkshop, Oct. 31, 1983.

‘“R. Davis, and C. Rich, Expert Systems: Fundamentals,Tutorial at the AAAI Conference, Aug. 22, 1983, p. 31.

“’Waltz, AI Magazine, op. cit., p. 61.“’see l%ctrom”cs, Nov. 3, 1983, pp.127-131, and Venture, No-

vember 1983, pp. 48-53.

Current systems, because of the limitationsof the concepts on which they are based, allsuffer from several serious weaknesses:144

●

●

●

●

●

They are highly customized to specific ap-plications and are useless in other, evenclosely related fields;Since the knowledge on which they arebased is collected over a period of time,often from a number of experts, there canbe inconsistencies in the programs thatare difficult to detect and repair;The systems are necessarily based on nar-row sets of rules and facts, therefore theirjudgments and recommendations can bemyopic and naive;Since all of the current systems containonly surface knowledge from which tomake inferences, if knowledge that is crit-ical for a given judgment is missing, theirperformance is poor;Few current systems have natural lan-guage interfaces, therefore they can bedifficult for the uninitiated to use.

Although many systems can provide expla-nation for the chain of reasoning that led toa given conclusion, these explanations are of-ten unsatisfactory because they are not tai-lored to the needs or understanding of individ-ual users. Neither do the explanations usuallyrefer to underlying principles, such as physi-ology or geology, so human experts find themunconvincing. 145

——“’Waltz, op. cit.14’Duda and Shortliffe, op. cit., p. 266. See also G. O. Barnett,

“The Computer and Clinical Judgement, ” editorial in The NewEngland Journal of Medicine, Aug. 19, 1982, pp. 493-494.

Table 22.—Representative Expert Systems

Type of RoutineExpert system Domain evaluation use

DENDRAL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mass spectroscopy interpretation Case studies YesMYCIN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Antimicrobial therapy Randomized trials NoINTERNIST-1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Internal medicine diagnosis Case studies NoCASNET . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Glaucoma assessment and therapy Case studies NoPROSPECTOR . . . . . . . . . . . . . . . . . . . . . . . . . . . . Geological exploration Case studies NoRI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Computer layout and configuration Case studies YesDigitalis Advisor . . . . . . . . . . . . . . . . . . . . . . . . . . Digitalis dosing advice Randomized trials NoPUFF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pulmonary function test interpretation Randomized trials YesMicroprocessor EXPERT . . . . . . . . . . . . . . . . . . . Protein electrophoresis interpretation Case studies YesHASP and SIAP . . . . . . . . . . . . . . . . . . . . . . . . . . . Ocean surveillance (signal processing) Case studies NoSOURCE’ R. 0. Duda and E. H. Shortliffe, “Expert Systems Research,” Science, Apr. 15, 19s3.

Ch. 3—Selected Case Studies in information Technology Research and Development ● 105

Research is underway to help alleviate someof these shortcomings, including the study ofimproved knowledge representation schemesand inference methods. Novel computer ar-chitectures should push back some of thelimitations now imposed on system size bycomputer speed. The most immediate andprobably far-reaching problem is the difficultyand expense of building and testing expertsystems, including the acquiring and vali-dating of rules and facts. Applied researchwork is progressing in the development of pro-gramming aids (tools, languages, environ-ments) to lower the time and cost of buildingexpert systems. Some current basic researchis concerned with the automatic acquisitionand structuring of knowledge. Incremental ad-vances can be expected over the next decade,but fundamental breakthroughs will be re-quired to expand the usefulness of expert sys-tems significantly beyond current limits.

Foreign National Efforts in ArtificialIntelligence R&D

Artificial intelligence is a key pursuit ofrecently initiated national, government fi-nanced R&D programs in Japan and GreatBritain. Britain’s effort is in fact a reaction tothe Japanese Fifth Generation Project 146 as itwas detailed in a conference held in October1981. ’47 The achievements of the objectives ofthese targeted national programs will requireunprecedented fundamental advances in basicresearch. Their ultimate success is much lesscertain than were previous national technicalefforts, such as the Manhattan Projector theManned Moon Landing. However, these na-tional AI research commitments are likely toproduce advances in AI science, technology,and commercial application.148

— . — .‘“A British delegation reported that the extent and cohesive-

ness of the Japanese plan, and the reaction to it that could beexpected from American industry, were a “major competitivethreat. ” A Programme for Advanced Information Technology,The Report of The Alvey Cou”ttee, Department of Industry,Her Majesty’s Stationery Office, London, October 1982, see p.5.

~d~The In~mation~ Conferene on Fifth Generation COmpUt-er Systems, Tokyo, Japan, Oct. 19-22, 1981.

“’OTA workshop, Oct. 31, 1983.

The Fifth Generation

The Japanese plan to build, from a base ofresearch that is almost exclusively borrowedfrom the United States and Western Europe,significant artificial intelligence systems withthe stated purpose of enhancing productivityin areas of the Japanese economy that thusfar have proven resistant to automation.149 Thetotal funding for the life of the project is notpublicly stated, but some sources estimatethat up to $1 billion to $1.5 billion may bespent over the next 8 to 10 years.150

The Fifth Generation Program is centrallymanaged by the Institute for New GenerationComputer Technology (ICOT), formed by theMinistry of International Trade and Industry(MITI) in 1982. ’61 The program will be funded,for the most part, by eight industrial compa-nies. The government has provided seed mon-ey and staffing; 42 of the total staff of 52 havecome from MITI’s Electrotechnical Labora-tory. 152 Spending thus far has been $2 millionin 1982, $12 million in 1983, and $27 millionin 1984.153

Overall, the functions that the Japanese seetheir Fifth Generation system performing in-clude: 154

1. problem-solving and inference,2. knowledge-base management, and3. intelligent interface (with users),These goals correspond roughly to previous-

ly discussed research in:

1. expert systems,2. knowledge representation, and

———lt~hese include a=iculture and fisheries, the office, and the

service industries; applications are expected, as well, in areasthat are already highly automated such as electronics andautomoble design and manufacturing. Fifth Generation Com-puter Systems, T. Moto-oka (cd.), op. cit., see p. 3.

IWP. Marsh, “The Race for the Thinking Machine,” iVew Su”en-tist, July 8, 1982, p. 85-87. Note, this is the highest estimateencountered.

15’’ ’Fifth Generation Computer System Under Development, ”Science & Technology in Japan, January/March 1983, p. 24-26.

152’’ ICOT: Japan Mobilizes for the New Generation, ” IEEESpectrum, November 1983, p. 48.

‘s’Personal communication from John Riganati, National Bu-reau of Standards, June 1984.

“’Science & Technology in Japan, op. cit., p. 24.


3. natural language understanding. ’55

The Fifth Generation R&D plan calls forthree phases of 3 to 4 years each:156

●

●

●

Phase I: Development of “basic technol-ogies” consisting of computer architec-tures and related software.Phase II: Development of prototypes ofthe inferencing and knowledgebase func-tional components, to correspond to thearchitecture and software developed inphase I.Phase III: A “total system” prototype isto be developed in the final phase: -

The reactions of experts in this country tothe Fifth Generation Program goals and meth-ods have been mixed, ranging from enthusi-astic to skeptical. However, there is agreementthat the goals the Japanese have establishedare quite ambitious for the time frame that isset for achieving them, and the program isunlikely to succeed in all of its objectives.

The Alvey Programme

Named for John Alvey, chairman of thecommittee that developed the plan for a Brit-ish response to the Japanese Fifth GenerationProgram, this academia-government-industrycooperative effort will concentrate on foursegments of information technology, one ofwhich is artificial intelligence, termed by theBritish, “Intelligent Knowledge Based” Sys-tems.157

The British see a manpower shortfall in AIR&D similar to that in the United States, sotheir program will initially have a strong ed-ucational component, focusing on the expan-sion of both academic and industrial trainingfacilities in the early part of the proposed 10year effort. As the program progresses, em-phasis will shift first to a broad effort in many

— .‘5SL. R. Harris, “Fifth Generation Foundations, ” DatanMti”orI,

July 1983, pp. 148-150, 154, 156.‘5GIbid., p. 25.ISTThe other ~eas me Software Engineering, Man/machine

Interfaces, and VLSI. A Programme for Advanced Informa-tion Technology, op. cit., p. 21.

aspects of basic AI research, and will later fo-cus on specific demonstration systems to bedeveloped in collaboration between the strength-ened university programs and industries in-terested in particular applications.158 Expectedmilestones include:159

●

●

●

●

●

The research community (numbered in1982 at some 150 people) is to grow by50 percent in 2 to 3 years and double inabout 5 years as a result of increases ingraduate student funding, and numbersof faculty and research positions;Computer equipment and software sup-port are to be expanded into an adequatebase for research in 2 to 3 years;An increased understanding will begained of knowledge representation andintelligent computer interface concepts,and knowledge based tasks and domainswithin 2 to 3 years;Within 5 years, there will be substantialprogress in understanding the applicationof knowledge based concepts using logicprogramming languages and parallelprocessors, expert systems and naturallanguage understanding systems;Over the 10 year time course of the pro-gram, progressively more sophisticateddemonstration systems will be developed,some early target applications beingteaching assistant programs, softwareproduction aids, and improved robot sys-tems; later applications might includemedical advisor systems, tactical decisionaids, full 3-D vision systems, and officemanagement and document productionsystems.

The government has pledged a total of $310million for the entire Alvey program, with pri-vate industry expected to contribute an addi-tional $230 million.l6O The plan calls for ex-penditures of about $39 million on IntelligentKnowledge Based Systems in the first 5 years,some $30 million of that going to universi-

‘6’ Ibid., p. 35.“gIbid., pp. 38-40.“OM. Peltu, “U.K. Eyes 5th Gen., ” Datamation, July 1983,

pp. 67-68, 72.

Ch. 3—Selected Case Studies in Information Technology Research and Development . 107

ties. ’G’ The Science and Engineering ResearchCouncil (SERC), roughly the equivalent of theAmerican NSF, will develop plans and dis-burse most of these funds.l62

The United Kingdom has a strong traditionin AI, particularly in the University of Edin-burgh and Imperial College, London. But inthe years immediately preceding the Alvey ini-tiative, efforts had been scaled back becauseof an unfavorable government report on the—-—

‘“A Program For Advanced Information Technology, op. cit.,pp. 47 and 49.

“’Ibid., p. 48.

prospects for artificial intelligence, and manyresearchers emigrated to the United States tocontinue work in American universities.163

Alvey Programme funding may reverse thesetrends, but there is disagreement among ex-perts concerning the ability of European uni-versities, in general, to respond flexibly to thegrowth of a field such as AI, where interdis-ciplinary research across traditional academicdepartmental lines is so crucial.

“’The OTA workshop, Oct. 31, 1983.

Chapter 4

Effects of Deregulation andDivestiture on Research

Contents

Page

Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Antitrust Laws, Deregulation, and Divestiture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Divestiture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Management of Research at AT&T . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .The-

BellModified Final Judgment and Bell Laboratories. . . . . . . . . . . . . . . . . . . . . . . . . . .Labs After Divestiture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Factors Affecting Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Stability of Earnings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Allocation of Research and Development Expenditures . . . . . . . . . . . . . . . . . . . . . . . .Basic Research. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Role of Bell Communications Research, Inc. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Availability of Research Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Policy Implications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 4 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

111

113116

120121122

123123125128129130131

134

Table No. Page

23.24.

Regional Bell Operating Companies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119R&D Intensities of Selected Major Telecommunication Firms . . . . . . . . . . . . . . . . . . 125


16.17.18.

Pre-Divestiture Bell System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117Post-Divestiture Organization of AT&T . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118Bell Operating Companies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

Chapter 4

Effects of Deregulation andDivestiture on Research

Findings

As a major source of information technol-ogy R&D—and as an organization that has re-cently undergone major legal, regulatory, andinstitutional changes-AT&T’s Bell Labora-tories merits special attention. In reviewingthe potential effects of the AT&T divestitureand of recent regulatory decisions on BellLabs, OTA made the following findings:

● Organizational changes within AT&TTechnologies and within Bell Labs indi-cate that AT&T is already preparing tospeed the development and marketing ofnew products. Other firms may also in-crease their development activities tomeet competition from AT&T.

• The effects on the research side are lessclear. AT&T has some incentives to con-tinue funding applied and basic researchat past levels, but these stand in tensionwith powerful new forces that could temptAT&T to direct more resources awayfrom research and into short-term re-search and development projects. Thereis little reason to think that AT&T’s com-petitors will perform more basic researchnow than they have in the past.

● The areas where AT&T will be the mostlikely to focus its competitive efforts arealso the areas where Bell Labs has beenresponsible for major scientific contribu-tions computer science, solid-state phy-sics, and photonics. Work in those areas,including basic research, is likely to con-tinue into the foreseeable future.

● A significant portion of Bell Labs’ re-search base has been moved to Bell Com-munications Research, Inc. (Bellcore), aunique new organization owned jointly bythe divested Bell operating companies.Bellcore’s role in basic research is stillunclear.

● It is possible to monitor research activi-ties over the next few years to determinewhether the quality or direction of basicresearch change in a deregulated environ-

ment. Because of the long-term nature ofthe work, however, it may take someyears for any changes to become evident.

The AT&T divestiture has been makingheadlines since January 1982, when AT&Tand the Department of Justice announced thesettlement of the Department long-standingantitrust suit. The divestiture marked the endof an era. Before divestiture, AT&T had beenthe nationwide provider of end-to-end telecom-munications services. AT&T’s system of Belloperating companies provided local service to85 percent of the telephones in the UnitedStates; the Long Lines division carried thevast majority of long-distance calls; the West-ern Electric subsidiary manufactured most ofthe equipment used in the system and leasedto end users. With assets of $150 billion andannual revenues of $69 billion, it was the big-gest communications company in the world.On January 1, 1984, the size of the corpora-tion was reduced to one-fourth as AT&T spunoff the Bell operating companies and gave uplocal telephone service.

While divestiture is indeed a dramatic event,the concern and publicity associated with ithave tended to obscure a related regulatorydecision: the Federal Communications Com-mission’s (FCC’s) decision in the Second Com-puter Inquiry (Computer II) detariffed the saleof terminal equipment, deregulated enhancedtelecommunication services, and permittedAT&T to sell these to end users through a sub-sidiary after Jan. 1, 1983.1 These changes inAT&T’s structure and markets have raisedsome important questions related to researchand development in telecommunications. Thequestion addressed here is how divestiture andderegulation will affect the functioning ofAT&T’s research arm, Bell Laboratories,

“’Enhanced communications” are services which require add-ing value to a transmission by altering the message in someway, as explained below.

111


viewed by some as the star of modern indus-trial and scientific research.

In 1982, before any of the changes associ-ated with divestiture and compliance withComputer II, Bell Labs had a budget of $2 bil-lion, facilities at 21 locations, and 25,000 em-ployees-3,000 with doctorates and 5,000 withmasters degrees. Bell Labs provided nearly allR&D leading to the manufacture of WesternElectric’s products, as well as systems engi-neering to support the Bell System generally.While the Labs’ principal role is in develop-ing products for sale or use by AT&T, about10 percent of the budget has been dedicatedto scientific research. The research has hadfallout applications in a wide variety of fields,from telephony and computer science to astro-

physics and health care. The research resultsand technical standards are widely publishedin scientific and technical journals. Bell Labsresearchers have made many fundamental ad-vances, inventing the transistor and other con-cepts at the base of the current generation ofcomputer and telecommunication technology.Among Bell Labs’ employees and alumni areseven Nobel laureates.

The Labs recently (1983) received its20,000th patent; this amounts to one patentper day since Bell Labs was incorporated in1925, and many other of its inventions havenot been patented. Traditionally, about 99 per-cent of AT&T’s R&D has been done internally.Very little technology has been bought fromoutside, although AT&T does enter into cross-

Photo credit: AT&T Bell Laboratories

Bell Labs developed the first 32 bit microprocessor: the dime-sized chip contains nearly 150,000 transistors and hasprocessing power comparable to that of today’s minicomputers.

Ch. 4—Effects of Deregulation and Divestiture on Research ● 113

licensing agreements. On the other hand, ithad been AT&T’s policy since the 1920s, anda legal requirement under the 1956 consent de-cree described below, to license its own patents ●

to other firms at reasonable cost. There arecurrently over 400 such licensing agreementsoutstanding in the United States and 200more with foreign firms. Arno Penzias, BellLab’s Vice President, has been quoted as say-ing that “without Bell Labs there would be noSilicon Valley.”2

Although that may be hyperbole, it is cer-tainly true that Bell Labs holds the basicpatents for the processes and products neededby many United States and foreign firms toget their start in microelectronics, computers,telecommunications, or other fields. The avail-ability of licenses and technical informationfrom Bell Labs greatly speeded developmentof the microelectronics industry.

Bell Labs’ R&D efforts are clearly importantto information technology generally. TheLabs’ budget makes up perhaps 15 percent ofthe R&D investment by information technol-ogy firms.3 Further, if Bell Labs is producingover 370 patents per year, then it accounts forperhaps 5 percent of U.S. patents in informa-tion technology fields.4 Anything that might

“’Bell Labs, Threatened Star of US Research,” Fortune, July5, 1982, P. 47.

‘See page 316. Investment in information technology by in-dustry in 1983 is estimated to be about $10.8 billion. Bell Labs’R&D budget of $2 billion is about 18 percent of the $10.8 bil-lion invested in IT R&D by large IT companies in 1982. How-ever, the $10.8 billion figure may be too low, as it does not in-clude R&D expenditures of many small firms.

‘This is an estimate based on approximately 5,180 Bell Labspatents as a fraction of 101,900 US patents in communicationsequipment and electronic components in 1963-1981. Data onU.S. patents from National Science Board, ~“ence Inol”cators,1982, U.S. Government Printing Office, 1983, p. 207.

reduce the scope or quality of research at BellLabs alarms observers who see the Labs as amajor contributor to the U.S. lead in informa-tion technology R&D.

The restructuring of AT&T creates pres-sures and incentives for Bell Labs that did notexist while AT&T was a regulated, end-to-endmonopoly. Because of competitive pressureson AT&T in the deregulated markets, BellLabs may choose to devote more of its re-sources to product development and to reducethe number of long-term research projectsleading to fundamental scientific discoveries.Such an event could be deleterious to the long-run competitive position of AT&T, and moreimportantly, might negatively affect the levelof U.S. R&D in information technology.

This chapter discusses the problems and op-portunities that the new post-divestiture en-vironment offers Bell Labs, and the possibleeffects that the changes in AT&T’s corporatestructure may have on research at the Labs,and throughout the telecommunication andcomputer industry. It focuses specifically onthe future stability of AT&T’s earnings, itsincentives to engage in research, and the pos-sible effects of deregulation on research else-where in the telecommunications and comput-er industry. Finally, it outlines some methodsfor monitoring the health of research at BellLabs and possible options for Federal Govern-ment action.

Before examining the effects on Bell Labs,it is necessary to briefly review the regulatoryand legal decisions leading to deregulation anddivestiture and to discuss the technologicaland market forces that drove them.

Antitrust Laws, Deregulation, and Divestiture

American Telephone and Telegraph is no General that resulted in the Kingsbury com-stranger to antitrust litigation. In order to mitment. In the commitment, AT&T agreed:avoid a threatened Government suit under the 1) to end its policy of aggressive mergers withSherman Antitrust Act in 1913, AT&T en- competing independent telephone companies;tered into negotiations with the U.S. Attorney 2) to allow the remaining independents to in-


terconnect with its long-distance system; and3) to get out of the telegraph business bydivesting itself of the Western Union tele-graph company.5 It removed AT&T from thetelegraphy market and significantly constrainedfuture purchases of competing telephone com-panies.

However, the actual effect of the Kingsburycommitment was to confirm AT&T as a regu-lated monopoly and to quell the competitionbetween Bell operating companies and inde-pendents which had grown up in the 1895-1913period. Under terms of the commitment, Bellcompanies and independents negotiated theborders of their service areas and exchangedtelephones where necessary to give each othergeographical monopolies. AT&T was acknowl-edged to control the entire long-distancenetwork, and the independents used that net-work as noncompeting partners in end-to-endservice.

The next major antitrust case against AT&T,in 1949, asked for an end to AT&T’s owner-ship of Western Electric and an end to all re-strictive agreements among AT&T, the BellOperating Companies, and Western Electric.The suit essentially sought the separation ofregulated monopoly services from equipmentsupply.

A negotiated settlement of the 1949 suit ledto a consent decree in January 1956. The con-sent decree imposed two important restric-tions on AT&T’s future activities. First, AT&Twas restrained from entering other lines ofbusiness, such as the sale of solid-state com-ponents or computers. It was restricted to pro-viding regulated common carrier service, withWestern Electric as its captive equipmentmanufacturer. AT&T was free to develop BellLabs technology, such as the transistor, foruse within its own system, but was forbiddento market these products to the public. Sec-ond, AT&T was required to license all patentscontrolled by the Bell System to any applicantat a “reasonable royalty” and to provide tech-

5Gerald W. Brock, The Telecommum”cati”ons Industry: TheDynamics ofi%farket Structure (Cambridge: Harvard Univer-sity Press, 1981), p. 155.

nical information along with patent licenseson payment of reasonable fees. This licensingprovision ensured that other firms could useBell technology outside of regulated telephonemarkets.6

Two major trends, each with a technologi-cal and a regulatory component, developedover the ensuing 25 years to make the line ofbusiness restriction of the 1956 consent decreeincreasingly unworkable. First was the devel-opment of technological alternatives in trans-mission and switching that greatly reducedthe cost of providing long-distance service andmade it economically attractive for competi-tors to challenge AT&T’s dominance of thelong-distance market. Second was the advancein computer microelectronics, which has beenleading to a convergence and interdependenceof communication and computation services.These technological changes, and the marketactivity that they generated, led to a numberof regulatory decisions that eroded AT&T’smonopoly position and gradually opened thetelecommunication transmission and equip-ment markets to competition.

The first chink in the long-distance monop-oly was FCC’s 1959 Above 890 decision,7 open-ing the microwave radio spectrum to privateusers. This led eventually to FCC’s approval,in 1969 and again in 1971, of MCI’s applica-tion for authorization to offer private line serv-ice via microwave. It was also in 1971 that theFCC made its Specialized Common Carrierdecision, 8 in which it concluded that a generalpolicy in favor of entry by new carriers intospecialized communications would serve thepublic interest. Long-distance service from“other common carriers” became more widelyavailable to the public in 1979 after a seriesof FCC and court decisions. By the end of1984, other carriers had captured 15 to 20 per-

5AT&T had been granting licenses and making available techn-ical information on its inventions before 1956. AT&T had de-veloped cross-licensing agreement with major manufacturerslike General Electric over the previous two decades. The pol-icy of licensing patcmts to smaller firms was in force in the1940s.

727 FCC (1959).*29 FCC 2nd 8 70 (1971).


cent of the long-distance market, as measuredby minutes of calls transmitted.9 Other carrierscan now claim relatively small numbers of sub-scribers, but they are principally the high vol-ume users. AT&T estimates that other carriersserve about one-third of the highest volumeresidential callers (those spending over $25 permonth) and one-half of high volume businesscallers (over $150 per month).*

In the terminal equipment market, theFCC’s 1968 decision in the Carterphone case10

was the first FCC action to allow consumer-owned terminal equipment to be attached tothe Bell system network. This decision, to-gether with an equipment registration pro-gram authorized by FCC in the 1970s, allowedmanufacturers other than Western Electric toenter the U.S. market, giving rise to the “in-terconnect” market for telephones and othercustomer equipment.

Meanwhile, the computer industry wasgrowing rapidly and without significant gov-ernment regulation. In order to determine howbest to deal with the policy questions thatwere already emerging from remote-accessdata processing, FCC initiated in 1966 its firstInquiry into Regulatory and Policy ProblemsPresented by the Interdependence of Comput-er and Communication Services and Facilities(Computer I Inquiry). The decision in Com-puter I, adopted in 1970, divided comput-er/communications services into two regulatedservices—pure communications and hybridcommunications-and two nonregulated serv-ices—hybrid data processing and pure dataprocessing. Under the terms of the 1956 Con-sent Decree, AT&T could provide pure and hy-— — — —

*FAA estimate, private communication, February 1985.*AT&T Communications briefing to OTA staff, August 17,

1984.IO% 13 F~ Zd, 420, 437 (1968). “Terrnina.1 equipment” or

“customer premises equipment” termin ates the telephone wireon the customer’s premises. The most common example is theordinary telephone. The terms are also used to refer to systemsof telephones, like the six button “key sets” used by many smallbusinesses, and to switching equipment, like the private branchexchanges (PBX) used to route calls inside large businesses.Modems (modulator~demodulat.ors that convert analog signalsto digital signals) interface between the telephone wire and acomputer and are considered terminal equipment, as are com-puter te rminals with built-in modems.

brid communications but could not provideany service or product that fell into the dataprocessing categories.

Throughout the 1960s and 1970s, AT&Thad been manufacturing and selling terminalsfor access to mainframe computers. Thesewere primarily built by the Teletype Corpora-tion, a subsidiary of Western Electric. Earlyterminals were clearly communication de-vices—they were only of use for sending infor-mation to a remote computer for processing.As microelectronics advanced, however, moreintelligence and power could be placed in ter-minals. It became increasingly difficult to de-termine at what point a terminal ceased to bea “hybrid communications” device and be-came a “hybrid data processing” device.

AT&T’s applications to the FCC for permis-sion to market new terminal equipment weresometimes challenged as being in violation ofComputer I rules and the consent decree. ”Further, AT&T was at a competitive dis-advantage because it had to go through a(sometimes lengthy) regulatory process beforeintroducing each new product, whereas theunregulated terminal suppliers (computermanufacturers) could introduce new productswhenever, at whatever price, they chose. Itwas clear that the combination of ComputerI rules, the consent decree, and the evolutionof technology were preventing AT&T from of-fering state-of-the-art terminal equipment tothe public.

FCC initiated its second inquiry, Comput-er II, in 1976 and issued a decision in 1980.That decision deregulated the sale of terminalequipment, both voice and data, and allowed

— — -l l F o r e x m ~ e, AT & T ’g rquegt for a tariff to sell the Data-

speed 40/4 was denied by FCC’s Common Carrier Bureau in December 1976. IBM and others objected that the terminal wouldbe in direct competition with terrnin sls built by computer man-ufacturers, and the Common Carrier Bureau agreed that theterminal’s storage and processing capabilities (designed to allowan operator to correct mistakes before sending data to the com-puter) violated FCC rules. The full Commission overturned thisdecision 9 months later. In its decision the Commission notedthat the Computer I rules were inadequate to deal with thechanging technology and that Computer II Inquiry then be-ginning would establish a new policy. See FCC Transmittal No.12449, 1977.


AT&T to offer this equipment for sale to thepublic through a subsidiary. Computer II alsoallowed AT&T to offer other enhanced tele-communication services through a subsidiary.12

AT&T created that subsidiary, AT&T Infor-mation Services or ATTIS (originally calledAmerican Bell), in June 1982.

Divestiture

Meanwhile, the Department of Justicebrought an antitrust suit against AT&T in1974, seeking many of the same goals as in1949. The suit alleged that AT&T monopolizedthe manufacturing, long-distance, and localservice markets; that it used its monopolypower in each market to strengthen its powerin the other markets; and that it attemptedto prevent competing equipment manufactur-ers and long-distance carriers from gaining ac-cess to the local networks. In January 1982,Department of Justice announced that it hadreached agreement with AT&T on changes tothe 1956 consent decree and in August 1982,Judge Harold H. Greene of the U.S. DistrictCourt for the District of Columbia approvedthe Modified Final Judgment. The Govern-ment’s case was dismissed upon acceptanceof the terms of the Modified Final Judgmentby all parties.

Under the Modified Final Judgment, localBell operating companies providing local ex-change telephone services were divested byAT&T, and spun off into seven regional hold-ing companies. AT&T retained ownership ofa nationwide intercity network composed ofits Long Lines division and the intercity fa-cilities of the Bell operating companies, andcontinued to own Bell Laboratories and West-ern Electric. The Modified Final Judgment

l~In it9 ~ond ComPuter Inquizy decision, the FCC distin-guished between basic and enhanced services. Basic serviceswere defined to be the transmission of information, whileenhanced services involved adding value to transmission bychanging or acting on the message itself in some way. As anexample, in voice traffic, a simple long-distance telephone callconstitutes basic service. Enhanced service would be providedif the carrier stores and forwards calls or provides recordedmessages for those who are calling. An enhanced data servicemight be one that provides protocol conversion so that non-compatible computers can communicate.

allowed AT&T to enter computer, computer-related, and information services markets incompetition with unregulated firms (althoughthere are still restrictions on AT&T’s actions;e.g., AT&T may not provide information serv-ices over its own lines for 7 years).

The breakup, according to Judge Greene, re-duces AT&T’s ability to rely on its monopolyat the local exchange to exact competitive ad-vantage in interexchange (long-distance), ter-minal equipment, and computer services mar-kets. AT&T’s long-distance market is stillregulated, but FCC regulation was not viewedby the court as so extensive, nor were barriersto entry seen as so high, that AT&T will beable to use its currently large share in thismarket to provide a competitive advantage inunregulated segments of the industry.

Figures 16 and 17 compare the predivesti-ture and post-divestiture organizational struc-ture of AT&T and the Bell operating compa-nies. Before divestiture the entire Bell systemexisted under a single corporate umbrella andthe firm was organized to provide end-to-endtelephone service. The Long Lines divisionprovided interstate long-distance services;Western Electric manufactured equipment foruse throughout the system; the 22 whollyowned Bell operating companies provided lo-cal and intrastate service; a small internationaldivision marketed AT&T equipment abroad.Bell Labs provided design and developmentfor Western Electric as well as research andnetwork system engineering for the rest of thesystem. The AT&T Information Systems sub-sidiary was created in 1982 in response to theComputer II decision.

As figure 17 shows, AT&T after divestitureis primarily comprised of AT&T Communica-tions, AT&T Technologies, AT&T Interna-tional, and the subsidiary, AT&T InformationSystems. AT&T Communications provideslong-distance service between local callingareas.13 AT&T Technologies includes the func-tions of Western Electric and Bell Labs. Itnow provides research and development, man-

laLATA—~~ A@ss and Tr~port Area-is the term nowused to identify a local calling area.

Ch. 4—Effect of Deregulation and Divestiture on Research • 117

Figure 16.—Pre.Divestiture Bell System a

Pacific/Nevada

aAll entitles report to AT&T

Southwestern Southern

ufacturing, and marketing of equipment andservices both in the United States and abroad.Western Electric no longer exists as an orga-nizational unit, but AT&T Technologies willcontinue to use it as a trade name. Bell Labsis the section of AT&T Technologies respon-sible for R&D.

AT&T Information Systems will market in-formation services, terminal equipment andcomputers to end users. Dealings betweenATTIS and the other AT&T entities, underrules of Computer II, must be at arm’s length.Information related to AT&T’s customer base,

for example, cannot be shared with ATTIS(unless it is also shared with competitors).

As shown in figure 18 and table 23, divesti-ture places the Bell operating companies intoseven regional holding companies, of approx-imately equal size in terms of assets and cus-tomer base. The seven jointly own and oper-ate Bell Communications Research (Bellcore),which provides technical and administrativeservices.

Judge Greene ruled shortly after the divest-iture that the name “Bell” and the familiar

Ch. 4—Effect of Deregulation and Divestiture on Research • 119

Figure

The seven regional Bell operating companies

1

2

3

4

5

6

7

Nynex

18.-Bell Operating Companiesa

Indiana BellMichigan BellOhio BellWisconsin Telephone

Southwestern BellSouthwestern Bell

U.S. WestMountain BellNorthwestern BellPacific Northwest Bell

Pacific Telesis

—

New EnglandTelephone

New YorkTelephone

Bell AtlanticC&P Telephone(4 companies)

Diamond StateTelephone

New Jersey BellBell of Pennsylvania

BellsouthSouthern Bell ISouth Central Bell

AmeritechIllinois Bell

Paclflc TelephoneNevada Bell

a The ~eglona~ Bell ~peratlng ~ompanle~ are h~ld,ng ~~rnpant~s for Bell operating companies that offer service In the States Inchcated above Within mOSt StateS IOcal

telephone service IS also provided by Independent telephone companies

SOURCE: AT&T.

Table 23.—Regional Bell Operating Companies

Total operating Value ofrevenue 1984 assets Net income embedded plant Access lines

Regions (millions) (billions) (millions) (millions) (thousands)

Ameritech . . . . . . . . . . . . . . . . . 8,900 16,26 1,037.1 14,409 13,970Bell Atlantic. . . . . . . . . . . . . . . . 8,732 16.26 1,054.5 14,596 14,011Bell South . . . . . . . . . . . . . . . . . 10,512 20.81 1,393.1 19,081 13,367NYNEX . . . . . . . . . . . . . . . . . . . . 10,006 17.39 1,029.8 15,186 12,658Pacific Telesis. . . . . . . . . . . . . . 7,895 16.19 977.1 14,493 10,717Southwestern Bell . . . . . . . . . . 8,859 15.51 887.9 14,112 10,189U.S. West . . . . . . . . . . . . . . . . . . 7,596 15.05 910.9 13,767 10,381

SOURCE: Bell Communications Re9earch, Inc., November 1984.

38-802 0 - 85 - 9


logo are the property of the Bell system–that abroad. The one exception to this ruling wasis, the Bell operating companies. AT&T may that the name of Bell Laboratories did notnot use the name or logo in the United States, have to be changed, although it is now calledalthough AT&T International may use it AT&T Bell Laboratories. -

Management of Research at AT&T

named and incorporated inBell Labs was1925, but it grew out of an in-house researchcapability which AT&T had maintained since1907. AT&T was a groundbreaker in bringingR&D out of the homes and private laboratoriesof individual inventors and into the industrialcontext. In many ways, research at AT&T wasa model for the modern industrial lab as it de-veloped in other industries.

Most of Bell Labs’ resources have been de-voted to design and development of productsfor sale or use in the Bell system. However,about 10 percent has traditionally been de-voted to research. “Research” at Bell Labs en-compasses those projects for which no specific,short-term benefit to the corporation is fore-seen. Most of the research is applied or di-rected systematically toward the solution ofparticular problems, but some resources havebeen devoted to basic research, sometimesleading to major scientific advances.

Before divestiture, Bell Labs’ work was sup-ported by the other AT&T entities. In 1982,and typically in the predivestiture era, abouthalf of the Labs’ support (54 percent) camefrom Western Electric, to cover costs of spe-cific design and development.14 In addition,Western paid another 3 percent to supportwork on products being developed under Gov-ernment contract. Another 11 percent camefrom Bell operating companies to pay for cen-tralized development of computer informationsystems.

The remaining 32 percent of Bell Labs budg-et was paid by AT&T for research and systems

14 Fiweg from ch~les River Associates, Impacts of theAT& Td”vestiture on Innovative Behavior, unpublished paperprepared for OTA, 1983, p. 17.

engineering. The majority of the funds usedfor research and system engineering came toAT&T from the Bell operating companiesunder the “license contract. ” The contract wasan arrangement under which the operatingcompanies were assessed up to 2.5 percent oftheir annual revenues to pay for their use ofAT&T technical and administrative services.About 30 percent of these funds, together witha contribution from the Long Lines division,were allocated to research.

Funding of research at Bell Labs was anal-ogous to some of today’s attempts at joint re-search funding, such as Microelectronics andComputer Technology Corporation (MCC) orSemiconductor Research Corporation (SRC).15

The operating companies were, in a sense, sep-arate user companies that contributed to thesupport of a central research facility for mu-tual benefit. The difference in this case wasthat the operating companies all existed undera single corporate umbrella, so that they hadlittle control over how their contributions werespent and no option of withdrawing from thejoint funding venture or establishing other ar-rangements.

A number of factors in the “climate” of BellLabs have been cited as contributing to itsachievements in fundamental research. Somehave pointed out that Bell Labs scientists hadaccess to state-of-the-art equipment, and werefree to focus on their research without the re-sponsibilities of teaching or serving on com-mittees that would be required in a universitysetting. Because of job security and the stabil-ity of funding, there was no need for research-ers to spend time pursuing grant support.

15FOr a description of these joint research ventures, .9- ch. 6.


There has been a tradition of staff interactionsacross disciplinary boundaries. Bell Labsmaintained an open publication policy-its re-searchers have published about 2,000 papersper year. With these advantages, Bell Labswas able to attract outstanding scientists andengineers to work in its research organiza-tion. l6

The Modified Final Judgment andBell Laboratories

In his opinion on the Modified Final Judg-ment, Judge Greene commented on the pro-posal that Western Electric and the Bell Lab-oratories be divested from AT&T. He notedthat the success of the Bell Laboratories inbasic and applied research (and the beneficialimpact of that research on the Nation’s eco-nomic position) was due to its relationshipwith the operating companies and the LongLines division. He argued that continued asso-ciation of the Labs with the AT&T entitiesproviding manufacturing and long-distanceservices would supply “the practical experi-ence that would be useful in stimulating theresearch operations. ”17

The possibility of negative effects on re-search at Bell Labs was considered in the ne-gotiations leading to divestiture, but was notconsidered a matter of highest priority. Ches-sler,l8 in summarizing the position of Govern-ment negotiators, indicates that they acceptedthe possibility that divestiture might lead toa reduction in basic research activities:

The competitive era in station equipment,interexchange communications, and informa-tion services under the [MFJ] will bring fortha great blossoming of progress in those areasof telephony. It was the thought of the frame-rs . . . that the blossoming will be so greatas to more than compensate for the loss ofpure research at Bell Telephone Laboratories,and the reduced incentives for innovation atthe Bell operating companies.

Judge Greene did not believe that incentivesfor innovation were being sacrificed or that

‘~TA, notes on interview with workshop participants.170pinion and Order, Aug. 11, 1982, p. 62.“Cited in “Bell Labs on the Brink, ” Science, Sept. 23, 1983.

divestiture per se would hurt the quality ofservice provided by the operating companiesor the research performed by of Bell Labora-tories. He noted-that the largest potential cus-tomers of Western Electric will be the divestedoperating companies, hence, Western Elec-tric’s association with Bell Laboratoriesshould provide an incentive to improve equip-ment and technology.19

The Modified Final Judgment sets aside the1956 Consent Decree and the requirement thatAT&T grant nonexclusive licenses for itspatents to any applicant. The elimination ofthis requirement makes it easier for Bell Labsto appropriate the potential benefit of newbreakthroughs, and therefore might be consid-ered an incentive to research. AT&T may nowgrant or deny licenses as it chooses, and maychange whatever royalty it chooses. Beforedivestiture, when revenues from local ex-change ratepayers were supporting Bell Labs’research, it made sense to require AT&T toshare the fruits of its monopoly financing withothers, according to Judge Greene. With thedivestiture of the-operating companies and thetermination of the license contract fee pay-ments, this rationale for required licensing iseliminated. Judge Greene also believed thatthe advance of technology and the dispersionof knowledge related to telecommunicationstechnology has reduced the dependence of es-tablished domestic firms and foreign compet-itors on information from Bell Laboratories.20

The Modified Final Judgment requires thatAT&T grant licenses to the divested operat-ing com-panics on all existing patents and allpatents issued for a period of 5 years follow-ing approval of the Modified Final Judgment.AT&T is also required to provide the operat-ing companies with nonpatentable technicalinformation that has been funded by the li-cense contracts. The operating companies willhave the right to sublicense AT&T patentsand technic-d information to those providingt h e m w i t h g o o d s a n d s e r v i c e s .

I’Charles River Associates, “Impacts,” p. 43.‘“Ibid.


Bell Labs After Divestiture

The most noticeable change resulting fromComputer II and divestiture is a reduction inBell Labs’ size. About 4,000 employees be-came part of AT&T Information Services(ATTIS). FCC has interpreted the ComputerII ruling that ATTIS and Bell Labs deal atarm’s length to mean that ATTIS employeesmust be kept separated from former Bell Labscolleagues, even though they are sometimeslocated in the same buildings. Another 3,000Bell Labs employees went to the newly createdBell Communications Research Inc. (Bellcore,formerly the Central Services Organization) ofthe Bell operating companies. This leaves BellLabs with about 18,000 employees, returningit to approximately the size it was in 1978.

Organizational changes taking place else-where in AT&T Technologies will also affectBell Labs. In 1983, AT&T Technologies wasorganized into “line of business” divisionsdefined by customer and product type. WithinBell Labs, development teams have been re-organized along the same line of business cat-egories in order to facilitate cooperation withmanufacturing. 21 Authority for managing de-

sign and development of products within BellLabs was given to executives running each lineof business division, as shown by the dottedlines in figure 17.

This is a major departure from previousAT&T policy wherein Bell Labs, Western Elec-tric, and AT&T shared this authority; unlikepractice at most firms, the old arrangementgave Bell Labs some control over a producteven after it went into production. The newarrangement was chosen to make developmentmore responsive to the needs of marketing andmanufacturing, and is a preparation to entercompetitive markets. Although the organiza-tional structure is new, it marks the continua-tion of a trend which began when the marketfor large private branch exchanges (PBXS)22

became competitive after 1968. Shortly there-

~lBrO utt~, “cold New World,” Fortune, June 27, 1983, P. 83.~~private Br~ch Exch~W is a generic term for the switch

used on the customer premises for routing calls within a build-ing or organization.

after, nearly all Bell Labs personnel workingon PBXs were collected in one Colorado Labfacility near the Western Electric facilitywhere PBXs were manufactured.

Figure 17 also shows that research at BellLabs remains independent from the lines ofbusiness in AT&T Technologies. However,sources in Bell Labs note that research isundergoing review and changes as a result ofderegulation and divestiture. The loss of re-search personnel to ATTIS and Bellcore causedsome realignment of research projects. Someother areas of research-for example, regu-latory economics and social psychology-havebeen judged unproductive or inappropriateand have been cut back. New research topics,such as robotics, are being undertaken.

A major change to Bell Labs’ funding sincedivestiture is the termination of the licensecontract revenues from the local operatingcompanies, funds that were specifically dedi-cated to research and system engineering.Under the current funding arrangements, re-search is supported by AT&T Headquarterswith funds provided by the AT&T companiesunder a “composite allocator. ” AT&T entitieswill be assessed for Bell Labs research (as wellas administrative functions of AT&T Head-quarters) according to their size, number ofemployees, and revenues. The allocation for-mula, under the Computer II rules, must bereviewed and approved by FCC to ensure thatAT&T allocates a reasonable portion of re-search costs to the unregulated portion of itsbusiness and does not subsidize it from regu-lated long distance revenues.23

Another major change for Bell Labs will bean increase in the work done on military proj-ects. AT&T Technologies is planning to in-crease the number of defense contracts, andthe design and development work will be donein Bell Labs. Although defense contracts wereonce very important to Bell Labs, they hadbeen reduced to a minor part of the R&D budg-et during the 1970s. In 1971, Bell Labs derived30 percent of its income from defense-related

*gSee FCC 83-600, Dec. 22, 1983 and FCC 83-123, Mar. 31,1983.


work; by 1976 the share was down to 2 per-cent and in 1983 about 3 percent.24

Growth of defense projects to an expected10 percent of Bell Labs’ budget should not bedifficult. Based on its previous work, AT&Thas strong ties with the Pentagon and a goodreputation for designing and building the— — . - —

24Marilyn A. Harris, “Bell Labs Looks to Military Research, ”Electrom”cs, Feb. 9, 1984.

kinds of large complex systems that the De-partment of Defense wants. As Solomon J.Buchsbaum, executive vice president for con-sumer systems, points out, “The military sideof government is a voracious eater of new tech-nology, and we are good at [providing] that. ”25

2s’’Bell Labs: The Threatened Star of U.S. Research, ” Bzzsi-ness Week, July 5, 1982, p. 49.

Factors Affecting Research

All the changes taking place in AT&T’s mis-sion, markets, and corporate structure cannotbut affect the activities of AT&T Bell Labs.The purpose of the Labs has always been toprovide research, systems engineering, andproduct design to support the corporate activ-ities of AT&T. As those activities have evolved,the role of the Labs has also changed. Of par-ticular concern to many observers is the wayin which deregulation and divestiture mightcause changes in the commitment to research,particularly basic research, within Bell Labs.

Several concerns have been voiced. WillAT&T, as a smaller corporation with a narrow-er revenue base, be able to support researchas it has in the past? What incentives doesAT&T have to allocate funds to research, andhow strong are they compared to incentivesto allocate more resources to development ofcompetitive products? How will changes re-lated to divestiture and deregulation affect re-search at other firms in the telecommunicationand computer industries? Could a reductionin the level of research at Bell Labs have a neg-ative effect on U.S. research generally, and ifso, what can be done about it? The remainderof this chapter addresses these questions.

Stability of Earnings

The future funding of research at Bell Labswill depend, at least in part, on AT&T’s suc-cess in the market. The combination of divest-iture and deregulation leave AT&T a smaller

firm.book

While the predivestiture AT&T had avalue of $150 billion, the new AT&T has

assets of only about $34 billion. However, thenew AT&T is expected to have a much morefavorable ratio of revenues to assets. Annualrevenues are now expected to be on the orderof $57 billion, compared with $69 billion forthe predivestiture firm. This is largely becauseAT&T will continue to provide long-distanceservice, which has traditionally been very prof-itable and is estimated to provide two-thirdsof the corporation’s profit base.26 Also, AT&Twill continue to manufacture telecommunica-tions equipment. Further, AT&T now has theopportunity to expand into potentially prof-itable computer-related markets.

While competitive computer markets are po-tentially profitable, they are also notable fortheir volatility over the past few years: newfirms and new products have had meteoric suc-cesses and catastrophic failures. This kind ofmarket may be dangerous for a firm which isunaccustomed to competition. AT&T has notbeen particularly successful in markets whereit has been open to competition in the past.After 1976, when customers were permittedto purchase their own private branch exchange(PBX) switching equipment from other man-ufacturers, AT&T’s market share fell sharply.Although AT&T is still the largest single man-ufacturer, it now has 24 percent of U.S. sales,

‘Peter Hall, “AT&T and the Great Divide, ” Hnanci”al World,Jan. 10, 1984.


compared to 100 percent 8 years ago. Majorcompetitors, specifically, Northern Telecom,Rolm, and Mitel have shares of 16,14, and 11percent, respectively .27

This loss of market share is due at least part-ly to AT&T’s higher prices and relative slow-ness in bringing new products to market. Boththese tendencies could be major disadvantagesin industries that are noted for rapid introduc-tion of new products and rapid obsolescenceof old ones. In part, slowness in bringing prod-ucts to market was related to the regulatoryprocess-a situation that has been eased sinceComputer II, but not eliminated.

AT&T has traditionally designed and manu-factured its products to extremely highstandards; they were expected to be highlyreliable with a long useful life. Such a strat-egy made sense when AT&T was the ownerof a huge nationwide network of transmissionand terminal equipment that had to be depre-ciated over 20 to 40 years. The higher costsof conservative design were made up by a longproduction run. Western Electric maintaineda price advantage over some other manufac-turers by producing large numbers of stand-ard products over many years.

Western Electric has often been at a costdisadvantage, however, in the case of newerelectronic products, the very ones that are thetarget of the competitive market. Small digitalPBXs for example, cost Western about 75 per-cent more to manufacture than those made bytheir lowest cost competitor, Mitel.28

AT&T has made a concerted effort to stream-line its manufacturing and to reduce costs.AT&T Technologies is reducing its work force,and several former Western Electric factorieshave been closed down or cut back. AlthoughAT&T did not sell integrated circuits andother electronic components to the public, itis the Nation’s 12th largest manufacturer.AT&T Technologies is now expanding thatmanufacturing capability, including construc-

‘7Northern Business Information, Inc., as cited in “ITl”s BigGamble” Business Week, Oct. 22, 1984.

*8Northem Business Information, as quoted in Bro Uttal,“Cold New World,” Fortune, June 27, 1983, p. 83.

tion of a new plant in Florida to make lowercost chips for use in computers and switches.

AT&T Technologies will continue to be amajor manufacturer of telecommunicationtransmission equipment, large central officeswitches, and terminal equipment. Potentialcustomers include Bell operating companies,independent telephone companies, and tele-communications agencies abroad.29

In addition, AT&T is now free to sell prod-ucts it developed but could not market to thepublic under the 1956 consent decree. It cannow market computers based on the UNIXoperating system, the 256 K-byte memorychip, and the 32-bit processor, all developedat Bell Labs. Its 3B computer series will of-fer a range of computers of varying size andcapability.

For the first time, AT&T is acquiring someof its new products, marketing talent, and dis-tribution channels through other firms. For ex-ample, AT&T acquired a 25 percent interestin Italy’s Olivetti Co. at a cost of about $260million.3o Olivetti is Europe’s largest wordprocessor and computer manufacturer. Theagreement is expected not only to supplyAT&T with Olivetti office equipment for theU.S. market, but also to provide a Europeandistribution system for AT&T products. AT&Thas also entered a joint venture with a Nether-lands electronics firm, Philips, to manufacturecentral office switching equipment for Europe.AT&T has also made agreements with a num-ber of smaller US. office computer manufac-turers for development of new office automa-tion equipment.

At the same time, and equally importantly,AT&T is developing its own marketing capa-bility. Before divestiture, Western Electricwas strictly a manufacturer and dedicatedvery few resources to marketing. One observernotes that a competitor, Northern Telecom,spends about 9 percent of manufacturing saleson marketing while Western Electric spent

~gKathl~n K. wiegner, “Prometheus is Unbound and Seek-ing His Footing, “ Forbes, Mar, 12, 1984, p. 143.

‘“Ibid.


only 1.2 percent in 1982.31 Now, AT&T Tech-nologies will be responsible for marketing allproducts not handled by ATTIS, and a mar-keting division and all the support functionsare being developed.

Allocation of Research andDevelopment Expenditures

As listed in table 24, AT&T and IBM havethe largest R&D budgets among the U.S.firms shown. The R&D intensity, that is, R&Das a percent of revenues, is based on totalsales, which before divestiture included reve-nues of the operating companies providinglocal telephone service. AT&T’s R&D inten-sity is a fairly low 3.3 percent when based ontotal revenues. Nordhaus cites historical evi-dence, however, to indicate that as a percent

~lBrO ut~, “cold New World, ” Fortune, June 27, 1983, P. 83.

of manufacturing sales AT&T spent approx-imately 9.8 percent of revenues on R&D in the1970s, as compared with an average of about2.8 percent for communication firms, and 1.9percent for manufacturers generally .32

On average, Bell Labs has spent about 10percent of its R&D budget on research. Amongthe other firms in table 24, both NorthernTelecom and IBM also claim to spend about10 percent on research. While Northern Tele-com is a competitive firm, it operates underthe corporate umbrella of Bell Canada, andshares the expenses of Bell-Northern Researchwith the regulated firm.

The general argument expressed by con-cerned observers is that AT&T may be forced,because of competitive pressures, to investmore of its R&D funds in developing salable

9Zchwle9 R,iVer Associates, Op. Cit., P. 7.

Table 24.–R&D Intensities of Selected Major Telecommunication Firms (1982)

R&D Expenses Sales R&D intensityCompany (millions of dollars) (percent)

(1) (2) (1)/(2)

AT&T . . . . . . . . . . . . . . . . . . . . . . . . . . . $2,126 a $65,093 3.3 ”/0COMSAT . . . . . . . . . . . . . . . . . . . . . . . .GTE . . . . . . . . . . . . . . . . . . . . . . . . . . . .Harris. . . . . . . . . . . . . . . . . . . . . . . . . . .ROLM . . . . . . . . . . . . . . . . . . . . . . . . . .United Telecommunications . . . . . . .Western Union . . . . . . . . . . . . . . . . . . .Zenith Radio. . . . . . . . . . . . . . . . . . . . .ITT . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Rockwell International . . . . . . . . . . . .General Dynamics . . . . . . . . . . . . . . . .IBM . . . . . . . . . . . . . . . . . . . . . . . . . . . .Motorola . . . . . . . . . . . . . . . . . . . . . . . .RCA . . . . . . . . . . . . . . . . . . . . . . . . . . . .General Electric . . . . . . . . . . . . . . . . . .L. M. Ericsson (1980). . . . . . . . . . . . . .Northern Telecom . . . . . . . . . . . . . . . .Plessey (fiscal year 1982) . . . . . . . . . .Siemens (fiscal year 1982) . . . . . . . . .Thomson-CSF (1980) . . . . . . . . . . . . . .The CIT-Alcatel Group (1981). . . . . . .Hitachi (fiscal year 1981) . . . . . . . . . .NEC (1981) . . . . . . . . . . . . . . . . . . . . . .Fujitsu . . . . . . . . . . . . . . . . . . . . . . . . . .‘{”~lud~~ $1,515 million spent by western Electric Co. and other subsidiaries, not worted in ATaT’s 10K.bCanadian dollars.SOURCES: Business Week, June 20, 1983; L. M. Ericsson Telephone Co., Anrrua/ Repoti, 1980; Plessy, Report and Accounts,

1982; Siemens, Annual Reporl 198142; Thomson4SF in 1980: The Year in Review; CITA/catel Group Review; Hitachi,1981 Annua/ Report; NEC Nippon Electric Co., Limited, Annual Report, 1981; and Fujitsu Limited, Anrrua/ Repoti,March 1981.

FROM: Charles River Associates, “Impact of the AT&T Divestiture on Innovative Behavior, ” unpublished report prepared forOffice of Technology Assessment, December 19S3,

22.3267.0

92.824.411.211.063.8

519.0222.0139.0

2,053.0278.0195.4781.0231.3241.4b

241.83,300(DM)3,600(FF)1,000(FF)

609.0227.8256.1

’41O12,066

1,719381

2,2491,0251,239

15,9587,3956,155

34,3643,7868,237

26,5002,779.43,035.5b1,723.9

34,600(DM)22,300(FF)1O,7OO(FF)15,996.14,820.22,769.9

5.42.25.46.40.51.15.13.33.02.36.07.32.42.98.38.0

14.09.5

16.19.33.84.79.2


products, and correspondingly less in fundingresearch projects that may lead to futurescientific breakthroughs. They note that thelicense contract fees described above, whichprovided a steady income source for Bell Labs,will no longer be available, and that researchfunding will depend on yearly corporate deci-sions. Any reductions, instability, or evenuncertainty about funds could have negativeeffects on the productivity of research projectsthat by their nature require long-term atten-tion and investment.

The arguments related to a possible changein AT&T’s policy toward research are basedon two major effects of the deregulation anddivestiture-AT&T will be operating in com-

petitive markets and it will be a smaller cor-poration.

Neither the theoretical nor the empirical re-lationships between market structure, firmsize, and innovative activity are straightfor-ward or well understood. It is not clear whetherinnovation is most likely to occur under con-ditions of competition or of monopoly. In ad-dition, although many support the view thatlarger firms have more incentives to innovate,there are many examples, especially in the in-formation industries, of small firms that growlarge due to extremely successful innovations.Further, most existing theory deals with “in-novative behavior” or R&D as a whole, ratherthan with the specific relationship of market


structure or firm size with the basic researchcomponent of R&D.

One theoretical argument, as proposed bySchumpeter and others, is that innovativebehavior is greater in monopolistic industriesthan in competitive ones because a firm withmonopoly power: 1) can prevent imitation andtherefore capture more profit from innovation,and 2) is better able to assemble the funds andbear the risks of R&D.33 On the other hand,critics of this position theorize that firms incompetitive industries are more likely to in-novate because new products or processes willhelp them to reduce costs or increase marketshare. In this view, monopolistic firms wouldbe slow innovators because they can continueto earn profits by continuing to produce thecurrent products. In addition, because themonopolistic firm is under less pressure tooperate efficiently, the results of innovativeactivity would be obtained at excessive cost.34

Real-world markets are characterized byvarying degrees of concentration rather thanextremes of pure competition or monopoly. At-tempts to empirically measure the relationshipbetween innovation and degree of industryconcentration have had mixed results.35 Forexample, Scherer36 found some evidence thatdominant firms in highly concentrated indus-tries are more innovative. However, in a laterstudy37 he found that the relationship variedgreatly depending on the industry, and thatthere were examples where higher levels of in-novation were associated with more competi-tive industries. In some cases, dominant firmswere only moderately productive innovators,but they were able to aggressively take advan-———. —

33 Summarized in Morton I. Kamien and Nancy Schwartz,Market Structure and Innovation (New York: CambridgeUniversity Press, 1982), p. 47.

SiMorri9 E. Morkre, “Innovation and Market Structure: ASurvey, ” Working Paper No. 82, Bureau of Economics, Feder-al Trade Commission, 1982, p. vi.

S~The 1itera~re is review~ in Morkre, “Innovation, p. 11.‘F. M. Scherer, “Firm Sizes, Market Structure, Opportunity,

and the Output of Patented Inventions, ’ American Econom”cReview, 55:11 19.

“F. M. Scherer, Industrial Market Structure and Econom”cPerformance (New York: Rand McNally, 1980), p. 431-432.

tage of innovations by other firms. IBM wasgiven as an example of such a firm.38

The empirical evidence on the effects of firmsize on innovation is less ambiguous than theevidence on market structure. R&D at smallfirms is sometimes more efficient than at largeones for R&D projects undertaken by bothlarge and small firms. 39 However, some proj -ects are simply beyond the reach of small firmsand there may be economies of scale for otherprojects. It appears that R&D intensity in-creases with firm size until firms reach annualsales of $250 million to $400 million (1978prices) and then level off.40 After reviewing theempirical evidence on firm size and innovation,Scherer concludes that an industry with amoderate degree of concentration and a vari-ety of firms of different sizes is most conduciveto innovation.

All things considered, the most favorableindustrial environment for rapid technologi-cal progress would appear to be a firm sizedistribution that includes a preponderance ofcompanies with sales below $500 million,pressed on one side by a horde of small, tech-nology-oriented enterprises bubbling overwith bright new ideas and on the other by afew larger corporations with the capacity toundertake exceptionally ambitious develop-ments. 41

After divestiture, AT&T is still many timesthe threshold level of size that empirical stud-ies have associated with maximum R&D. Itwill still be the dominant firm in a telecom-munications industry that fits well Scherer’sdescription of the environment most favorablefor innovation. Thus, though the details ofAT&T’s R&D may change, there are no con-vincing theoretical arguments or empiricalevidence related to market structure or firmsize that would predict a lessening of its in-novative activity.

3BIbid., p. 432.‘gMorkre, p. vi.40Charles River Associates, “Impacts,” p. 88.ilscherer, lndust~”~ Market Structure, P. 422.


Scherer and others have concluded that theimportant determinant of innovation may notbe market structure or firm size but rather therichness of innovative opportunities opened upby the underlying base of scientific knowledge.Advances in science related to semiconduc-tors, computers, software, satellites, micro-wave transmission, fiber optics, and lasers pro-vide a rich set of technological opportunitiesupon which to base innovations in telecommu-nications and information technology.42

Basic Research

Economic literature on “innovation,” how-ever, does not deal adequately with the effectof firm size or market structure on contribu-tions to the knowledge base that supports in-novation. The expected effect of competition,as noted above, is investment in developmentof new products and services, which will re-duce cost or improve market share in the shortterm. Investments in research, especially basicresearch, may not pay off until many yearsafter the initial investment is made.

One unique characteristic of Bell Labs is itsreputation for doing basic research.43 In gen-eral, only a few firms in the information indus-tries have spent much on in-house basic re-search. In 1981, out of 110 firms doing R&Din information technology, only seven did anybasic research at all, according to the NationalScience Foundation.44 Speaking of Bell Labsone observer from Bell-Northern Researchnoted, “Most other organizations are looking

— . — - .42Charles River Associates, Impacts, p. 85.48National Science Foundation, Reseamh and Development

in Industry, 1981, NSF 83-325 (Washington, DC: U.S. Gover-nment Printing Office, 1983), p. 3. The National Science Foun-dation defines basic research as “original investigations for theadvancement of scientific knowledge not having specific com-mercial objectives, although such investigations may be in fieldsof present or potential interest to the firm.

“Information technologies in this case includes firms in thefollowing categories: office, computing, and accounting ma-chines (SIC 357); communications equipment (SIC 366); elec-tronic components (SIC 367). See “Table B-33–Number of R8zDPerforming Companies Conducting Basic Research By Indus-try: 1981,” p. 38 in National Science Foundation, Reseamh andDevelopment in Industry, 1981, NSF 83-325 (Washington, DC:U.S. Government Printing Office, 1983).

at how to exploit technology, not at how topush it forward.’’”

When AT&T gets more experience as a com-petitive firm, will it continue to do basic re-search, or will it begin to behave as it appearsother competitive firms do, and dedicate moreresources to product-oriented research and de-velopment? Some observers, including Nord-haus, believe that AT&T will now “tilt muchmore toward a conventional equipment man-ufacturer, and it will therefore have a rela-tively greater incentive to invest in R&D thatwill enhance its equipment sales and profits”and relatively less incentive to invest in basicresearch.46

At the present time, AT&T’s managementhas voiced a commitment to continuing funda-mental research, recognizing that advances inscience are necessary to advances in technol-ogy. In testimony before the Senate Com-merce Committee, AT&T President Charles L.Brown called Bell Labs the “jewel” of the Bellsystem, and pointed out that “basic researchhas been the root of Bell Laboratories success

and will continue to be the root of it. Wedo not intend to skimp on it . . . . This is some-thing we have as a basic tenet.”47

It is probably true that more than half a cen-tury of reliance on internally developed tech-nology will not be quickly tossed aside. OneBell Labs spokesman said that the “corporateculture” of AT&T is completely oriented to-ward doing basic research in-house. The forcesof habit and tradition may resist some pres-sures to shift too many resources to develop-ment.48 Most of the technologies that will becommercially important to AT&T in the fu-ture–computer science, photonics, and solid-state physics—are the very areas where BellLabs has made ongoing contributions to basic

46John A. Roth, executive VP, Bell Northern Research, ascited in “Bell Labs the Threatened Star of US Research,” BusJ”-ness Week, July 5, 1982.

4’William Nordhaus, cited in “Bell Labs on the Brink,”Science, Sept. 23, 1983, p. 1267.

47Testimony of Charles L. Brown Before the Senate Commit-tee on commerce, Saence and Transportation, March 1982.

‘sInterview, March 1984.

Ch. 4—Effects of Deregulation and Divestiture on Research • 129

science. It is highly unlikely that AT&T willabandon research in these areas. Further, aswas pointed out in the case studies in chapter3, the boundaries between basic and appliedresearch in these fields are sometimes veryfuzzy. Bell Labs’ researchers are likely to makesome contributions to the advancement ofscience even in pursuit of commercial ends.

There are some tangible and intangible ben-efits of performing basic research that are asadvantageous to AT&T now as they were be-fore Computer II or divestiture. For example,Nelson49 points out that research often yieldsdiscoveries and inventions in unexpectedareas. The wider a firm’s scope of activities,the higher the proportion of these unantici-pated outcomes it will be able to use. Thus,diversified firms can realize higher rates of re-turn from research, and engage in more of itthan firms with narrow product lines. Prior todivestiture, AT&T was a vertically integratedfirm which could make use of research resultsin a large number of areas. Although the sizeof the firm is now reduced, AT&T is now ina position to diversify in other areas, and willcontinue to benefit from research results.

An additional benefit to funding basic re-search is that a reputation for achievementsin basic science gives the Labs a certain pres-tige, credibility, and glamour, even if its chiefbusiness is not basic research. Although thesebenefits are not quantifiable, they are usefulin attracting qualified scientists and engineers.

Divestiture and Computer II changed therules under which AT&T funds research, andthere has been speculation that the new rulesmay bring about a reduction in the amountspent on research over the long term. As arate-base regulated monopoly, AT&T was ableto spread the costs of basic research overmany ratepayers. The license contract fee wasessentially a “tax” on telephone calls. Therevenues generated provided a regular sourceof income that could be counted on year after

‘gRichard R. Nelson, “The Simple Economics of Basic Scien-tific Research, ” Journal of PoliticaJ Economy 67, 3 (June): pp.297-306.

year.50 AT&T was free to use those funds muchas a government might use tax revenues, al-locating some portion of those revenues tosupport activities that were for the generalgood but provided no immediate commercialbenefit. While some research eventually paidoff in discoveries useful to AT&T, some neverpaid off at all. Many research results that werenot of direct benefit to AT&T were made avail-able to others through licenses of patents orthrough scientific and technical publication.

As a competitive firm, AT&T must now sup-port its research through a different internalfunding mechanism. An important aspect ofthe deregulation and divestiture rules is thatAT&T will be watched closely by FCC to besure that it allocates a reasonable portion ofresearch costs to the nonregulated portion ofits business. Before divestiture, most of thecost of research was paid for by the Bell oper-ating companies and the Long Lines division.Under the new “composite allocator” devel-oped by AT&T and approved by FCC, approx-imately 50 percent of research costs will bepaid by AT&T Communications and 50 per-cent will be paid by AT&T Technologies andAT&T Information Systems.

Role of Bell CommunicationsResearch, Inc.

Another unknown factor in the future oftelecommunications research is the role of BellCommunication Research, Inc. (Bellcore), thetechnical services organization owned by theregional holding companies. The scope andquality of Bellcore’s research effort is still un-known. One of Bellcore’s jobs will be to testand evaluate products and equipment for theBell operating companies. In order to do thisproperly, Bellcore will have to stay ahead ofthe manufacturers, anticipating the state-of-the-art and doing some basic research. Accord-ing to Alan G. Chynoweth, Vice President for

‘OAT&T points out that the license contract payments werenot completely guaranteed income. Occasionally a Stab regu-latory comrnission would disallow a portion of a BOC’S licensecontract payment.


Applied Research, “Everything we do will bechosen because of its relevance to the long-term needs of the telephone companies. We’resmaller than Bell Labs. We have to be moreselective. But in those areas we select to beexpert in, we’ll dig very deeply. ”51 Among theareas where research will be done are mathe-matics and computer science, materials, solid-state science, fiber optics and photonics, andswitches.

Nearly half of Bellcore’s technical person-nel came from Bell Labs. To the extent thatformer Bell Labs research personnel will stillpursue the same sorts of problems at Bellcore,the value of their research contributions hasnot been lost to the Nation. It remains to beseen whether the creation of Bellcore will havea positive or negative impact on basic researchin areas related to information technology. Theresearch agendas of Bellcore and Bell Labs willnaturally overlap in certain areas. It is possi-ble that this duplication of effort will be inef-ficient and may reduce the quality of U.S. re-search in information technology overall. Onthe other hand, it maybe that the creation ofthis new center of initiative will have a stim-ulating effect on research.

The regional Bell operating companies arethe owners of Bellcore and have control overhow funds are spent. Under the current ar-rangement, they all contribute to certain“core” projects, but each is able to limit itsinvestment in “noncore” projects it does notbelieve to be beneficial to its own business.52

Funding priorities for Bellcore will dependpartially on actions of State regulatory com-missions. Before divestiture, a few State com-missions sometimes disallowed part of a Belloperating company’s payment for support ofBell Labs on the grounds that research did notbenefit the telephone ratepayers of that State.Support of research at Bellcore may face thesame sort of problem.

The growing competition among its ownersmay also affect Bellcore’s future. Althoughthey provide regulated telephone service onlywithin their assigned geographic areas, the re-

5’Lee Dembart, “Dividing Bell Labs: Breakup to Put the Bestto New Test, ” Los Angeles Times, Sept. 6, 1983, pp. 1,3.

‘*Remarks of Irwin Dorros at seminar “Research at Bell” heldApr. 5, 1984, Massachusetts Institute of Technology, Programon Research in Communications Policy.

gional operating companies are creating sub-sidiaries to enter other lines of business.Among the enterprises already under way arecomputer sales and repair, computer softwaresales, office equipment sales, cable televisioninstallation, and real estate development. Inmany cases the regionals are providing goodsand services in nationwide markets, in directcompetition with one or more of the others.Other ventures are being planned, subject toapproval by Judge Greene’s court, under termsof the divestiture.

The regional Bell operating companies havemany common R&D goals because the majori-ty of their business will continue to be the pro-vision of regulated local and interstate tele-phone service. However, there are a growingnumber of areas where their interests divergeor where one company wishes to withholdinformation from some or all of the others.Bellcore is still developing an organizationalstructure to deal with this situation. It is pos-sible that the growing competition betweenthe owners could encourage them to jointlyfund basic research at Bellcore but to turn toother labs for development of products neededfor the competitive market. At this point it is im-possible to say what Bellcore’s long-term re-search agenda will be. Bellcore will be an in-teresting experiment in jointly funded R&D.It remains to be seen how much of Bellcore’sresources the regional Bell operating compa-nies will be willing or able to spend on basicresearch with possible long-term payoffs.

Availability of Research Results

Even if Bell Labs continues to perform re-search at the current levels, it has fewer in-centives to make the results available to others.It has maintained a fairly open policy, en-couraging its scientists to publish results,present papers, and consult informally withother researchers. Some of the research re-sults, as well as some of the patents, were oflittle direct value to AT&T because it was per-mitted only to provide regulated common car-rier service. But some of them were of im-mense value to firms in related fields and evento AT&T’s competitors.

Now, according to Bell Labs Vice PresidentArno Penzias, AT&T will “have the opportu-nity and motive to use our own technology. ”


Photo credit AT&T Bell Laboratories

An experimental, interactive computer-based systemis helping Bell Labs engineers design

integrated circuits.

However, he emphasized that in the area ofbasic research, Bell Labs is still part of thescientific and technical community where com-munication and trading of information is vital.In order to benefit from the results of researchelsewhere, it will have to continue to share itsresearch results. In order to keep good scien-tists on the staff, it will have to allow themto publish.

Only a small number of basic research proj-ects lead to results that have an obvious ap-plication. In some of those cases AT&T wouldprobably get patent protection before pub-lishing the results. In other cases, the pub-lished paper may report a discovery withoutgiving details of how to duplicate it. This typeof protection has been used by many labs, in-cluding Bell Labs in the past. Penzias notedthat at IBM the number of papers publishedper dollar of research is about the same as BellLabs, even though IBM is a competitivefirm.53

A policy of complete openness of researchresults may be transferred to Bellcore. Its in-terest is to see that research results are dis-seminated widely so that manufacturers canuse them to produce the best and lowest costproducts for use by the Bell operating com-panies. Bellcore itself, under its current char-ter, will not be able to manufacture productsor otherwise benefit from any discoveries orinventions resulting from its research. There-fore, it may establish a publication and licens-ing policy even more open than Bell Labs’ hasbeen in the past.

‘9Arno Penzias, remarks at a seminar “Research at Bell, ” heldApr. 5, 1984, Massachusetts institute of Technology, Programon Research in Communication Policy.

Policy Implications

The recent divestiture and the entry ofAT&T into competitive markets poses newchallenges for U.S. policy toward the telecom-munication and information industries. The1979-83 period in which the divestiture andComputer II decisions were announced andimplemented was also a period of intensivecongressional debate about telecommunica-tions. Bills have been introduced to modify theCommunications Act of 1934, to deregulateparts of the industry, or to force some versionof AT&T divestiture.64 Many of the policy is-sues raised in this legislative debate have nowbeen addressed by FCC in Computer II andthrough settlement of the Department of Jus-tice suit. Speculation over the effects of the

“For example, S. 898, H.R. 5158, as introduced in the 97thCongress, are only two bills which proposed modifying the 1956consent decreq creating a subsidiary of AT&T to enter new un-regulated markets, and stimulating competition in terminalequipment.

new policies have added to the uncertainty andchange in the information industry. Severalyears under the new rules will be necessarybefore all the effects can be assessed.

Similarly, the full effects of deregulation anddivestiture on the quality and direction of re-search at Bell Labs will only become clear asthis “shakedown” period goes on. Neither his-tory nor economic theory seem to be of muchhelp in foreseeing the future of research at BellLabs. There appear to be only a few thingsthat government can do about major changesin research at Bell Labs. Clearly, in the post-divestiture era, decisions about the fundingand nature of research will be in the hands ofAT&T management. This is not new. Deci-sions about research have always been man-agement decisions, in AT&T as throughoutU.S. industry.


It is possible that, in the new climate cre-ated by deregulation and divestiture, AT&Tmanagement will make decisions about re-search that will allow the quality or quantityof Bell Labs’ research to decline. In that case,there may be a role for limited government ac-tion. Some regulatory or funding policiesmight be developed to stimulate or facilitateresearch. These policy changes, discussed laterin this section, might be aimed at AT&T alone,but might also be applied more generically toraise the quality of research in industry, uni-versities, and government.

However, it would be premature to intro-duce policy changes without evidence that thecurrent institutional arrangements are inade-quate, or that the U.S. research capability isin jeopardy. The first step of government ac-tion might be to monitor Bell Labs’ researchover the next several years to see whether thequality of research actually changes. Themonitoring effort might be expanded to in-clude the whole range of industry and univer-sity research in information technology. Itwould not be difficult to develop an analyti-cal framework and a set of criteria for meas-uring the vigor or quality of research. A num-ber of possible measures are suggested below.While none of them is decisive in isolation, to-gether they might give a picture of the healthof research at Bell Labs and at other researchorganizations. 65

For example, it would be possible to moni-tor the funds that AT&T allocates to researchover the next few years. Dollar amounts seemvery objective and quantifiable, but alone arenot a sufficient gauge of the quality or direc-tion of research effort. For example, if all basicscience were dropped and the research effortsteered toward more applied projects, the totalamount spent for “research” might remain thesame. This criterion may be useful, but can-not be used in isolation.

Another measure would be the number ofpapers by Bell Labs scientists published inprestigious scientific and technical journalseach year. A decline in the number of paperscould be an indication that the amount of re-search is declining, perhaps, or that AT&T issignificantly limiting publication in order to

“Some of the measures listed have actually been used infor-mally by Bell Labs management to monitor the strength of re-search in the Labs.

protect possible commercial advantage stem-ming from certain types of research.

In addition to monitoring the number of pa-pers published, it might also be possible to ex-amine the quality of the journals in which theyappear. Although this measure is subjective,it should reflect the quality of Bell Labs workas viewed by other members of the scientificcommunity. Researchers in all fields have aclear idea which of their journals is the “best.”To the extent that Bell Labs work continuesto be published in the same sorts of journalsas now, it may be evidence that the qualityof results remains unchanged. A shift to pub-lication in less prestigious journals might in-dicate a decline in quality.

The vigor of research can also be measuredby its usefulness to other researchers. Thus,a possible measure of the continuing value ofBell Labs research would be the number oftimes their work is cited in papers publishedby scientists in universities and other labs. Inaddition, the attitude of the scientific commu-nity toward Bell Labs could be monitored byits ability to attract and retain well-qualifiedresearch workers.

There is the possibility that, even with a re-duction of basic research at Bell Labs, researchin the information field generally will not suf-fer. Scientific research may simply move toother laboratories. Any decline in quality offundamental research would certainly make itharder for Bell Labs to attract and keep a staffof qualified scientists. Top graduate studentswould choose to work at other firms or at uni-versities. To the extent that researchers con-tinue to work in the same scientific fields andare equally productive in their new surround-ings, there may be no noticeable effect on U.S.basic research.

To get a complete picture of the effects ofderegulation and divestiture on the state of in-formation technology basic research in theUnited States as a whole, it would be neces-sary to monitor the research performed through-out industry and at university labs as well.Even then, it would be extremely difficult toattribute observed changes in the U.S. re-search environment to changes occurring atBell Labs. As noted in chapter 2, the state ofinformation technology research is in flux andchanges will occur with or without Bell Labs.

Ch. 4—Effects of Deregulation and Divestiture on Research . 133

It may be possible, however, to trace somecausal factors and to, at least, draw reason-able inferences.

One difficulty with the proposed studies isthe collection of relevant data over an ex-tended period of time. While the data neededare not extensive, they include items thatfirms do not currently report to any Federalagency (except that the FCC will continue tobe concerned with AT&T’s research budget).Some special effort and cooperation on thepart of industry and the Federal Governmentwould be needed to collect and analyze the nec-essary information.

It is difficult to say who might be bestsuited to carry out the studies mentionedabove. One possibility is the FCC, which is re-sponsible for oversight of many aspects ofAT&T’s business. However, many of the firmsand institutions engaging on information tech-nology research are not regulated by the FCCand it may not be appropriate for the Commis-sion to study them. Other possibilities mightbe the National Telecommunications and In-formation Administration (NTIA), which hasan interest in the health of U.S. informationR&D, or the National Science Foundation(NSF), which monitors the state of R&D andbasic research in a number of fields. Yetanother possibility might be an independentresearch group outside of government—per-haps one created by a university or industryassociation. Most of the technical and scien-tific journals needed for bibliometric studiesare already in the database of the Library ofCongress. These analyses might be performedby the Congressional Research Service, an in-dependent research group, or one of the agen-cies mentioned above.

For some of the studies mentioned abovedata may only be available several years afterthe actual research has been done, and in manycases meaningful conclusions can be drawnonly after data for 5 or 10 years have been ana-lyzed. If there is a reduction in basic researchat Bell Labs, the trend might have been underway for several years before the data indicatea change. By that time, it might be difficultto effect any correction in the trend.

If it is determined that changes in basic re-search at AT&T have had a major effect onthe U.S. research capability, and that Govern-ment action is warranted, there is a questionof what can be done. Basically, it appears thattwo general approaches might be considered.Regulatory policies might be changed to mod-ify the rules under which AT&T operates, giv-ing it greater incentives to perform basic re-search or requiring it to do so. More broadly,consideration could be given to implementingfunding policies that might stimulate more ba-sic research throughout industry and in uni-versity laboratories.

In the regulatory area, for example, it mightbe possible to allow some subsidy for basic re-search. At the present time, FCC is workingunder the terms of the divestiture and Com-puter II to make sure that AT&T does not usethe revenues it earns in the regulated marketto support research or development that leadsto advantages in the nonregulated market. Forthis reason FCC must approve the “compos-ite allocator” developed by AT&T to allocateresearch costs among the various AT&T en-tities. The economic theory is that a competi-tive firm should pay its own R&D costs with-out shifting them to regulated ratepayers.

On the other hand, when AT&T was per-mitted to use such a cross-funding arrange-ment, it used the funds to create a highlyrespected and productive research organiza-tion that presumably benefited the Nation asa whole. If experience over the next few yearsshows that it is impossible for AT&T to main-tain Bell Labs’ quality without additionalfunding, and if it is determined to be in thenational interest that such an organization bemaintained, then additional funds must beprovided. They could come from a direct Fed-eral subsidy or from some kind of cross-fund-ing. The former is not likely to be politicallyacceptable; the later is increasingly complexas the long-distance telecommunication mar-ket becomes more competitive. By its ownestimate, AT&T now provides only 69 percentof total long-distance capacity.56 AT&T israpidly losing the market power it once had

“AT&T, private communication, Apr. 30, 1984.

134 ● information Technology R&D: Critical Trends and Issues

to control prices throughout the industry. Ifit were required to raise the price of long-distance calls to provide greater support forbasic research, it would be placed at a com-petitive disadvantage with respect to otherlong-distance carriers that do not support re-search. Development of a mechanism by whichall long distance carriers contribute to fund-ing basic research would be difficult in the in-creasingly competitive long distance market.

An alternative regulatory approach mightbe to stimulate basic research by allowingmore cooperation between ATTIS and BellLabs. About 4,000 former Bell Labs employ-ees were moved to ATTIS when it was createdin 1982. Expertise in some research areas hasbeen lost to Bell Labs through this transferand through the subsequent transfer of 3,000employees to Bell Communications Research.The FCC’s interpretation of Computer II rulesdo not allow the exchange of market and net-work information between ATTIS and BellLabs and they also prohibit the joint develop-ment of certain products, especially computersoftware.

In the future, easing this requirement to theextent of allowing ATTIS and Bell Labs to co-operate on certain types of research, mightallow greater cross-fertilization among the tworesearch organizations.

Policies to stimulate basic research gener-ally might include such incentives as addi-tional grant support from National Science

Chapter 4

“Bell Labs on the Brink, ” Science, Sept. 23, 1983,pp. 1267-1269.

“Bell Labs: The Threatened Star of U.S. Re-search, ” Fortune, July 5, 1982, pp. 46-52.

Brock, Gerald W., The Telecommunications Indus-try: The Dynann”cs of Market Structure (Cam-bridge: Harvard University Press, 1981).

“Competing with the New AT&T,” Venture, Jan-UWY 1984, pp. 54-57.

Charles River Associates, Impacts of the AT&TDivestiture on Innovative Behavior, unpub-lished paper prepared for Office of TechnologyAssessment, 1983.

Foundation (for example), direct support ofbasic research through direct Federal sub-sidies, or tax incentives for industries thatengage in basic research. Such policies mightbe applied not only to Bell Labs, of course, butto Bellcore or to other university and indus-try research organizations. Over the next fewyears, while the health of research is beingmonitored, it might be possible to structuresuch programs to stimulate research and todevelop “trigger” mechanisms for puttingthem into place if results indicate that thequality of research is declining.

In conclusion, it is still too early to tellwhether the quality or direction of research atBell Labs will be adversely affected by deregu-lation and divestiture, or whether any changesin its research would have major repercussionsfor U.S. research as a whole. There are severalpossible measures for monitoring the healthof basic research at Bell Labs, but evidenceof change may not be apparent for severalyears. While Bell Labs is a major contributorto the sciences related to information technol-ogy, it is not the only important player. Togain a true picture of the effects of deregula-tion and divestiture it would be worthwhile toexpand such studies to monitor the state ofbasic research throughout industry and atuniversity labs as well. It will be importantto begin collecting information soon in orderto fully document the transition from the pre-to post-deregulation environment.

References

Dembart, Lee, “Dividing Bell Labs: Breakup toPut the Best to New Test,” Los Angeles Times,Sept. 6, 1983, pp. 1,3.

Hall, Peter, “AT&T and the Great Divide, ” Finan-cial World, Jan. 10, 1984, pp. XX.

Harris, Marilyn A., “Bell Labs Looks to MilitaryResearch,” Electrom”cs, Feb. 9, 1984, pp.102-104.

Kamien, Morton I., and Nancy Schwartz, MarketStructure and Innovation (New York: Cam-bridge University Press, 1982).

Morkre, Morris E., “Innovation and Market Struc-ture: A Survey, ” Working Paper No. 82. Fed-


eral Trade Commission, Bureau of Economics,1982.

National Science Board, Science Indicators, 1982(Washington, DC: U.S. Government PrintingOffice, 1983).

National Science Foundation, Research and Devel-opment in Industry, 1981, NSF 83-325 (Wash-ington, DC: U.S. Government Printing Office,1983), p. 3.

Scherer, F. M., “Firm Sizes, Market Structure, Op-portunity, and the Output of Patented Inven-tions, ” American Economics Review, 55:1119.

Scherer, F. M., Industrial Market Structure andEconomic Performance (New York: Rand Mc-Nally, 1980).

Uttal, Bro, “Cold New World, ” Fortune, June 27,1983, pp. 81-84.

“Why AT&T Will Lose More Long Distance Busi-ness, ” Business Week, Feb. 13, 1984, pp. 102-104, 110.

Wiegner, Kathleen K., “Prometheus is Unboundand Seeking His Footing, ” Forbes, Mar. 12,1984, PP. 141-148.

38-802 0 - 85 - 10

Chapter 5

Education and Human Resourcesfor Research and Development

Contents

PageFindings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

The Concern About Manpower . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

The Relationship Between Manpower Development and Economic Growth . . . . . . . . . 140

Identifying Particular Manpower Problems and Solutions . . . . . . . . . . . . . . . . . . . . . . . . 141

The Range of Manpower Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145

The Supply of Manpower. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

The Problems in Higher Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

Elementary and Secondary Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

The Federal Role in Manpower Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

The Societal Context for Determining Federal Manpower Policy . . . . . . . . . . . . . . . . . . 161

Chapter 5 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

TablesTable No. Page2A. Scientists and Engineers Engaged in R&D Per Labor Force

Population, By Country: 1963-82 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14525. BLS Manpower Estimates.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146

FiguresFigureNo. Page20.21.22.23.

24.25.26.

27.28.29.

Employed Scientists and Engineers by Field, 1981 . . . . . . . . . . . . . . . . . . . . . . . . . . . 143Distribution of Scientists and Engineers by Primary Work Activity, 1981 . . . . . . . 144Percentage of Firms Reporting Available Jobs for Scientists and Engineers . . . . . 148Comparison of Growth in Engineering UndergraduateEnrollment and Number of Faculty, 1973-80 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150Share of All S/Es Employed in Educational Institutions by Field, 1981 . . . . . . . . . 151Educational Gifts by Computer Vendors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153Selected Indicators of the Overall Quality of Mathematicsand Science Education in the United States. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154Americans Science Teacher Shortage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156Selected Indicators of Shifts in Undergraduate Science/Engineering Education . . . 157A Quarter Century of Student Aid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

Chapter 5

Education and Human Resourcesfor Research and Development

Findi

OTA found that major Federal actions de-signed to affect the supply of manpower toperform research and development (R&D) inthe area of information technology would ap-pear to be unwarranted at this time. Forecastsof future manpower needs in this area are re-plete with uncertainty. Moreover, developingmanpower for the specific areas where poten-tial shortages might tentatively be predictedwould be particularly hard to accomplishthrough broad Federal actions. The educa-tional backgrounds and skills required to meetthese potential shortages are at once both toobroad and too narrow to be developed at theFederal level. In addition, given the length oftime required to develop skills and the rapidchanges taking place in the area of informa-

ngs

tion technology, Federal action, taken now,might prove to be inappropriate in the future.

A number of legislative proposals have beenmade that are designed to increase the futuresupply of highly qualified scientific and tech-nical manpower. These proposals differ con-siderably in terms of their goals, their targets,their costs, and their scopes. Given the highlevels of uncertainty that surround the presentmanpower debate and the number of compet-ing uses to which the Nation’s limited educa-tional resources might be profitably put, themost prudent course might be to adopt thosepolicies that would provide for the greatestamount of flexibility and the broadest rangeof skills.

The Concern About Manpower

In the United States today, there is a grow-ing and widespread belief that the Nation’spoor economic performance is inextricablylinked to the relative decline in the size andthe quality of its technical work force. Notingthat Japan and West Germany, our major in-ternational competitors, have four times asmany electrical engineers and computer scien-tists, per capita, as the United States, manyof the people who hold this view fear that, asthe economies of the developed world becomemore technologically intensive, and thus asR&D becomes more critical to their success,the United States will increasingly lose itsability to compete. Typical of this perspectiveis the statement made by Representative Mar-garet Heckler during hearings on EngineeringManpower Concerns, when she said:1

To maintain its technological edge in worldmarkets the United States must reemphasizescience and engineering on our agenda of na-tional priorities. When the Soviets launchedSputnik I, a remarkable engineering accom-plishment, the United States rose to the chal-lenge with new dedication to science andtechnology. Today, our technology lead isagain being challenged, not just by the Sovi-et Union, but by Japan, West Germany, andothers.

The negative consequences of having ashortage of manpower in information technol-ogy R&D, it is argued, may be particularlysevere. Given the speed with which the fieldis changing, even a temporary shortage mightimpair the ability of information technologyindustries to remain at the frontiers of re-

‘Representative Margaret Heckler, Opening Statement, Hear-ings On Manpower Concerns, before the Committee on Scienceand Technology, House of Representatives, 97th Cong., 1stsess., Oct. 6-7, 1981, p. 4. For a more recent statement of thisperspective, see also, Hearings on Mathematics and %“ence Ed-

ucation, before the Committee on Education and Labor, Houseof Representatives, 98th Cong., 1st sess., Jan. 26-28, 31, 1983;see also, Amen”ct-i Competitive Challen~ Report of Business-Higher Education Forum, 1983.

139


search. And, because the information technol-ogy industry represents the fastest growthsector of the economy, failure to keep pace inthis industry may have serious consequencesfor the Nation’s economy as a whole.

Concerned about the state of available hu-man resources for high-technology jobs, spokes-men from business, government, and educa-tion have called on the Federal Governmentto undertake a number of significant educa-tional measures and reforms. These measuresrange widely in terms of goals, targets, costs,and scope. Some of them, for instance, focuson a specific curriculum area, such as mathand science; others emphasize educationalinfrastructure—the training of teachers andthe need for equipment; while still others seekto foster new modes of cooperation betweenbusiness, government, and educational insti-tutions. Given our Nation’s limited resourcesand the growing number of demands andstresses that are being placed on our educa-tional system at all levels, choices and deci-sions will have to be made about which goalsto pursue and about which measures to adopt.

Notwithstanding the widespread discussionand concern about the poor state of the Na-

tion’s manpower resources, there has beenvery little systematic effort to clearly identifyand characterize the nature and the extent ofthe problems. Before making any major poli-cy decisions designed to affect the supply ofmanpower, therefore, it will be necessary tohave a greater understanding of: 1) what weknow and don’t know about the relationshipbetween manpower development and econom-ic growth; 2) what we know and don’t knowabout this relationship as it relates in particu-lar to R&Din information technologies; 3) therange of projections about the future supplyof and demand for manpower in this field; 4)the ability of the present institutional struc-ture to accommodate these manpower needs;5) the role of the Federal Government in thedevelopment of manpower; 6) the societal con-text in which, today, decisions about educa-tion will be made; and 7) the range of Federalalternative strategies and options for meetingfuture manpower needs in the area of informa-tion technology R&D. The following discus-sion provides a preliminary basis for such anunderstanding.

The Relationship Between ManpowerDevelopment and Economic Growth

The assumption of a positive relationship be-tween the size and quality of a nation’s workforce and its economic wealth is not a new one.Over 100 years ago, for example, the BritishGovernment sponsored a parliamentary com-mittee to investigate the causes of rapid in-dustrial growth in the United States. Likemany of our recent studies of economic growthin Japan, the British parliamentary commit-tee attributed much of America’s industrialsuccess to the superior education of the Amer-ican worker.2

‘Report Fmm the Select Committee on Su”enti”fi”c Instruction,Parliamentary Papers, 15, (1867-1868) Q 6722, as cited in Wil-liam Abernathy, Kim B. Clark, and Alan Kantrow, Industzz”alRemissanm: l%niucing a Competitive Future for Amen”ca (NewYork: Basic Books, 1983).

Indeed, ever since the beginning of the in-dustrial revolution, economists and other so-cial observers have argued that a skilled andeducated work force is the most productive.Writing as early as 1776, Adam Smith, for ex-ample, pointed out that “the skill, dexterity,and judgment with which it’s [the nation’s] la-bor is generally applied,” is the primary fac-tor determining the size of “the fund whichoriginally supplies it with all the necessitiesof life. ”3

— — —‘Adam Smith, The Wealth of iVati”ons (New York: The Mod-

em Library, 1937), p. lviv.As modem societies became more technologically ad-

vanced, an increasing amount of attention was paid tothe development of the labor force. Anticipating the ef-fect that technology would have on society, the Germansociologist, Max Weber, pointed out for example, that,in an advanced industrial society, the organization ofhuman relations could no longer be left to chance. In-stead, human beings become factors of production-their

Ch. 5—Education and Human Resources for Research and Development . 141

Present government policies designed to af-fect the supply of manpower are also based onthis assumption. However, as the followingdiscussion illustrates, while we can identifysome general linkages between education,manpower, and economic development, our un-derstanding of causal relationships, or of rela-tionships in specific situations, is extremelylimited.

While acknowledging that having qualifiedmanpower is critical to the success of a na-tion’s economy, social and economic analystsare still unable to fully account for, or to com-pletely explain, the nature of the relationshipbetween the size and the skill level of the la-bor force and economic growth and develop-ment. As the economist, Nathan Rosenberg,has noted:4

———.—..relationships w be structured in accordance with therequirements of industrial progress. And the Americaneconomist, Thorstin Veblen, writing in the 1930s, wentso far in his discussions of technology and society as tosuggest that, for technology to develop to its full poten-tial, the technical expert-the engineer-wodd have toplay a key role in society’s decisionmaking process. JayWeinstein, Sociolo@I’echnology: Foundations of PostAcademz”c fi”ence, Transaction Books, 1982, p. 32;Thorsten Veblen, The Theoxy of the Leisure Class (NewYork: The Modem Library, 1934).

4Nathan Rosenberg, Inside the Black Box–Technology andEconorm”cs (Cambridge, MA: Cambridge University Press,1982), p. 8.

One of the central historical questions con-cerning technical progress is its extreme var-iability over time and place. . . . Clearly thereasons for these differences, which are notyet well understood, are tied in numerouscomplex and subtle ways to the functioningof the larger social systems, their institu-tions, values, and incentive structures. Theexplanation of these differences is intimatelytied to such even larger questions as why so-cial change occurs and why economic growthproceeds over time and place.

Not only is each piece of the puzzle difficultto solve; the whole problem is subject to thevagaries of external events, well beyond ouranticipation and calculation. Nor are the toolsof analysis particularly refined. For—althoughdemographers may tell us something aboutpopulation trends; sociologists somethingabout the institutions and processes in whichindividuals are recruited, educated, and trainedfor work; economists something about thepoint at which, and the rate of exchange bywhich, the supply and the demand for laborare brought into a state of equilibrium-his-torians are sure to remind us that it is, moreoften than not, a unique set of circumstancesthat has had the most significant effect on aparticular outcome.

Identifying Particular Manpower Problems and Solutions

Our limited knowledge of the role of humanresources in economic growth and technologi-cal change is clearly evident in our efforts toidentify and analyze specific manpower prob-lems and solutions. For although manpowerspecialists might agree that having sufficientqualified manpower is critical to a nation’seconomy, they do not necessarily agree aboutthe number of people who are required to meetthe employment needs of a particular sector;about the kinds of skills and experience thatmight be required to perform particular kindsof jobs; or about the way in which these skillsmight best be obtained or developed.5

6National Institute of Education, Education, Productivity,and the National Economy, A Research Iru”tiative, December1981; see also Edwin Mansfield, Education, R&D, and Produc-tiw”ty Growth, revised, University of Pennsylvania, Jan. 31,1982.

To identify future manpower needs in a par-ticular area, policymakers have traditionallyrelied on economic and other forecasting meth-odologies. While useful as policymaking tools,these methodologies are subject to a numberof problems and weaknesses which stem,among other things, from imperfect data,weak forecasting models, and ill-founded as-sumptions. 6 To be most useful, forecastingmethods need to be flexible and responsive.Acknowledgment should be made of the limi-tations of these methodologies, and effortsshould be undertaken to verify their resultsby conducting frequent surveys and by per-forming case studies designed to determine

6R. H. Bezdek, Long-Range Forecasting of Manpower Re-qu.z”rements (New York: Institute of Electrical and ElectronicsEngineers, 1974).


the changing skill requirements that are linkedto the emergence of new technologies.

It should be noted, moreover, that manpow-er forecasts can themselves produce a pendu-lum effect, underminingg the validity of the pro-jections. This effect results from both the longperiod of time that it takes for people to pre-pare for afield of work, and from the fact that,once committed to a career path, people rarelychange their plans in midstream to adapt tonew circumstances. Upon hearing predictionsof an impending manpower shortage in a par-ticular field, for example, an inordinate num-ber of students may seek to pursue such a ca-reer, hoping that when they have finished theireducations, jobs will be plentiful and competi-tion will be in short supply. The resulting man-power glut will appear only later; but predic-tions of it may induce a number of studentsto avoid the field, leading to another shortagein the future.

It should also be remembered that man-power predictions can be interpreted differ-ently by different kinds of people. Economistsmight describe a shortage, for example, whenthey see a rapid increase in wages due to a gapbetween the supply and demand for labor.Businessmen might consider that there is ashortage of manpower when they are dissat-isfied with the quality of preparedness of thepool of people from whom they have to selectemployees. New graduates may interpret amanpower shortage to mean that they face lit-tle competition in seeking employment.

The problem of predicting manpower needsin the area of information technology R&D iseven more complicated, because the field isnew and in a rapid state of flux. There is, forexample, very little historical basis for iden-tifying who the people are who might typicallyperform R&D tasks in the area of informationtechnology; what skills they should possess inorder to perform these tasks most effectively;or what their optimum career patterns mightbe.

It is only very recently that either the Na-tional Science Foundation (NSF), the keyagency mandated to monitor the supply and

demand of engineers and scientists in theUnited States, or the Bureau of Labor Statis-tics, have begun to treat computer scientistsas a distinct group. NSF, for example, has onlyrecently stopped labeling everyone who worksin a computer-related field as working in thearea of computer theory, a heading that wasitself a subcategory of mathematics. And,even today, NSF does not list a departmentof computer science under that heading if thedepartment’s name appears in a combinedform and if the words “computer science” ap-pear second in that combination.7 Even whenthe appropriate statistics have been collected,moreover, they have often been subject to avariety of interpretations.8

It is also difficult to determine not only whoor how many people are working on or withthese technologies but also who or how manypeople are performing specific R&D tasks inthis area. NSF estimates that in 1981,3.1 mil-lion scientists and engineers were employedin the United States. Of these, 47 percent wereemployed as engineers (including engineers doing management jobs), and 13 percent wereworking as computer specialists (see fig. 20).9

NSF reports, moreover, that 34 percent of allscientists and engineers are involved in R&Dactivities”10 (see fig. 21). As table 24 illustrates,compared to other countries, this is a high pro-portion of R&D scientists and engineers rela-tive to the total labor force. Figures are notavailable, however, for the percentage of scien-tists and engineers who specifically performR&D tasks in the area of information tech-nology.

Use of aggregated data based on broad skillcategories or outmoded technologies reducesthe value of manpower demand forecasts. Acategory such as computer programmer, forexample, is much too broad to use for forecast-ing manpower demand in R&D. Further sub-

7Kent K. Curtis, “Computer Manpower-Is There a Crisis?”(Washington, DC: National Science Foundation, January 1983).

‘Ibid.‘Su”ence Inch”catms 1982: An Analysis of the State of U.S.

Sa”ence, Engine, and Tdnology, National Science Board,1983, p. 63.

IOIbid., p. 66.


Figure 20.-Employed Scientists and Engineers byField, 1981

Engineers47%

Mathematical

Physicalscientists

8 %

Environmental

Psychologlsts 30/o

40/0

alncludes earth sclentlsts, oceanographers, and atmospheric scientists.NOTE: The total number of scientwts and engineers m 1981 was 3.1 mdlion.

SOURCE: Science Indicators, 1982,

division into types such as entry level, applica-tions, and systems programmers based onexisting job specifications, while useful, willprobably not suffice for long, since the mix ofcomputer-related skills required for R&D is es-pecially sensitive to technological innovation.11

Some of the most important skill categories,for example, lie at the frontiers of informationtechnology, and these do not show up in thebroad categories based on aggregated data.The problem of identifying these skills, there-fore, is not just one of substituting one set ofstatic descriptors for another; it is a problemof gathering information about new skills overtime and in response to changing conditions.

The task of identifying research and devel-opment workers in the field of informationtechnology is complicated, moreover, by thefact that the traditional distinctions that havealways been made between the tasks that are

——. ——llAb~ hfowshowi~, “on Predicting R&D Skill [email protected]

for Information Technology,” paper prepared for the Office ofTechnology Assessment, February 1984.

entailed in R&D and those entailed in produc-tion are becoming increasingly blurred in thisarea. This problem is clearly evident in thecase of software engineering. Calculations offuture manpower needs that focus specificallyon R&D activities are, therefore, particularlydifficult to make in the area of informationtechnologies.

Questions also arise with respect to how theskilled technical workers, who provide supportto the R&D process but who do not performthe most highly skilled tasks, might best befactored into manpower projections. Theseworkers might include, for example, all ofthose who maintain, troubleshoot, repair, andsometimes fabricate sophisticated equipment,those who build and help develop prototypesof new products, draftsmen and nondegree de-signers, computer system operators, and tech-nical writers. Since the skills that these work-ers require may be more easily obtained ormay be more easily substituted—either byother workers or by technology-than theskills required for the more highly technicaljobs, the manpower projections for this sec-tor of the R&D process, and the policy impli-cations that might be drawn from them, mayalso be quite distinct. Policies designed to af-fect the availability of skilled technical work-ers, for example, might call for some generaleducational changes at the elementary and/orsecondary levels whereas those that are de-signed to influence the supply of manpowerin a highly technical area such as artificial in-telligence, or software engineering, might callfor targeted incentives at the university orgraduate studies level.

Defining R&D manpower and estimatingmanpower needs for this area becomes evenmore troublesome the further one looks intothe future. Over the long run the future de-mand for manpower, for example, is likely todepend on the extent to which and the speedat which the new technologies are deployedthroughout society. However, the rate and de-gree of their deployment will depend, in turn,on the kinds of social variables that are mostoften left out of forecasting models, and thatare the most difficult to predict.


Figure 21 .–Distribution of Scientists and Engineers by Primary Work Activity, 1981

and computing, consulting, other, and

SOURCE: Science Indicators,

Ch. 5—Education and Human Resources for Research and Development ● 145

Table 24.–Scientists and Engineersa Engaged in R&D Per Labor Force Population, By Country: 1963.82

Country 1965 1968 1972 1975 1979 1982— —.S/Es a engaged in R&D per 10,000 labor force population

France . . . . . . . . . . . . . . . . . . . . 21.0 26.4 28.1 29.3 31.6 NAWest Germany. . . . . . . . . . . . . . 22.7 26.2 36.0 41.0 47.7 NAJapan . . . . . . . . . . . . . . . . . . . . . 24.6 31.2 38.1 47.9 50.4 NAUnited Kingdom . . . . . . . . . . . . 19.6 20.8 30.4 31.2 33.2 b NAUnited States. . . . . . . . . . . . . . . 64.1 66.9 57.9 55.5 57.9 63.8U.S.S.R. (lowest) . . . . . . . . . . . . 44.8 53.5 66,5 78.2 84.4 89.8U.S.S.R. (highest) . . . . . . . . . . . 48.2 58.8 73.2 87.5 95.5 102.4

S/Es a engaged in R&D (in thousands)France . . . . . . . . . . . . . . . . . . . . 42.8 54.7 61.2 65.3 72.9 NAWest Germany. . . . . . . . . . . . . . 61.0 68.0 96.0 103.9 122.0 NAJapan . . . . . . . . . . . . . . . . . . . . . 117.6 157.6 198.1 255.2 281.9 NAUnited Kingdom . . . . . . . . . . . . 49.9 52,8 76.7 80.5 87.7 b NAUnited States. . . . . . . . . . . . . . . 494.5 550.4 518.3 532.7 620.2 716.9U.S.S.R. (lowest) . . . . . . . . . . . . 521.8 650.8 862.5 1,061.8 1,216.4 1,340.4U.S.S.R. (highest) . . . . . . . . . . . 561.4 715.2 950.1 1,187.6 1,377.4 1,340.4

Total Iabor force (in thousands)

France . . . . . . . . . . . . . . . . . . . . 20,381 20,744 21,817 22,310 23,059 NAWest Germany . . . . . . . . . . . . . . 26,887 25,968 26,655 25,323 25,573 NAJapan . . . . . . . . . . . . . . . . . . . . . 47,870 50,610 51,940 53,230 55,960 NAUnited Kingdom . . . . . . . . . . . . 25,498 25,378 25,195 25,798 26,464 NAUnited States. . . . . . . . . . . . . . . 77,178 82,272 89,483 95,955 107,050 112,383U.S.S.R . . . . . . . . . . . . . . . . . . . . . 116,494 121,716 129,722 135,767 144,201 149,215alncludes all scientists and engineers engaged in R&Don a full-time equivalent basis (except for Japan whose data include persons Primarily employed in R&D ex.ciudlng soc!al scientists, and the United Kingdom whose data include only the Government and industry sectors)

blg78NA—Not available.

SOURCE National Science Foundallon,Sc~ertce /ndicators 1982

Future manpower requirements in software changes or their future impacts is extremelyengineering, for example, might be significant- difficult, given the newnessofthe field and thely reduced as new software tools are developed fact that they are dependent on a number ofand introduced, if new institutional practices social variables-e.g., the willingness of indi-are adopted to improve the efficiency of those viduals and institutions to both adopt andworking in the area, and if more and more ap- adapt to technological changes-variablesplacations software are developed on personal that are themselves notoriously unpredictable.computers by end users. 12 Predicting these

“Ibid.

The Range of Manpower Predictions

Given the problems involved in identifyingfuture manpower requirements, it is not sur-prising that there has been considerable con-troversy over and discrepancy between manyof the projections that have been made aboutthe need for high-technology manpower. Amongthose making forecasts, the consensus hasbeen the greatest with respect to projected

shortages of Ph.D.s to teach at the universitylevel in the fields of engineering and computerscience. To a somewhat lesser degree, man-power experts concur that the future growthin demand for computer scientists will be ex-traordinary. They disagree, however, aboutwhether or not the educational system, as itexists today, can effectively respond to meet


that demand. Agreement is lowest with regardto whether or not there will be a future short-age of engineers.

The following summaryprovides some senseof the range of projections. Because theseforecasts are based on different assumptions,methodologies, baselines, and timeframes, itis impossible to compare and contrast themanalytically as a whole. Evaluations abouttheir reliability and accuracy are quite depen-dent, therefore, on judgments about the va-lidity of their methods and assumptions.

Generally speaking, however, it can be saidthat manpower forecasts are more reliable thegreater the number of and the more refined theunderlying analysis. 13 The least sophisticatedforecasting methodology, for example, mightbe one based solely on survey data, or simplyextrapolating trends on the basis of the present.A much more comprehensive approach, on theother hand, might be one that takes into ac-count such things as changes in the relativeprices of capital and labor, and/or that positsa set of alternative assumptions about thefuture. The most ambitious forecasts are thosethat try to factor into their analysis the im-pact of technological change.14

Perhaps the most widely referred to, andamong the more sophisticated projections, arethose that have been put forward by the Bu-reau of Labor Statistics (BLS). These projec-tions cover a period of 10 to 15 years. Theyare not only based on a set of alternative eco-nomic scenarios, positing different rates ofgrowth; to some extent, they also seek to taketechnological change into account. Moreover,BLS has consistently sought to improve itsmethodology by systematically evaluating theaccuracy of its own projections. Such evalua-tions show that BLS has had more success indetermining how technology might effect fu-ture job growth than it has in identifying atwhat point such changes in employment pat-terns might take place. Past projections have,

l~Henry M. hvin and Russell W. Rumberger, me Edu~ti”ozI-al Implications of High Technology, The National Institute ofEducation Report #83-A4, 1983.

141bid.

moreover, tended to exaggerate the growth oftechnical occupations and underestimate thedecline of certain traditional jobs. One usefulmeasure of the accuracy of these projectionsis the recent finding that 60 percent of the1980 forecasts fell within a 10 percent rangeof the actual employment level for that year.15

BLS projects the rate of growth and the fu-ture demand for manpower in given occupa-tional categories. The Bureau’s most recentprojections for those categories most relevantto information technology are listed in table25.16 For each category, there are three pro-jections—high, medium, and low—each corre-sponding to one of the three economic scenar-ios used by BLS in developing their forecasts.

Because it does not make predictions aboutthe future supply of manpower, the Bureau ofLabor Statistics does not predict labor short-ages or labor surpluses per se. However, re-view of the recent articles in the BLS pub-lication, Occupational Outlook Handbook,suggests that there will be a multi-tiered mar-

‘51bid.‘K!onversation with Tom Nardone, Manpower Economist, Bu-

reau of Labor Statistics, Department of Commerce, May 15,1984.

Table 25.—BLS Manpower Estimates

Base year: 1982 1995

Electrical and electronic engineers:320,000 . . . . . . . . . . . . . . . . . . . . . . . LOW 531,000

Moderate 528,000High 540,000

Computer specialists:Programmers:

226,000 . . . . . . . . . . . . . . . . . . . . . . . LOW 465,000Moderate 471,000High 480,000

Systems analysts:254,000 . . . . . . . . . . . . . . . . . . . . . . . LOW 469,000


Technicians:55,000 . . . . . . . . . . . . . . . . . . . . . . . . Low 106,000


Computer operators:211,000 . . . . . . . . . . . . . . . . . . . . , . . Low 366,000

Moderates 371,000High 378,000

SOURCE: Tom Nardone, Manpower Economist, Bureau of Labor Statistics,Department of Commerce, personal communication, May 15, 19S4.


ket for these job categories, with a shortageof people with some specific skills and a sur-plus of those with others.

The National Science Foundation also makesprojections of the supply and demand for sci-entific, engineering, and technical personnel.These projections are developed using a multi-step model, similar to the one used by BLS.Like the BLS model, for example, the NSFmodel tries to anticipate how technologicalchange might affect future manpower needs.In one way, however, the NSF model goes fur-ther than that of BLS: its alternative scenariosposit different levels of defense spending aswell as different levels of economic growth.17

In its recent report, Science Indicators 1982,NSF pointed out that computer specialistsaccounted for almost 45 percent of the totalgrowth of scientific employment over the pe-riod 1976-81. Matching this growth againstthe future supply of computer science person-nel, it predicted a future shortage in this area.Indicators for the future supply and demandof engineers were more mixed, however. Forwhile these indicators revealed a shortage in1981, they also suggested that by mid to late1982, the situation already appeared to beshifting back towards a balance between thesupply of and the demand for engineers ingeneral.

A recent survey conducted for NSF raisessome questions about the degree to whichmanpower shortages may in fact materializein the future, even in the area of computerscience. Nearly one half of the 351 firms sur-veyed, for example, reported fewer openingsfor scientists, engineers, and technicians dur-ing the 1982-83 recruiting year than in the1981-82 period. These results are broken downby area in the following figure18 (fig. 22).

Focusing on manpower needs for defense,the U.S. Air Force, in The Regional Planningand Evaluations Systems (ROPES) Project,forecast manpower needs for 30 States and 70major cities. The States were selected for anal-

‘7Levin and Rumberger, op. cit., pp. 12-13.‘“Ibid.

ysis because they are major centers of defenseactivity. The ROPES study found that theneed for all skill groups involved in “the use,operation and repair of computer equipmentwill grow at an alarming rate throughout the1980s and that these skills will be particularlyaffected by increased expenditures for de-fense. ” Although the Air Force study groupdid not project the supply of manpower forthese areas, they concluded, based on the pro-jected rate of growth in demand, that the fieldof computer science, and perhaps electrical andmechanical engineering, may be areas of po-tential national shortage.

A 1982 study by Betty M. Vetter of the Sci-entific Manpower Commission summarizes in-formation on the present and future supplyand utilization of scientists and engineers inthe United States. Vetter concludes that, ex-cept in the field of computer science, the sup-ply of scientists appears sufficient to fill near-term demands. For engineers, she notes thatthe number of new graduates at the baccalau-reate level has been rising since 1975, but that“a high level of demand has not only fueledthat increase, but has utilized so many engi-neering graduates at the baccalaureate levelthat graduate enrollments of U.S. studentshave not climbed commensurately and short-ages of Ph.D. engineers have become serious,at least at academic institutions. ” She con-cludes, however, that there is no general agree-ment about the adequacy of the future supplyof engineers, even when considering particu-lar specialties.

Industry projections of future manpowerneeds are quite inconsistent with one another.Derived, as they are, by stakeholders, theirconclusions must be regarded with some de-gree of caution.

Basing its conclusions on a survey of 815manufacturing facilities, the American Elec-tronic Association (AEA) predicts that through1987, the need for technical professionals willgrow by 69 percent; while the need for techni-cal paraprofessionals will increase by 60 per-cent. Respondents to their survey estimate,moreover, that in 5 years they will need to


Figure 22. —Percentage of Firms Reporting Available Jobs for Scientists and Engineers

Engineers Scientists(Percent of firms) (Percent of firms)

o 20 40 60 100 0 20 40 60 100

Petroleum

Electronic

Electrical

Computer

Civil

Chemical

Mechanical

Industrial

Metallurgical

Manufacturing

SOURCE: National Science Foundation.

employ more than 100,000 each of new tech-nical professionals and new technical para-professionals. l9 The occupational groups forwhich they foresee a tremendous amount ofgrowth during this period—defined as over100 percent total increase—are software engi-neers, electronic engineering technologists,and computer analysts/programmers.20

To project the future supply of these key oc-cupational groups, AEA uses data that as-sume that U.S. colleges will continue to in-crease the number of Bachelor of Science/Computer Science degrees at the same rate asthey have over the past 5 years. Comparingprojected supply and demand, the AEA reportconcludes that by 1987 there will be a short-

*UAEA, p. 10.

Computer science

Geology/geophysics

Chemistry

Systems analysis

Physics

Mathematics

Biochemistry

Biology

age of 113,406 computer scientists and elec-trical engineers. Discounting employment re-lated to defense, the shortage would be 81,780.

More skeptical about the likelihood of animpending shortage of electrical engineers isDavid Lewis, Council Chairman for the CareerActivities Council of the Institute of Electricaland Electronic Engineers, Inc. (IEEE). Not-ing that the engineering profession is made upof people who are trained in a range of dis-ciplines, projections of shortages, he says, failto take into account the extent to which elec-trical engineers can be substituted for by engi-neers trained in other specialties. He has ex-pressed concern, moreover, that an uncriticalacceptance of such predictions might lead toa surplus of electrical engineers, a situationnot dissimilar to the one that existed for aero-nautical engineers in the early 1970s.21


The Supply of Manpower

Although manpower projections tell ussomething about the number of people whowill be needed and who may appear to fill ex-isting high-technology positions, they say verylittle about the quality of skills and experiencethat the people who are available might bringto these jobs. To evaluate the quality of ourexisting and future supply of manpower forR&D in information technologies, we have tolook at the major source of this manpower—at the Nation’s educational institutions.

Formal educational institutions are, ofcourse, neither the only nor necessarily themost significant institutional setting for man-power training and development. At one pointin history, for example, it was the family thatdominated in preparing its members for eco-nomic roles. Later, as society became moretechnologically advanced, a somewhat moreformal system of apprenticeship emerged. For-mal schooling became especially importantduring the age of industrialization.

Today, as we move towards what has beencharacterized as a high-technology society,businesses have themselves become involved,both formally and informally, in performingeducational tasks. This has been particularlytrue in the area of information technology,where the larger corporations like IBM, Xerox,and Digital Equipment Corp. have set up theirown educational centers. Moreover, informaltraining takes place and is diffused within thebusiness community as people, trained in largecompanies, move on to form new companiesof their own.

Recognizing that a number of differentkinds of institutions are presently involved inthe development and training of future man-power, this chapter will nonetheless focus onthose that are a part of the formal educationalsystem. For it is chiefly within the context ofthese institutions that the Federal Govern-ment plays out its role in manpower devel-opment.

The Problems in Higher Education

While the American university system hasalways been renowned for the number of schol-ars and the amount and quality of researchthat it has generated, today many people arebeginning to question whether universities cancontinue to effectively perform all of theirtraditional roles. And, although almost allareas of university education have sufferedfrom the problems of increased educationalresponsibilities and increased educationalcosts, the problems that universities face ap-pear to be particularly acute in the areas thatgenerate manpower to perform R&Din infor-mation technologies. University departmentsin these areas are having an especially diffi-cult time because, given their limited funding,they are finding it almost impossible to com-pete for manpower and other resources in what

is becoming a rapidly growing and wide-openhigh-technology market.

The difficulties are well illustrated in thecase of engineering and computer science edu-cation, where a large proportion of facultypositions are unfilled and where the numberof Ph.D.s graduating each year has droppedsubstantially. The problem in this area is notone of attracting highly qualified undergrad-uate students.22 Over the past decade under-graduate enrollments in these areas havegrown at a tremendous rate—by 80 percent inthe case of engineering29 and by 20 percent in

ZZJeme MCDel-rnOtt, “’1’echnical Education: The Quiet Cri-sis,” High ‘1’echnology, November/December 1982, p. 87.

ZgJerrier A. Haddad, “Key Issues in U.S. Engineering Edu-cation, ” NAE Bridge, summer 1983, p. 11.


the area of computer science.24 And, given thegrowing popularity of electrical engineeringand computer science and the limitations that,in almost all engineering schools and depart-ments of computer science, are now beingplaced on the number of admissions, the highqualifications of new entrants are withoutprecedent. 25

Rather, as figure 23 illustrates, the problemat universities-has been one of recruiting suf-ficient faculty members to support this enroll-——.— .—

““AS Students Flock to Computer Science Courses, Colleges~r~ble TO Find Professors,” T..e Chronicle of Higher Edu-cation, Feb. 9, 1981.

‘5John Horgan, “Technology ’84 Education,” IEEE Spec-trum, January 1984. [Data on Admission Illustrations.]

Figure 23.—Comparison of Growth in EngineeringUndergraduate Enrollment and Number

of Faculty, 1973-80

100

90

80

30

20

10

01973 1974 1975 1976 1977 1978 1979 1980

Year (fall term)

NOTE: Faculty numbers drawn from ASEE’s Research and Graduate Studydirectory were adjusted on the basis of enrollment data to include thesame number of schools as in EMC’S enrollment survey.

SOURCE: Engineering Education, November 1982.

ment. According to the American Council onEducation, 1,583 teaching positions were va-cant in the Nation’s 244 accredited depart-ments of engineering during the 1980-81 aca-demic year.26

The gap would probably be much greater,moreover, were it not for the sizable numberof foreign engineers who teach in Americancolleges and universities. A recent survey ofengineering schools found, for example, that25 percent of all junior faculty members inengineering received their bachelor’s degreeoutside of the United States.27

The shortage of faculty members in the fieldof engineering and computer science has beenattributed to the fact that industry, by offer-ing higher salaries and other, nonmonetary in-centives, has been able to draw a number ofacademics and students away from universi-ties.28 The extent of the problem is illustratedby figure 24, which shows that, in contrast toother areas of science, only a relatively smallproportion of the Nation’s engineers and com-puter scientists are employed in academia.29

Discrepancies between the salaries earnedby scientists working in industry and aca-demia have, in fact, been quite extensive. Ithas not been atypical, for example, for an in-experienced electrical engineer with a bache-lor’s degree to earn more than an assistantprofessor of engineering with a Ph.D.30

*g”As Students Flock to Computer Sciences Courses,” op. cit.““AS Students Flock to bmputer Science Courses,” op. cit.*a’’ Supply of Engineering Faculty,” Electronz”c Market

Trends, January 1982, p. 14:It should be noted in this regard that the continued

supply of foreign faculty will depend to some extent onthe fate of The Immigration Reform and Control Act,a bill that was recently passed by the Senate and thatwould require fonigners to return home after graduationfor at least 2 years. An amendment maybe attached tothe bill, however, allowing certain students studying inhigh-technology fields to stay.

‘eScience lti”cators, op. cit., p. 123.‘McDermott, op. cit., p. 47.

Ch. 5—Education and Human Resources for Research and Development • 151

Figure 24.—Share of All S/Es Employed inEducational Institutions by Field, 1981

All S/Efields

Mathematicalscientists

Lifescientists

Socialscientists

Physicalscientists

Environmentalscientists

Computerspecialists

Engineers

(Percent)

o 10 20 30 40 50I I I I I I I 1 I

.

I

II I

SOURCE: Science Indicators, 1982.

university campuses to industry is the poorcondition of most university research facilities.In an effort to reduce costs, for example, manycolleges and universities have failed to pur-chase, maintain, and upgrade their buildingsand equipment. As a result of such decisions,university instrumentation inventories arenow nearly twice as old as those of leadingcommercial labs.31 The cost of adequately im-proving these facilities in the area of engineer-ing education alone has been estimated to be,at a minimum, between $1¼ billion to $2 bil-lion.32 This cost, moreover, is rapidly escalat-ing with inflation.

Problems of university instrumentation areparticularly serious in those fields where thecost of equipment is especially high and whereit plays an essential, if not an integral, partof the educational and research processesthemselves. This is the case, for example, inthe area of artificial intelligence. Only very fewuniversities can afford the cost or have thespace and facilities available to house and sup-port the kinds of sophisticated equipment re-

81John Brademas, “Gradua~ Education: Signs of Trouble,”Sa”ence, vol. 223, Mar. 2, 1984, p. 881.

‘zHaddad, op. cit.

quired to perform R&D at the leading edge ofthe field.33

Other institutional factors that have beencited as reasons for the exodus from academiato industry include uncertainty of tenure,heavy teaching loads, inadequate funds andinstitutional support for research, and in-creases in educational fees compounded by di-minishing financial aid for students.

It is difficult to assess the extent to whichthese problems will persist in the future. Thenumber of students studying for Ph.D.s inengineering significantly decreased over thecourse of the decade 1972-82. Educators of en-gineering suggest that, as a result, there willnot be enough new faculty members to replaceeven those who die or retire.34 If this kind oftrend continues, there will probably be a fac-ulty shortage in the future. However, there aresome indicators that point to a reversal of thistrend. In 1982, for example, the number of en-gineers earning doctorates increased for thefirst time in 8 years. 35 In 1983, the number in-creased again—from 2,888 to 3,023.3’

On the other hand, the academic manpowerproblem may be more difficult to overcome inthe field of computer science where the per-centage of faculty leaving for industry is twotimes that of any other field of engineering.37

Even in this area, however, there is some anec-dotal evidence reported by NSF to suggestthat the number of graduate students study-ing in this area is now increasing.38

It is possible, however, that faculty short-ages could become even more critical in thefuture in some specific areas, limiting theamount of research and the amount of teach-ing that can be done in these fields. This is par-ticularly true, for example, in an area such asartificial intelligence (AI).39 Because the field

8WTA Case Study on Artificial Intelligence.MNAE Bridgq op. cit., p. 121.sb~”aa Indz”catirs, op. cit., p. 123.‘Manpower Comments, vol. 21, No. 3, April 1984, p. 24.sTC@i9, Op. cit., p. 8.‘Conversation with Kent Curtis, National Science Foun-

dation.S90TA ewe study on Artifi”a”al InW”gence.

38-802 0 - 85 _ 11


of AI is so specialized and because the size ofthe research community is so small to beginwith, the number of qualified faculty memberscannot be multiplied rapidly enough to meetthe rising number of students who are now be-ginning to enter the field. The problem is likelyto worsen if, as might reasonably be expected,the greater commercialization of AI applica-tions together with enhanced military inter-est in the field lead to increased competitionfor those trained in artificial intelligence in thefuture. The effect of a faculty shortage maybe particularly acute in an area such as thiswhere the required skills and knowledge arebest taught in an apprenticeship-type sit-uation.

To alleviate the present faculty shortage,more than half of all engineering colleges havehad to eliminate courses, and two-thirds areincreasingly substituting graduate assistantsfor full professors for some coursework, a sit-uation that could adversely affect the teachingof advanced technical courses. A number ofuniversities have also increased the teachingload.40 It has been estimated, for example, thatover the past 10 years the teaching burden ofthe average professor of engineering has in-creased by 40 percent.41 While such actionsmay ameliorate the problem in the short term,in the long term they may—to the extent thatthey discourage people from pursuing teach-ing careers-actually exacerbate it.

Compounding their problems of loss of fac-ulty and deterioration of equipment, schoolsof engineering and departments of computerscience will also have to find ways to modern-ize their curricula to meet the educationalneeds of a high-technology society undergo-ing rapid technological change. It has beenestimated, for example, that given the speedat which technological change is occurring,today’s engineering education, acquired overthe course of 4 or 5 years, will be obsolete inabout the same amount of time.42 Up-to-datecoursework, moreover, will have to take into

@McDemot,t,, op. Cit., P. ‘-41~nfowor~d, May 30, 1983, P. 32.4“’The Changing Face of Engineering,” Electronics, May 31,

1983, p. 127.

account the changing style of doing engineer-ing and computer science. To be effective, forexample, the engineers and computer scien-tists of the future may have to be adept incommunications as well as in technical skills.43

As one young computer scientist described it,

The old stereotype of an engineer was aguy with a slide rule hooked to his belt andwho was not into sports, not into team play.

The jobs that we have to do now are not1-man year projects but are 5- to 10- to 20-man-year projects. Integrated circuits, evenlinear ICs, are so complex that they are nolonger designed by one engineer, but byteams. . . this concept may be new to us, butit’s not to the Japanese.

The mounting problems that engineeringschools are facing would appear to be havinga negative effect on the quality of educationthat they provide. In the last 3 years, for ex-ample, 30 percent fewer engineering schoolswere given full 6-year accreditation. Moreover,there has been an increase of over 70 percentin the number of institutions that have beenasked to show cause why their accreditationshould not be revoked. 44 Many university offi-cials agree with this assessment. In a surveyof engineering colleges conducted by the Amer-ican Council of Education, 49 percent of therespondents reported that the quality of edu-cation provided by their institutions had ei-ther greatly or moderately declined.45 Concernsabout the decline in the quality of engineer-ing education are also echoed in the industrialcommunity, where a number of companieshave independently taken steps to improve thepool of engineering manpower.46

The current situation is not, however, set instone. To the contrary, the American educa-tional system is today in the midst of consid-erable change. In adopting policies to affectthe future supply of manpower, therefore, pol-icymakers should be reminded of the “pendu-lum effect.” In making decisions about the fu-

4sIbid,, pp. 127-128.44McDermott, op. cit., p. 90.4’High Technology Manpower in the West: Strategies for Ac-

tion, op. cit., p. 12.*For a discussion of many of these efforts see Chapter 6: New

Roles for Universities in Information Technology R&D.


ture, they need to take into account not onlythe projections of future manpower needs butalso how present reactions to those projectionsmay, in fact, undermine their validity.

Concerned about the present state of high-technology manpower training in the UnitedStates, a number of businesses and corpora-tions, for example, have already taken it uponthemselves to improve the situation. In par-ticular, they have adopted a number of meas-ures that are designed to keep faculty in aca-demia and to encourage students to pursueacademic careers. Bell Labs has recently be-gun a program that, over time, will provide $2million to doctoral candidates studying infields related to telecommunications. The Gen-eral Electric Co. has donated $2 million toengineering schools to be used primarily byteachers and Ph.D. candidates. IBM has es-tablished 150 graduate and postdoctorate fel-lowships in engineering, computer science, andinformation systems. And Hewlett-Packardrecently initiated a student loan forgivenessprogram that allows students to write off apart of their loans in accordance with a prear-ranged schedule for each year that they teach.

Companies are also providing funds forequipment and instrumentation. Digital Equip-ment Corp. and IBM together have donated$50 million for equipment to MIT. Sizable con-tributions have also been made to a numberof other universities by IBM, Hewlett-Pack-ard, Apple Computer Inc., Wang Laborator-ies, Inc., NCR Corp., and Honeywell, Inc. (seefig. 25).

In addition, businesses and corporations are,more and more, joining with universities incooperative ventures to overcome perceivedshortages of scientific and technical manpow-er. The goal of training manpower, for exam-ple, was the major reason for establishing theMicroelectronics and Information SciencesCenter at the University of Minnesota, and themost important inducement in gaining indus-try support. These cooperative ventures, aswell as a number of others, are discussed insome detail below, in chapter 6.

Figure 25.—Educational Gifts by Computer Vendors

IBM $50 million in cash and equipmen to 20advance research in CADICAM.

• Co-donation of $50 million in equipment to MITto research data transfer.

● $15 million pledge to Brown University to establish institute for research in informationand scholarship.

● $2.4 million in graduation fellowships to scienceand engineering students.

Digital • $ co-donation of $50 million in equipment to MIT.EquipmentCorp. •-$1.6 million to Boston University to fund new

compute science programs.● Total of $45 million in fiscal 1982 to highereducation.

Apple ● $21 million in equipment to California schoolscomputer, grades K-12 for “Kids Can’t Wait” Program.Inc. Company wants a Similar nationwide pro-

gram, but wants Federal tax deductibility firstvia so-called “Apple Bill"

● $500,000 to Brown University in form of 5OLisa Computer.

Hewiett- ● Approximately $22 million, mostly in equipment,Packard to universities in fiscal 1983.co.Wang ● Total of $3.7 million in equipment and $458,000Laboratories, in cash to 23 universities and secondaryinc. schools in 1982.NCR • $140,000 in equipment to Michigan StateCorp. University.

● $170,000 in equipment to Cornell University.• Several gifts of companies NCR systems to

other universities.Honeywell, ● S220,000 to Arizona State University.Inc. ● $30,000 to United Negro College Fund.

● Total of $3 million in education contributionsin 1982.

SOURCE: Compurerworld, July 25, 1983.

The States have also taken steps to improveand foster the development of manpower fora knowledge-intensive society. Not only havethey joined in cooperative ventures betweenindustry and the universities, they have alsotaken steps to enhance the research environ-ment and to improve faculty salaries.

While most State and corporate efforts toimprove manpower training have been focusedon the most prestigious universities, therehave been a number that have sought to im-prove the education and training provided toless advantaged colleges and universities. Ex-xon, for example, recently donated $1.8 mil-lion to six predominantly black engineeringcolleges, to be spent during the course of thenext 3 years. IBM has lent several of its em-ployees to teach in schools that serve “disad-vantaged students. ”


Elementary and Secondary Education

To determine whether or not there will beenough highly qualified people available toperform R&D in information technology, onemust look not only at institutions of highereducation, but also at elementary and second-ary schools. Generating the pool of studentsfrom which university and graduate studentsare drawn, these institutions have an effect onboth the number and the quality of studentswho desire to pursue fields of study that mightlead to work in this area.

A number of reports have been released re-cently that raise serious questions about theability of elementary and secondary schoolsin the United States to adequately prepare theNation’s youth to work in a knowledge-inten-sive society. While differing in the focus oftheir concerns, all of these reports find thatour schools are inadequately preparing Amer-ican students for their futures. As evidence,they point to declining test scores, the needfor colleges and business to provide remedialeducation and training programs, the level offunctional illiteracy among the general popula-tion and the Nation’s poor showing in in-ternational comparisons of student achieve-ment.47

Of particular concern in all of these studiesis the overall poor quality of math and scienceeducation. A number of indicators give causefor such concern (see fig. 26). Enrollments inmath and science courses, for example, aregenerally low. Only one-third of all U.S. highschool graduates have completed 3 years ofmathematics, and less than 8 percent havecompleted a calculus course.48 Moreover, ingrades kindergarten through six, studentsspent only approximately 20 minutes per dayon science lessons. And by 10th grade, onlyabout half of all students study any science

B. Education in Reports on and Recommendations for Change, Issue Brief #

IB83106 Congressional Service, Library of Congress,NOV. 17, 1983.

“High School Science Problems Gain Spot-light,” & News, May 24, 1982, p. 39.

Figure 26.— Selected Indicators of the OverallQuality of Mathematics and Science Education

in the United States

Scholastic Aptitude Test score averages from1963 to 1981

500 i 1

480

460

440

420

01963 1965 1967 1969 1971 1973 1975 1977 1979 1981

Years

Proportion of 1980 high school leavers who have takenvarious courses

Algebra I

Algebra II

Geometry

Trigonometry

Calculus

Physics

Chemistry

o 10 20 30 40 50 60 70 80

Percent

Average scores on science tests among school children aroundthe world (1973)

England

Scotland

Germany

Japan

USA

o 10 20 30 40 50 60

Percent of maximum score


at all.49 In contrast, students in many otherindustrialized countries spend about threetimes more class time on math and sciencethan the most scientifically oriented Ameri-can student.50

Student attitudes about science and mathalso reflect problems in the education of thesecourses at the elementary and secondaryschool levels. Science, for example, is the sub-ject that is preferred the least by most Amer-ican students.51 Moreover, although most chil-dren regard scientists themselves as beingintelligent and dedicated, they consider thework that scientists do as being “boring, dull,and monotonous. ”52 Mathematics is not muchmore popular. While it is the favorite subjectof 45 percent of students in the third grade,it is rated tops by only 18 percent of those inthe 12th.53

Overall, student performance in these areasreflects the lack of interest and low enrollmentin these subjects. In the three studies con-ducted by the National Assessment of Educa-tional Progress, which were designed to assessthe achievements of precollege students in anumber of areas over the years 1969, 1973, and1977, all age groups showed significant aver-age declines on mathematical applicationswhich involve the use of mathematical knowl-edge, skills, and understanding to solve prob-lems.54 In the area of science, the results weresimilar. There was a downward trend for allage groups, from the first to the second assess-ment, although this decline appeared to be di-minishing for 9 and 13 year olds in the thirdassessment.55

As in the case of higher education, many ofthe problems in precollege math and scienceeducation have been attributed to the short-age of high-quality teachers to teach in theseareas. According to the Director of the Nation-

“Herbert J. Walberg, “Scientific Literacy and Economic productivity in Intematimal Perspective, ” Dzmdalus, p. 12.

‘OHeylin, op. cit.*] Walberg, op. cit.5$ Heylin, op. cit..“Walberg, op. cit.“Sa”ence Indicators, op. cit., p. 74.551bid.

al Science Teachers Association, since 1971there has been a dramatic decline in the num-ber of people training to teach in these areas—79 percent in the case of math and 64 percentin the case of science (see fig. 27).

The lack of new entrants into the teachingprofession is not surprising insofar as thereare few incentives to draw well-qualified indi-viduals into the field.56 With the financial prob-lems facing schools today, a career in teachingis no longer considered to be secure. And it isconsidered by many to be less likely to pro-vide the rewards of public status or personalesteem. Since teaching salaries are by nomeans competitive with those in private indus-try, many of the most qualified teachers—especially those with training in math andscience—are giving up teaching to sell theirskills on the open market. Moreover, well-qualified women–a traditional source of high-quality educators—are particularly likely toseek out the wider range of employment op-portunities now available to them.57

The actual extent of the present shortage ofqualified math and science teachers is a mat-ter of some debate, however. A survey of ed-ucational placement offices conducted by theNational Science Teachers Association found,for example, that no less than one-half of theNation’s math and science teachers were hiredon an emergency basis.58 The Association forSchool, College and University Staffing hasreported, moreover, that there is a consider-able nationwide shortage of teachers in thearea of mathematics, physics, and computerprogramming on the other hand, a recentstudy conducted by the General AccountingOffice concluded that the gaps in the available— . .——.

‘The New Sa”entist, July 14, 1983.In explaining the shortage of teachers, much of the at-

tention has focused on salary differentials between teach-ing and industry. While salary levels are no doubt im-portant, one recent study found that internal moralefactors, such as the potential for personal growth, areeven more so. Edward B. Fiske “Teacher Fulfillment PutAbove Pay, ” The New York Times, Oct. 4, 1983, p. Cl.

“J. Myron Atkin, “Who Will Teach in High School?” Amer-ica Schools: Pubh”c and Pn”vate, Daedalus, summer 1981.

6 8 T h e New Scjentjst, Op. cit.

W’eacher SupplyDemand 1984, Association for School, Col-lege, and University Staffing.


Figure 27.—America% Science Teacher Shortage

A 1982 survey reveals just how many new science and mathematicsteachers in the U.S. do not hold the appropriate qualifications to dotheir jobs. This is partly a result of a fall in the numbers of those

840/0460/o

43 ”/0

3

Pacific Mountain 6 3 % — 40%

1971 base

1971 1973 1975 1977 1979

1-0

0-75

0-50

0-25

0

SouthAtlantic.

EastSouth

Central

1971 1973 1975 1977 1979

- Qualified teachers available for employment_ Accept ing teaching posi t ions

Teachers taken on in 1982 to teach science and math not qualifiedin those subjects (top left). Decline in number of new teachers be-

SOURCE: New Scientist, July 14, 1963.

information are so great that whether or notthere are shortages of math and science teachers,and whether or not the quality of technicalteaching has declined in recent years, cannotbe determined.60

Since a strong foundation in math and sci-ence is, in most cases, essential to a future ca-reer in information technology R&D, thesefindings raise a number of questions about thefuture supply of manpower in this area. Theiranswer, however, is not particularly clear orstraightforward. For although the data showthat the level of math and science understand-ing and competency for the student popula-tion as a whole has declined, it has not declined

60New D&@”ons for Federal Programs to Aid Mathematicsand Su”ence Teachi”ng, GAOIPEMD-84-5, Mar. 6, 1984.

achieving the appropriate qualifications. Another important factor isthe far more attractive salaries offered to graduates in the sciences,whether qual i f ied teachers or not . -

Teachers(all disciplines)

Those enteringgovernment orindustry, bydiscipline:

Humanities

Social sciences

Biological sciences

Agricultural sciences

Accounting

Health professions

Chemistry

Mathematics

Computer sciences

Earth sciences

Engineering

M e n

W o m e n

1 1 I 1 I 1 I12 14 16 18 20 22 24 26 28

Thousand dollars

tween 1971 and 1980 (above left). Starting salarlis for new graduatesentering careers in government or Industry in 1982 (above r!ght).

among those students who are most likely topursue careers requiring a strong backgroundin math and Science. 61University math and sci-ence departments and schools of engineeringreport, moreover, that they are continuing torecruit high-quality students, many of whomare being drawn from the humanities depart-ments with the idea of having better futurejob prospects62 (see, e.g., fig. 28). On the otherhand, given the diminishing supply of high-quality educators in these fields and thepredicted decline in the school age popula-

el~”ena In~”ca~rs, Op. cit., P. 77”ufiti J. A-, stud~t Quali”tyin the ti”ences and Ew”-

neering Opku”on of Sam”or Academi”c Offi”a”als, Higher Educa-tion Panel Report No. 58, American Council on Education, Feb-ruary 1984.


Figure 28.—Selected Indicators of Shifts in Undergraduate Science/Engineering EducationChart 1. Perceptions of academic officials about change overlast five years In quality of undergraduate science/engineering

(S/E) students, 1982

Percent

o 10 20 30 40 50 60 7(-I-- --Type of 1 I I I I I

Ins t i tu t ion NO S i g n i f c a n t c h a n g e

A l l i n s t i t u t i o n s s i g n i f i c a n t i m p r o v e r n e n t

Doctorategrant ing

grant ing

Publ ic lycontrol led J

Private lyc o n t r o l l e d

100 largest S/Ebaccalaureate

SOURCE: National Science Foundation and American Council on Education,

tion,63 a more serious problem may emerge inthe future.

Just as the situation in the case of highereducation, the situation in elementary and sec-

—WI’he number of Americans graduating from high school is

projected to drop from a peak of 3.2 million of 1977 to a lowof 2.3 million in 1992, following a roller coaster pattern throughthe end of the century. High Schcnd Graduates: Projtwti”ons forthe Fifty States (1982-2000), Western Interstate Commissionfor Higher Education, Boulder, CO.

Chart 2. Opinions of academic officials about shift in distributionof most ab le undergraduate s tudents between sc ience/eng ineer ing

(S/E) and other fields over the last five years, 1982

Percent

o 10 20 30 40 50 60 70 80

T y p e o f institution I No change

All institutions S h i f t a w a y S/E trends

Doctorategranting

Nondoctorategranting

Publlclycontrolled

Privatelycontrol led I

100 largest S/Ebaccalaureate . I

ondary education is not static and changes aretaking place that might affect the long-termsupply of scientific and technical manpower.The public’s response to the numerous reportson the condition of American education hasbeen immediate and widespread, involvingparents, teachers, businessmen, and govern-ment officials at all levels.

The Federal Role in Manpower Development

As the largest employer of scientists and en-gineers, the Federal Government has a directinterest in whether or not the educational sys-tem can provide an adequate supply of tech-nically trained people. Its role in the develop-ment of manpower, however, has traditionallybeen indirect. For it is the States, and not theFederal Government, that have had the pri-mary responsibility for educational policies.

Given the Constitutional limitations on theFederal Government’s role in education, theresponsibility for developing manpower hasalways been shared by a number of differentsocial institutions ranging from the family tothe business community. As American societyhas become more technologically advanced,however, the Federal Government has been in-creasingly called onto play a more significant


role. Pressure on it to be more active in thisarea is particularly strong today, as the Na-tion seeks to maintain its place in a highlytechnical and competitive world environment.

While conscious of the economic benefits as-sociated with having a skilled labor force,Americans did not originally adopt a formalsystem for transmitting vocational and tech-nical skills, in the earliest years of their his-tory, when agriculture was the dominant modeof production.64 Instead, most formal educa-tional institutions were designed to serve gen-eral social and political functions, while gen-eral vocational skills were left to be passed onmore or less informally, by family membersor through apprenticeship systems.65

In particular, there was little effort given to,or even much concern for, the development ofscientifically trained manpower. When scien-tists were needed in the United States, theywere brought over from Europe. Thus, it wasonly in 1902 that the Federal Government firstbecame involved in the development of scien-tific manpower. Concerned at this time aboutthe security dangers entailed in relying almostexclusively on foreigners for scientific exper-tise, Congress established the Army Corps ofEngineers at West Point.66

It was only with the rapid industrializationof society, at the end of the century, that edu-cation came to be really valued in economicand technical terms.67 And then, as Americans

‘For a discussion of American education in the preindustrialperiod, see Bernard Bailyn, Education in the Fornu”ng of Amer-ican Sou”ety (New York: W. W. North, 1980); Lawrence Cremin,TradI”tiorIs in Amen”can Education (New York: Basic Books,Harper, 1976); and Rush Welter, Popular Education andDemocratic Thought in Amerz”ca (New York: Columbia Univer-sity Press, 1962).

‘A. Hunter Dupree, Sci”ence in the F’dmd Governm ent: His-toxy of Pd”a”es and Actk”tiee to 1940 (Cambridge, MA: BelknapPress of Harvard University Press, 1957).

sdlbid. Initi~ly trained in the s~s neCeSSSry b cm outexplorations of the West and to conduct surveys of the east-ern coastline, these West Point graduates played a key role inmany of the scientific ventures carried out by the Federal Gov-ernment up until the end of the 19th century.

o~David K. Cohen and Barbara Newfeld, “The Failure of HighSchools and the Progress of Education,” Amen”ca’s Schools:Public and Pn”vate, Daedalus, spring 1981.

came to believe that special technical knowl-edge was the key to prosperity in the modernage, secondary educational institutions wererestructured to prepare American youth foran increasingly differentiated set of economicroles. Not only were vocational courses addedto the educational curriculum, but the schoolsthemselves were remodeled to conform to theprevailing business standards of efficiency.The business community played a major rolein bringing about these changes. Concernedabout strikes, labor turnover, and increasingworker absenteeism, they hoped that school-ing would socialize a growing number of im-migrant youths for the workplace.68 69

The growing enthusiasm for scientific andtechnical knowledge also had an impact on thenature of higher education. Accustomed totraining gentlemen as preachers, lawyers, anddoctors, the universities began to expand theirroles and to train people in the more vocationalapplications of education.70 Efforts to move inthis direction met with considerable resistancefrom traditional academics, however, whowere disdainful of the study of experimentalscience and even more so of the teaching ofthe “useful arts.”71 72

The transformation of the university systemwas greatly facilitated by the passage of Fed-eral legislation establishing the land grant col-leges. Provided for under the Merrill Act of1862,73 these colleges, open to children of allbackgrounds, were established to provide edu-cation in practical fields such as agriculture,engineering, home economics, and business

‘David Tyack and Elizabeth Hansot, “Conflict and Consen-sus in American Education, ” ti”a /Wwls: Public and Pri-vate, Daedalus, summer 1981.

‘sIbid.~OEmest L. Boyer ~d Fr~ Heckinger, Higher Education in

the Nati”on’s Serw”ce (Washington, DC: Carnegie Foundationfor the Advancement of Teaching, 1981); and Edward Shils,“The Order of Learning in the United States From 1865-1920,”Minerva, vol. 21, No. 2, summer 1978.

nDa~d Noble, Am~”ca by Detu”gn: &i”eXlce, T~OIOKY, ‘dthe Rise of Corporate Capitalism (New York: Alfred A. Knopf,1977), p. 24.T2Boyer ad HW~@r, op. cit..; ~d ShilS, oP. cit.

7~his law provided land to the States, the proceeds of whichwere to be used to teach in the fields of agriculture and mechanical arts. Subsequent legislation provided Federal finan-cial support for research and the operation of the landgrantcolleges.


administration. Through their agriculture ex-periment stations and their service bureaus,their activities were designed to serve thestate .74

The impact of the Merrill Act on develop-ment of scientific and technical manpower isclearly evident in the case of the engineeringprofession. Before its passage, State legisla-tures had been reluctant to invest in techni-cal education. Responding to the offer of Fed-eral grants, however, they quickly sought toestablish the new types of schools, while pri-vate colleges, caught up in the movement, alsoestablished departments of engineering.75

Schools of engineering expanded rapidly,thereafter, numbering 110 by 1886. The num-ber of engineering students similarly increased,from 1,000 in 1890 to 10,000 in 1900.7’ Asmore and more engineers were educated in for-mal institutions, there was a greater empha-sis in engineering on science. Moreover, withthe establishment and growth of these insti-tutions, a profession was developed and withit a means of preserving, transmitting, andincreasing an evolving body of engineeringknowledge.77

The real impetus for manpower develop-ment, and for a strong Federal role in it, came,however, after World War II, when advancedtechnology had proven to be critical not onlyfor the Nation’s economic growth, but also forits defense. It was in recognition of this fact,for example, that the National Science Foun-dation was created in 1950 with the task ofimproving the Nation’s potential in scientificresearch and in science education.78

The philosophical basis for establishingNSF, and the rationale for including the de-velopment of scientific manpower within its

“Clark Kerr, The Uses of the Um”versity (Cambridge, MA:Harvard University Press, 1972).

7 6 N o b l e , op. c i t . , p p . 38+9.

“Edwin T. Layton, Jr., The Revolt of the En@”neers: %cialR.esponsibih”ty and the Amen”can Engineering Pmfesm”on (Cleveland, OH: The Press of Case Western Reserve University, 1971).

771bid.‘oThe Natjon~ ~ience Foundation and Pre-COflege Science

Education: ~950-1975, report prepared for the Subcommitteeon Science and Technology, U.S. House of Representatives, 94thCong., 2d sess., by the Congressional Research Setice, Libraryof Congress, January 1976.

organizational mission, was explained by Van-nevar Bush in Science—the Endless Frontier,his report to the President on a program forpostwar scientific research. About the need forscientific manpower, he said,79

. . . Today, it is truer than ever that basic re-search is the pacemaker of technologicalprogress. In the 19th century, Yankee me-chanical ingenuity, building largely on the ba-sic discoveries of European scientists, couldgreatly advance the technical arts. Now thesituation is different.

A nation which depends on others for itsnew basic scientific knowledge will be slowin its industrial progress and weak in its com-petitive position in world trade, regardless ofits mechanical skill.

Provoked by the successful launching of theSoviet spacecraft Sputnik, defense considera-tions also motivated the passage of the Na-tional Defense Education Act of 1958 (NDEA),which was aimed at improving instruction inmathematics, science, and foreign languages.Under this law, funds were provided on amatching basis to public schools, and as long-term loans to private institutions, for neededequipment in these instructional fields, for cur-riculum development, for guidance counseling,for vocational education in defense-relatedfields, and for teacher training in foreign lan-guage instruction. The passage of NDEAresulted in substantial increases in Federal aidto education (see fig. 29). Since Federal dollarshad to be matched by State and local fundsunder provisions of the act, the overall invest-ment in NDEA programs was large. Between1958 and 1961, $163.2 million in Federal fundswere dispersed. Approximately 75 percent ofthese funds were directed to the developmentof science curricula.

Although the U.S. educational system is or-ganized on a local basis, the Federal Govern-ment has, over the years and in response tochanging technological developments, come toinfluence the development of scientific andtechnical manpower in a number of ways. Par-ticularly in the area of science and math-ematics, the Federal Government has, for

7e1bid., p. 19.


Figure 29.—A Quarter Century of Student Aid

25 years ago, the National Defense Education Act was signed

Student loans

NDEA, 1958National Defense Student Loans, for

which about $30 million was appropri-ated in the first year, were provided at aninterest rate of 3 percent. The law toldcolleges to give “special consideration”to outstanding students who planned tobecome teachers in elementary andsecondary schools and to those with abackground of high achievement inscience, mathematics, engineering, or a

1983The program, renamed National Direct

Student Loans and transferred to theHigher Education Act, provided loans atan interest rate of 7 percent. About $193million was appropriated for fiscal 1983.The debts are forgiven for borrowers whobecome teachers in schools serving dis-advantaged students and those who teachthe handicapped.

International educationNDEA, 1958 1983

The law authorized support The programs have beenfor fellowships, research, and modified and incorporated inuniversity centers for foreign- the Higher Education Act, withlanguage and area studies, increased emphasis on support- k

.., . . . . ,putting emphasis on the training the education of nonspe-ing of specialists in languages cialists as well as specialists in“needed by the federal govern- foreign language and areament or by business, industry, studies. About $26 million wasor education in the United spent on those programs inStates.” Some $5 million was fiscal 1983.

~ ,, *

appropriated in the first year.n

.,=[3

,

fellowships for students plann- vamped and transferred to the Iing to become college teachers Higher Education Act in 1972. and grants to encourage the The Education Department’sestablishment and expansion principal graduate-fel lowship of graduate schools. About program provided support for $5.3 million was spent in the women and for blacks andfirst year. other minorities. About $10 roil-

l ion was prov ided in f i sca l t

1983. , f

$6,718

1959 ’60 ’61 ’62 ’63 ’64 ’65 ’66 ’67 ’68 ’69 ’70 ’71 ’72 ’73 ’74 ’75 ’76 ’77 ’78 ’79 ’80 ’81 ’82 ’83

YearsSOURCE: Compiled by the American Council on Education, based on data from the U.S. Department of Education. Totals do not include Veterans Educa[

or Social Security Benefits for college students.lion Belnefits

Ch. 5—Education and Human Resources for Research and Development • 161

example, provided support for curriculum de-velopment and the production of new coursematerials. In addition, it has also lent consid-erable institutional support to colleges anduniversities, in the form of direct grants,research grants and contracts, equipmentgrants, and the establishment of institutes andspecialized instrumentation centers. The Fed-eral Government has, moreover, assisted in-dividuals through training programs and fel-lowships as well as by providing scholarshipsand other forms of financial aid.

By targeting scholarships or fellowships forspecific fields of study, the Government can

also influence the direction of manpower de-velopment. 80 This was particularly true, for ex-ample, during the years following Sputnikwhen there was a widespread belief that thenormal pool of technically trained graduatescould not meet the manpower needs of the ci-vilian space and the military ICBM programs.Where a large number of people have been in-volved, the Government has tended to provideonly generalized incentives. Where the num-ber is much smaller and the area of expertisequite specific, it has sought more to narrowlytarget its support.

‘“Ibid.

The Societal Context forFederal Manpower

DeterminingPolicy

The role of the Federal Government in de-velopment of scientific and technical manpow-er is not only indirect, it is also complicated.It is complicated by the fact that the Nation’seducational system-the primary institutionalmeans by which the Federal Government canaffect manpower— serves a number of societalgoals in addition to, and at times competingwith, those that directly relate to manpowerdevelopment.

Today, given the growing concerns aboutthe United States’ declining position in the in-ternational economy of high-technology goodsand services, the American educational sys-tem is being called on to develop the Nation’sscientific and technical manpower as a meansof gaining for the United States a greater com-petitive edge. In examining this issue, thischapter has posed a number of questionsabout: 1) the causal relationships between edu-cation and economic performance, and 2) theextent to which there will actually be a criti-cal shortage of manpower in the future to per-form R&D in such high-technology areas asinformation technology. Before policy altern-atives can be adopted to address the manpowerproblem, it is necessary to inquire further andto ask first whether or not Federal efforts to

move the educational system more in the di-rection of manpower ‘development can beachieved without sacrificing other equally, ifnot perhaps more important, educational andsocietal goals.

Education has always played a particularlyimportant role in American society. For whileeducational institutions are publicly supportedin many societies, in no other country havethey been established with such deliberate pur-pose and public expectation, or been conceivedof as being such an integral part of the politi-cal, social, and economic order. Contrastingthe attitude of Americans towards educationwith that of Europeans, for example, Alexisde Tocqueville, the well-known French com-mentator on American society, noted in 1831:81

Everyone I have met up to now, to what-ever rank of society they belong, has seemedincapable of imagining that one could doubtthe value of education. They never fail tosmile when told that this view is not univer-sally accepted in Europe. They agree in think-ing that the diffusion of knowledge, useful forall peoples, is absolutely necessary for a peo-

a] Alexis de Tocqueville, Journey to Amen”ca, translated byGeorge Lawrence, J.P. Mayer (cd.) Anchor Books, 1971.

.


ple like their own, where there is not propertyqualification for voting or for standing forelection. That seemed to be an idea takingroot in every head.

The goals that Americans have sought toachieve through education have changed overtime and in different historical circumstances.82

In the earliest years of American history, forexample, education was considered essentialfor the survival of the new democratic nation.Later, with the need to acculturate immi-grants into society and to unite a divided Na-tion in the aftermath of the Civil War, it wasconsidered the means for building a nation ofcitizens. At the turn of the century, educationwas expected to train and socialize Americanyouths to participate in a modern, industrial-ized society. In the 1960s, Americans saw ineducation a way of providing equal access tosocial and economic opportunity.

Today, much of the national discussionabout education has focused on the goal ofmeeting the manpower needs of a high-tech-nology economy. Unlike earlier periods ofAmerican history, however, when the educa-tional system was itself undergoing tremen-dous expansion and enjoying a period of con-siderable prosperity, today the educationalsystem is having to take on a multitude of newtasks at a time when it faces the prospect ofshrinking economic and human resources. Thustoday, given limited social and economic re-sources for education, the choice to take Fed-eral actions to increase the supply of man-power for R&D in the area of informationtechnology might be made at the expense ofaddressing other educational problems or ofmeeting other educational and societal goals.

Some of the most important of the newtasks that educational institutions will haveto perform are those that relate to the emer-gence of an “information society. ” In the re-cent OTA study, Information Technology andIts Impact on American Education, it wasfound, for example, that the growing use ofinformation technology throughout society iscreating new demands for education and train-——— ——

esw~ti~, op. cit.; cohen and Newfield, op. cit.; ‘1’’ya~, oP. cit.

ing in the United States and is increasing thepotential economic and social penalty for thosewho do not respond to those demands. More-over, the study found that the informationrevolution is creating new stresses on manysocietal institutions, particularly those suchas public schools and libraries that tradition-ally have borne the major responsibility forproviding education and other informationservices.

New demands will also be placed on the ed-ucational system by virtue of the changes thatare now taking place in the American econ-omy. Programmable automation, for example,as the recent OTA study, Computerized Man-ufacturing Automation: Employment, Educa-tion, and the Workplace, points out, is one ofthe economic forces that is currently reshap-ing the roles for and the values being assignedto education, training, retraining, and relatedservices such as career guidance and job coun-seling. This study concludes, moreover, thatthe inadequate capacities of the present in-structional system may constrain the estab-lishment of strategies to develop adequateskills for programmable automation.

Educational institutions will also be calledonto perform a number of cultural tasks. Fore-casts of demographic trends suggest, for ex-ample, that Hispanics, Asians, and other cul-tural groups with specialized educational needswill soon comprise a major portion of the schoolpopulation. Moreover, the increase in the num-ber of school-age children living with a singleparent or two working parents will force schoolsto provide more supervision and socialization.In addition, a growing number of adults are turn-ing to the educational system to fill in the non-vocational, liberal arts gaps in their educationalbackgrounds. 83 Thus, instead of cutting back ontheir educational activities, schools and univer-sities will, most likely, have to cater to a grow-ing number and variety of educational needs.

The role of the educational system as ameans of providing equitable access to eco-nomic and social opportunities could become

‘a’’More Adults Return to College h Study the Liberal Arts,”Chronicle of Higher Education, Apr. 25, 1984, p. 1.


all the more important in an age of high tech-nology. The Bureau of Labor Statistics, for ex-ample, has predicted that the number of low-skilled jobs will constitute an increasinglylarger proportion of the total work force thanthe highly technical jobs.84 Other social observ-ers have suggested, moreover, that the wagesearned by low-skilled workers will decline rela-— — . —

BiBLS, Levin and Ruxnberger, op. cit.

tive to those earned by technical workers.ab

Under such circumstances, the competition foreducational credentials may become more in-tense in the future, and the issue of equitableaccess may become at least as important as,if not more important than, the issue of man-power development.

8’Bob Kuttner, “The Declining Middle, ” The AtlanticMonthly, July 1983.

Chapter 5 References

“Academic Gap Between High and Low AchieversNarrowing, ” Chem”cal & En~”neering News,Feb. 29, 1983.

America’s Competitive Challenge, Report of theBusiness-Higher Education Forum, 1983.

“Comrnentary: The Science Education Stampede,”Science & Government Report, Oct. 1,1983, pp.7-8.

“Computer People: Yes, They Really Are Differ-ent, ” Business Week, Feb. 20, 1984, pp. 66-77.

Education, Productivity, and the National Econ-omy: A Research Initiative, National Instituteof Education, December 1981.

“Education for a High Technology Future: TheDebate Over the Best Curriculum, ” School Re-search Forum, Educational Research Service,May 1983.

“Employment Up in Manufacturing for Scientists,Engineers, Technicians, ” National ScienceFoundation, Mosai”c, March/April 1983.

“Engineers Need More Learnin30, 1983, p. 32.

g,” Infoworld, May

Hem”ngs on Mathematics and Science Education,before the Committee on Education and Labor,House of Representatives, 98th Cong., 1st sess.,Jan. 26-28, 31, 1983.

“Is Shortage of Engineers a Matter of Defini-tion?” Science, vol. 223, p. 800.

“Labor Market Slackens for Science and Engineer-ing Graduates, ” National Science Foundation,Mosm”c, January/February 1983.

Manpower Concerns, Hearings, before the Com-mittee on Science and Technology, House ofRepresentatives, 1st sess., Oct. 6-7, 1981.

“Manpower Surveys Continue to Disagree, ” Elec-tronics, July 28, 1983.

“More Foreign Students Enroll, But Only SomeStay to Work, ” Physics Today, October 1983.

New Directions for Federal programs to Aid Math-ematics and Science Teaching (GA0/REMD—84-5, Mar. 6, 1984).

“NSF Survey Shows Academic EmploymentGains for Scientists and Engineers, ” AFIPSWashington Report, September 1983, p. 4.

“Recent Reports on Education: A Digest, ”Change, November/December 1983.

Science and Engineering Education for the 1980’sand Beyond, National Science Foundation andthe Department of Education, October 1980.

Science Indicators 1982: An Analysis of the Stateof U.S. Science, Engineering and Technology,National Science Board, 1983.

“Supply of Engineering Faculty,” Electronic Mar-ket Trends, January 1982.

“The Changing Face of Engineering, ” Special Re-port, Electrom”cs, May 31, 1983, pp. 125-148.

The National Science Foundation and Pre-CollegeScience Education: 1950-1975, report preparedfor the Subcommi ttee on Science and Technol-ogy, U.S. House of Representatives, 94th Cong.,2d sess., by the Congressional Research Serv-ice, Library of Congress, January 1976.

“The Science Education Stampede, ” Science &Government Report, Oct. 1, 1983, p. 6.

“Q and A on Job Market for Science, Engineer-ing, ” Science & Government Report, vol. XII,No. 14, Sept. 1, 1983.

Abernathy, William, Clark, Kim B., and Kantrow,Alan, Indust~”al Renm”ssance: Producing a Com-petitive Future for America (New York: BasicBooks, 1983).

Adler, Mortimer, “The Reform of Public Schools, ”


The Center Magazine, September 1982, pp.12-33.

Aiken, James N., Teacher Supply/Demand 1983:A Report Based Upon an Opinion Survey ofTeacher Placement Officers, Association forSchool, College and University Staffing, 1983.

Atkin, J. Myron, “Who Will Teach in HighSchool?” America Schools: Public and Private,Daedalus, summer 1981.

Bailyn, Bernard, Education in the Forming ofAmerican Society (New York: W. W. North,1980).

Bezdek, R. H., Long Range Forecasting of Man-power Requirements (New York: Institute ofElectrical and Electronics Engineers, 1974).

Blakesless, Sandra, “Teacher Fulfillment PutAbove Pay,” New York Times, Oct. 4, 1983, p.C1.

Botkin, James, Dimancescu, Dan, and Stata, Ray,“High Technology, Higher Education, andHigh Anxiety,” Technology Review, October1982, pp. 49-52.

Boyer, Ernerst L., and Hechinger, Fred, HigherEducation in the Nation’s Service, the CarnegieFoundation for the Advancement of Teaching,Washington, DC, 1981.

Brademas, John, “Graduate Education: Signs ofTrouble, ” Science, vol. 223, No. 4639, Mar. 2,1984, p. 881.

Cohen, David, and Neufeld, Barbara, “The Fail-ure of High Schools and the Progress of Edu-cation, America's Schools: Public and Private,Daedalus, summer 1981.

Cooper, Edith Fairman, U.S. Science and Engi-neering Education and Manpower: Background:Supply and Demand; and Comparison with Ja-pan, the Soviet Union, and West Germany, Con-gressional Research Service, Library of Con-gress, May 24, 1982.

Cooper, Edith Fairman, United States Supply andDemand of Technology, Part I, Current Situa-tion and Future Outlook, Congressional Re-search Service, Library of Congress, Nov. 6,1981.

Crernin, Lawrence, Traditional in American Edu-cation (New York: Basic Books, Harper, 1976).

Curtis, Kent K., “Computer Manpower-Is Therea Crisis?” National Science Foundation, Jan. 4,1983.

Dennison, E. F., SIower Economic Growth (Wash-ington, DC: Brookings Institution, 1979).

Dollar, Bruce, “What Is Really Going On inSchools,” Social Policy, fall 1983, pp. 7-19.

Dupree, A. Hunter, Science in the Federal Govern-ment: A History of Policies and Activities to1940 (Cambridge, MA: The Balknap Press ofHarvard University Press, 1957).

Evangelauf, Jean, “Salaries of New Assistant Pro-fessors Vary More Than $8,000 Across Disci-plines, ” The Journal of Higher Education, Feb.29, 1984, pp. 15-17.

Evangelauf, Jean, “Top Students Move to ScienceStudies, Leave Humanities,” Chronical of High-er Education, Feb. 11, 1984, p. 1.

Finn, Chester F., Jr., “American Education Re-vives, ” Wall Street Journal, July 7, 1982.

Fox, Jeffrey L., “The Uneven Crisis in Science Ed-ucation,” Science, vol. 221, p. 838.

Ginzberg, E., and Voita, G. J., “The Service Sec-tor of the U.S. Economy, ” Scientific American,VOl. 299, March 1981, pp. 48-55.

Haddad, Jerrier A., “Key Issues in U.S. Engineer-ing Education, ” NAE Bridge, summer 1983.

Harris, Marilyn A., “Manpower Surveys Continueto Disagree, ” Electronics, July 28, 1983, pp.108-110.

Herman, Ros, “Science Slops in the Land of Op-portunity,” The New Scientist, July 14, 1983,pp. 110-113.

Heylin, Michael, “High School Science ProblemsGain Spotlight,” Chemical and EngineeringNews, May 24, 1982, pp. 39-41.

Hubbard, Pat Hill, “Technical Employment Pro-jections: 1983-1987, ” paper prepared for theSymposium on Labor-Market Conditions forEngineers, National Academy of Sciences,Washington, DC, Feb. 2, 1984.

Irwin, Paul M., “Education in the 98th Congress:Overview, ” Issue Brief # IB83055, Congres-sional Research Service, Library of Congress,Oct. 28, 1983.

Jordan, K. Forbis, “Precollege Math and ScienceEducation: Issues and Proposals, ” Issue Brief# IB82092, Congressional Research Service, Li-brary of Congress, Aug. 1, 1983.

Jordan, K. Forbis, “Teachers for PrecollegeMathematics and Science Programs, ” Congres-sional Research Service, Library of Congress,July 28, 1982.

Kerr, Clark, “The Uses of the University” (Cam-bridge, MA: Harvard University Press, 1972).

Kutscher, Ronald E., “Future Labor Market Con-


ditions for Engineers, ” paper presented to theSymposium of the National Academy of Sci-ence, Feb. 2, 1984.

Kuttner, Bob, “The Declining Middle,” The Atlan-tic Monthly, July 1983, pp. 60-72.

Layman, John W., “Overview of the Problem, ”Physics Today, September 1983.

Layton, Edwin T., The Revolt of the Engineers:Social Responsibility and the American Engi-neering Profession (Cleveland, OH: The Pressof Case Western Reserve University, 1971).

Levin, Henry M., and Rumberger, Russell W.,Forecasting the Impact of New Technologies onthe Future of the Job Market, Project Report#84–A4, National Institute of Education, Feb-ruary 1984.

Levin, Henry M., and Rumberger, Russell W., TheEducational Implications of High Technology,The National Institute of Education Report #83-A4, 1983.

McDermott, Jeanne, “Technical Education: TheQuiet Crisis,” High Technology, November/De-cember 1983, pp. 87-92.

McDonald, Kim, “Engineering Deans Ask Con-gress to Give Their Field Equal Standing WithScience in Mission of NSF, ” The Chronical ofHigher Education, Feb. 29, 1984, p. 13.

Mansfield, Edwin, Education, R&D, and Produc-tivity Growth, revised, University of Pennsyl-vania, Jan. 31, 1982.

Miller, Harry G., Dean, School of Technical Ca-reers, Southern Illinois University at Carbon-dale, Carbondale, IL, “Higher Education forEmployment: Considerations of Degree Deter-mination for Emerging Technical Careers. ”

Missimer, William C., Jr., “Business and Indus-try’s Role in Improving the Scientific and Tech-nological Literacy of America’s Youth, ” T.H.E.Journal, February 1984, pp. 89-94.

Moshowitz, Abbe, “On Predicting R&D Skill Re-quirements for Information Technology,” un-published paper prepared for the Office of Tech-nology Assessment, February 1984.

Noble, David, America by Design: Science, Tech-nology, and the Rise of Corporate Capitalism(New York: Alfred A. Knopf, 1977).

Olson, Larry, Future Technical Management Re-quirements and Availability in InformationTechnology R&D: United States and Compari-son With Japan, France, and the United King-dom, unpublished paper prepared for the Officeof Technology Assessment, 1983.

Porter, Beverly Fearn, and Czujko, Roman, “Sci-entific Employment in a Tightening Economy, ”Physics Today, February 1983.

Prewitt, Kenneth, “Scientific Illiteracy and Dem-ocratic Theory, ” Daedalus, spring 1983, pp.49-64.

Richie, Richard W., Hecker, Daniel E., and Bur-gan, John, “High Technology Today and To-morrow: A Small Slice of the Employment Pie, ”Monthly Labor Review 106, November 1983,pp. 50-58,

Rosen, S., “Human Capital: A Survey of EmpiricalResearch, ” Research in Labor Economics,(Greenwich, CT: JAI Press, vol. 1, pp. 2-39.

Rosenberg, Nathan, Inside the Black Box–Tech-nology and Economics (Cambridge, MA: Cam-bridge University Press, 1982).

Schultz, T. W., Investment in Human Capital(New York: Free Press, 1971).

Setlzer, Richard, J., “Impact of Technology onEmployment Probed,” Chemical & EngineeringNews, Aug. 1, 1983, pp. 23-24.

Shils, Edward, “The Order of Learning in theUnited States from 1865 to 1920, ” Minerva, vol.21, No. 2, summer 1978.

Silverstri, George T, Lukosiewicz, John M., andEinstein, Marcus E., “Occupational Employ-ment Projections Through 1995, ” Monthly La-bor Review 106, November 1983, pp. 37-49.

Smith, Adam, The Wealth of Nations (New York:The Modern Library, 1937).

Stedman, James B., Education in America: Re-ports on Its Condition and Recommendationsfor Change, Issue Brief #83106, CongressionalResearch Service, Library of Congress, Aug. 5,1983.

Tyack, David, and Hansot, Elizabeth, “Conflictand Consensus in American Education, ” Amer-ica Schools: Public and Private, Daedalus,summer 1981.

Upthegrove, William R., “Engineering Manpowerand Education: A Report of the Business-High-er Education Forum, ” paper presented at theSymposium on Labor-Market Conditions forEngineers, National Academy of Science, Feb.2, 1984.

Vanski, Jeane E., “Projected Labor Market Bal-ance in Engineering and Computer SpecialityOccupations: 1982 -87,” prepared for the Sym-posium on Labor-Market Conditions for Engi-neers held at the National Academy of Sciences,Washington, DC, Feb. 2, 1984.


Veblen, Thorsten, The Theory of the Leisure Class tions of Post Academic Science, Transaction(New York: The Modern Library, 1934). Books, 1982.

Walgerg, Herbert J., “Scientific Literacy and Eco- Welter, Rush, Popular Education and Democraticnomic Productivity in International Perspec- Thought in America (New York: Columbia Uni-tive,” Daedalus, spring 1983, pp. 1-28. versity Press, 1962).

Weinstein, Jay, Sociology/Technology: Founda-

Chapter 6

New Roles for Universities inInformation Technology R&D

Contents

PageIntroduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

A Conceptual Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

Forces Driving New Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

Impacts of New University Arrangements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173Expected Benefits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174Potential Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175

New Roles for Universities: Selected Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176Massachusetts Institute of Technology Microsystems Industrial Group . . . . . . . . . . 177Microelectronics and Information Sciences Center University of Minnesota . . . . . . . 178Rensselaer Polytechnic Institute Center for Industrial Innovation . . . . . . . . . . . . . . . 180Stanford University Center for Integrated Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . 182Microelectronics Center of North Carolina. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184Semiconductor Research Corporation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189Microelectronics and Computer Technology Corporation . . . . . . . . . . . . . . . . . . . . . . . . 193

Chapter preferences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195

TablesTable No. Page26. Stanford University, Center for Integrated Systems Research Topics . . . . . . . . . . . 18527. A Distribution of SRC Funding by Region, March 1983 . . . . . . . . . . . . . . . . . . . . . . 192

FiguresFigure No. Page30. A Conceptual Framework: New Roles for Universities

in Information Technology R&D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17031. Stanford University, Center for Integrated Systems Patent Policy . . . . . . . . . . . . . 18332. Communications System Linking MCNC Participating Sites. . . . . . . . . . . . . . . . . . . 18733. MCNC Participating Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18734. Working Relationships, Microelectronics Center of North Carolina . . . . . . . . . . . . . 188

Chapter 6

New Roles for Universities inInformation Technology R&D

Introduction

Throughout history, new institutional ar-rangements have been created to satisfychanging needs. As one of society’s major in-stitutions, the university, too, has evolved.Particularly in the case of information tech-nology R&D, new roles for universities aredeveloping. Other major players–industry,State and local governments and the FederalGovernment–are involved along with the uni-versity in the formation of new institutionalrelationships.

The new institutional arrangements be-tween university and industry, as well as thoseamong university, industry and government,are driven by many factors. These include theneed for new knowledge and the application

of advancing technologies in new products andprocesses; the efficient use of high-technologymanpower and ensuring a renewable supplyof manpower resources; economic survival andfuture industrial growth; and maintenance ofnational security and defense.

Since these efforts are essentially in theirformative stages, it is difficult to draw con-clusions now about their long-term impacts.In establishing a framework for analysis andpolicy options, OTA developed a series of sev-en case studies (described later in this chapter)that were selected as examples of the rangeof new institutional relationships. Taken to-gether, they illustrate several of the key issuesin today’s debate.

A Conceptual FrameworkOTA created a conceptual framework to

analyze the changing institutional R&D rela-tionships being played by the Nation’s aca-demic institutions–one that emphasizes thepivotal role being played by them. Figure 30outlines this framework and focuses on theconnections among university, industry, andgovernment in terms of education, research,and economic development. At the same timethere are forces converging on these institu-tions that, while creating new opportunitiesand strengthening connections among univer-sity, industry and government, also are cre-ating strains and producing tensions.

While education, research and development,and economic development are separate func-tions, they, like the institutions that fosterthem, are becoming increasingly interrelated.For example, concepts in advanced computerarchitecture taught in a university programor course are directly dependent upon the rap-

id advances in research and development atboth the university and in the industry. Sim-ilarly, the development of new industries andsubsequent economic growth are directly tiedto the products coming from the university—highly trained technical graduates and newknowledge, new processes, and new applica-tions—as well as to the advances and offshootscoming directly from industry.

In examining the institutional players interms of their relationship to education, re-search, and economic development, we see thatboth universities and industry are directly in-volved in the creation of new knowledgethrough research, that both universities andState and local governments are directly con-cerned with the educational process and theprovision of a renewable supply of trainedgraduates, and that both State and local gov-ernments and industry are directly concernedwith economic health and growth.

169


Figure 30.–A Conceptual Framework: New Roles for Universities in Information Technology R&D

*Changing nature of R&D: increasingcomplexity; high cost of research.

*Dwindling resources: faculty andstudents drawn to industry; obsoletefacilities.

UNIVERSITY

New knowledge T r a i n e d m a n p o w e rinnovation and expertise

-

INDUSTRY ECONOMIC GROWTH

1

STATE AND LOCALGOVERNMENT

joint lndustry- ‘ Economic development; need for new● International competitionnational R&D efforts.

Products,servlces,industries; strengthening the tax base.

*Need for cooperation to reduce high * Expanding Iimited resources forrisks, high cost of R&D.

jobs, revenueeconomic development , educat ion, andpublic services,

* Forces driving new Institutional relationships

SOURCE Off Ice of Technology Assessment

There are also important indirect relation-ships. The educational program and knowl-edge resources of the university can have animpact on the economic well-being of the re-gion. Industry is involved in education indi-rectly through the training and resources ithas to offer to the other players. Industry isalso a consumer of talent and new ideas, theproducts of the university. States and local-ities are finding that they need to be concernedwith research not only because the creation ofnew knowledge can lead to the developmentof new processes and products by industry,but because the quality and scope of univer-sity research efforts can strengthen the educa-tional program and in turn provide the region

with a renewable supply of highly trainedmanpower.

There is also a national dimension to thistriad. These relationships and interactions areaffected by and in turn affect national issuesand the Federal role. The strength and effec-tiveness of the educational system, the qualityof research, and the level of economic growthand industrial innovation and productivity de-termine, in part, the Nation’s national secu-rity and economic well-being. The role of theFederal Government has also been both directand indirect. For example, the passage of theMerrill Act in 1862 fundamentally affected thenature of research and education in the univer-

Ch. 6—New Roles for Universities in Information Technology R&D ● 171

sities, as did the direct collaboration amonggovernment, industry and academic research-ers during World War 11.1 Recently, the Fed-eral Government’s role has become more in-direct, increasing authority to State and local— - — —

‘See ch. 6, “The Provision of Education in the United States,”in Information Technology and Its Impact on American Edu-cation, OTA, 1982.

government for various discretionary educa-tion programs, providing a positive climate forindustrial joint ventures, and encouragingjoint sponsorship of R&D through tax in-centives.2

‘See ch. 2, “The Environment for Research and Developmentin Information Technology in the United States. ”

Forces Driving New Relationships

One can argue that the institutional frame-work has been in place a long time, even thoughthe interconnections may not have been sharp-ly defined. Why, then, are the relationshipsamong the university, industry, and State andlocal government increasing in strength andactivity today? Although there are many fac-tors that could be considered, several forcesappear to be critical.

One of the principal factors has been thechange in the direct and indirect roles playedby the four participants shown in figure 30.The changing Federal role in education, re-search, and economic growth has shifted cer-tain areas of responsibility to State and localgovernments, academia, and the private sec-tor. While Federal R&D funding has stayedroughly constant in real dollars over the pastdecade, recent increases have been targeted tospecific areas, such as defense. The consolida-tion of federally funded discretionary pro-grams in education has increased local andState decisionmaking and control.

Universities have been constrained by theresources available for support of research, fac-ulty, students, and facilities.3 The rapid obso-lescence of laboratory and research tools, cou-pled with the highly complex and sophisti-cated nature of the equipment now needed foradvanced information technology research, re-sults in capital costs beyond the reach of mostacademic institutions.4 The retention and at-— . . . . —. . —-

‘W. R. Lynn and F. A. Long, “University-Industrial Collab-oration in Research, ” Technology in Society, vol. 4, 1982, p. 199.

‘For example, in 1970 a need for $200 million in new instru-mentation in the Nation’s university research laboratories wasidentified: a decade later the accumulated need is estimated to

traction of top-quality faculty and the recruit-ment of advanced-level students, many ofwhom are being drawn to industry, are criti-cal problems.5 Further, as information technol-ogy R&D advances, multidisciplinary effortsare required to achieve new breakthroughs.G

Not unexpectedly, universities have had toseek new ways to operate educational and re-search programs.

Given the resources for R&D within the ma-jor information technology corporations,’ it islogical to ask why industries would initiate orbe responsive to new institutional research re-lationships. The change in the scope of the in-formation technology industry from a nationalto a global arena has been a critical factor.Competition within the industry has expandedfrom the U.S. to a new situation where thecompetition derives from nationally coordi-nated industry-government efforts world-wide.8

be more than $1 billion. National Research Council, Revitaliz-ing Laboratory Instrumentation (Washington, DC, NationalAcademy Press, 1982).

‘Louis Branscomb, former Chairman of the National ScienceBoard, points to needs for advanced degree training in com-puter science, electrical engineering, polymer science and

. materials engineering-a problem which requires both fellow-ship support and the strengthening of university instructionaland research facilities, At the same time, he points out thatincentives must be provided to make university research careersas attractive as offers from industry-principally through theprovision of research and equipment support. See: L. Branscomb,“The Computer’s Debt to Science,” Perspectives in Computing,vol. 3, No. 3, October 1983, p. 18.

61 bid, pp. 13-15. See also ch. 3, Case Studies on AdvancedComputer Architecture, Fiber Optics, Software Engineering,and Artificial Intelligence.

‘See table 23, R&D Intensities of Selected Major U.S. Tele-communications Firms, 1982, ch. 4, Divestiture.

Wee ch. 7. Information Technology R&Din the United King-dom, France, and Japan, for examples of such efforts.


Based on the perception that industrialgrowth, productivity and competitiveness aredependent on new knowledge and innovation,and a renewable supply of highly trained man-power, industries have turned increasingly tothe universities. The increased cost of R&Dmakes cooperative efforts highly desirableamong the industries themselves, and amongcosponsored efforts with universities. Such co-operation goes beyond cooperation amonglarge companies. It includes cooperation be-tween large and small companies, and amongbusiness, academia, and government.9 In sum,given the rapid advances in technology, theescalating costs of R&D, and the global inten-sity of competition, intranationa. 1 cooperationis seen as a means of maintaining internationalcompetitiveness, 10 and universities are seen ascornerstones of the cooperative effort.

Over the past two decades, several regionsof the United States have developed stronglocal economies based on high-growth, tech-nology-based firms that are engaged in sys-tematic development and commercializationof new products, processes, and services.These firms, and the industries they represent,have provided a major source of new jobs inthe manufacturing sector.11 12 Thus States, as

‘See for example, testimony by Erich Bloch, Vice President,IBM, and Chairman, Semiconductor Research Corp. “In orderto cope with increasing competition in the world market, thesemiconductor industry must increase its efforts in researchand development. At the same time the research tasks are be-coming more complex and more capital-intensive; lead time isincreasing and the shortages of sufficiently trained manpowermake the staffing of needed projects difficult. For all these rea-sons, some research efforts are beyond the affordability of in-dividual companies. ” Hearings Before the Subcommittee on In-vestigations and Oversight and the Subcommittee on Science,Research and Technology, House of Representatives, 98thCong., 1st sess., June 29-30, 1983, Japanese Technological Ad-vances and Possible United States Responses Using ResearchJoint Ventures, p. 46.

IOW. B. Norris, “How to Expand R&D Cooperation. ” 13usi-ness Week, Apr. 11, 1983, p. 21. Keynote Address “Coopera-tion for Improving Productivity, ” San Diego, July 20, 1983,IEEE Task Force on Productivity and Innovation.

“U.S. Congress, Office of Technology Assessment, Technol-ogy, Innovation, and Re~”onal Economic Development, Censusof State Government Im’tiatives for High Technolo~ IndustrialDevelopment-Backgmund Paper OTA-BP-STI, May 1983; En-couraging High-Technology Development—Background Paper,OTA-BP-ST1-25, February 1984.

12Tot~ employment in the information technology industryexperienced considerable growth in the decade between 1972and 1982, despite economic recessions. See ch. 9.

well as local communities have become activecompetitors in seeking to attract high-technol-ogy firms. Just as U.S. industries have hadto acknowledge the international change in thecompetitive forces for their products, so toohave State and local governments had to rec-ognize that the competition for high-technol-ogy industry is interregional. The intensiveState bidding for location of the Microelec-tronics and Computer Corporation (MCC) issuch an example, with some 60 mayors and27 governors involved. Notes Arizona Gover-nor Bruce Babbitt,

The great MCC bidding war marks a spe-cial chapter in American industrial history.State and local governments across the coun-try have discovered scientific research andtechnological innovation as the prime forcefor economic growth and job creation. Andlocal officials have also uncovered a broadbase of public interest that can be translatedinto support for aggressive action programs.With the exception, perhaps, of the post-Sputnik era, such grassroots enthusiasm forscience and technology has not been seensince the Gilded Age of the 19th century,when communities vied to finance the trans-continental railroads. *

To attract such industry, incentives such astax breaks, donations of real estate, venturecapital for industry and funding for educa-tional programs have been provided. ’3 Not allState and local high-technology initiativeshave focused on education, nor does everyState or locality have equal resources on whichto draw. However, a strong educational baseis seen as a way of becoming more competi-tive. In a survey of 691 high-technology firms,completed for the congressional Joint Eco-nomic Committee, “the importance of skilledlabor points up the necessity of linking Stateand local development efforts with a region’suniversities in order to attract high-technology

*B. Babbitt, “me Sta@s and the Reindustriahzation Of Amer-ica,” Issues in Science and Technolo~, fall 1984, p. 85.

“For examples of such efforts, see the report by the NationalGovernor’s Association, Zkwhnology and Growth: State lm”tia-tives in Technolo~”cal Innovation (Washington, DC: NationalGovernors’ Association, 1983), pp. 23-45.

Impacts of New University Arrangements

While we can find examples of State and lo-cal high-technology initiatives and numerousexamples of long-standing university-industryinteractions including industry’s support forresearch through gifts of funds or equipment;cooperative research grants and contracts; theuse of university consultants, exchange of per-sonnel between universities and industries and

other arrangements,15 the university today isin a special position. Universities are beingcourted by all of the principal actors and manyare initiating programs of their own. Most im-

Foundation, ResearchRelationships (Washington, DC: National Science Foundation,1982).


portantly, they are the linking element in mul-ti-institutional R&D relationships.

The new institutional relationships takemany forms; some efforts represent new andlargely experimental ways of working togeth-er; many efforts that are being developed arenot really new, but are evolving from previ-ous efforts and relationships. However, all ofthe efforts that OTA examined involve a setof agreements whose principal characteristicsare multidisciplinary arrangements and com-mitments to research with long-term objec-tives. While several of the ventures have beeninitiated by one or few individuals, the nego-tiated agreements themselves are made at theinstitutional level. It is the level of commit-ment and the extent of the involvement thatdiffers from previous university-industryefforts.

It is too early to know with any certaintythe benefits and costs of the new university-centered activities. However, in breaking newground, the university arrangements, cospon-sored efforts, and high-technology State andlocal initiatives have generated high expec-tations amid questions of appropriateness.l6

The number of meetings, conferences, hear-ings, and publications on this subject has beensignificant. 17 The debates over these re la t i on -ships have involved university leaders andacademicians, governors, congressmen, andcorporate executives.

-.— — .— —“This concern has been most focused on biotechnology, where

se~reral university-industry agreements have involved largesums of funding, over multiyear periods, and where a majorresearch unit of the university is involved with a single com-pany with varying agreements for industry participation oncampus, and on some agreement to delay publication or pro-vide exclusive licensing to processes and products developedduring the duration of the research agreement.

174’ Academe and Industry Debate Partnership, ” Science, vol.219, No. 4481, January 1983, pp. 150-151; T. W. Langfitt, S.Hackney, A. P. Fishman, et al., Partners in the Research Enter-prise, University-Corporate Relations in Science and Technol-ogy (Philadelphia: University of Pemsylvania Press, 1983). U.S.House of Representatives, University/Industry Cooperation inBiotechnology. Joint hearings of the Subcommittee on Inves-tigations and Oversight and the Subcommittee on Science, Re-search, and Technology, June 16-17, 1982.

Expected Benefits

The increased interactions among the uni-versity-industry-government triad depicted infigure 30 underlie potentially successful ap-proaches to several critical problem areas:

●

●

●

Research and new knowledge: The cou-pling between university and industrymay match needs of both. Industry getsaccess to the research and knowledge thatresides in the university. Similarly, uni-versity researchers can benefit from thepool of industry expertise. The academiccommunity obtains R&D laboratory facil-ities and research tools, as well as fund-ing to undertake research. With increasedinteraction, the university has an oppor-tunity for better understanding of the in-dustry’s practical concerns, and, con-versely, industry may get a closer look atthe university’s research findings, speed-ing technology transfer.Education and manpower: With the in-creased opportunities for research, and astrengthening of the academic researchprogram, the educational program canalso be affected positively. Incentivesthat would attract and retain top-levelfaculty and advanced graduate studentsare derived from higher levels of supportof new facilities and research. Moreover,as a result of interacting with industrypersonnel, students can make more in-formed decisions about their future em-ployment. The combination of top-levelpersonnel, adequate facilities, and a vig-orous research agenda can strengthen theeducational program, as new courses aredeveloped and learning opportunities in-crease. The university products can thenfeed back into both industry and the com-munity.Economic growth: New institutional ef-forts are aimed at a strengthened researchbase and a renewable source of highlytrained manpower, which are needed byindustry for its economic growth. This, inturn, can strengthen the regional econom-

Ch. 6–New Roles for Universities in Information Technology R&D ● 175

ic base through new jobs, new spin-off in-dustries, and the continued developmentof entrepreneurial efforts to fill newniches. 18

Potential Costs

The increased interactions among academia-industry-government also raise questions oflong-term impact:

● Research and new knowledge: Closer in-teraction may lead to subtle changes inthe setting of research goals both in termsof the selection of topics to be studied andin the shortening of time horizons for re-sults. Breakthroughs in fundamental re-search often require decades of study.Thus some problems may be overlookedif the institutional time frames for re-search are 3, 5, or even 10 years. How willtopics not of direct interest to industrybe covered in such institutional ar-rangements?

In addition, industry’s traditional em-phasis on secrecy is in direct conflict withacademic practices. The new arrange-ments have involved extensive negotia-tion regarding patent and licensing agree-ments.l9 Thus far, most universities haveresisted industry pressure to limit accessto research results, maintaining their tra-ditional role “to protect and to foster anenvironment conducive to free inquiry,the advancement of knowledge and thefree exchange of ideas. “2° The conflict be-tween openness and control is of continu-ing concern,21 and will require solutionson a case by case basis.

“For example, in a recent study of the Route 128 High Tech-nology Industrial Corridor, Massachusetts’ advantage h at-tracting and supporting industries has resulted from the richuniversity environment. “Entrepreneurs in the electronics fieldscome mainly from the staffs of universities and their researchlabs, and from other high tech firms (already established).” N. S.Dorfman, “Route 128: The Development of a Regional HighTechnology Economy, “ Research Policy, 12:6, December 1983,p. 309; see also table 1, p. 301, ibid.

1’Fowler, “University-Industry Research Relationships: TheResearch Agreement, ” Journal of College and University Law,9:4, 1982-83.

‘“A. B. Giarnetti, “The University, Industry and CooperativeResearch, ” Science, vol. 218, December 1982, pp. 1278-1280.

“D. Nelkin, “Intellectual Property: The Control of ScientificInformation, ” Science 216:14, May 1982, pp. 704-708.

Education and manpower: It is possiblethat these new, highly visible, excitingventures will cause competition betweenresearch and education, drawing facultyaway from teaching, and recruiting stu-dents from other areas. There is the dan-ger that these new efforts will skew thebalance among programs and capture un-equal attention and support from univer-sity administration.Economic growth and development:While there are many other joint indus-try-university activities involving smalland large businesses, a variety of academ-ic institutions, and individual facultymembers, the industry sponsors and ma-jor participants in both the multidiscipli-nary university centers and the industrycooperative ventures have mainly beenthe large information technology corpora-tions. Fewer numbers of smaller compa-nies have joined projects as ‘‘associate’(in contrast to fully participating) mem-bers. Full membership includes access toresearch as well as personnel exchanges,and active participation in planning re-search, selecting proposals for fundingand evaluating ongoing programs. Thus,cooperative joint-industry ventures amongthe information technology giants mayput the smallest, entrepreneurial compa-nies at a disadvantage.

The costs of increased interaction come fromtwo directions. First, in coming together, eachof the institutions may lose some measure ofautonomy and relinquish aspects of their tradi-tional roles. Conflict is inevitable, for exam-ple, between the university’s need for opennessand industry’s need for protection of proprie-tary interests. Another conflict may developif the traditional distinctions that have sepa-rated the use of public funds from private in-terest and gain are blurred.22

Second, there is the “cost” of nonparticipa-tion. Most of the debate has focused solely on

22’’ WeighinK the Social Costs of Innovation, ” Science, vol.223, Mar~h 1384, p. 1368. A suit involving the University ofCalifornia and its research, raises the question of the legalityof spending public funds for research that allegedly benefitslarge agribusiness more than small farmers and laborers.


the generic institution without recognition ofthe differences within each of the institutionalcommunities. Thus, it has not looked verymuch at how the establishment of universityindustry relationships may affect a smaller orless prestigious research institution, nor howjoint-industry ventures may affect smallerbusinesses, nor how local and State initiativesplace some other regions or other institutionswithin the region in an inequitable position.

There is concern that the already existingdifferences between the Nation’s top-tier andsecond-tier universities may grow even great-er as the competition for industrial resourcesand partnerships accelerates. Thus, there isconcern for the needs of the range of the Na-tion’s universities. Even though programsmay be less ambitious in scope and scale, theneeds for sophisticated equipment, advancedresearch facilities, highly trained and knowl-edgeable faculty, and advanced-level students

New Roles for Universities:

University-industry-government informationtechnology R&D efforts demonstrate a varie-ty of approaches which have been only recent-ly implemented. Because these efforts are es-sentially in formative stages, an assessmentof their effectiveness and impact on R&D ispremature. Yet it is clear that the efforts ex-amined in this chapter provide important ex-amples of new directions and major commit-ments. The case studies provide examples ofhow problems or barriers raised by these newrelationships are being addressed, as well asthose issues which are not yet resolved. Thecase studies also provide an understanding ofthe motives and factors stimulating change,how the institutional players are responding,and the role played by incentives and direc-tions from the Federal Government. Thus,they provide a framework for analysis and thedevelopment of policy options.

While each of the case studies is unique, sev-eral themes emerge:

remain critical. Therefore, a diversity of effortsand approaches needs to be explored and sup-ported.

At the State and regional level, the abilityto compete for new industry, for research cen-ters of excellence, and for expert manpowermay also become equity issues. As noted ear-lier, while the number of high-technology cen-ters has increased in recent years, the competi-tion among State and regional localities isbecoming fierce. The ability of States to assuresignificant support for new facilities, addition-al faculty and graduate students, the availabil-ity of venture capital, and the cooperative ef-forts of business and academic leaders arecritical factors in attracting new institutionalresearch ventures. Moreover, as a result of suc-cessful bids, regions expect to attract otherhigh-technology companies while the univer-sities hope to attract senior faculty and thetop graduate students.

●

●

●

●

Selected Case Studies

Institutional arrangements involve long-term, multiyear commitments with agree-ments that include facilities, equipment,and human resources.These arrangements bring together multi-disciplines, multi-institutions, and multi-funding resources to support wide-rangingresearch, educational, and developmentefforts.These arrangements involve leadershipand support of individuals at the highestlevels of the university, corporations, andgovernment.While Federal funds continue to supporta significant portion of the research at theuniversity centers, the Federal Govern-ment had a limited role in developmentof the institutional arrangements, influ-encing them by providing limited fundsfor startup activities, by creating taxcredit incentives, and by its supportivepolicy towards joint ventures.

Ch. 6–New Roles for Universities in Information Technology R&D ● 177

Massachusetts Institute of TechnologyMicrosystems Industrial Group

MIT’s experience with joint industry ar-rangements is extensive and reaches virtuallyevery program area in the Institute. Morethan 300 joint programs are currently spon-sored at MIT by industrial firms. Federalsources nonetheless provide the bulk of fund-ing for research. In 1983, only 10 percent ofMIT’s sponsored research was funded by in-dustry.23 It is estimated that support from in-dustry will not grow beyond 15 or 20 percent.However, MIT faculty point out that indus-try involvement is important because it pro-vides exposure to and understanding of indus-trial concerns, motivations, and needs. Thisinteraction is seen as critical for the future ofmost students who graduate from the univer-sity to work in the information technologyfield. Equally important, it provides necessaryexpansion of academic research concerns thatwould otherwise be guided primarily by theinterests of particular Federal funding agen-cies. 24

While there have been long and well estab-lished industry-university research ties atMIT, the formation of the Microsystems In-dustrial Group breaks new ground. Increasedindustrial involvement in the MIT Microsys-tems program was stimulated by the pro-gram’s need for advanced state-of-the-artequipment and laboratory facilities. A propo-sal to reach out to industry for help in devel-oping these facilities was made by membersof the faculty, who argued that the amountneeded (originally estimated at $10 million)could not be supported by any Federal pro-gram or by the university itself. The advancedresearch Very Large Scale Integration (VLSI)laboratory and facilities are supported by in-dustry sponsors; contributions are estimatedto be $10 million (half of the actual cost of ren-ovation and operation). The rest of the cost isbeing recovered through overhead charges oncurrent contract research. In agreements nego-tiated with industry, full member companies-— —-— ———

“Kenneth A. Smith, “Industry-University Research Pro-grams,”’ Physics Today, vol. 37, No. 2, February 1984, p. 24.

“Ibid., p. 24.

contribute $250,000 annually for 3 years, andassociate member companies contribute $50,000annually for 5 years.

Thus far, 18 companies have joined thegroup. 25 Full member companies can send onetechnical staff member to MIT to work in theMicrosystems research program annually for3 years. Each visiting professional submits aplan of proposed research topics, and these arematched with an appropriate faculty memberor research group. More than a dozen indus-try people have participated in the program.The opportunity to work in a new area of in-terest, “get caught up in the MIT atmos-phere” and interact daily with students andfaculty members is viewed very positively .2’

Research projects under way have been de-veloped by faculty and reflect their traditionalroles as principal investigators. The directorand faculty meet with the industry memberadvisory group, who provide information andadvice. Even more directly, the technical peo-ple from industry have contributed to the re-search efforts, and have broadened the viewof faculty and students. The director of theMicrosystems Industrial Group explains:“These are smart people with different back-grounds than my University colleagues. It isvery important for those of us who do researchto have the industrial viewpoint in front ofus. ‘ ’27

U n d e r s t a n d i n g g r o w s b o t h g e n e r a l l ythrough interaction with the Council of mem-ber companies, who offer advice and guidance,and in the process of working out specific vis-iting relationships. This understanding helpsfaculty, students, and the academic program.But it is also clear to faculty and administra-

z5Full member Compwies included AT&T Bell labs, Di@~Equipment, General Electric, General Motors, GTE, Harris,IBM, Raytheon, and United Technologies. Associate membercompanies are Analog Devices, GCA, Genrad, NCR, Polaroid,Sanders Associates and Teradyne.

*’Personal communication, March 1984. According to PaulPenfield, Director, MIG, these experiences provide industry as-sociates unique opportunities for professional growth and de-velopment, and appear to be a way for a company to retainhighly valued employees.

“Personal communication. Paul Penfield, Director, Microsys-tems Industry Group.


tors who have responsibility for university-industry research that, even though time andcompetitive pressures are high, the academicintegrity of both the research and educationalprogram must not be compromised. Thus, insome cases, decisions are made not to under-take certain projects, for example, when theproprietary stakes are too high, or the time-frames are inappropriate or if scientific ex-change is jeopardized.28 From industry’s pointof view, the Microsystems Industrial Group,as well as other similar efforts such as the Cen-ter for Integrated Systems at Stanford, areworking because the research effort isfocused. 29

New institutional relationships can benefitboth university and industry if the agreementsmeet the needs of the partners. In analyzingthe aspects of such negotiations at MIT, theAssociate Provost and Vice President for Re-search identifies the fundamental issues to beaddressed:

1.

2.

3.

4.

the relevance of a proposed line of inquiryto the essential missions of the universityand the industry—maintaining a balancebetween the pursuit of research as an inte-gral part of the educational process andindustry’s need for useful knowledge tobe applied in the development of prod-ucts, processes and services;the organization of a program that meetsthe different time constraints of industryand the university—accommodating themultiyear efforts of graduate studentswith the shorter time pressures of themarketplace;the issue of proprietary rights versusopenness—achieving openness and freeexchange of research results while protec-ting the industrial partners’ proprietaryrights;the issue of patents and copyrights—determining licensing agreements thatadvance scientific and technological dis-coveries in ways that are most likely to

‘“Personal communication, George Dummer, MIT, March1984.

*gPersonal communication, Bill Nelson, GTE, April 1984.

5.

benefit the public and the research par-ticipants and institutions; andthe issue of conflict of commitment—assuring that faculty are primarily com-mitted to the university: its research andits educational programs.30

Microelectronics and InformationSciences Center (MEIS)University of Minnesota

The Microelectronics and Information Sci-ences Center (ME IS) is a joint endeavor be-tween the University of Minnesota’s Instituteof Technology and Minnesota industry. It wascreated to establish a center of excellence inthese sciences as well as to meet local indus-try’s technical manpower needs. Such joint ef-forts are not new to Minnesota.31 The impetusfor the Microelectronics and Information Sci-ences Center came from Minnesota industries—Control Data, Honeywell, 3-M, and SperryCorp. —who committed $6 million to launchthe effort. The Minnesota State legislatureallocated an additional $1.2 million. Currentoperation is at $2.5 million a year matched by$4.0 million in external grants and contracts.

Faculty members, university officials, cor-porate executives and center administratorshave worked together to define the directionsfor research and educational programs, thecenter’s operation, and the university-industryinterface mechanisms. This negotiation tooktime to work out, and programs were phasedin gradually over a several-year period.32 The

‘°K. A. Smith, “Industry-University Research Programs, ”op. cit., p. 25.

31 For example, in the early 1970s such a joint effort initiatedthe development of the Minnesota Educational Computing Con-sortium to provide instructional time-sharing capability to theState’s colleges and universities, as well as the elementary andsecondary schools. See: a case study of “Minnesota Schools andthe Minnesota Educational Computing Consortium, Informa-tional Technology and Its Impact on American Education,(Washington, DC: U.S. Congress, Office of Technology Assess-ment, OTA-CIT-187, 1982) pp. 214- 221.

32The Center’s slow start has been criticized by some. How-ever, the benefits of taking the time to work out an arrange-ment that suited the needs of both the university and the spon-soring industries outweighed the costs of delay. Personalcommunication, Dr. Martha Russell, March 1984,


results of the negotiations are embodied in theME IS Center goals:

1. to sponsor and conduct research at thefrontiers of microelectronic and informa-tion sciences;

2. to strengthen the course offerings of theUniversity of Minnesota in these sciences;and

3. to provide active interplay between uni-versity researchers who seek discoveryand industrial firms that apply those re-search results to the development andmarketing of innovative products andservices.

Research

Through the sponsorship of interdisciplin-ary research projects, the center links faculty,students, and industry. Proposals for researchare submitted by faculty members. ME IS-sponsored research is reviewed annually bytechnical experts at the university and sup-porting companies. In 1983, projects in 3-DIntegrated-Circuits, Processor Array Conceptsfor Engineering, Design Automation and Soft-ware Engineering, and Ultrasmall ElectronicResearch received ME IS seed funding, total-ing $625,000; additional research funds of$4,312,000, principally from Federal grants,was obtained. In 1984, MEIS has awardedboth seed and matching funding to three inte-grated team efforts in Intelligent Systems Re-search, II I-V Semiconductor Materials andHigh Speed Devices, and High-PerformanceIntegrated Circuits. Another planned effortwill include a project on Artificially StructuredMaterials.

Renovation and development of laboratoryfacilities has been directly tied to the researchefforts. The University is planning a new Com-puter Science and Engineering building whichwill house both offices and laboratories. ME ISco-owns, with Argonne National Laboratory,a Synchrotrons X-Ray Beamline Facility, lo-cated in Stoughton, WI. In addition, MEISand the University share the newly remodeledmicroelectronics laboratory and the VLSIengineering design laboratory.

Education

Strengthening the educational program hasfocused on increasing the number of facultymembers, starting new courses, and attractingtop-quality graduate students. Eighteen newgraduate and undergraduate courses havebeen added in computer science, electricalengineering, materials science and chemicalengineering; seven new faculty members havealready been recruited through a 3-year cost-sharing program with the university, andplans call for hiring an additional four mem-bers in Computer Science as well as a direc-tor; 54 graduate students and four post-doc-toral assistants were supported by MEISfunds in the five departments receiving ME ISresearch sponsorship; 16 fellowships will beavailable for 1984-1985.33

Technology Transfer

The exchange of knowledge and technologybetween MEIS member companies and theuniversity community has been a major goalof the Center. Through direct scientist-to-scientist interaction, it is anticipated that thetime between discovery and application willbe shortened. Faculty, students, and industrytechnical staff have worked jointly on projects,in some cases using industry’s state-of-the-artfacilities for design, special fabrication ortesting. A major assumption is that graduatestudents serve a key role in the transfer oftechnology between industry and university.After the first year of graduate study in thedoctoral program, students work in the re-search laboratories of the industry sponsors,learning what drives industrial use of innova-tion in science and technology, and bringingtheir recently acquired knowledge and skillsto the task. Research projects developing fromthese experiences expand the involvement offaculty, students, and the industrial scientists.

Like other joint efforts, the center has fos-tered the exchange of ideas through confer-

—“Microelectronics and Information Sciences Center, 1983 An-

nual Report, February 1984.


ences and seminars. ME IS technical reportsand newsletters have also been widely dissem-inated. Center participants from both academ-ia and industry point out that this open ex-change has been facilitated by concentratingon long-term research areas conducted over a5 to 7 year period. The Center has thus faravoided the issues regarding exclusive re-search and proprietary information.

Continued ability to recruit high-qualitygraduate students, as well as recruiting andretaining excellent faculty, is critical to thelong-term stability and growth of the program.Stable funding and full implementation of pro-grams are anticipated by 1985. In addition,the Center expects to attract additional Stateand private support.

Rensselaer Polytechnic Institute Centerfor Industrial Innovation

The RPI Center for Industrial Innovationis the result of a focused University initiativethat has involved key participants from aca-demia, industry, and New York State govern-ment—including the governor. This initiativewas based on the experiences of RPI’s threeestablished Centers for Interactive Graphics,for Manufacturing Productivity, and for Mi-croelectronics. With a $30 million interest-freeloan from the State and an additional $30 mil-lion commitment by RPI, construction of a fa-cility to house these Centers is under way.These Centers involve more than 100 arrange-ments and agreements with industry, includ-ing support for ongoing research through indus-t r y a f f i l i a t e s , s p e c i f i c r e s e a r c h a n dproblem-solving agreements, continuing edu-cation and training, adjunct industry-facultyarrangements, faculty-industry consulting, in-dustry fee payments, and gifts or loans ofequipment and software. However, it was notthe quest for industrial partnerships, butrather the desire to improve the undergrad-uate engineering education program, thatserved as the initial catalyst for these activi-ties. 34

— .— ——-— ...—“G. M. I.ow, “The Organization of Industrial Relationships

in LJni\’ersities, ” Partners in the Research Enterprise, ‘I’. W’.I.angfitt et al. (wis. ), (Philadelphia: Uni\’cv-sity. of Pennsylvaniapl.(=$s, 1 983), P. 71

The effort to improve RPI’s educational pro-gram was begun in 1975-76 and resulted inseveral new interrelated directions: expansionof the graduate program (from approximately500 to a goal of 2,400 graduate students bythe year 2000); an institutional commitmentto research through the expansion of facultyand facilities as well as of the number of stu-dents, and a revision of the undergraduate cur-riculum to overcome the lack of hands-on engi-neering experiences.

Center for Interactive Graphics

The first step was the creation of an interac-tive computer graphics laboratory designed asa service facility for undergraduates. This wasbased on the belief that an important emerg-ing tool for engineering was the interactivecomputer graphics terminal. The facility andthe applications have grown beyond the origi-nal classroom to a Center for InteractiveGraphics. The growth was due not only to theincreased use in almost all engineeringcourses, but also to the decision to combineresearch with practice as the means for keep-ing up to date with the advancing technology.

The Center for Interactive Graphics was cre-ated in 1978, with initial funding from the Na-tional Science Foundation. From its inception,it was intended to involve industry, and toshare research results with industry. A meas-ure of its success is that the Center has grownfrom 20 supporting companies with $20,000annual fees to 35 companies with $40,000 an-nual fees. An early concern that it would notbe possible to keep up with the continually ad-vancing hardware and software has been re-duced: companies have been willing to donatetheir latest equipment. Just recently, for ex-ample, the Center received a $3 million equip-ment grant from IBM.3S

Center for Manufacturing Productivity

The Center for Manufacturing Productivityand Technology Transfer was the result of adeliberate decision to train students in areas

—— —..-——‘ Personal communication, Dr. Christopher I,eMaistre, Direc-

tor Center for Industrial Innovation and ~lssistant Dean, Schoolof I’engineering, March 1984.


Graduate student and faculty member using aCAD/CAM workstation, in the Center for

Interactive Computer Graphics, RPI

that would be needed in the decade of the1980s and at the same time to meet industry’sresearch needs so that industrial fundingwould be available. Support from Federalsources was not available. Thus, the Dean ofthe School of Engineering made initial con-tacts with executives from General Electric.He and two GE executives traveled to Europeto see how industry and universities wereworking together. RPI’s Center was modeledafter one at the University of Aachen in WestGermany.

With a commitment from GE, the Centerstarted in May 1979. Presently there are eightfounding companies and five affiliate compa-nies who support the operation of the Centerand its research.36 In addition, the Center en-gages in contract work and involves under-graduates and graduates in “real life, realtime” industrial problem solving, all under thedirection of faculty and a project manager. Theintent is to create experiences that are directlyrelevant for students’ entry into the industrialworld. The Center reports to and receives guid-—- - . -—.

“Member companies include General Electric, General Mo-tors, Boeing, Norton, IBM, Alcoa, Digital Equipment Corp.,and United Technologies. Affiliate companies include Kodak,Cincinnati Milacron, Fairchild Republic, Fairchild Schlumber-ger, Altech, and Timex.

ance from a board of advisers comprised offounding member company representatives.

Center for Integrated Electronics

The Center for Integrated Electronics alsohas industrial sponsors, who support researchefforts which are fully open with no restric-tions on disclosure.37 The Center is equippedwith $8 million in hardware, much of it do-nated, from companies such as IBM, Calma,and Computervision. RPI also provides “in-cubator space” for small fledgling companieson campus and provides administrative sup-port to help companies while they find venturecapital, develop management capability, andbegin to grow.

In addition, RPI has created an industrialpark, located 10 miles south of the institute,on a 1,200-acre parcel of land owned by RPI.With strong leadership from then RPI Presi-dent George Low the Institute committed $3million for initial preparation of the site in1982. National Semiconductor was the firstcompany to move in, with several others fol-lowing. RPI is currently constructing its ownbuilding in the park to provide startup spacefor companies that are not yet large enoughto be on their own.

Center for Industrial Innovation

All of these activities led to the Center forIndustrial Innovation. The RPI President andthe Chief Executive Officers of GE and Kodak,along with other corporate executives, metwith the Governor to push for a State Tech-nology Initiative to be funded through theState legislature. The arguments, as in otherindustrial States, were that the smokestackindustries were dying, new technology indus-tries were locating elsewhere, and that to over-come this, new catalysts were needed. RPIargued that it had the necessary infrastruc-

. —

3TFounding members include Harris COW., Computervision,Digital Equipment Corp., Eastman Kodak, General Electric,Raytheon, Polaroid, GTE, IBM, Phoeti Data Systems, EatonCorp., and AIR Products. Affiliate members include Sperry,Xerox, Hewlett-Packard, Perkin-Elmer, Fairchild Semiconduct-or, BTU Corp., Matheson, I’IT, and the PEW Memorial Trust.


ture in place and that what was needed wasa $30 million interest-free loan.

The ground has been broken for the Centerat RPI and the university expects completionof the facility in September 1986. RPI expectsto put an additional $30 million into the facil-ities. In return for the State loan, RPI will pro-vide an outreach program to 2-year collegesand to industry to upgrade the level of tech-nological expertise.

While there has been strong support forthese activities and agreement that they havehelped the Institute, their industrial orienta-tion does cause concern to some faculty. TheCenter argument is, however, that there is ahealthy balance between uncommitted re-search support and both focused engineeringand applied projects. There is evidence thatthe program brings together a blend of re-search and application for students, and thatthe quality of the instructional program hasbeen significantly improved.38

Stanford University Centerfor Integrated Systems

Like the MIT, RPI, and University of Min-nesota efforts, the institutional relationshipdeveloped between Stanford University andindustry breaks new ground. In May 1983,more than the construction of a $15 million fa-cility to house the Center for Integrated Sys-tems (CIS) was being celebrated. Accordingto participants from the faculty, universityadministration, and industry, this project andothers like it are part of a new willingness byindustry and the academic community to be-come “allies in basic research.”39 According toWilliam Hewlett, “CIS is a clear and distinctanswer to three major problems that face theUnited States–the failure of our national pro-grams of basic research to keep pace with theneeds of our universities and industries, theneed to strengthen our system of education,and the challenge to U.S. trade and technol-ogy posed by foreign countries.”40

38Low, op. cit.“F. H. Gardner, “Special Report: The Center for Integrated

Systems, ” Hewlett-Packard Journal, November 1983, p. 23.‘“I bid.

Planning Activities

As early as 1977, Stanford engineering fac-ulty discussed the idea of a center for inte-grated systems research, a multidisciplinaryendeavor involving the interaction of peopleknowledgeable about integrated circuits withanother group knowledgeable about computerand information systems. “From the outsetit was clear that a collaboration, more intensethan had ever before occurred, between ICtypes and systems engineers needed to evolve.”41

Furthermore, such a center could vertically in-tegrate the research process, with a state-of-the-art facility for design, fabrication and test-ing of VLSI chips. This fast-turnaround facil-ity would allow a systems designer, in collab-oration with an IC designer, to createexperimental devices in a shorter time thanever before.

The faculty took their idea to the Dean ofthe School of Engineering, and subsequentlya formal proposal was submitted to the uni-versity. By January 1980, the Center for In-tegrated Systems was under way with approv-al from the Board of Trustees. Executives ofHewlett-Packard, TRW, Xerox, and Intelformed a development committee to raise funds.By March 1981, 10 corporations agreed to con-tribute $750,000 each, spread over a 3-yearperiod. By 1983-84, an additional 10 sponsorsbrought the total to 20, with each also agree-ing to provide $100,000 annually for educa-tion, research, and administration of the facil-i t y .42

Formulating New Policies

Not unexpectedly, the most controversialaspect of the plan was not the facility, but theintention to involve industrial companies assponsors of the Center and offer them “facili-tated access” to the research program. Of con-

——- —- —.—. .“John G. Linvill, Director Industrial Programs, Center for

Integrated Systems, Stanford University. Personal communi-cation, April 1984.

‘zCorporate sponsors are General Electric, Hewlett-Packard,TRW, Northrop, Xerox, Texas Instruments, Fairchild, Honey-well, IBM, Tektronix, Digital Equipment Corp., Intel, 11’”1’,GTE, Motorola, United Technologies, Monsanto, Gould/Amer-ican Microsystems, Inc., North American Philips/SigneticsCorp., and Rockwell International.

Ch, 6—New Roles for Universities in Information Technology R&D . 183

cern to faculty members and corporate spon-sors was how patent ownership, licensing, andintellectual property rights would be deter-mined. The companies’ initial posture was in-sistent on exclusive proprietary rights to in-ventions which involved their own people. Theuniversity, on the other hand, argued that re-search in the Center was funded principallywith Federal support and that the Univer-sity’s legal obligation was to make sure thatany resulting patents would be brought intothe stream of commerce as quickly as possi-ble, with any company capable of commercial-ization having the right to bid and obtain alicense.

The successful resolution of the issue cen-tered on categorizing the patent in terms ofthe inventor (see fig. 31, CIS Patent Policy).In addition, if a corporate visiting scholar de-velops a patentable product jointly with aStanford faculty member or student, the com-pany may request a 90-day delay of publica-tion to get the patent filed; it also gets freebut nonexclusive rights to exploit the product.CIS sponsor companies have also agreed to across-licensing plan, sharing any inventionsdeveloped at CIS. While intellectual propertyrights have generated much discussion andhave taken time to work out, it appears un-likely that highly commercial applications willresult, given the basic nature of the research

under way at CIS. Moreover, both sponsorsand faculty point out that it is in their inter-est that the Center produce new basic knowl-edge as well as students trained as broadly aspossible.

Another area of concern focused on the ques-tion of research direction: would the natureand direction of basic research be distorted byindustrial sponsorship? The answer appearsto be that this is unlikely, given the traditionof independent research teams led by principalinvestigators. At the same time, the invest-ment of industrial sponsors is not insignifi-cant, and there is the sense that faculty willbe receptive to good problems posed by indus-try. There is also the sense that research ques-tions can be shaped to examine fundamentalissues likely to be of interest to all.43 Thereis additional concern that there will be pres-sures to keep research secret. Such pressuresare likely to be strongly resisted; the numberof seminars, publications, and open meetingshave demonstrated the University’s and theCenter’s intent to maintain openness.

Finally, there is the question concerning theadvantage corporate sponsors have over non-participating companies. The questions here—. —..—

“John Linvill, Director, CIS, notes, “We gain much morethrough access with each other. The watchword is not isola-tion, but interaction. ” Personal communication, March 1983.

Figure 31. —Stanford University, Center for Integrated Systems Patent Policy

CIS Patent PolicyDisposition of Rights

VisitingScholar

. .

p-~ -.---z=

1. Stanford takes title and Inventor receives nonexclusive, fullv paid Iicense Including right to sublicense2. Upon request of inventors, Stanford will authorize a petltlon by Inventor’s employees to U.S. Government for greater r!ghts than 1

SOURCE: HewlettPackard Journal, November 19S3.

38-802 0 - 85 - 13


relate not just to this project but to other proj-ects as well. Proponents of the Center arguethat the fundamental nature of the universityhas not been changed, that everything the uni-versity does is open to dissemination-a wayof life that undergirds every relationship—nomatter “whom we get money from. ”44 Whilesponsoring companies have access to graduatestudents, and may seem to have an advantagein recruiting them, the networks betweenStanford faculty and their contacts in hun-dreds of companies in the Silcon Valley remainstrong. In addition, if a CIS research teamwishes to include a nonparticipating companyon a research project, they can do so. Reachingthis agreement, noted several participants,was a hard fight to win.

Nonetheless, the controversy persists, andquestions are likely to remain.45 The negoti-ated agreements concerning patents, and thedisposition of licenses are seen as experimentswhich may or may not work out and whichmay need to be revised as research and workprogresses. The experimental nature of thecenter is not limited to intellectual propertyarrangements, note participants, but includesas well the social and organizational arrange-ments built around cooperation—both amongthe various departments and among the cor-porations.

Center Operation

While the Center itself is still evolving interms of the agreements for intellectual prop-erty rights, the working relationships betweencorporate sponsors and faculty, and the facil-ities under construction, the research and in-volvement of the faculty and students are wellunder way. CIS research projects, funded prin-—. —..—-. . —.-

441 bid.45A recent article in the New York Times highlights the con-

tinuing questions and controversy. While the Vice Provost notesthat CIS is an “innovative setup for Stanford, ” a professor ofhistory argues, “It’s potentially very dangerous for a univer-sity to give privileged space and privileged access to informa-tion to particular companies. There is a danger that research-ers will create relationships that are likely to influence whatthey study and what they do not study. It is a threat to theautonomy of the university. ‘“ See: R. Reinhold, “Stanford andIndustry Forge New Research Link, ” New York Times, Feb.10, 1984.

cipally by the Federal Government, total $12million a year. (see table 26: CIS ResearchTopics). Seventy-one faculty members, repre-senting seven departments, are affiliated withthe center, and 30 Ph.D. and 100 MS degreecandidate students a year are being trained.It is the people who are the most importantoutput of the Center for Integrated Systems:“To the extent that we educate people withthe right background, have them do interest-ing research of significance to the Nation’sproblems but at a fundamental level, and dothis in close collaboration with the industrieswhich need such people, we will significantlymodify the nation’s productivity and competi-tiveness. “46

Microelectronics Center ofNorth Carolina (MCNC)

Universities play a critical role in the Micro-electronics Center of North Carolina (MCNC).As a multi-institutional R&D effort, MCNCcombines the resources of Duke University,North Carolina A&T State University, NorthCarolina State University, University of NorthCarolina at Chapel Hill, and University ofNorth Carolina at Charlotte, as well as the Re-search Triangle Institute. These institutions,along with the MCNC core organization, thenew MCNC research facility, and the commu-nications system linking all the facilities pro-vide a concentration of resources for educa-tion, research, technology transfer, andindustrial development.

Established in July 1980, MCNC is orga-nized as a private, nonprofit corporation toassist North Carolina’s development of mod-ern electronics and related high-technology in-dustry. MCNC has been funded primarily withState grants, as a result of strong leadershipfrom the governor and support from the leg-islature. Thus far, $43 million has been allo-cated by the North Carolina General Assem-bly for constructing, equipping, and operatingMCNC. An additional $34 million for new cap-ital facilities at the participating institutionshas also been provided.

“John Linvill, personal communication, April 1984.

Ch. 6–New Roles for Universities in Information Technology R&D ● 185—

Table 26.—Stanford University, Center for Integrated Systems Research Topics

Although its building won’t be completed until early 1985,the Center for Integrated Systems is already coordinating a$12-million-a-year basic research program at Stanford, fundedlargely by the Federal government. Here’s a sampler of re-cent and current topics investigated by CIS affiliated facultyin various existing labs.Computers● Research on streamlined-instruction -set microprocessor

(featuring partnership of computer and IC engineers)● Research in VLSI systems● Knowledge-based VLSI project● Image understanding● IntelIigent task automation● Network graphics● Partitionable computer systems. Analysis and verification of high-order language programs● Study of very high-speed integrated circuits (VHSIC-Phase

3)● Real-time communications systems: design, analysis, and

implementationŽ Structured design methodology for VLSI systems● Logical methods for program analysis● Ultra-concurrent computer systems● Silicon compilation● Data base theory● Computer languages for VLSI fabricationInformation Systems● Multiple user channels and information theory● Computational complexity, efficiency, and accountability

in large-scale teleprocessing systems● Multiplexed holographic reconstruction methods for 3D

structures● Information theory and data compression● Signal processing and compression● Statistical data processing, system modeling, and re-

liability● Algorithms for Iocating and identifying multiple sources

by a distributed sensor network● Fast algorithms for improved speech coding and recog-

nition● Dual-energy digital subtraction radiography for noninvasive

arteriographyIntegrated Circuits● Computer modeling of complete IC fabrication process● Integrated electromechanical and optical sensor arrays for

Optacon II (a reading aid for the blind)● BME Center for Integrated Electronics in Medicine (to pro-

duce implantable telemetry systems for biomedical re-search)

● Computer-aided design of IC fabrication processes for VLSIdevices

● Submicron device physics and technology● Development of multichannel electrodes for an auditory

prosthesis● Study of transdermal electronics for an auditory prosthesis

● Multilevel-metal interconnection technology● Biomedical silicon sensors● Fast turnaround laboratory for VLSISolid State● Ion implantation and laser annealing in semiconductors

and related materials● Laser and electron beam processing of semiconductors● Characterization of high-speed semiconductor device

materials using advanced analytical techniques● Ion implantation and laser processing of 3-5 compound

conductors● Defects at electrode-oxide and electrode-silicon interfaces

in submicron device structures● Microstructure fabrication using electron beams of con-

ventional and very low energies● Advanced concepts in VLSI metalIization● Advanced packaging concepts for VLSI● Studies of surfaces and interfaces of 3-5 compounds and

Si:silicides● Silicon photocells in thermophotovoltaic energy con-

version. Investigation of metalIic impurities introduced into SiO2

and Si by various candidate VLSI metallization systems● Modeling of emitters● Structural and bonding studies of practical semiconduc-

tor layersSpace Telecommunications and Radioscience● Establishment of a Center for Aeronautics and Space In-

formation Sciences at Stanford University (Funded by theU.S. National Aeronautics and Space Administration. Thefocus will be on applying VLSI techniques to develop newhardware and firmware for space instrumentation and com-mand and control.)

● Communication satelIite planning centerMaterial Science and Engineering● Atomic-level physics modeling of the thermal oxidation

process● Fabrication and properties of muItilayer structures● Computer simulation of surface and fiIm processes● Photoelectronic properties of 2-4 heterojunctions● Photoelectronic properties of zinc phosphide crystals,

films, and heterojunctions● Photovoltaic heterodiodes based on indium phosphide● Preparation and properties of CdTe evaporated fiIms com-

pared with single-crystal CdTeGinzton Laboratory● Superconducting thin fiIms, composites, and junctions● Acoustical scanning of optical images● Research on nondestructive evaluation● Evaluation of machining damage in brittle materials● Optical and acoustic wave research● High-frequency transducers● Research on acoustic microscopy with superior resolution● Study of properties of material by channeling radiation● Surface acoustic wave MOSFET signal processor

SOURCE Hewlett Packard Journal, November 1983


While the extent of State support is signif-icant, such a role is not new for North Caro-lina. The precedents for government-industry-university cooperation span two decades. Thedevelopment of the Research Triangle Park,the Research Triangle Institute, the establish-ment of the North Carolina Board of Scienceand Technology, and now MCNC, are seen asmodels of government-industry-university co-operation to develop new technology-based in-dustries. 47

New Facilities for Education,R&D, and Technology Transfer

The MCNC facility, under construction sinceMay 1982, will house core MCNC staff, visit-—.

“U.S. Congress, of Technology Assessment, Technol-ogy, Innovation, and Development, Censusof State Government for High Technology Indus-trial Development–Background Paper, OTA-BP-STI-21, May1983, p. 56,

ing engineers and scientists from industry,and visiting faculty and graduate studentsworking on special research projects. Thel00,()()()-squar~foot,” $30 million facility has ca-pability for performing advanced manufactur-ing processes, including high-density inte-grated circuit fabrication, system design, anddesign tool research.

The MCNC $6.5 million communicationssystem, scheduled for completion by 1985, willput in place a ISO-mile microwave networklinking the educational and research activitiesat MCNC, the universities, and the ResearchTriangle Institute (see fig. 32, MCNC Commu-nications System). In the first phase hookupof the system, computer science students atDuke University in Durham and UNC atChapel Hill take classes (which originate fromDurham) together without leaving their owncampuses. Similarly, courses on Computer

MCNC dual source electron beam/r.f. metal evaporator for next generation integrated circuit manufacturing research

Ch. 6—New Roles for Universities in /formation Technology R&D . 187

Figure 32.—Communications System Linking MCNC Participating Sites

MCNC Communications System

● Unique high-performance microwave system.● 152 miles from end to end. \● 2 two-way, full-motion video channels.. 7 program-origination locations.. 192 high-speed data paths.

SOURCE Mlcroelectronlcs Center of North Carollna, Research Triangle Park, NC

Graphics and VLSI Design originate fromChapel Hill.

Like the MCNC research facility, the devel-opment of the telecommunications system re-quired funding beyond the reach of each of theindividual institutions. “Before the center wascreated, each of our participating universitieshoped to develop its own major microelectron-ics program. But the financial realities and thedifficulties of attracting talent from the lim-ited talent pool, soon made it apparent thatthe only way to develop a first class programwas to join forces and work together. ”48

MCNC Working Relationships

MCNC is more than a consortium of univer-sities sharing resources and interacting withindustry. The participating institutions arelinked together by MCNC (see fig. 33, TheMCNC Community) and MCNC as an organi-zational entity and actor bridges the interestsand functions of the industry and universitycommunities (see fig. 34, Working Relation-ships). As members of industry work alongwith university researchers, the applied re-

—..— . .“D. S. Beilrnan, President of MCNC, “New Initiatives in Mod-

em Electronics, ” address before the Materials Research SocietyAnnual Meeting, Boston, Nov. 14, 1983.

Figure 33.— MCNC Participating InstitutionsMicroelectronics Center of North CarolinaMCNC and the participating institutions

The MCNC community

SOURCE: Microelectronics Center of North Carolina, Research Triangle Park,NC.

search and technology projects tie togetherthe commercial technology activities underway in industry and the basic research beingconducted at the universities.

Since 1980, more than 30 new faculty mem-bers in microelectronics-related disciplineshave been recruited. In contrast to the recenttrend of faculty leaving universities for em-


Figure 34.—Working Relationships, Microelectronics Center of North Carolina

SOURCE: Microelectronics Center of North Carolina, Research Triangle Park, NC

ployment in industry, 20 of the 30 new faculty staff of the Center number more than 70, withhave come directly from industry .49 Full-time 12 having joint institutional appointments.

Moreover, the combined microelectronics-re- that was able to compete with be- lated manpower resources at the participating

cause it offered industry-level salaries, access to an advanced institutions consist of over 150 faculty mem-state-of-the-art research and manufacturing facility, a strong bers and 450 graduate students, a significantuniversity R&D environment, and a thriving high-technologyindustrial center. pool of talent.

Ch. 6–New Roles for Universities in Information Technology R&D • 189

MCNC spokesmen argue that MCNC’s abil-ity to bridge university/industry concerns isenhanced by its structure as a nonprofit “neu-tral” facility. The Center’s permanent staff,specialists with joint university and Center ap-pointments, resident scientists and engineersrepresenting industrial affiliates, visitingscientists from other national centers, andgraduate students involved in special projectsphysically work together in the MCNC micro-electronics manufacturing research complex.

In contrast to MCC (see below), a joint ven-ture owned by its corporate companies, MCNCleaders envision this effort as providing tech-nology functions that are not now provided byeither MCC, other planned industry joint de-velopment programs, or other joint industry-university collaborative efforts in the micro-electronics field. “All of the industrial joint de-velopment programs or companies are spon-sored by the first tier of large electroniccompanies. The perceived need for these large-company joint efforts to remain competitivegreatly amplifies the need for similar supportto the larger number of second-tier and evolv-ing smaller companies in the electronics indus-try. “5° MCNC expects to involve a broadersegment of companies, in part because of lowerfee levels, and because of substantial, continu-ing State support. The industrial affiliate pro-gram is just getting under way. By 1985, theCenter expects to have 20 industry affiliates.”

Affiliates can come and use the MCNC fa-cility to develop products, can participate inresearch and educational programs (tuitionand fees are provided for three staff membersat a time for each affiliate), are represented onan advisory council, participate in semiannualreviews, and in the process have increased ac-cess to faculty and students. Nonexclusive,nontransferable licenses for intellectual prop-erty rights are available to affiliates on a pre-ferred royalty basis. It is expected that themajority of cooperative research will be openlydisseminated.

—.—..——““D. S. Beilman, p e r s o n a l c o m m u n i c a t i o n , A p r i l 1 9 8 4 .“I+~ach industrial affiliate pays an annual $250,000 fee.

The establishment of MCNC appears to bean example of effective university-industry-government collaboration. Some point outthat it is an example of how State leadershipand initiative can be the driving force in pull-ing together traditionally independent publicand private academic institutions and forgingnew relationships to attract significant indus-try participation in education, research, tech-nology transfer, and industrial development.What has made MCNC work? In reflecting onthe experiences thus far, the president ofMCNC lists three basic requirements:

1.

2.

3.

the need for a long-term strategic ap-proach to substantial funding. This in-cludes building upon existing programsand investments, as well as obtaining atleast a 3-year commitment from all mem-bers of the collaborative effort as a de-pendable commitment to common goals;the need to structure in-depth interactionamong the limited talent available. Fullparticipation by personnel from industry,universities, and government is necessaryfor understanding each other’s perspec-tive and for crystallizing and mutually ac-cepting responsibility for important com-mon goals; andthe need to plan for and accelerate thetransfer of research into technology andto promote R&D in progressively morescientific endeavors while making use ofall basic related investments.”

MCNC spokesmen are confident of MCNC’Sfuture. Continued support for two-thirds of itsoperation are expected to come from the State.Industrial and Federal support are anticipatedto cover the remaining third.

Semiconductor ResearchCorporation (SRC)

Under the aegis of the Semiconductor Indus-tries Association, the Semiconductor ResearchCorp. (SRC) was formed in 1981, to establish

— ————“D. S. Beilmen, “New Initiatives in Modern Electronics, ”

op. cit.

—

190 • Iformation Technology R&D: Critical Trends and Issues

a cooperative organization that would spon-sor university research needed by the indus-try. SRC was incorporated in February 1982.As noted by Robert Noyce, of the Intel Corp.,“The semiconductor industry is fiercely com-petitive and that competition has resulted inthe vitality and success of the industry. ”53 Thecooperative institutional arrangement thatmakes SRC possible represents a change forthe semiconductor industry.

In the case of that industry, the forces stim-ulating cooperation centered around threeissues: research, manpower resources, and in-ternational competition. There was growingconcern that the industry’s basic research ef-forts, the foundation of its future well-being,were increasingly directed towards the solu-tion of near-term problems; that industry re-search efforts were often duplicative and re-dundant, as each corporation tried to stay ontop of the competition; and that fierce com-petition and the squeeze for profits, combinedwith increasing costs of R&D, created disin-centives and high risk for long-term industryresearch efforts.54

At the same time, there was growing recog-nition that the Nation’s research universitieswere underutilized resources that industrycould turn to for long-term basic research andcreation of new knowledge. There was also con-cern that the pool of experienced and trainedmanpower was being “overgrazed” by the in-dustry itself, and that both faculty and ad-vanced graduate students were leaving univer-sities, irresistibly drawn to industry .55

In view of the growing competition for semi-conductor products, and the increased and co-ordinated R&D efforts undertaken by its for-eign competitors, it was argued that the U.S.

. — — —“R. Noyce, “Competition and Cooperation–A Prescription

for the Eighties, ” Research Management, March 1982, pp.13-17.

“R. M. Burger, and L. W. Summey, “An Update on the Semi-conductor Research Corporation”, QIE Conference Proceedings,1983, pp. 51-59

“Robert Noyce uses the analogy of the “Tragedy of the Com-m o n s ’ where the commonly held resources, in this case, re-

search and manpower are exploited by industry self-interest,op. cit., p. 15.

semiconductor industry had to increase R&D.56

Moreover, more complex and more capital-intensive research tasks, coupled with increas-ing lead time and a perceived shortage of suf-ficiently trained manpower, created additionaldifficulties to undertake such research, partic-ularly for a single company .57 And drawing onthe examples of other nations, such as Japan,joint coordinated efforts are seen as ways ofassuring long term competitiveness throughcooperation.

These factors resulted in four major objec-tives for SRC:

1. increasing semiconductor research in theUnited States;

2. sharing research efforts among industrysponsors;

3. strengthening and upgrading research inthe universities; and

4. attracting more students to this field ofstudy and improving the quality of edu-cation.

Since its formation in 1982, SRC has grownfrom 10 to 40 companies, of varying size, com-panies that manufacture or purchase semicon-ductor devices for manufacturing other prod-ucts, or companies that manufacture equip-ment or materials for use by the semiconduc-tor industry .58 Membership fees are tied to acompany’s IC sales or purchases worldwide,with annual fees ranging from $60,000 up. All— . -- - .

‘)’’ E:. f?loch, prepared statement, Hearings, Japanese Techno-lo~”cal Advances and Possible United States Responses UsingJoint Research Ventures, Subcommittee on Investigations andOversight and the Subcommittee on Science, Research, andTechnology, U.S. House of Representatives, June 29-30, 1983.

“Ibid., p. 46.58SRC membership, as of September 1984, includes: Advancwl

Micro Devices, Inc., AT&T Technologies, Inc., Burroughs Corp.,Control Data Corp., Digital Equipment Corp., E.I. du Pent deNemours & Co., Eastman Kodak, Eaton Corp., E-Systems, Inc.,GCA Corp., General Electric Co., General Instrument Corp.,General Motors Corp., Goodyear Aerospace Corp., GTE Lab-oratories, Inc., Harris Corp., Hewlett-Packard Co., Honeywell,Inc., IBM Corp., Intel Corp., LSI Logic Corp., Monolithic Mem-ories, Inc., Monsanto, Co., Motorola, Inc., National Semicon-ductor Corp., Perkin-Ehner Corp., RCA, Rockwell International,Silicon Systems, Inc., Sperry Corp., Texas Instruments, Inc.,Union Carbide Corp., Varian Associates, Inc., WestinghouseElectric Corp., Xerox Corp, and Zilog, Inc.. In addition, the fol-lowing companies are in the Semiconductor Equipment andMaterials Institute, Inc.: Micro Mask, Inc.; Pacific Western Sys-tems, Inc.; Probe-Rite, Inc.; Pure Aire Corp.


member companies have equal privileges: ac-cess to all sponsored research through semi-nars, annual meetings, and newsletters andreports, access to research data bases andlicense rights, as well as an expanded recruit-ing base.

The SRC 1984 budget is over $15 million,up from $6 million in 1982, and $10 million in1983. Currently, approximately $12 million isavailable for university research projects, anamount which “substantially increases totalavailable funding for basic research in semi-conductor technology. ”59 Spokesmen point outthat SRC has promoted research in engineer-ing, mathematics, and the physical sciencesunderlying semiconductor technology. Majorareas of focus established by the industrialboard members and the technical advisoryboard are:

• Microstructure Sciences:—Materials, Phenomena, and Device

Physics,—Microsciences,—Device Fabrication Technologies.

● Systems and Design:–Design Automation,–System Component Interactions.

● Production and Engineering:–Reliability, Quality Assurance, and

Testing,—Packaging,—Manufacturing.

Impacts

In planning the research activities to beundertaken, several different levels of fundingand effort were envisioned by board members.SRC has awarded individual university re-search projects, as well as several contracts

—.—. ——-—“Erich Bloch, Former Chairman of the Board, SRC, estimates

that the semiconductor industry allocates 3 to 5 percent of itsR&D budget to basic research–approximately $35 million to$50 million annually. He notes that the R&D tax credit wasan important factor in the decision to proceed with the forma-tion of SRC. Moreover, if there were a differentially larger taxincentive for industry-sponsored university research, there couldbe an even broader expansion of industry funding of univer-sity research in the future. Testimony before the Subcommit-tee on Taxation and Debt Management, Senate Committee onFinance, Feb. 24, 1984.

for major research “centers of excellence” andmajor research projects. In its initial solicita-tion for proposals from the universities, SRCreceived 166 proposals from 63 universities.In the first. year of operation, eight universitiesreceived research contracts. In 1983, morethan 30 universities, involving approximately100 researchers and 125 graduate students, re-ceived $10 million for research through 47 con-tracts with SRC.60 By May 1984, 34 universi-ties involving 125 faculty and research staffand 202 graduate students were supported by$12.275 million in SRC research funding.

It has been SRC policy to distribute the con-tracts for centers and individual research proj-ects on a broad geographic basis among lead-ing research centers as well as to universitieswhose expertise in these areas is not as wellestablished (see table 27, “Regional Distribu-tion of SRC funding”). Thus, SRC efforts mayhave an impact that goes beyond the specificresearch projects: in helping to expand a uni-versity’s research capabilities, it may help itattract high caliber faculty and graduate stu-dents. Moreover, in the long run the SRC sup-port may contribute to additional university-industry cooperation, and new high-technol-ogy industrial development.

SRC research “centers of excellence” includeCornell University, University of California atBerkeley and Carnegie Mellon University. Ma-jor research programs are supported at theNorth Carolina consortium (MCNC), Massa-chusetts Institute of Technology, Clemson,Stanford, Rensselaer, and the University ofCalifornia at Santa Barbara. Over the next fewyears, there are plans to support 8 to 10 moreof these centers and programs, and to to con-duct research into design of microstructure,properties of silicon material, computer-aideddesign and automation of design, lithography,beam processing, fault tolerance, micropack-aging and cooling, three-dimensional siliconstructures, and manufacturing systems re-search.

‘(’N. Snyderman, ‘‘I ndustry Observer,Electrom”c .%’ews, ,Jan-uary 1984.


Table 27.—A Distribution of SRC Funding by Region,January 1985

Region/institution Funding

New EnglandMIT . . . . . . . . . . . . . . . . . . . . .$ 976,110Yale . . . . . . . . . . . . . . . . . . . . . 211,258Brown . . . . . . . . . . . . . . . . . . . 104,000Vermont . . . . . . . . . . . . . . . . . 86,220

Middle Atlantic:Cornell ... ... ... ... ... .. .$1,776,651CMU . . . . . . . . . . . . . . . . . . . . 1,414,580RPI . . . . . . . . . . . . . . . . . . . . . 800,000Penn State . . . . . . . . . . . . . . . 126,489Rochester . . . . . . . . . . . . . . . . 119,000Johns Hopkins . . . . . . . . . . . 114,589Columbia . . . . . . . . . . . . . . . . 89,901

North CentralIllinois . . . . . . . . . . . . . . . . . . . 750,607Michigan . . . . . . . . . . . . . . . . 337,155Minnesota . . . . . . . . . . . . . . . 206,755Notre Dame . . . . . . . . . . . . . . 126,000Iowa . . . . . . . . . . . . . . . . . . . . 118,013Purdue . . . . . . . . . . . . . . . . . . 102,870Wisconsin . . . . . . . . . . . . . . . 93,500

South Atlantic:MCNC . . . . . . . . . . ... ......$ 900,000Clemson . . . . . . . . . . . . . . . . . 215,424Georgia Tech.... . . . . . . . . . 190,553Auburn . . . . . . . . . . . . . . . . . . 152,342North Carolina . . . . . . . . . . . . 149,744Mississippi State . . . . . . . . . 135,000Florida . . . . . . . . . . . . . . . . . . 71,864

Mountain:Arizona . . . . . . . . . . . . ......$ 503,530Arizona State.... . . . . . . . . . 100,914Colorado State . . . . . . . . . . . 84,000Texas A&M . . . . . . . . . . . . . . 50,100

Pacific:Stanford .. .. .. .. .. .. .. .. .$1,511,990Berkeley (UC)... .. .. .. ... .$1,350,000Santa Barbara . . . . . . . . . . . . 450,000Southern California . . . . . . . 101,943UCLA . . . . . . . . . . . . . . . . . . . 96.307

$1,377,588

$4,441,210

$1,734,900

$1,814,927

$ 738,544

$3,510,240

Total . . . . . . . . . . . . . . . . . .. .. ..’. .. .. .$13,617,409SOURCE: Semiconductor Reseach Corp. Reasearch Triangle, NC.

In its short span of operation, SRC providesan example of a joint industry approach in themanagement and coordination of informationtechnology R&D and in the establishment ofnew relationships between industry and aca-demia for the conduct of research efforts.Thusfar, SRC member companies have been ableto agree on research priorities. Ongoing re-search projects have focused onVLSI circuitprocesses and Computer-Aided Design, aimedat commercially relevant results over a 3-to5-year period. In developing a list of potentialresearch topics for longer term research (e.g.,research needs in GaAs), SRC workshops have

involved both university researchers and in-dustry participants. Beginning with abroadarray of potential research needs, the groupswere able to reach consensus on research topicpriorities and areas for future focus.

There is other evidence that SRC’s approachhas fostered closer links between industry andacademia. In addition to the technical advi-sory board, composed of member companyrepresentatives, SRC has established indus-trial mentors for each contract. With recoin-mendation from the technical advisory board,an industry engineer or scientist in an SRCmember company becomes the direct contactpoint for each of the SRC contracts. The in-dustrial mentor can help identify importantproblem areas, and may from time to time beable to provide direct technical assistance tothe university research community. Throughtopical research meetings, additional industry-university contacts are strengthened.

Program reviews of SRC Centers of Excel-lence cover a wide spectrum of technical inter-est and are designed to attract abroad repre-sentation from the industry and researchcommunity. Member companies may also par-ticipate in SRC activities by assigning an em-ployee to participate in the management of theSRC program at Research Triangle Park. In-dustry assignees may also become research-ers in residence, spending at least 3 monthsto a year, working in the university laboratorywith one or more of the university research-ers. This may foster technology transfer inways that are not accomplished through thedissemination of reports, newsletters, and con-ference results.

There is no question that SRC has providedadditional research opportunities for the uni-versity community and that these opportuni-ties have reached a range of institutions. Re-search results have been freely disseminated.The ownership rights to the patents are heldby the universities. So far, only one patent hasresulted from SRC-sponsored research. Sev-eral researchers have indicated their apprecia-tion of lack of bureaucratic hassle in the SRCcontracting process, and find the yearly re-ports and reviews helpful.


The eventual impacts of SRC will be seenin how well it meets the needs of both univer-sities and the semiconductor industry. For themember companies, the usefulness of researchresults, in both the short and long term maybecome factors in their continued support. Forthe industry as a whole, new knowledge andnew manpower are important, as well as theattraction of additional researchers to newfields of study in the future. How effective isthe interface between the university and theindustry over the long term? A sign of success,at least interim success, note SRC spokesman,is the increase in the number of member com-panies and the continued support of the ini-tial companies who have signed on each of thethree years of operation.

Microelectronics and ComputerTechnology Corporation (MCC)

The Microelectronics and Computer Tech-nology Corporation (MCC) is an R&D joint-venture owned and operated by 20 U.S. com-puter and semiconductor companies.61 Theidea for MCC was conceived by William C.Norris, President of Control Data Corp., andin his view “MCC represents a cooperative ef-fort to develop a broad base of fundamentaltechnologies for use by members who will eachadd their own value and continue to competewith products and services of individual con-ception and design. ”62 While these companieshave traditionally avoided cooperation, “inthis period of scarce resources, however, andat a time when this country’s leading positionin technology is being challenged by foreigncompetitors, refusal to cooperate is no longertenable. ’63

Governed by a Board of Directors composedof representatives of each shareholder com-

‘lShareholder companies include Advanced Micro Devices,Allied Corp., BMC Industries, Control Data Corp., DigitalEquipment Corp., Eastman Kodak, Gould, Hank Corp., Honey-well, Lockheed, Martin Marietta Aerospace, Mostek, Motorola,National Semiconductor, NCR, RCA, Rockwell, Sperry Corp.,Boeing, and 3-M.

62W. C. Norris, “Cooperation for Improving Productivity, ”keynote address, Prepatory Meeting for the White House Con-ference on Productivity, San Diego, CA, July 20, 1983.

63W, C. Norris, “How to Expand R&D Cooperation, ” Busi-ness Week, Apr. 11, 1983, p. 21.

pany, MCC began formal operations in Janu-ary 1983, with the selection of its chief execu-tive officer and the development of a plan forR&D. A technology Advisory Board of share-holder representatives provides advice in de-veloping the research strategies, in evaluatingnew program proposals, and in monitoring ex-isting programs.

With a $50 million to $60 million annualbudget, four long-range advanced technologyprograms are expected to cover a 6- to 10- yeartime Span.64 In defining the areas of research,the shareholder companies “came to concen-trate on areas in which they believed accom-plishments were necessary to make quantumjumps in the performance of the next genera-tion of computers. ”65 The programs include:

●

●

●

●

Packaging: A 6-year program to advancethe state-of-the-art in semiconductorpackaging and interconnect technology,with a focus on technologies compatiblewith automatic assembly at both the cir-cuit and system level.Software Technology: A 7- to 8-year pro-gram to develop new techniques, proce-dures and tools that can be used to im-prove the productivity of the softwaredevelopment process by one or two ordersof magnitude.Computer-Aided Design and Manufactur-ing (CAD/CAM): An 8-year program to im-prove CAD/CAM technology and to de-velop an integrated set of tools that willhave particular application to complexsystems and the complex VLSI chipsfrom which they will be built.Advanced Computer Architecture: This 10-year effort will focus on artificial intelli-gence, new techniques for database man-agement, human interface with comput-ers, and parallel processing.

In addition to forming a comprehensiveagenda for research, MCC has selected a sitefor operation and hired staff. While still in tem-

84B. R, Admiral, President, Microelectronics Computer Corp.,personal communication, May 1984.

‘i’M. A. Fischetti, “MCC: An Industry Response to the Jap-anese Challenge, ” IEEE Spectrum, November 1983, pp. 55-56.


porary quarters, more than 173 professionalshave been brought into the operation. Origi-nally, the staffing plan was to draw senior andhighly trained technical professionals from theparticipating companies, with only about 25percent expected to come from the outside. Inactuality, 40 percent of the professionals havecome to MCC from the shareholder companies.There is some concern that MCC will attractsenior faculty members from the universitiesand put a strain on available manpower re-souces, particularly in areas such as artificialintelligence.

66 MCC officials recognize that if

they hire away the university faculty, they willcompromise the universities’ ability to producehighly trained top-quality graduates-the verypeople they need for the future.67

So far, the MCC strategy appears to beworking; of the talent on board, the majorityare from industry, and the remainder fromacademia and government. Full operation, ex-pected by late 1985, will bring the total num-ber of professionals to 350. At full strength,MCC will be looking for the “brightest grad-uates, ” and it is expected that many will comefrom nearby educational institutions–theUniversity of Texas and Texas A&M Uni-versity.

Not surprisingly it was these universities,and their promised commitment to develop amajor, first-class computer science and micro-electronics program, as well as strong supportfrom the State officials and the business com-munity, that led to the decision to locate MCCheadquarters in Austin, after conducting asearch of 57 cities in 27 States. It has beennoted that few cities in Texas—or anywhereelse—could put together the incentives thatwere offered. The University of Texas at Aus-tin promised to construct a $20 million office-laboratory facility, to be leased to MCC. Thirty

ea~e the OTA Case study on Artificial Intelligence.@Tin a recent interview, Admiral Inman discussed this issue.

“I have a standard rule that I will not recruit from universities.If I am approached by someone on a faculty, my requirementis that they go up the chain and say they are going to leaveto go to industry. I can’t have it both ways–to encourage theproduction of additional topquality graduate students, and tohire away university talent. ” See: J. A. Turner, “Big-Spend-ing U. of Texas Aims for the Top in Computer Science, ” Chron-icle of Higher Education, Apr. 4, 1984.

new faculty positions and 75 new graduatefellowships would be supported. In addition,there was a commitment of at least $1 milliona year for maintenance and support for re-searchers, and $5 million for purchase of lab-oratory equipment at UT-Austin. After MCCselected the Austin site, an anonymous donormade available $8 million, with the provisothat other private sources match that amount.The university then matched that total, usingfunds from the Permanent University Fund(derived from revenues from oil leases on landowned by the University). The result is 32 newendowed chairs at the University of Texas, 10of which are in microelectronics and computersciences.68

The developments in the academic commu-nity, the development of MCC, and the devel-opments in the fast-growing high-technologycorridor between Austin and San Antonio 69

have drawn national attention. The potentialfor economic growth, quality education andcutting-edge research are cited as the realcause for excitement.70 Texas leaders point outthat these high-technology initiatives (e.g.,MCC, the university programs) are just the be-ginning of the State’s commitment to hightechnology development. It is recognized thatnot only the universities, but the entire Stateeducational infrastructure have to be strength-ened and supported over the long term. Theimprovement of the State’s elementary andsecondary schools has been addressed by theGovernor as well as MCC’s director, AdmiralBobby Inman and other leaders in industry,who are concerned that without improvementTexas public schools represent a deterrent torecruiting engineers and other highly trainedspecialists. Among the recommendations of aspecial panel headed by Texas industry leader,H. Ross Perot, are increased teacher salaries,

68A11 32 ch~rs me timed at strengthening the university’sscience and engineering programs. Eight disciplines are the f~cus of this effort: chemistry, mathematics, molecular biology,physics, computer engineering, manufacturing, systems engi-neering, materials science, and microelectronics.

‘gFor examples of recent economic development, see J. R. Lint+backer, “Letter from Austin: Texas Cash Fuels ElectronicsBoom. ” Electronics, June 15, 1983, pp. 95-96.

‘“J, Kraft, “The Japaning of Texas, ” 14’ashington Post, Apr.17, 1984.


State aid to equalize school spending amongrich and poor districts, and strengthened cur-riculum requirements—at a cost estimated atnearly $1 billion in new taxes.

The ultimate test for MCC will be its abilityto draw sufficient talent to conduct the R&Dnecessary to keep its member companies inter-nationally competitive. MCC officials and cor-porate sponsors are confident that this can beaccomplished. Some observers are less confi-dent that MCC will be able to transfer its tech-nology to individual corporate efforts. As

Chapter 6

“Academe and Industry Debate Partnership, ”Science, vol. 219, No. 4481, January 1983,pp.150-151.

“The Academic-Industrial Complex, ” Science, vol.216, May 28, 1982, pp. 960-961.

“Artificial Intelligence (I): Into the World, ” Sci-ence, vol. 223, February 24, 1984, pp. 802-805.

“Cooperation is the Key: An interview with B. R.Inman, ” Communications of the ACM, vol. 26,No. 9, September 1983, pp. 642-645.

“The Challenges: Designing the Next Generation, ”IEEE S’ectrurn, November 1983.

Probable Levels of R&D Expenditures in 1984:Forecast and AnaZysis (Columbus, OH: BattelleMemorial Institute, December 1983).

“Texas Uses Oil to Fuel Research, ” Science, vol.220, April 22, 1983, pp. 390-391.

“Tomorrow’s Computers: The Quest, ” IEEE Spec-trum, November 1983.

“Semiconductor Research Co-op Eyes 4 MegabitMemory Chip,” Ekctrom”c News, July 18, 1983.

“Weighing the Social Costs of Innovation, ” Sci-ence, vol. 223, Mar. 30, 1984, pp. 1368-1369.

Ashford, N. A., “A Framework for Examining theEffects of Industrial Funding on Academic Free-dom and the Integrity of the University, ” Sci-ence, Technology and Human Values, vol. 8,Issue 2, spring 1983, pp. 16-23.

Brademas, J., “Graduate Education: Signs ofTrouble, ” Science, vol. 223, No, 4639, March1984, p. 881.

Branscomb, L. M,, “The Computer’s Debt to Sci-ence, ” Perspectives in Computing, vol. 3, No. 3,October 1983.

noted earlier, MCC originally intended to drawits staff principally from the member compa-nies, thereby speeding technology transfer.Since recruitment has drawn more heavily onoutside sources, MCC will have to find otherapproaches if it is to accomplish this goal.

While it is too soon to assess the impactsof the MCC joint venture, and related activi-ties at the University of Texas and Texas A&M,they do provide an example of how academia,business, and government can join forces tocreate new institutional arrangements.

ReferencesBrown, T. L. “University-Industry Relations: Is

There a Conflict. ”Burger, R. M. and Surnney, L. W., “An Update on

the semiconductor Research Corporation, ” QIE(Quab”tyin Electrom”cs) Conference Proceedings,1983.

David, E. E., “Supporting Research with a Com-mercial Mission, ” Change, vol. 14, September1982, pp. 25-29.

David, E. E., “The University-Academic Connec-tion in Research: Corporate Purposes and SocialResponsibilities, ” Journal of the Patent OfficeSociety, vol. 64, April 1982, pp. 209-218.

Dineen, G. P., “Why Cooperative Research is aViable Strategy at Honeywell. ” Presentation atConference on Cooperative Research Ventures,sponsored by Control Data Executive Forum.Apr. 4-5, 1984, New York City.

Dorfman, Nancy S, “Route 128: The Developmentof a Regional High-Technology Economy, ” Re-search Poh”cy, vol. 12, No. 6, December 1983, pp.299-316.

Fowler, D. R., “University-Industry Research Re-lationships,” Research Management, vol. 27,No. 1, January/February 1984, pp. 35-41.

Fowler, D. R. “University-Industry Research Rela-tionships: The Research Agreement, ” The Jour-nal of College and University Law, vol. 9, No.4, 1982-1983, Pp. 515-532.

Giametti, A. B., “The University, Industry andCooperative Research, ” Science, vol. 218, De-cember 1982, pp. 1278-1280,

Gray, P. E., “The Role of the University in GlobalTechnological Change, ” In GlobaZ Techno]o~”-


cal Change: A Strategic Assessment, June 21-23, 1983. Industrial Liaison program of the Mas-sachusetts Institute of Technology (Cambridge,MA: MIT, 1983).

Hutt, P. B. , “University/Corporate ResearchAgreements,” Technology in Society, vol. 5,1983, PP. 107-118.

Inman, B. R., “Research and Development JointVentures” testimony before the Subcommitteeon Monopolies and Commercial Law, House Ju-diciary Committee, Sept. 28, 1983.

Kraft, J. “The Japaning of Texas, ” Washz”ngtonPost, Apr. 17, 1984.

Langfitt, T. W., Hackney, S., Fishman, A. P., etal., Partners in the Research Enterprise,Um”versity-Corporate Relations in Science andTechnology (Philadelphia: University of Penn-sylvania Press, 1983).

Location of High Technology Firms and Re~”onalEconorm”c Development, a staff study preparedfor the Joint Economic Committee, Subcommitt-ee on Monetary and Fiscal Policy, U.S. Con-gress (Washington, DC: U.S. Government Printing Office, 1982).

Lynn, W. R. and Long, F. A., “University-Indus-trial Collaboration in Research, ” Technology inSociety, vol. 4, 1982, pp. 192-212.

National Research Council, Rew”tdizing Labora-tory Instrumentation, the report of a workshopof the Ad Hoc Working Group on Scientific In-strumentation, Mar. 12-13, 1982 (Washington,D. C.: National Academy Press, 1982).

National Science Foundation, Unz”versity/industryResearch, Relationslups: Selected Studies, 14thAnnual Report of the National Science Board(Washington DC: U.S. Government Printing Of-fice, 1982).

National Science Foundation, Um”versity-IndustryResearch Relationslups: Mflhs, Reah”ties andPotentials, 14th Annual Report of the NationalScience Board (Washington D. C.: U.S. Govern-ment Printing Office, 1982).

Nelkin, D., “Intellectual Property: The Control ofScientific Information, ” Science, vol. 216, No.14, May 1982, pp. 704-708.

Noble, D. F., America By Design: Science, Tech-nology, and the Rise of Corporate Capitah”sm(New York: Alfred A. Knopf, 1977).

Norns, W. C., “Cooperation for Improving Produc-tivity, ” Keynote Address, IEEE Task Force onProductivity and Innovation, San Diego, July20, 1983.

Norris, W. C., “How to Expand R&D Coopera-tion, ” Business Week, Apr. 11, 1983, p. 21.

Noyce, R., “Competition and Cooperation–A Pre-scription for the Eighties, ” Research Manage-ment, vol. , No. , March 1982, pp. 13-16.

Ploch, M. “Micros Flood Campuses,” High Tech-nology, vol. 4, No. 3, March 1984, pp. 47-49.

Prager, D. J. and Omenn, G. S., “Research, In-novation, and University-Industry Linkages, ”Science, vol. 207, No. 254, January 1980, pp.379-384.

Schmitt, R. W., “Building R&D Policy fromStrength, ” Science, VO1.220, June 3, 1983, pp.1013-1015.

Schmitt, R. W., “National R&D Policy: An Indus-trial Perspective, ” Science, vol. 224, June 15,1984, pp. 1206-1209.

Shils, E., “The order of Learning in the UnitedStates from 1865 to 1920: The Ascendancy ofthe Universities, ” Minerva, vol. 16, No. 2, sum-mer 1978, pp. 159-195.

Smith, K. A., “Industry-Univeristy Research Pro-grams, ” Physics Today, vol. 37, No. 2, February1984, Pp. 24-29.

Sproull, R. L., “Protectionism and the Univer-sities,” Science, vol. 219, No. 4581, January1983.

Thomas, L. “The Value of Basic Science, ” ReviewMagazine of the Um”versity of Rochester.

Turner, J. A., “Big Spending University of TexasAims for the Top in Computer Science. ” Ckom”-cle of Higher Education, vol. XXVII, No. 6, Apr.4, 1984.

U.S. Congress, Committee on Science and Tech-nology, U.S. House of Representatives, Hear-ings before the Subcomnu”ttee on Investigationsand Oversight and the Subcomzm”ttee on Sci-ence, Research, and Techmdogy: Japanese Tech-nological Advances and Possible Um”ted StatesResponses Using Research Joint Ventures, June29-30, 1983 (Washington, DC: U.S. GovernmentPrinting Office, No. 45, 28-3770, 1983).

U.S. House of Representatives, Hearings onUm”versity/Industry Cooperation in Biotechnol-ogy, Subcommittee on Investigations and Over-sight and Subcommittee on Science and Tech-nology, of the Committee on Science, Researchand Technology, June 16-17, 1982.

U.S. Congress, Office of Technology Assessment,Commercial Biotechnology: An InternationalAnalysis, OTA-BA-218 (Washington, D. C.: U.S.Government Printing Office, January 1984.


U.S. Congress, Office of Technology Assessment,Informational Technology and Its Impact onAmerican Education, OTA-CIT-187 (Washing-ton, D. C.: U.S. Government Printing Office, No-vember 1982).

U.S. Congress, Office of Technology Assessment,Technology, Innovation, and Regional Eco-nomic Development: Census of State Govern-ment Initiatives for High-Technology IndustrialDevelopment—Background Paper, OTA-BP-STI-21., (Washington, DC: U.S. GovernmentPrinting Office, May 1983).

U.S. Congress, Office of Technology Assessment,Technology, Innovation, and ReD”onal Eco-nonu”c Development: Encoura~”ng High Tech-nology Development—Background Paper 2,OTA-BP-STI-25 (Washington, DC: U.S. Govern-ment Printing Office, February 1984).

Wilson, K. G., “Science, Industry, and the NewJapanese Challenge, “ Proceedings of the IEEE,vol. 72, No. 1, January 1984, pp. 6-18.

Chapter 7

Foreign Information TechnologyResearch and Development

ContentsPage

Information Technology Research and Development 201201

for International Markets 202

International Trends inInternational Trade . .Adapting TechnologyMultinational Corporations . . . . . . . . . 206Technology Exchange Agreements . . . . . 207Patents . . . . . . . . . . . . . . . . . . . . . . . . . . 210Scientific and Technical Literature . . 211Science and Engineering Students . 212

Implications for U.S. Information Technology R&D Policies 214S c i e n c e a n d T e c h n o l o g y P o l i c y G o a l s . , . 215Government Role in Information Technology Research and Development 216Government/Industry/University Institutional Arrangements for Information

Technology Research and Development ., 217Industry Participation in Information Technology Research and Development 219

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221

Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222The Size of Japanese Participation in Information Technology Markets 225Government . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226University . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243

France . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . .,, 246The Size of French Participation in Information Technology Markets 247The Political Environment for French Information Technology

Research and Development . . . . . 248Social Environment for French Information Technology

Research and Development. . . . . . . . 252Government . . . . . . . . . . . . . . . . . . . . . . . 252University. . . . . . . . . . . . . . . . . . . . . . . . 257Industry . . . . . . . . . . . . . . . . . . . . . . . . 259

The United Kingdom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261The Size of U.K. Participation in Information Technology Markets . . . . . . . . . . . . . . 263Government . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263University. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270European Strategic Program for Research in Information Technology . . . . . . . . . . . . 272

TablesTable No. Page28. Computer Production and Apparent Domestic Consumption

of Six Leading Supplier Nations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , 20129.U.S. Computer Trade: Origins and Destinations,

Flow Value, and Annual Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20430. World Trade in Telecommunications Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20431. Aggregate Trends in U.S. Telecom Equipment Trade . . . . . . . . . . . . . . . . . . . . . . . . . 20432. Company R&D Performed Abroad by Foreign Affiliates

of U.S. Domestic Companies by Selected Industry: 1975 and 1981 . . . . . . . . . . . . , 20733. International. Technology Agreements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20834. Number of U.S. Patents Granted to Selected Foreign

Countries in All Product Fields and in CommunicationEquipment and Electronic Components, 1963-81 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210

35. Share of Foreign Patenting in the United States for theThree Most Active Countries by Selected Product Fields: 1981 . . . . . . . . . . . . . . . . 211

36.U.S. Owned and Foreign Owned U.S. Patents in Information Technologies . . . . . . 21237. Distribution of Foreign Students With Percentage

of All Foreign Students by Field of Study, for Selected Years: 1954/55-1981/82 . . 21338. Doctoral Degrees Awarded to Foreign Students as a Percent

of All Doctoral Degrees from U.S. Universities by Field: 1959-81, . . . . . . . . . . . . . . 21439. Average Annual Growth Rates in the Engineering Industry . . . . . . . . . . . . . . . . . . . 21840.R&D Performed in the Business Enterprise Sector

by Source of Funds: 1970, 1975, and 1979 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22041. Percent of Private Industrial R&D in Selected Industries: 1969-79 . . . . . . . . . . . . . 22042. R&D Expenditure by Type of Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22443. Japan Development Bank Loans for Development of Technology.. . . . . . . . . . . . . . 24044. Department of Trade and Industry R&D Programs . . . . . . . . . . . . . . . . . . . . . . . . . . 266

FiguresFigure No. Page35.U.S. Computer Trade Imports by Source; Exports by Destination. . . . . . . . . . . . . . 20336. Sources of U.S. Imports and Destinations of U.S. Exports

of Telecommunications Equipment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20537.U.S. Bilateral Trade Position in Telecommunications, With Selected Countries . . . 20638. Share of Foreign Patenting in the United States

for the Three Most Active Countries in 1981 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21139. index of International Cooperative Research by Country . . . . . . . . . . . . . . . . . . . . . 21240. Estimated Ratio of Civilian R&D Expenditures

to Gross National Product for Selected Countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . 21741. Trends in the Production Composition Ratio

of Major Consumer Electronics Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22542. Japanese Share of World Production of Consumer Electronics Products in 1980. . 22643. Japanese Information Technology Industry Sales . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22744. Integrated Circuit Relative Production Share . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22845. Japanese Government Organization for

Information Technology Research and Development. . . . . . . . . . . . . . . . . . . . . . . . . . 22946. Japanese Government Support for Information Technology . . . . . . . . . . . . . . . . . . . 23347. Cooperation Between Research Participants

for the Fifth-Generation Computer Systems Project . . . . . . . . . . . . . . . . . . . . . . . . . 23448. Concept Diagram Showing How Research and Development

Are to Progress in the Fifth Generation Computer Systems Project . . . . . . . . . . . . 23549. Chart Showing the Recommended French Government/Industry/University

Cooperation Relationships for Information Technology R&D . . . . . . . . . . . . . . . . . . 251

Chapter 7

Foreign Information TechnologyResearch and Development

International Trends inResearch and

Several trends demonstrate that the UnitedStates is experiencing greater international in-terdependence in the area of information tech-nology research and development. They in-clude: 1) the large and growing world marketfor computer and communications products;2) the increasing adaptation of informationtechnology products and standards for inter-national markets; 3) the growing number ofmultinational information technology firms;4) the increasing number of international tech-nology exchange agreements; 5) the increas-ing percentage of U.S. information technology-related patents granted for foreign inventions;6) the greater utilization of foreign contribu-tions in U.S. scientific and technical journals;7) and the growing number of foreign studentsenrolled in technical and scientific programsat U.S. universities.

These trends indicate two significant fac-tors, both of which make foreign organization

Information TechnologyDevelopment

and activities relevant to U.S. R&D efforts.First, they indicate a growing number of linksbetween other nations’ R&D efforts and thoseof the United States. Second, these trendspoint to a growing participation of foreign na-tions in information technology innovationand markets, which has led to a relative de-cline in the U.S. market share. Thus, theUnited States, which in the past has developedpolicies for its internal markets that werelargely unaffected by foreign manufacturers,may now need to take greater account of for-eign information technology research and de-velopment efforts.

International Trade

World trade in computer products is grow-ing rapidly. For each of the major supplier na-tions, overseas shipments are a steadily ris-ing share of both total output and consump-tion. Table 28 shows this trend towards a glo-

Table 28.—Computer Production and Apparent Domestic Consumptiona ofSix Leading Supplier Nations

1982 Percent 1982 Percentproduct ion change ADC a change($ million) 1981-82 ($ million) 1981-82

United States. . . . . . . . . . . . . . .Japan . . . . . . . . . . . . . . . . . . . . .France . . . . . . . . . . . . . . . . . . . .West Germany. . . . . . . . . . . . . .United Kingdom . . . . . . . . . . . .Italy . . . . . . . . . . . . . . . . . . . . . . .

Total . . . . . . . . . . . . . . . . . . . .Estimated share of world total

(percent) . . . . . . . . . . . . . . . . .

$33,550 b

7,1 79C

3,834d

3,5111,929d1,076

$51,079

89

12.30/o $26,888 15.3’%021.0 6,276 10.2

–5.0 4,720 8.47.5 3,789 5.3

11.8 2,898 NA11.0 1,343 3.2

10.80/0 $45,914 NA

80aAPParent Domestic consumption (ADc) iS production ITIi I_IUS exports Plus ‘m P o r t sbEstimated by Bureau of Industrial EconomicsCDOeS not Include parts

‘Preliminary.

SOURCE U S /ndustr/a/ Out/ook, 1984, Bureau of Industrial Economics, U.S Department of Commerce, 1984

201


bal computer market as it has evolved over thelast few years, and figure 35 illustrates ma-jor sources and destinations. This trend re-sults in part from the rapid rise in demand forcomputer-related products in the developingworld, a steady demand in traditional marketsfor products that incorporate informationtechnology, and increasing overseas activitiesof multinational subsidiaries.1

For the United States, increased global par-ticipation in the information technology mar-ket has meant a rapid rise in imports and adecreasing world market share. During the pe-riod 1978-82, U.S. imports of computers andcomputer-related products rose by approxi-mately 30 percent. (See table 29.)

The telecommunications market is also be-coming internationalized as equipment man-ufacturers look beyond maintaining tradi-tional markets (the national telecommu-nications service monopolies, or PTTs) towardexpanding international trade.2 Table 30 illus-trates this trend and summarizes the currentpositions of the United States, the UnitedKingdom, France, and Japan.

The internationalization of the telecommu-nications market, as in the case of the com-puter market, has weakened the relative U.S.position in telecommunications trade. Al-though U.S. exports have increased at a rateof 13 to 18 percent per year, a continuing in-crease in foreign imports (24 to 30 percent ayear) has diminished the U.S. trade surplus(table 31). Japan supplied about 50 percent ofU.S. imports, resulting in a U.S. trade deficitwith Japan of $250 million (figs. 36 and 37).3

—. —.. .——..‘High Technology Industries: Profiles and Outlooks: The

Computer Industry, U.S. Department of Commerce, Interna-tional Trade Administration, 1983, p. 22.

‘High Technolo~ Industn”es: Profiles and Outlooks, The Tel-ecoxnmuxu”cations Industry, U.S. Department of Commerce, In-ternational Trade Administration, 1983, p. 18.

‘Although the French, the British, and the Japanese are in-creasing their participation in information technology markets,particularly in the computers and telecommunications areas,the degree to which this trend is linked to information technol-ogy R&D remains unknown. The traditional skills needed forsuccess in the marketplace range from basic research, to ap-plied R&D, to production and distribution, and to marketingskills; it is therefore difficult to attribute success in the mar-ketplace solely to R&D efforts or to any other single factor.See ch. 2 for a more complete discussion.

Adaptations of Technology forInternational Markets

The growing international trade in informa-tion technology products has led to increasedefforts to develop international standards forinformation technology products in order toallow access to foreign markets and to allowinterconnections of services. For example, fol-lowing a recent meeting of the Commission ofEuropean PTTs (CEPT), European countriesagreed to develop technical standards not onlyfor basic equipment such as telephone hand-sets, but also for videotex systems and othersophisticated data communications systems.The CEPT program will also suggest otherareas where national telecommunications prac-tices might be standardized. This could even-tually lead to a unified European network ofapproximately 400 million subscribers. TheEuropean Program for Research and Develop-ment in Information Technology (ESPRIT)has a group working on international stand-ards specifically designed to enable variousEuropean-manufactured products to commu-nicate with each other.

In markets where standards do not exist, in-formation technology products, such as com-puter software, must be tailored for interna-tional sale. Because personal computerhardware has proliferated worldwide withouta parallel growth of indigenous software com-panies, many American software companiesare developing products for the internationalmarket.

For example, Lotus Development Corp. hasbeen tailoring its software packages to the lan-guage and idioms of other nations. The LotusInternational Character Set enables the pro-gram to generate different currency signs anddifferent versions of international day anddate displays. Although the cost of convertingprograms for international markets can bequite high, Lotus Corp. reportedly believesthat the return on its investment will also besubstantial. They expect that internationalsales will eventually generate between 30 and40 percent of the company’s income.4

— . — —4Mich~el Schrage, “Firms See Boom in Software, ” The Wiwh-

ington Post, Mar. 4, 1984, p. H, 4.

—

Ch. 7—Foreign Information Technology Research and Development ● 203

Figure 35.—U.S. Computer Trade (SIC 3573) Imports by Source; Exports

Other countries25.80/0

I —

by Destination

United Kingdom15.20/o

NOTE: SIC 3573 includes: Computing equipment (equipment, peripherals, and services).

SOURCE: High Technology Industries: Profiles and Outlooks: The Computer Industry, International Trade Administration, U.S. Department of Commerce, 1983.


Table 29.—U.S. Computer Trade @W 3573): Origins and Destinations, Flow Value, and Annual Growth1981-82, percent of value) (millions of U.S. dollars)

1982 ImportsJapan . . . . . . . . . . . . .$ 729Canada . . . . . . . . . . . . 404Hong Kong . . . . . . . . 151Mexico . . . . . . . . . . . . 123United Kingdom . . . . 90West Germany . . . . . 79Spain . . . . . . . . . . . . . 78France . . . . . . . . . . . . 64

Total ... ... ... .. .$2,140

(+88.20/o)(+ 0 . 0 % )(–21.60/o)(+27.00/o)(+ 12.2%)(+12.90/o)(+25.3%)(– 2.6%)(+29.9%)

United States

Imports Exports1978 . . . . . . $ 755 $4,1941981 . . . . . . 1,646 8,4931982 . . . . . . 2,295 8,9571983 . . . . . . 4,100 10,3001984 . . . . . . 6,470 12,360Growth

1982 Exports

United Kingdom ... .$1,374 (+15.3%)Canada . . . . . . . . . . . . 1,103 (–11.2%)West Germany . . . . . 958 (– 6.20/o)France ., . . . . . . . . . . 841 (+ 7.0%)Japan . . . . . . . . . . . . . 777 (+ 8.30/o)Netherlands . . . . . . . . 380 (+ 14.0%)Australia. . . . . . . . . . . 344 (+ 0.0%)Italy. . . . . . . . . . . . . . . 298 (– 4.1%)

Total ... ... ... .. .$9,040 (+ 4.5%)(1978-82) . +29.8% +21.2`%

NOTE: SIC Code 3573 includes: Computing equipment (equipment, peripherals, and services).SOURCE: H/gh Techrro/ogy /mWstr/es: Profiles arrd Outlooks: The Computer Industry, International Trade Administration, U.S. Department of Commerce, 1983.

Table 30.–World Trade in Telecommunications Equipment (SIC 3661)(millions of U.S. dollars)

1977 1981

Principal producer countries Imports Exports Balance Imports Exports Balance

Japan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $ 24 $ 363 +$ 339 $ 4 6 $ 911 +$ 665Sweden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 458 +422 65 776 +71 1West Germany . . . . . . . . . . . . . . . . . . . . . . . . 104 562 +458 128 809 +681Netherlands. . . . . . . . . . . . . . . . . . . . . . . . . . . 126 228 + 102 128 398 +270France . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 168 +111 86 320 +234United States . . . . . . . . . . . . . . . . . . . . . . . . . 129 257 + 128 494 653 + 159Canada . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 80 –13 143 298 + 155United Kingdom . . . . . . . . . . . . . . . . . . . . . . . 91 247 + 156 235 331 +96Belgium/Luxembourg . . . . . . . . . . . . . . . . . . . 76 248 + 172 118 262 + 144Italy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 97 +47 101 143 +42

Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 786 2,708 + 1,922 1,544 4,901 +3,357NOTE: SIC Code 3881 includes: Telephone and telegraph apparatus.SOURCE: High Tec/rno/ogy hrdusfries: Prof//es and Outlooks: The Telecommunications Industry, International Trade Administration, U.S. Department of Commerce, 1983.

Table 31.–Aggregate Trends in U.S. Telecommunications Equipment Trade (SIC 3661)(millions of U.S. dollars)

1977-831972 1977 1979 1980 1981 1982 1893 growth rate

Exports . . . . . . $76 $257 $448 $557 $653 $725 $850 +22.1 “/0Imports . . . . . . 86 129 319 421 494 635 790 +35.3%Balance . . . . . . –10 + 128 + 128 + 136 +159 +90 +60 – 11.1%NOTE: SIC Code 3881 includes: Telephone and telegraph apparatus.

SOURCE: U.S. /ndustrla/ Out/ook, 1983, Bureau of Industrial Economics, U.S. Department of Commerce, 1983.

International marketing is also an impor- West Germany. Like Lotus, Microsoft hastant component for Microsoft, a U.S. software revised many of its software programs forcompany whose overseas market accounted foreign use. The company has closely tailoredfor approximately one-third of its estimated its software products for the Japanese mar-$75 million 1983 revenue. Microsoft already ket by offering phonetic Japanese versions ofhas development operations in Japan and sub BASIC, and it has translated its Multiplansidiaries in the United Kingdom, France, and program (business applications program) and

Ch. 7—Foreign Information Technology Research and Development Ž 205

Figure 36.—Sources of U.S. Imports (1982) and Destinations of U.S. Exports (1982) of TelecommunicationsEquipment (SIC 3661)

Japan45.20/o

N e t h e r l a n d s

West Germany 2.4%

United Kingdom 2.9% Canada . . \ 22.50/o \ / “

I- ! - - - - -

Kingdom/

I

4.60/o4 3 . 0 %

NOTE: SIC 3661 includes: Telephone and telegraph apparatus.

SOURCE: High Technology /ndustries: Profiles and Outlooks: The Te/ecornmunicatio~s Industv, International Trade Administration,U.S. Department of Commerce, 1983


Figure 37.– U.S. Bilateral Trade Position inTelecommunications with Selected Countries (1981)

Japan Can U.K. Ger Neth Mex

+ 1 5 0

+ 100

+ 5 0

(Mil $) O

- 5 0

- 1 0 0

– 150

– 200

- I m p o r t sExports

SOURCE: High Technology Industries: Profiles and Outlooks: The Tebcommuni-cations Industry, International Trade Administration, U.S. Departmentof Commerce, 19S3.

its MSX operating system (home computerprogram) for the Japanese market.

Multinational Corporations

Rising innovation costs and the accompany-ing size of financial risks, as well as increas-ing equipment costs, have intensified the needfor expanding production and have forcedmany manufacturers beyond the limitationsof domestic markets. For a variety of reasonsincluding tariffs and other forms of protectivelegislation that place imported products at acompetitive disadvantage, multinational firmshave attempted to capture specific foreignmarkets through the establishment of foreignsubsidiaries.

Many U.S. firms have opened productionand R&D facilities in foreign nations. DigitalEquipment Corp., for instance, operates sixplants in Europe and three more in the FarEast. Hewlett-Packard, Wang, Data-General,Datapoint, and Texas Instruments are U.S.minicomputer manufacturers that also oper-ate foreign production facilities. Apple Com-puter has plants in Ireland and Singapore.

Amdahl and Trilogy Systems, manufacturersof plug-compatible mainframes, have openedfacilities in Ireland, intended in part to supplythe Common Market.

Many foreign firms operate subsidiaries inthe United States and other foreign countries.For example, Japan’s NEC Corp. has estab-lished three subsidiaries in the United Statesand has won major contracts to supply U.S.manufacturers with Japanese technology. Inaddition, NEC has subsidiaries in Germany,the United Kingdom, and countries in Africaand South America.

This growth of international activity has notonly led to increased trade in information tech-nology products, but has also encouraged theperformance of R&D by firms in various na-tions. After establishing foreign subsidiaries,many companies find that R&D is necessaryto support local manufacturing when the re-quirements or standards for the foreign mar-ketplace are significantly different.

U.S. companies are performing an increas-ing amount of research and developmentabroad. Since 1975, total R&D conducted byU.S. subsidiaries overseas has more than dou-bled and in 1981 amounted to $3.2 billion-9percent of total U.S. private R&D funding.Table 32 illustrates the increasing amount ofelectronics R&D which is performed abroad byforeign affiliates of U.S. companies. In con-trast, in 1979, total expenditures for elec-tronics research and development performedby U.S. affiliates of foreign companies in-creased to $148 millions

Many governments actively encourage for-eign subsidiaries not only to establish produc-tion facilities, but also to conduct R&Din theirnations. The United Kingdom has, for in-stance, implemented a series of incentives forforeign firms to innovate. In addition to pro-viding financial incentives for establishingmanufacturing facilities, the United King-dom’s Support for Innovation Program (SFI)

‘Science Indicators, 1982, National Science Board, NationalScience Foundation, 1983, p. 25.

Ch. 7—Foreign Information Technology Research and Development . 207

Table 32.—industry R&D Performed Abroad byForeign Affiliates of U.S. Domestic Companies by

Selected Industry: 1975 and 1981(millions of U.S. dollars)

PercentIndustry 1975 1981 increase

Food and kindred products. . . 23 66 187Chemicals and allied products 269 651 142

Industrial andother chemicals . . . . . . . . 85 275 124

Drugs and medicines . . . . . . 184 376 104Stone, clay, and

glass products . . . . . . . . . . . 7 15 114Primary metals . . . . . . . . . . . . . 9Fabricated metals. . . . . . . . . . . (a) 26 b N—AMachinery . . . . . . . . . . . . . . . . . 331 585 77Electrical equipment . . . . . . . . 245 455 86

Electronic components . . . . 7 47 571Transportation . . . . . . . . . . . . . . 412 893 117

Motor vehicles and othertransportation equipment . 373 791b 112

Aircraft and missiles . . . . . . 39 l o 2b 161Professional and scientific

instruments . . . . . . . . . . . . . . 49 101 106Other manufacturing

industries . . . . . . . . . . . . . . . . 105 147 40Nonmanufacturing industries . 4 12 200

Total ... ... ... ... ... .. .$1,454 $3,157 117

alncluded in the other manufacturing industries 9r0upbEstimatedNA—Not available

SOURCE: Science /rrdicafors, 1982, National Science Board, National ScienceFoundation, 1963

offers grants of up to 33 1/3 percent towardsthe cost of significant research and develop-ment of high technology products. About 200information technology firms are located in“Silicon Glen” in Scotland. U.S. firms thereinclude IBM, Honeywell, NCR, Hewlett-Pack-ard, Digital Equipment, and National Semi-conductor.

Multinational information technology cor-porations are also forming cooperative inter-national research and development arrange-ments. This new type of internationalarrangement is exemplified by the recentestablishment of Europe’s first multinationalresearch institution for information technol-ogy. Following an agreement made last Sep-tember, Europe’s three largest computer man-ufacturers, France’s Compagnie des MachinesBull SA, Britain’s International ComputersLtd. (ICL), and West Germany’s Siemens AG,will operate a jointly run and jointly financedEuropean Computer Research Centre (ECRC).

Located in Munich, close to a number of elec-tronics firms, the ECRC will begin operationswith an initial capital investment of $655,000and a staff of four researchers from the threefirms. The number of researchers is expectedto reach 30 to 35 by 1985, and approximately50 within 2 years.

Technology Exchange Agreements

Table 33 illustrates some of the recent tech-nology exchange agreements between U. S.,European, and Japanese companies. Suchtechnology exchange is seen by firms as a wayto spread the risk in large projects, to enterinternational markets where politics or nation-al market specifications hinder entry into do-mestic markets, and to allow competitors topursue a dominant position in a specificmarket.6

An example of a technology exchange agree-ment is the reciprocal development and mar-keting agreement for office telecommunica-tions equipment between AT&T and Ing. C.Olivetti & Co., a major European supplier ofoffice automation equipment. In accordancewith this agreement, AT&T will increase itsstake in Olivetti over the next 4 years toacquire 25 percent ownership in the company.The arrangement gives AT&T access to Oli-vetti equipment such as workstations, wordprocessors, typewriters, and data processingsystems for domestic marketing. In turn, Oli-vetti will market AT&T communications con-trollers for voice, data and networking applica-tions, and a variety of micro-computers.7

Other technical partnerships include thoseof LM Ericsson of Sweden with Honeywell,and Atlantic-Richfield in the United Stateswith Thorn EM I in the United Kingdom. InItaly, Italtel is cooperating with Telettra andthe U.S. company GTE in the public switch-ing field. The British firm, ICL, has links withMitel, and Plessey (another U.K. company),

—— — ——“’Bulls on Skis, ” The Economist, Feb. 4, 1984, p. 76.‘In addition to its technical exchange agreements with Oli-

vetti, AT&T has also made exchange agreements with Philips.


Table 33.—international Technology Agreements

Sector/partners a Date Technology Agreement

Communications:Hitachi and Western Electric

Fujitsu and Ungermann-Bass (CA)

NTT and HughesSperry-Univac and Mitsubishi

Motorola and NEC

Exxon Office Systems andMitsubishi

ATT and PhilipsHoneywell and Ericsson

GTE and Italtel

Plessey and Stromberg Carlson

General Instruments and Thomson

Micro V and Jeumont-Schneider

Data processing:Honeywell and NECExxon Office Systems and ToshibaTRW and Fujitsu

Sperry and MitsubishiVertimag and Teijin (Osaka)

Amdhal and FujitsuDrexler Technology and Toshiba

Olivetti and DocutelOlivetti and StratusNixdorf and AuragemPhilips and MicomATT and Olivetti

Fortune and Thomson

Tandy Corp. and Matra

Matra and Tymshare

Cii Honeywell Bull and Trilogy

Cii Honeywell Bull and Honeywell

AMD and IBM

1981

—

——

1982

—

19831983

1982

1982

1983

1982

1984——

1982—

1982—

19821982198319771983

1982

1982

1982

1980

—

—

Communication equipment

Local networks

Telecommunications by satelliteLocal networks and

communication and dataprocessing system

Portable paging systems

Telecommunications equipment

PBXTelecommunications and office

automationPBX

PBX

Videocommunication andteledistribution by cable

PBX

Main frame computersOffice automationData processing

Office automationHigh-density magnetic memory

Computers and peripheralsSmart card

Office automationMinicomputersMinicomputersOffice automationData processing, office

automation, communicationsMicrocomputers

Microcomputers

Terminals

Main-frame computers

General data processing

Computer-aided design

Patent exchange—technologicalagreement

Industrial and commercialagreement

—Technological agreement

Agreement for manufacture andcommercialization in Japan

Technological agreement

Commercial agreementTechnological and commercial

agreementTechnological and commercial

agreementPurchase of Stromberg-Carlson by

Plessey

Technological and commercialagreement

Technological and commercialagreement with partialacquisition of Micro V andcreation of a joint subsidiary

Technological agreementTechnological agreementJoint venture now controlled 100

percent by FujitsuTechnological agreementTechnological and commercial

agreementTake-over by FujitsuTechnological and industrial

agreementControl of Docutel by OlivettiControl of Stratus by OlivettiControl of Auragem by NixdorfPurchase by PhilipsATT acquires 25 percent of Olivetti

Thomson acquires 17 percent ofFortune


Commercial joint venture in theUnited States

Technological agreement with CiiHoneywell Bull having minorshare in Trilogy

Technological and commercialcooperation

Commercialization by IBM of Catia


Table 33.—lnternational Technology Agreements—continued

Sector/partnersa Date Technology Agreement

Electronics and components:/rite/ and NECTexas Instruments and FujitsuHewlett-Packard and HitachiWestern Electric and NTTIBM and NTTZilog and Toshiba

Western Digital and SiemensUnited Technologies and AEG

TelefunkenPhilips and SigneticsPhilips and MagnavoxPhilips and Sylvania

GCA and Matra

Harris and Matra

Motorola and Thomson

/rite/ and Matra

Thomson and RCARhone-Poulnec and SiltecRhone-Poulnec and DysanSagem and MotorolaThomson and Diasonic

19821979

————

1982

1982

197519771980

1982

1981

1978

1981

19711983

——

1983

VLSI microprocessors and circuits Technological agreementIntegrated circuitsIntegrated circuitsIntegrated circuitsIntegrated circuitsMicroprocessorsIntegrated circuits

Custom-made semiconductors

ComponentsConsumer electronicsConsumer electronics and tubes

Microelectronic equipment

Components and integratedcircuits

Components and integratedcircuits

Integrated circuits

Color tubesSiliconMagnetic disksBubble memoriesMedical instrumentation

Technological agreementTechnological agreementTechnological agreementTechnological agreementTechnological agreementProduction and commercial

agreementProduction and commercial

agreementPurchase by PhilipsTake-over by PhilipsPurchase by Philips


Technological, industrial, andcommercial agreement



Technological agreementJoint ventureTechnological agreementTechnological agreementTechnological and commercial

agreementaThe technologically dominant partner is italicized.

SOURCE: Office of Technology Assessment. Compiled from Research and Deve/opmertf /n Electronics.’ USA-France, 1982/7983, French Telecommunications Council, 1984.

now owns Stromberg-Carlson. In addition,France and the United States have arrangedjoint ventures whereby American semiconduc-tor firms have exchanged technical know-howfor access to the French markets-particularlyto the French telecommunications market(normally well protected by the French PTT).These joint ventures in which the French part-ners hold controlling interests include Thom-son and Motorola, Saint-Gobain and NationalSemiconductor (in a firm named Eurotechni-que), and Matra and Harris.

U.S. companies also have technical ex-change agreements and other business rela-tionships with Japanese firms. Intel Corp., forexample, has a 5-year cross-licensing, cross-compatibility, and technology exchange agree-ment primarily in the area of controllers andperipheral equipment with NEC Corp. AmdahlCorp. currently uses a semiconductor chip de-veloped and manufactured by Fujitsu in its

U.S.-manufactured computers. Moreover,when Amdahl had difficulties raising capitalfor expansion, Fujitsu bought 40 percent ofAmdahl Although Amdahl might have beensold to another American firm, Amdahl man-agement preferred to sell its stock to Fujitsuin order to facilitate technology exchangeagreements, which involve cross-licensing,financing, and information exchange.8 FujitsuLtd. has also recently agreed to supply TexasInstruments with gate array technical know-how; Texas Instruments will produce the Jap-anese gate arrays and ship them back toFujitsu.

Other joint technical agreements betweenAmerican and Japanese information technol-ogy firms include Sperry’s high technology co-operative agreement with Mitsubishi, whichcovers joint activities in manufacturing, re-

8Patricia Keefe, “Many U.S. Firms Have Japanese Ties, ”ComputerWorld, May 2, 1983, p. 73.


search and development, and marketing ofcomputer systems. Mitsubishi also has a jointagreement with IPL Systems to develop anIBM-compatible processor. The technical ex-change agreement between the two firms com-bines Mitsubishi’s” computer-aided design andlarge scale integration technology with IPL’sdesign expertise. Under the agreement, bothfirms are granted the right to market thejointly developed products. Mitsubishi andWestinghouse have also arranged a joint ven-ture to design and manufacture integrated cir-cuits. In addition to technical exchange agreements between private firms, the U.S. Depart-ment of Defense encourages Japan to trans-fer defense-related electronics technologies tothe United States. In 1981, for example, theU.S. Government asked Japan to provide ad-vanced very large scale integration (VLSI)technology to enhance air and antisubmarinedefense capabilities.

Patents

A large number of U.S. patents are grantedto foreign individuals and corporations (seetable 34). Foreign patenting activity in theUnited States has been related both to in-creased foreign inventive activity and to agrowing interest in the U.S. information tech-nology market. Moreover, studies have shownthat foreign patenting activity in the UnitedStates by selected OECD countries correlatessignificantly with industrial R&D in thosecountries. This correlation is especially highin the electrical and electronics industries.9

Foreign patenting in the communicationequipment and electronic components cate-gory was as much as 40 percent of the totalnumber of U.S. patents granted during 1979-81, while the percentage of U.S. owned foreignpatents in the same category was 13 percent.l0

— . — —9Keith Pavitt, “Using Patent Statistics in Science Indicators:

Possibilities and Problems, ” Z%e Meani”ng of Patent Statistics,National Science Foundation, 1979.

*’The fields that have relatively high percentages of U. S.-owned foreign patents are the areas corresponding to U.S. di-rect investment and research activity abroad. It is possible thatU.S. laboratories abroad supported R&D that resulted inpatented innovations. Sa”ence hh”cators, 1982, National Sci-ence Board, National Science Foundation, 1983, p. 14.

Table 34.-Number of U.S. Patents Granted toSelected Foreign Countries” in All Product Fieldsand in Communications Equipment and Electronic

Components (1963-81)

Communications equip-ment and electronic

Country of inventor All fields componentsUnited States . . . . . 865,124 101,914Foreign . . . . . . . . . . 369,519 41,242

West Germany . . 91,359 7,850Japan . . . . . . . . . . 77,450 13,013United Kingdom . 51,138 5,976France . . . . . . . . . 35,244 4,455Switzerland . . . . . 21,622 1,258Canada. . . . . . . . . 20,241 1,773Sweden . . . . . . . . 13,368 1,007Italy . . . . . . . . . . . 11,958 847Netherlands . . . . 11,103 2,701 b

U.S.S.R. . . . . . . . . 5,111 454Belgium . . . . . . . . 4,459 360Austria . . . . . . . . . 4,080 273Australia . . . . . . . 3,585 198Denmark . . . . . . . 2,520 161Mexico . . . . . . . . . 1,075 21Other foreignc. . . 15,206 895

Total . . . . . . . . . 1,234,643 143,156%ountrles were selected on the basis of being in the top 10 of at least one ofthe Standard Industrial Classifications.

blndicates ranking among the top five foreign countries in this PafllCLIlar Prod-uct field.

cother foreign Includes patents granted to foreign COUntrieS not shownseparately.

SOURCES: Office of Technology Assessment; compiled from information in Of-fice of Technology Assessment and Forecast, U.S. Patent andTrademark Office, Indicators of the Patent Output of U.S. /ndustry/V(fW3+31), 19S2; in Science Mlcafors, 19S2, National Science Board,National Science Foundation, 19S3,

Table 35 and figure 38 show that Japan hasthe largest number of foreign U.S. patents incommunications equipment and electroniccomponents, although West Germany hadbeen the foreign leader in this field through-out the 1960s and mid-1970s. Since 1970, Ja-pan has doubled its patent activity in commu-nications equipment and electronic comp-nents, food and kindred products, primarymetals, and professional and scientific in-struments.11 Data for the period 1970-81, pre-sented in table 36, show that the percentageof U.S. patents in information technologyareas decreased more than 20 percent whilethe Japanese share of U.S. patents increasedby over 2oo percent.12

*’Science Indicators, 1980, National Science Board, NationalScience Foundation, 1981, p. 21.

IZThe Office of T~hnology Assessment and Forecast, U.S.Patent and Trademark Office.


Table 35.—Share of Foreign Patenting in the United States for theThree Most Active Countries by Selected Product Fields (1981)

Total West United OtherProduct field foreign Germany Japan Kingdom foreign

Percent of foreign

Chemicals, except drugs and medicines . . . . .Drugs and medicines . . . . . . . . . . . . . . . . . . . . .Nonelectrical machinery . . . . . . . . . . . . . . . . . . .Electrical equipment, except

communications equipment . . . . . . . . . . . . .Communications equipment and

electronic components . . . . . . . . . . . . . . . . . .Motor vehicles and other equipment

except aircraft . . . . . . . . . . . . . . . . . . . . . . . . .Aircraft and parts. . . . . . . . . . . . . . . . . . . . . . . . .Professional and scientific instruments . . . . .

100100100

282326

272227

1114

9

344038

100 21 37 8 34

100 18 44 9 29

100100100

262822

344243

9108

312127

Number of patents

Chemicals, except drugs and medicines. . . . .Drugs and medicines . . . . . . . . . . . . . . . . . . . . .Nonelectrical machinery. . . . . . . . . . . . . . . . . . .Electrical equipment, except

communications equipment . . . . . . . . . . . . .Communications equipment and

electronic components . . . . . . . . . . . . . . . . . .Motor vehicles and other transportation

equipment except aircraft . . . . . . . . . . . . . . .Aircraft and parts. . . . . . . . . . . . . . . . . . . . . . . . .Professional and scientific instruments . . . . .

5,3381,2888,166

1,520300

2,088

1,452288

2,240

566182731

1,800518

3,107

2,541 535 952 202

279

852

3.027 534 1,338 876

1,652777

4,100

429216892

563323

1,760

15175

329

509163

1,119

SOURCE: Office of Technology Assessment and Forecast, U S Patent and Trademark Office, Indicators of Patent Output of U.S Industry (1%53-81); in Science Indicators,1982, National Science Board, National Science Foundation, 1983.

Figure 38.—Share of Foreign Patenting in the United from 45 to 37 percent in the United Kingdomstates for the Three Most Active Countries (1981) and from 32 to 28 percent in France. Overall,

C o m m u n i c a t i o n s e q u i p m e n t from 1971 to 1981, U.S. patenting in Canada,and electronic components Japan, and in the European Economic Com-

munity declined approximately 40 percent.

Scientific and Technical Literature

U.S. utilization of foreign engineering andtechnical literature is growing. Between 1973and 1980, U.S. citations of foreign researchfindings in engineering and technology fieldsincreased by 4 percentage points and by 7 per-centage points in the field of mathematics. Al-though the U.S. utilization of foreign researchhas grown, U.S. use of foreign research

SOURCE. Science /rrd/carors, 1982, National Science Board, National Sctence literature is lower than other nations’ use ofFoundation, 1983 foreign research literature.14

While the foreign patenting activity in the The number of jointly authored articles by

United States has been increasing, U.S. pat- scientists and engineers from different coun-

ent activity abroad has been decreasing over —– —-the past decade. Over the past 10 years, the ‘3Sa”ence In&”cators, 1982, National Science Board, National

Science Foundation 1983, p. 15.U.S. proportion of foreign patents decreased “Ibid, p. 12.


Table 36.—U.S..Owned and Foreign-OwnedU.S. Patents in Information Technologiesa

U.S. patent Percent ownership Percentownership 1970 1981 change

United States . . . . . . . . . 76 58 –23.7Japan . . . . . . . . . . . . . . . . 6 19 +216United Kingdom . . . . . . . 4 4 0France . . . . . . . . . . . . . . . 3 4 +33West Germany . . . . . . . . 5 8 +60alnformation technologies included here comprise SIC 357—Office c0mPutin9and accounting machines; and SIC 365-367—communications equipment andelectronic components.

SOURCE: The Office of Technology Assessment and Forecast, U.S. Patent andTrademark Office. Information technology numbers were calculatedfrom data developed under the support of the National Science Founda-tion, Science Indicators Unit.

tries is also increasing. International co-authored engineering and technology-relatedarticles as a percentage of institutionally co-authored articles have risen from 13 percentin 1973 to 16 percent in 1980. In 1980, morethan 40 percent of all jointly authored articlesin mathematics were international collabora-tive efforts.16 Moreover, the United Kingdom,France, and West Germany had a greater per-centage of internationally co-authored articles(as a percentage of all institutionally co-authoredarticles) than the United States. Japan and theUnited States had the lowest percentages ofinternationally co-authored articles. (See fig-ure 39.)

Science and Engineering Students

The number of foreign students in scientificand technical fields in U.S. universities is in-creasing. In mathematics and computer sci-ence the number of foreign students enrolledin U.S. universities was 22,620 in 1981 -82.16

(See table 37.) Table 38 illustrates the largeproportion of doctoral degrees awarded to for-eign students in mathematics and computerscience during 1981. In 1982 non-U.S. citizenswere awarded 38 percent of the 542 doctoratesin electronics and electrical engineering and54 percent of the 72 doctorates in computerscience. l7 Although some of the foreign engi-neering and mathematics students choose to

161bid, p. 31.la~”ence ~~”cator9, 1980, National Science Bored, Nation~

Science Foundation, 1981, p. 240.17’’ Washington Newsletter,” Ehwtrom”cs, Jan. 12, 1984, p. 70.

Figure 39.—index’ of International CooperativeResearch by Country

(Percent)

o 5 10 15 20 25 30 35 40 45 50I I I I I [ I I I I 1

UnitedStates

J a p a n

U.S.S.R.

France

C a n a d a

UnitedK ingdom

West

Ge rmany

lobtained by dividing the number of all articles which were Written by sCif3n-tists and engineers from more than one country by the total number of articlesjointly written by SIE’S from different organizations regardless of the countryinvolved.

NOTE: Based on the articles, notes, and reviews in over 2,100 of the influentialjournals carried on the 1973 Sc/errce Cltatforr Index Corporate Tapes ofthe Institute for Scientific Information.

SOURCE: Science Indicators, 1982, National Science Board, National ScienceFoundation, 1963.

remain in the United States (if permitted), alarge number of them return to their nativecountries.

Although the number of foreign graduatescience and technology students in U.S. univer-sities has been increasing, the number of U.S.students enrolled in technical programs at for-eign universities has been decreasing. Thenumber and percent of U.S. graduate studentsstudying abroad was highest during 1971, butnow only constitutes about 1 percent of U.S.graduate students.18 The decline of U.S. grad-uate students studying in foreign universitiescould be attributed to employment considera-tions and cost of living differences. Currently,as other industrialized nations’ technical ca-pabilities improve, the low number of graduatestudents abroad could inhibit the U.S. abilityto keep abreast of the latest foreign researchmethods and developments.

“Science Inculcators, 1982, National Science Board, NationalScience Foundation, 1983, p. 29.

Ch. 7—Foreign Information Technology Research and Development • 213

C o r - a m

00

.

0 0 0.

.

- C o m a ) . nU ) ” N - . - - -

c

- c)d

0

0“

I

0

r -

0 0 0 0o

N N T -

0 0 0 3 . ( = . =

- c

m c - - J

00

0

0

0 0 0 0 0 0

- m

- - - - -

- -

0 0 0

- -

Co

“

.m C--J

c n

.-- m-c-+----

o - m m - o

0

0

C-J

)

n?) c-o

- Jk

J

214 ● Information Technology D: Critical Trends and Issues

Table 38.—Doctoral Degreesa Awarded to Foreign Students as a Percent ofAll Doctoral Degrees from U.S. Universities by Field 9-81)b

Field 1959 1963 1967 1971 1975 1979 1981

Science and engineering . . . . . . . . . . 14.8Physical sciences . . . . . . . . . . . . . . 12.6

Physics and astronomy . . . . . . . 14.9Chemistry . . . . . . . . . . . . . . . . . . . 10.5Earth sciencesc . . . . . . . . . . . . . . 17.1

Mathematical sciences. . . . . . . . . . 13.4Mathematics . . . . . . . . . . . . . . . . . AComputer sciences . . . . . . . . . . .

Engineering . . . . . . . . . . . . . . . . . . . 24.5Life sciences . . . . . . . . . . . . . . . . . . 17.6

Biological sciences. . . . . . . . . . . 15.5Agriculture and forestry . . . . . . . 24.9

Social sciences . . . . . . . . . . . . . . . . 10.7Psychology . . . . . . . . . . . . . . . . . . 5.5Other social sciences. . . . . . . . . 15.5

Conscience total . . . . . . . . . . . . . . . . . 6.0All fields . . . . . . . . . . . . . . . . . . . . . . 11.7

15.514.014.512.219.915.2NANA

20.817.516.521.011.64.0

17.66.5

12.3

17.515.716.914.617.214.6NANA23.719.616.232.413.54.4

19.97.4

14.0

18.716.718.615.614.617.5NANA29.818.214.333.613.75.6

19.38.0

14.4

22.122.927.619.822.124.3NANA

42.119.514.937.413.75.8

20.28.7

16.2

21.121.025.719.716.625.526.721.346.816.512.135.212.94.0

22.410.1

16.1

22.122.026.121.217.130.833.726.351.519.811.137.613.03.9

24.011.017.2

a percen t of those whose citizenship is known.

bFiscal year of doctorate,

clncludes oceanography.

NA—Not available.

SOURCE: Doctorate Record File, Special Tabulations, unpublished data; National Science Foundation, ScierIce Indicators, 1982, National Science Board, National ScienceFoundation, 1983

Implications for U.S.R&D

Given the strong links between other na-tions’ information technology research and de-velopment activities and those of the UnitedStates, and economic competition from coun-tries such as the United Kingdom, France, andJapan in information technology innovationand international markets, foreign R&D ini-tiatives are a concern for the United States.For the U.S. Government to participate in in-ternational research and development and, atthe same time, successfully compete withthese nations’ R&D initiatives, the UnitedStates will need to understand other nations’economic and social goals for technological de-velopment, government, and industry roles ininformation technology R&D, and the signifi-cance of targeted national information tech-nology R&D programs.

Although the philosophy of industrial com-petition is prevalent in other economies-par-ticularly in the United Kingdom and Japan–

Information TechnologyPolicies

these industrialized nations have coordinatedtheir R&D efforts at a national level. Indus-trial and governmental cooperation is viewedas a means to achieve common national tech-nological objectives and enhance the competi-tiveness of the entire national informationtechnology industry in global markets. Coordi-nation for R&D includes efforts at the nationallevel to disseminate information on technologi-cal developments, share research results, anddivide research activities among enterprises.Moreover, in coordinating R&D efforts be-tween government, university, and industryparticipants, these nations have attempted tolink trade competitiveness strategies moreclosely with R&D policies.lg

In the United States, competition in tech-nological development among firms is viewed

‘gWilson P. Dizard, “U.S. International Information Trade, ”The Information Society, vol. 2, No. 3/4, 1984, p. 189.


as a variant of other forms of competition,such as competition in production efficiency,product quality, and marketing. Consequent-ly, the United States relies heavily on openmarket competition and private initiative tospur industrial R&D.20 Moreover, trade com-petitiveness factors are not always consideredin the formulation of U.S. R&D policy.

Since other industrialized nations target in-formation technology research and develop-ment programs at the national level, the ques-tion has been raised whether the United Statesshould adopt a similar national industrialstrategy for technological development. Al-though a number of nations that pursue coor-dinated industrial policies have relativelyweaker overall economic performances thanthe United States, those that target informa-tion technology as a national priority may im-prove their competitiveness.

In response to foreign coordinated nationalR&D programs, a number of legislative op-tions (several of which are modeled on foreigninitiatives) for coordinating and targeting in-dustrial sectors have been proposed in theUnited States. There is also evidence, pre-sented in this report, that the United Statesmay already be responding in ways particu-larly suited to its social, governmental, andeconomic traditions. For instance, the govern-ment-industry technology transfer activitiesstimulated by the Stevenson-Wydler Technol-ogy Innovation Act of 1980 (Public Law 96-480) (described inch. 2) and the recently form-ing university-industry cooperative joint ven-tures (described in ch. 6) may be indicationsthat the United States is beginning to developindigenous mechanisms to pursue commontechnological objectives.21

If, however, the United States is to developcoordinated national industrial strategies inresponse to other nations’ targeted efforts inthe area of information technology, the use ofother nations’ national R&D organizations

—— ———‘“Jack Baranson and Harold B. Malmgren, “Technology and

Trade Policy: Issues and an Agenda for Action, ” Bureau of In-ternational Labor Affairs, U.S. Department of Labor, and theOffice of the U.S. Trade Representative, 1981, p. 5.

“See for example, Jan Johnson, “America Answers Back, ”Datamation, May 15, 1984, p. 4057.

and activities as models should be carefullyconsidered. As a result of major differences inthe historical, cultural, and economic charac-teristics of each nation, there appear to be dif-ferences in their respective approaches toscience and technology policies, as well as gov-ernment and industry participation in infor-mation technology R&D. Moreover, theUnited Kingdom, French, and Japanese na-tional information technology research and de-velopment programs differ in overall goals andorganization and vary significantly from cur-rent U.S. R&D efforts.

Science and Technology Policy Goals

Although the United Kingdom, France, andJapan each developed science and technologypolicies in part to strengthen and modernizetheir economies, each of these nations variesin its conception of what technology policyshould comprehend and what its objectivesshould be.22 In some nations policy is aimedat strengthening the competitiveness of tar-geted industries. Other nations are concernedwith developing information technologies forsocial needs or national security applications.In other nations, the perception of technologypolicy is much broader, and constitutes a partof a more general plan of how the economyshould be structured in the future. Many ofthe goals for science and technology policy ofthe United Kingdom, France, and Japan arerooted in history. At the finish of World WarII, each of these nations’ societies and econo-mies were severely damaged. These nations’governments therefore perceived a need to ac-tively promote the growth of a high technol-ogy industry in order to aid their ailingeconomies.

Since World War II and particularly in morerecent times, science and technology policieshave become increasingly politicized. This re-flects the view that science and technology islinked to nations’ economic well-being (e.g.,trade, productivity, and employment) and so-cial welfare (e.g., quality of life, education, andtraining). Moreover, the widespread belief that

“Jack Baranson and Harold Malmgren, op. cit., p. 6.

38-802 0 - 85 - 1 5


“the nation that dominates the informationprocessing field will possess the keys to worldleadership in the twenty-first century, ”23 hascaused nations to look to the development ofinformation technology for their future well-being. Evidence of the movement of scienceand technology policies into the political arena,for instance, can be found in both the recentThatcher (United Kingdom) and Mitterand(France) political platforms in which informa-tion technology research and developmentfunding and programs were emphasized. How-ever, these two leaders differ in their policiesfor technological development. These differ-ences are important in understanding andevaluating the roles of these governments ininformation technology research and devel-opment.

The various current approaches to scienceand technology policy and the different objec-tives of each nation’s research activities under-lie the distinct goals of each country’s nationalresearch program. For example, Japan’s con-cerns lie in developing information technologyfor improving Japanese society (which entailsdeveloping an information-based infrastruc-ture) and improving its world trade positionin information technology products. Conse-quently, Japan’s goals for its national infor-mation technology research program, theFifth-Generation Computer Systems Project,is to develop a fifth-generation computer forsocial applications and to develop a technologi-cal knowledge base which will enable Japanto maintain and improve the volume of infor-mation technology exports that is so vital tothe Japanese economy.

Like Japan, the United Kingdom is also con-cerned with its economic survival in worldmarkets, as exports also play an importantrole in the U.K. economy. The basic goal,therefore, for the U.K. Programme for Ad-vanced Information Technology is to improvethe competitiveness of the U.K. informationtechnology industries in the world market inorder to reverse its negative balance of infor-mation technology trade.

‘gRobert E. Kahn, “A New Generation in Computing, ” 1~~~Spectrum, November 1983, p. 36.

French governmental interest and efforts toexpand advanced information technology re-search and production are directed at two ma-jor goals: strengthening France’s internationalcompetitiveness and the development of an in-formation technology-based infrastructure forthe preservation and continued developmentof French culture and society. ReflectingFrench national goals, France’s national infor-mation technology research and developmentprogram, La Filiere Electronique, has as itslong-term goals: to place France on a techno-logical level closer to that of the United Statesand Japan; create a trade surplus in informa-tion technology products; create new jobs;assure a sound technological base; and accel-erate the production of information technol-ogy products.

The Europeans, through the European Eco-nomic Community (EEC), are also concernedabout their basic economic survival in worldmarkets. Consequently, Europe’s major goalis to establish a strong technological basethrough collaborative research on variouslong-range projects that may not be adequate-ly funded within individual nations. The Euro-pean Strategic Program for Research in Infor-mation Technology (ESPRIT), involves the 10EEC member countries. ESPRIT’s major ob-jective is to keep Europe competitive with theUnited States and Japan in advanced infor-mation technology fields.

Government Role in InformationTechnology Research and Development

The level of government funding for re-search and development varies widely. Theratios of civilian research and development ex-penditures to gross national product (GNP)presented in figure 40 show that the UnitedStates devotes a lower proportion of its GNPto civilian R&D than Japan, but a higher pro-portion than the United Kingdom. An exami-nation of the annual growth rates of nationalresearch and development expenditures forelectrical and electronics industries revealsthat the United States lagged behind most of


Figure 40.— Estimated Ratio of Civilian R&DExpenditures to Gross National Product (GNP) for

Selected Countriesa

2.5

2.4

2.3

2.2

2.1

2.0

1.9

2 . 6

1.8

1.7

1.6

1.5

1.4

1.3

1.2

1.1

1.0

● ☛ ● ☛

.9

o~

1961 63 65 67 69 71 73 75 77 78 81 83

Years

Natlonal expenditures excluding Government funds for defense and sPace

R&D

SOURCE Scierrce Indicators, 7982, National Science Board, National ScienceFoundation, 1983

the industrially developed nations during the1970s.24

The United Kingdom, France, and Japansingled out information technology as an areafor government promotion and support. Thiscontrasts sharply with the U.S. Government,which at the present time has not singled outor targeted any specific industry or technol-ogy for government support. Although in mar-ket economies governments use a more or lessstandard set of policies to support researchand development,25 other nations differ from

ZtR,easOns ~or the relative declining growth rate Of us. GOV-ernment R&D funding during this period include major de-creases in funding for defense and space wsearch and develop-ment, while civilian research and development was heldconstant. Curnmtly, U.S. budget projections reflect increasingR&D funding for defense purposes.

‘Slnternational Competitiveness in Electronics (Washington,DC: U.S. Congress, Office of Technology Assessment, OTA-ISC-200, 1983), p. 381.

the United States in the extent to which eachcoordinates and targets these policies to sup-port information technology research and de-velopment. Policy measures for R&D supportinclude low interest loans, direct subsidies, oractual government contracts. Other incentivesto stimulate information technology R&D in-clude tax incentives, national developmentbanks which channel funds specifically to in-formation technology industries, public sectorprocurement, and merger/antitrust policies.

To varying degrees the governments of theUnited Kingdom, France, and Japan have es-tablished institutional mechanisms to facili-tate coordinated research and developmentpolicymaking. Coordinated government inter-vention on the part of these industrialized na-tions is reflected in each of the governments’structure which centralizes the responsibilitywithin one or a few government ministries, andin the establishment of planning councils.Moreover, the coordinated efforts of the Brit-ish, French, and Japanese extend much fur-ther than the U.S. pluralistic and decentralizedapproach to R&D support, to the coordinationof government, industry, and university infor-mation technology research and developmentobjectives and activities.

Government/Industry/UniversityInstitutional Arrangements for

Information Technology Researchand Development

A major difference between the institutionalarrangements for U.S. information technologyresearch and development and those of theUnited Kingdom, France, and Japan is there-lationship between government and industry.In the United States, government tends to beviewed as a regulator and enforcer of laws andsocial policies; U.S. industry generally per-ceives its relationship with government as ad-versarial. Moreover, the U.S. Government,which traditionally avoids involvement in theprivate sector, relies largely on private initia-tives for risk-taking, innovation, job creation,and the generation of profits and capital.

Although this adversarial relationship be-tween the public and the private sectors ex-


Table 39.—Average Annual Growth Rates in the Engineering Industry

United UnitedStates Japan Germany France Kingdom

1970-79 1970-79 1971-79 1970-79 1969-78

Research scientists and engineers: a

Aerospace . . . . . . . . . . . . . . . . . 0.3Electrical and electronics . . . . –0.4Machinery . . . . . . . . . . . . . . . . . 5.1Transport equipment . . . . . . . . 3.3

Total expenditure:Aerospace . . . . . . . . . . . . . . . . . – 1.1Electrical and electronics . . . . 0.1Machinery . . . . . . . . . . . . . . . . . 5.5Transport equipment . . . . . . . . 5.4

Government funds:Aerospace . . . . . . . . . . . . . . . . . – 1.7Electrical and electronics . . . . –2.2Machinery . . . . . . . . . . . . . . . . . 1.0Transport equipment . . . . . . . . 3.1

industry and foreign funds:Aerospace . . . . . . . . . . . . . . . . . 0.6Electrical and electronics . . . . 2.3Machinery . . . . . . . . . . . . . . . . . 6.3Transport equipment . . . . . . . . 5.9

—6.72.58.5

—5.85.2

11.0

————

————

–5.33.26.73.8

–2.35.89.83.9

–1.07.1

10.010.1

4.14.98.53.3

1.14.13.91.8

2.64.70.97.2

1.80.1—6.9

5.07.7—7.4

–2.60.93.60.2

–0.53.11.5

–0.3

–3.19.91.13.0

(b)

–0.7–3.4d

– 1.4a~aPan not in FU1l.Tirne Equivalents.

bLarge increases from a very IOW start.cFrom abroad 4.91 per annum.dElectrical and electronics, For R&D purposes the following subclasses are identified with this subgroup: ISIC 3832 and ISIC

638 nec.

SOURCE: OECD Science and Technology Indicators, OECD, 1964.

ists in most countries, the governments of theUnited Kingdom, France, and Japan have his-torically participated to a greater degree in in-dustrial and technological development. Forcultural, historical, and political reasons, theseindustrialized nations often regard govern-ment-industry relations as a partnership orperceive government as an institution forguiding and supporting targeted industries.Government-industry coordination has begunin some nations with the establishment of mul-tipartite advisory groups representing govern-ment and private industry.26 Formal and in-formal advisory councils have become in somenations, particularly in Japan, important for-ums for government-industry-academic con-sultations on industry policy and implemen-tation. Particularly in information technology,these councils orchestrate joint research, de-velopment, and marketing schemes among in-formation technology firms and government.

‘Franklin Delano Strier, “On Economic Plannin g, Japan andWest Germany Have a Better Idea,” The Center Magazine, Jan-uary/February 1984, p. 36.

To coordinate more fully both their targetedpolicies and information technology R&D, theUnited Kingdom, France, and Japan have re-cently adopted major national programs.These programs, which have no counterpartin the United States,27 have been establishedto pursue research projects cooperatively be-tween government, private industry, and uni-versities. The scale of funding for all the na-tional programs is large and roughlycomparable, representing a major commit-ment of between half a billion and several bil-lion dollars over the first 5 years.28 This fund-ing is magnified because the companies whichreceive government research funds usually arerequired to match the funds.

“Some individuals contend that the Department of Defense(DOD) programs, such as VHSIC, are similar to the nationalresearch and development programs of the United Kingdom,France, and Japan. However, there are major differences be-tween DOD programs and these other nations’ programs.

28Trudy E. Bell, “Tomorrow’s Computers-The Teams and ThePlayers,” IEEE Spectrum, November 1983, p. 46.


Each nation, in structuring its cooperativeresearch program, has looked to successfulorganizational examples both domesticallyand in other countries, borrowing organiza-tional concepts and applying them in innova-tive ways. For example, the Institute for NewGeneration Computer Technology (ICOT),with its central research center, is unique toJapan. ICOT is organized for close coopera-tion of research activities among government,industry, and university participants. Japan’scustomary approach has been to have eachparticipating research institution or companyconduct research individually. ICOT is alsocontracting with outside companies and lab-oratories for some of the research and devel-opment—a technique often used for U.S. de-fense contracts, but unusual in Japan.

The institutional arrangements for the U.K.Programme for Advanced Information Tech-nology are also unique, although the AlveyCommittee closely modeled the program orga-nization on Japan’s Ministry of InternationalTrade and Industry (MITI) and the U.S. De-fense Advanced Research Projects Agency(DARPA). This new organization is intendedin part to stimulate the transfer of basic re-search from academic environments to in-dustry.

The ESPRIT program shares both similari-ties and differences with organizations suchas ICOT and the Alvey Programme. Likethese other new institutional arrangements,ESPRIT research is undertaken by teams ofuniversity, government, and industry scien-tists. In contrast to other nations’ cooperativeprograms, ESPRIT represents a joint ventureat the international level in which each proj-ect must involve researchers from at least twocountries.

The success of these national programs inachieving both national aims and researchgoals remains undetermined. However, thesenew institutional arrangements raise some in-teresting questions for the United States. Howwill the cooperative research programs of theUnited Kingdom, France, and Japan altertheir traditional research structures and serve

as models for future research projects? Whatare the relative strengths of these new re-search programs versus traditional U.S. re-search environments? To what degree willthese new national research programs affectU.S. technological development and marketshare?29

Industry Participation in InformationTechnology Research and Development

Funding of industrial information technol-ogy research and development varies from na-tion to nation. Table 40 represents industrialR&D funding patterns in industrialized na-tions and shows that industry has been a moredominant contributor in Japan (although theJapanese Government also provides a greatdeal of support through indirect subsidies)than in any other nation, including the UnitedStates.303’ The U.S. Government, in contrast,supported approximately half of the U.S. re-search and development activities in 1970.However, by 1979, U.S. industry had increasedits share of industrial funding to 67 percent,approximately as much as the French indus-try’s 71 percent support of its own research.

Table 41 illustrates industrial research anddevelopment support for electrical and elec-tronics and computers categories in the indus-trialized nations. Japan and West Germanyhad the highest proportion of industrial R&Din the electrical and electronics category. Inthe computer category, however, the UnitedStates had the greatest proportion of indus-trial R&D followed by the United Kingdom,France, and then Japan.

‘gIbid.300ECD Science and Technology Inal”cators, OECD, 1984, p.

119.31 Sa”ence Indicators, 1982, National Science Board, National

Science Foundation. 1983. K). 9.,*


Table 40.—R&D Performed in the Business Enterprise Sector by Source of Funds (1970, 1975, and 1979)

National currency(in millions) Percent

Country and source 1970 1975 1979 1970 1975 1979— — —France . . . . . . . . . . . . . . . . . . . . . . . . . .

Total domestic . . . . . . . . . . . . . . . . .Business enterprise . . . . . . . . . .Government . . . . . . . . . . . . . . . . .Private nonprofit . . . . . . . . . . . . .Higher education . . . . . . . . . . . . .

From abroad. . . . . . . . . . . . . . . . . . .Japan . . . . . . . . . . . . . . . . . . . . . . . . . . .

Total domestic . . . . . . . . . . . . . . . . .Business enterprise . . . . . . . . . .Government . . . . . . . . . . . . . . . . .Private nonprofit . . . . . . . . . . . . .Higher education. . . . . . . . . . . . .

From abroad... . . . . . . . . . . . . . . . .United Kingdoma . . . . . . . . . . . . . . . . .

Total domestic . . . . . . . . . . . . . . . . .Business enterprise . . . . . . . . . .Government . . . . . . . . . . . . . . . . .Private nonprofit . . . . . . . . . . . . .Higher education.. . . . . . . . . . . .

From abroad... . . . . . . . . . . . . . . . .United Statesb . . . . . . . . . . . . . . . . . . .

Total domestic . . . . . . . . . . . . . . . . .Business enterprise . . . . . . . . . .Government . . . . . . . . . . . . . . . . .Private nonprofit . . . . . . . . . . . . .Higher education. . . . . . . . . . . . .

From abroad... . . . . . . . . . . . . . . . .West Germanyc . . . . . . . . . . . . . . . . . .

Total domestic . . . . . . . . . . . . . . . . .Business enterprise . . . . . . . . . .Government . . . . . . . . . . . . . . . . .Private nonprofit . . . . . . . . . . . . .Higher education . . . . . . . . . . . . .

From abroad . . . . . . . . . . . . . . . . . . .

8,322.48,007.35,310.02,689.0

4.63.7

315.1895,020.0894,193.0876,608.0

17,585.0NANA827.0680.3647.7431.2216.5NANA

32.618,067.018,067.010,288.07,779.0

———

7,114.07,090.06,146.0

939.05.0

—24.0

15,616.514,393.59,965.84,376.8

47.23.7

1,223.01,684,847.01,683,200.01,654,502.0

28,698.0NANA

1,647.01,340.21,255.5

841.4414.1NANA

84.724,187.024,187.015,582.08,605.0

———

14,469.014,005.011,397.02,596.0

12.0—464.0

26,260.024,460.018,723.05,674.0

58.05.0

1,800.02,664,913.02,662,698.02,624,843.0

36,807.0935.0113.0

2,215.02,324.32,138.81,459.0

679.7NANA185.5

38,226.038,226.025,708.012,518.0

———

20,720.020,070.015,650.04,400.0

20.0—650.0

100.096.263.832.3

.1—3.8

100.099.997.9

2.0NANA

.1100.095.263.431.8NANA4.8

100.0100.056.943.1———

100.0——————

100.092.263.828.0

.3—7.8

100.099.998.2

1.7NANA

.1100.093.762.830.9NANA

6.3100.0100.064.435.6———

100.096.878.817.9

.1—3.2

100.093.171.321.6

.2—6.9

100.099.998.5

1.4——

.1100.0

92.062.829.2NANA

8.0100.0100.067.332.7———

100.096.975.521.2

.1—3.1

alg70figure9 forthelJnited Kingdom are from 1989, and 1979 figures are from 1978.bcurrent expenditures plus depreciation onlY.c1970figures for West Germany are from 1969.NA—Not separately available.NOTE: Details may not add to totals becauseof rounding.

SOURCES: OECD Scierrce artd Technology Indicators, VOLBOECD, 1982; Research arrd Development in/ndust~, National Science Foundation, 1981; in Science/n.dicators, 19S2, National Science Board, National Science Foundation, 1983.

Table 41 .—Percent of lndustrial R&D in Selected Industries (1989-79)

United States Japan West Germany United Kingdom France

Industry 1970 1979 1970 1979 1969 1979 1969 1978 1970 1979

Six-industry total . . . . 71.2 70.2 61.5 55.6 71.1 72.4 58.4 60.7 NA 61.7Aerospace . . . . . . . . . . . . 11.8 8.6 – – .1 .6 1.1 5.4 7.8 8.8Electrical and

electronics . . . . . . . . . 19.5 17.6 24.9 23.4 29.3 26.9 20.4 16.0 16.5 20.2Instruments . . . . . . . . . . . 5.3 7.8 2.3 2.9 1.6 2.1 3.1 1.9 NA 1.2Machinery . . . . . . . . . . . . 14.3 6.0 8.7 7.0 }7.0 }16.6 6.6 NAComputers . . . . . . . . . . . “ 11.5 2.9 2.8 !; 5.2 NA : : ;Chemicals group a . . . . . 20.3 18.7 22.7 19.5 33.1 26.2 22.1 25,6 24.2 23.4alncludes chemicals and allied products and petroleum refining industries.SOURCES: OECD Science and Technology /nd/caters, vol. B OECD, 1982; Research and Development in Industry, National Science Foundation, 1981; in Science hr-

dicators, 1982, National Science Board, National Science Foundation, 1983.


In addition to direct funding, governmentscan support or discourage industrial researchand development with tax incentives, fiscaland monetary policies which affect interestrates and the availability of capital, regulatorypolicies, procurement practices, patent pol-icies, and antitrust policies.32 Moreover, ac-——— —.. —.

32Science Indicators, 1982, op. cit., p. 10.

tions of a national government to define thestructure of its information technology indus-try will have profound but unpredictable influ-ence on industrial R&D. For example, Francehas recently nationalized some major informa-tion technology firms, whereas the UnitedKingdom and Japan currently are moving inthe opposite direction by privatizing their tele-communications entities.

Conclusions

Cultural, social, and institutional differencesamong nations profoundly influence the wayin which technological innovation occurs andunderlie the considerable differences in R&Dpolicies in the United Kingdom, France, andJapan. Therefore, many of these nations’ suc-cessful research and development policies andendeavors may not be easily transferable orapplicable to U.S. R&D environments. Never-theless, cross-cultural comparisons of these in-dividual nations, which have developed differ-ent methods of addressing similar publicpolicy issues and technologies, can be usefulto an individual society, such as the UnitedStates, in devising a conceptual framework fordeveloping information technology domesticand international R&D policies.

International comparisons also provide auseful method for evaluating the status of U.S.information technology research and develop-ment activities and expenditures. However,making comparisons is difficult. Differencesexist among countries in definitions, concepts,data collection methodologies, and statisticalreporting procedures.33 These problems areparticularly prevalent in the area of informa-tion technology .34 Although several interna-

tional organizations such as OECD and theITU have initiated the development of uni-form definitions and standards for informationtechnology products and facilitated the ex-change of information on nations’ R&D poli-cies and activities, the U.S. Government cur-rently does not have a designated agency oroffice within an agency to analyze and moni-tor foreign information technology R&D pol-icies and practices.

The economic importance of industrial com-petitiveness and its close reliance on techno-logical development emphasizes the impor-tance of developing a program for periodicmapping of the pattern of technological advan-tages and disadvantages relative to foreigncompetitors.35 Such surveys could help to alertgovernment and industrial representatives tochanges in the technological underpinnings oftheir competitive positions. These surveyscould also be broadened to encompass non-technological factors affecting competitive-ness or future technological developments.These could include current or prospectivechanges in foreign government R&D policiesthat influence the competitiveness of their pro-

— — — ———-—————— gories. Moreover, each nation has its own methods for categoriz-

tsscjence ~n~.catom, 1980, National Science Board, Nation~ ing information technology products and research and devel-Science Foundation, 1981, p.4. opment expenditures and activities.

34The U.S. SIC Codes, for example, do not have categories 3%ee S. E. Goodman and M. R. Kelly, “We Are Not Alone:for information technology, but rather information technology A Sample of International Policy Challenges and Issues, ” Theproducts are encompassed within different and separate cate- Information Society, vol. 2, No. 3/4, p. 250-268.


ducers in export markets. Such a broader sur-vey could provide the basis for a more effec-tive assessment of changes in the current andprospective competitiveness of foreign R&Defforts than a narrow focus on technologicalcapabilities alone. 3 6 A s o n e a n a l y s t s t a t e d :

Most governments, and most especiallythe U.S. Government, are organized to reflectprimarily domestic interests and have greatdifficulty in dealing adequately with one ofthe consequences of science and technology—the gradual blurring of the distinction be-tween domestic and international affairs . . . .

—-“Bela Gold, “Technological and other Determinants of the

International Competitiveness of U.S. Industries, ” JEh’ETransactions on En&”neering Management, vol. LM 30, No. 2,May 1983, p. 58.

Perhaps the most important observation[is] . . . that the general character of changesbrought about by science and technologytends, overall, to lead to increased interna-tional interaction and integration, with cor-respondingly reduced relevance of nationalborders. The parallel spread and diffusion oftechnological competence that is eroding thedominance of one or a few nations in scienceand technology makes it imperative that werecognize the nature of the underlying changestaking place as we attempt to develop pol-icies to deal with the specific implications ofany given technology, or to influence the di-rection of development of technology itself.37

“Report by Eugene Skolnikoff entitled “Impact of Scienceand Technology on the International System, ” in “Overviewof Intimational Science and Technology Policy” hearings beforeCommittee on Foreign Affairs, House of Representatives, 98thCong., 1st sess., Aug. 2,3 and Sept. 21, 1983, p. 317.

Japan

Given Japan’s minimal domestic natural re-source base and its high dependence on othernations for food, energy, and raw materials,the Japanese Government treats science andtechnology policy as a means of spurring over-all economic growth and enhancing Japan’scompetitive position internationally. Japan’spolicies focus on maintaining a long-term, highvolume of exports in order to gain technologi-cal and, thus, market leadership in a broadspectrum of high technology, high value-addedproducts. This perception in Japan has led toa consensus in the nation’s government, busi-ness, financial, and academic communities tocontinue strengthening the nation’s technolog-ical base.

The coordination of science and technologypolicy to the promotion of economic develop-ment is rooted in Japan’s postwar recovery ef-forts. During the postwar recovery period, var-ious science and technology institutions andlaboratories were established on the assump-tion that they would help stimulate an eco-nomic recovery. Another major component ofthis strategy was to import and improve tech-nology from the United States and Western

Europe. Following the enactment of the lawon Foreign Capital in 1950, the Japanese havesigned more than 36,000 licensing agreementscosting approximately $12 billion.38 Agree-ments between Japanese and foreign firmswere made under strict government supervi-sion partly to control the outflow of foreignexchange and partly to concentrate technologi-cal resources into certain key industries. Prod-ucts manufactured with these imported tech-nologies initially served to develop theJapanese domestic market, bringing about aGNP growth exceeding 10 percent through-out the 1950s. For example, transistor tech-nology imported and commercialized by theJapanese in the early 1950s provided a foun-dation for the modern electronics industry.

Although the general trend of importingtechnologies has been receding, the imports oftechnologies related to electronic computersincreased by 16 percent over fiscal year 1980,with those relating to software accounting forover 166 imports. Broken down into sectors

——. -———3%mard Lynn, “Japanese Technology: Successes and Strat-

egies,” Current History, November 1983, p. 366.


of industry, out of a total of 2,142 cases, thenumber of imports of foreign technologies dur-ing fiscal year 1980 in the electrical industryamounted to 413 cases.39

In concert with the policy of importing tech-nology, Japan has sought to develop its ownindigenous research base. The ratio of nationalR&D expenditure to national income has risenfrom less than 1 percent during the first halfof the 1950s to 1 percent in 1957, 1.5 percentin 1960, 2 percent in 1971, and 2.36 percentin 1981. According to official plans, this per-centage will be increased to 2.5 percent by1985 and to 3 percent by 1990.40 According tostatistics released by the Japanese Prime Min-ister’s Office, all Japanese R&D, publicly andprivately funded, in the area of informationprocessing (which includes software and com-puter systems development only) totaledY 158.6 billion ($687 million) in 1979 andY 164.6 billion ($713 million) in 1980.4142

In Japan, a far lower percentage of total re-search funds is provided by government (Ja-pan 27.7 percent, United States 51.1 percent,United Kingdom 51.7 percent) than in othernations.43 In part, this difference can be attri-buted to the high level of expenditure on de-fense research by Western governments (ap-proximately 15 percent of total research funds)relative to the small amount spent by the Jap-anese Government (0.7 percent). In some areasof information technology R&D such as inte-grated circuit development, low military R&Dexpenditure has helped Japanese industry. Ingeneral, military areas demand the higheststate-of-the-art standards, regardless of costs.

—‘g’’ Import of Foreign Technologies in Japan, ” Science and

Technology in Japan, April 1982, p. 27,‘“’’ Summary of fiscal year 1981 White Paper on Science and

Technology, ” Science and Technology Agency, Tokyo, ForeignPress Center, 1981.

“Barry Hilton, “Governme nt Subsidized Computer, Software,and Integrated Circuit Research and Development by JapanesePrivate Companies, ” Scientific Bulletin, Office of Naval Re-search Far-East, U.S. Department of the Navy, vol. 7, No. 4,October-December 1982.

42All Japanese yen figures are converted into U.S. dollars ac-cording to foreign exchange rates as of June 1, 1984, where+231 = $1.

“’’Science in Japan, “ Nature, vol. 305, Sept. 29, 1983, p. 361.

Therefore, these military developments some-times result in expensive products which areso specialized that civilian or consumer ap-plications can be limited. This is often the casewith integrated circuit development in theUnited States. On the other hand, Japan hassucceeded in developing integrated circuitproducts solely for commercial application.

Taking into account all funds spent on de-fense, the government of Japan still contrib-utes significantly less to total scientific re-search expenditure than other countries.44

More specifically in the area of informationprocessing, the Japanese Government R&Dexpenditure in 1979 accounted for 8.2 percentin 1979 and 6.2 percent in 1980 of the totalJapanese information technology R&D ex-penditures.45 In Japan, this government R&Dfunding is concentrated in the national univer-sities (13.5 percent), national research insti-tutes (13 percent), with as little as 1.5 percentof government funding channeled to privateindustrial laboratories. Because governmentR&D funding, which is the major supporterof academic basic research, is relatively lim-ited, reasons for Japan’s perceived ineffective-ness in basic research can be clearly under-stood. As a result, current improvements inthe Japanese academic environments for basicresearch as well as increases in funding levelsfor overall basic research are high prioritieson the Japanese policy agenda.

Because Japanese Government R&D fund-ing is small and for the most part channeledinto university and national research insti-tutes, Japanese industry funds constitute ap-proximately 70 percent of R&D activities. Ap-proximately 28 percent of all Japanese indus-trial R&D funding is devoted to informationtechnology R&D. Although the Japanese Gov-ernment does not directly make use of thevitality offered by private enterprise, the lack

44These low expenditures are also the result of the JapaneseGovernments current large budget deficits.

46Barry Hilton, “Government Subsidized Computer Software,and Integrated Circuit Research and Development by Japanese’Private Companies, ” Scientific Bufletin, Office of Naval Re-search Far-East, vol. 7, No. 4, October-December 1982.


of government funding has intensified compe-tition in the area of information technology re-search and development. This competitive ef-fort is exemplified in Japan’s computer andsemiconductor industries which, in order tosurvive, must develop and efficiently producehigh quality products as quickly as possible.

This “privatized” environment for R&D hasgiven Japan advantages and disadvantages inits information technology R&D efforts. Oneresult of the large percentage of privatelyfunded R&D is the lack of basic fundamentalresearch activities. Because the R&D is sub-sidized mainly by private firms, basic creativeresearch, which is high-risk and long-term, issometimes ignored in favor of cost-efficient,developmental, applied, commercialized R&D.This continued preoccupation with R&D ef-forts that bring quick economic results hasresulted in a trend which places less impor-tance on basic, innovative studies. The Japa-nese Science and Technology Agency (STA),for example, published a list of 15 basic dis-coveries in the fields of recombinant DNA andcomputer technology (superconductivity, op-tical fibers, lasers, Josephson junctions, tun-nel diodes, and transistors). Japan was respon-sible for only two of the breakthroughs listed;America for nine; the United Kingdom and theNetherlands for four. This bias in Japan’soverall research expenditures toward appliedresearch and prototype development is re-flected both in government-supported R&Dand private sector research expenditure (seetable 42).

On the other hand, in the area of develop-ment, the application of basic research results,

Table 42.— R&D Expenditure by Type of Activity

Basic Applied Development1970 . . . . . . . . . 18.9 28.2 52.91974 . . . . . . . . . 15.0 21.7 63.31975 . . . . . . . . . 14.2 21.5 64.31977 . . . . . . . . . 16.2 25.1 58.71978 . . . . . . . . . 16.6 25.1 58.41979 . . . . . . . . . 15.6 25.9 58.5SOURCE: Kagaku Gijulsu Hydron (Indicators of Science and Technology), Kagaku

Gijutsu-Cho (Science and Technology Agency) 1981. Note: This tablecovers all R&D, public and private.

the Japanese privatized R&D efforts are high-ly successful. Japan has clearly outstrippedmost Western nations in processing technol-ogies and incremental engineering-rapidlyrefining existing designs and ideas by makingthem smaller, lighter, faster, and cheaper. Jap-anese engineers, for instance, reengineered the16 K RAMs with finer features to produce a64 K RAM within a 2-year period.

Examples of Japanese strengths and weak-nesses in information technology R&D are welldocumented in the area of software develop-ment. Software development, currently be-lieved to be crucial for future information tech-nology development, can be classified intovarious categories. Japan, with its industrialemphasis on applied R&D, has concentratedits efforts in production process-control soft-ware which has wide commercial industrial ap-plications and which will reap significant eco-nomic benefits, both domestically in terms ofthe productivity and capacity utilization of in-dustry, as well as internationally, in terms ofthe benefits of trade and technology transfer.However, in other categories of software de-velopment such as computer-aided design(CAD), the United States is technologicallymore advanced than Japan. Most of this tech-nological lead resulted from billions of dollarsthat have been allocated for aerospace and de-fense basic research. As a result of this basicresearch for U.S. defense purposes, the U.S.computer simulation models and 3-D designprograms are among the most sophisticatedin the world.

Currently, Japanese industry is beginningto experience some difficulties with its empha-sis on appliedborrowed technology. Japan hasbeen slowly catching up to western technologi-cal innovations and has less input from foreignbasic research patent licenses on which to baseits refinements. Furthermore, Western firmsare expressing a disinclination to sell patentsto Japan, as they see the reengineered Japa-nese products competing with their own prod-ucts. Moreover, Japanese firms that havesometimes neglected basic research, have few-er technological innovations worth offering


Western companies when they wish to inquireabout the possibility of cross-licensing. In ad-dition to the fear of the decreasing amount ofinnovative ideas which Japan can buy and per-fect and the fear of being excluded from futureU.S. and other Western nations’ technical de-velopments, national pride is also forcing Ja-pan to put more effort into basic R&D.

As a result of some of these difficulties, theJapanese Government is beginning to placemore emphasis on basic research activities.This movement towards increased basic re-search is reflected in both government and in-dustrial R&D activities as well as in the cur-rent Japanese Government’s institutionalmechanisms for influencing and funding indus-trial research. Because the government mech-anisms which influence and fund informationtechnology R&D are mostly aimed at promot-ing R&D in the private sector (and many takethe form of informal cooperation), it is diffi-cult at times to disassociate government andprivate sector initiatives and roles in informa-tion technology R&D activities. However, forpurposes of clarity and comparison, the roleof government, universities, and industry envi-ronments for the conduct of information tech-nology R&D will be separately described. Be-fore discussing these environments in detail,it is also important to understand the size ofJapanese participation in information technol-ogy markets.

The Size of Japanese Participationin Information Technology Markets

Utilizing its basic technology policy whichhistorically dictated that Japan reengineer im-ported technological innovations, Japanese in-dustry has developed a very strong positionin world information technology markets. Inmany areas where Japan has managed to cap-ture a substantial percentage of the world in-formation technology market, it can largelybe attributed to Japanese industry’s strongcapabilities in product development, market”ing strategies, and quality control.

Beginning in the 1950s, Japanese informa-tion technology industry efforts focused or

microelectronics. Over the last three decades,there have been major shifts in Japanese con-sumer electronics production. Figure 41 illus-trates the shift from the production of radios,to television sets, to audio equipment, and fi-nally to videotape recorders. This productionprogression is particularly interesting in termsof technology because it not only illustratesthe steady restructuring of an industry tohigher and more complex technologies, butalso illustrates the changing position of Jap-anese information technology industry interms of global competition.46

In each shift in Japanese consumer electron-ics production, industry has been dependenton foreign technological innovations. For in-stance, Bell Laboratories supplied transistortechnology, RCA licenses made Japanese colortelevision production possible, and CorningGlass supplied glass tube technology. Perhapsmore than any other of Japan’s industries, thedevelopment of Japan’s consumer electronicsindustry is the result of imported technologythat competitive Japanese firms adapted, im-proved, and drove costs down. Figure 42 illus-trates the Japanese share of world consumerelectronics productions. The total value of pro-duction, second only to that of the UnitedStates, reached Y 8,683 billion ($37.6 billion)—or approximately 150 times that of 1955—making electronics one of Japan’s major in-dustrial sectors.

‘James C. Abegglan, and Akio Etori, “Japanese TechnologyToday, ” Scientific American supplement, 1983, p. J, 18.

Figure 41.—Trends in the Production CompositionRatio of Major Consumer Electronics Equipment

-1951 1955 1960 1965 1970 1975 1980 ’82Years

SOURCES: Annual Data on Japan’s Electronics Industry, 1983 Edition, ElectronicsIndustry of Japan, Tokyo, 1983; in James C. Abeggian and Akio Etori,“Japanese Technology Today, ” Scientific American, supplement,1983.


Figure 42.—JaDanese Share of World Productionof Consumer Electronics Products (1980)

SOURCES: Japan Electronics Industry Development Association; in Gene AdrianGregory and Akio Etori, “Japanese Technology Today: The ElectronicRevolution Continues,” Scientific American, supplement, 19S4.

Recognizing the primacy of computers andtelecommunications as growth sectors, as wellas conforming to the trend towards more ad-vanced technology production, Japanese in-dustry has focused on a technology key tothese areas-integrated circuits. As in the caseof consumer electronics, Japan imported basicsemiconductor technology and in the early1970s initiated the production of integratedcircuits. Although exports were insignificantduring the beginning production years, Japa-nese industry focused on lowering costs andimproving quality. Integrated circuitry pro-duction has grown in value terms at approxi-mately 25 percent per year. Figure 43 illus-trates the growth in the Japanese informationprocessing, computer, and integrated circuitindustries between 1974-81. By 1976, Japanaccounted for a 40 percent share of the worldmarket for 16 K RAMs, and in 1978, FujitsuLtd. was the first to announce the commercialproduction of the 64 K RAM. Japanese com-panies as well as U.S. manufacturers are the

first to produce 256 K RAMs which will beavailable in 1984-85. Figure 44 illustrates Jap-anese integrated circuit production relative toU.S. production.

Perhaps the most significant step in thistechnology development sequence is the recentintroduction by Japanese firms of one of thefastest supercomputers worldwide. These newcomputers, manufactured by Fujitsu andHitachi, represent a major step in Japan’sgovernment-sponsored national effort to buildfifth-generation computers.”

Government

The large percentage of private R&D fund-ing would indicate the tremendous importanceof Japanese industry in the Japanese successin world information technology markets.However, it is a mix of government support,a favorable and stable political structure, aswell as freedom from national security expend-itures, that have combined with the ratherunique Japanese sociology to create a periodof economic growth in the area of informationtechnology. The nickname for the Japaneseeconomy “Japan Inc., ” which was given someyears ago, may be said to be a realistic evalua-tion of the Japanese Government and privatecorporations during postwar Japan, when Ja-pan sought to catchup with the industriallyadvanced nations. The term “Japan Inc. ” isstill used today but in most cases this wordappears to reflect a misunderstanding of therelationship between the Japanese Govern-ment and industry.

The Japanese Government does not controlindustrial R&D through funding mechanismsor specific policies that must be adhered to,but rather there is a participatory partnershipamong different segments of government andindustry, based on pragmatic decisions, mu-tual respect, working within a framework ofcommon goals. The councils and industrialassociations have long been proposing to theJapanese Government to increase its research

47 Phillip J. Hilts, “Japanese Firms Build Two Fastest Com-puters, ” The WAirJgton Post, Feb. 7, 1984, p. A,l.


2,00(

E

1 ,00(

Figure 43.—Japanese Information Technology Industry Sales

Information processing industry

I Computer industry 2,104.0

IC industry

1,651.8

959.8

245.3

58Q.O

1974

933.9P

275.1

7

541.2

1975

1,123.0

307.0

! ,

618.9

1976

1,34U.4

412.6

1977

460.2

19791978

2,535.6

2,166.9669.8

596.6

,295.6

1980 1981

Years

NOTE: Information Processing Industry sales figures for 1981 not available.

SOURCE: Survey on Specified Services, Production Statistics in Computer 1982 Japan Infer. Processing Center.

and development support, but governmentfunding has only increased 1.4 percent annual-ly over the past 10 years. Although the gov-ernment does promote some industrial R&D,industry has always been a larger investor.Government-industry relations in Japan havebeen broadly discussed, and it is often mis-understood that such relations are largely dueto the Japanese Government subsidy of indus-trial R&D.

Japan’s information technology firms arefiercely competitive and the government’s roleis seen as a means for providing an orderlyframework for coordinating private industrialdevelopment. But when any coordination or

intervention is decided on, it is undertakenthrough the development of a consensusamong private enterprise and government.This consensual decisionmaking process be-tween industry and government, frequentlyaccomplished informally as well as throughformal institutional structures, is in manyways the most important factor affecting deci-sions on R&D projects and funding for infor-mation technology.

The major function of the Japanese Govern-ment is to select, or to guide the selection oftechnologies to be targeted, to reduce the eco-nomic risks normally associated with develop-ing new technologies, and to assist companies


Figure 44.—integrated Circuit (IC) Relative Production Share

SOURCES: Electronics, newspaper articles; in James C. and Akio “Japanese Technology Today, ” supplement, 1982

to achieve large scale production. The directfinancial support for R&D provided by theJapanese Government to targeted industriesis in many instances less important than thefact that the industry has been singled out bythe government as a “target” sector. There aretangible and intangible benefits which flow tosuch industries. A targeted sector gains pres-tige and public respect. Private banks aremore willing to extend credit, customers andsuppliers will tend to give preferred treatment,and government officials in various agenciesare also likely to be more responsive to the par-ticular needs of target-sector companies.

In addition to the informal consensual deci-sionmaking, targeting and funding of informa-tion technology R&D is accomplished throughmajor formal government institutions: Min-istry of International Trade and Industry(MITI), Science and Technology Agency(STA), Ministry of Education, Culture, andScience (MOE), Ministry of Posts and Tele-communications (MPT) which has nominalcontrol over Nippon Telegraph and Telephone

(NTT), and the Ministry of Finance (FOC)through the Japan Development Bank (JDB).The major Japanese Government organiza-tions directly involved with information R&Dare illustrated in figure 45.

Ministry of International Tradeand Industry (MITI)

Perhaps the most misunderstood agencywithin the Japanese Government, the Minis-try of International Trade and Industry(MITI) has been credited by many in theUnited States as being much more pervasivethan it is in reality .48 Taking into account thescale of the Japanese economy, the complex-ity of international markets and the rapidchanges in technology, MITI alone cannot anddoes not completely control industry. A casein point is MITI’s failed attempt during the1970s to consolidate the Japanese automobile— —

Tsuruta, “The Myth of Japan Inc.,” July 1983, p. 43-48, and Robert C. Christopher,

“Don’t Overestimate Tokyo Industrial Aid,” The New YorkTimes, Jan. 30, 1984, p. A, 21.

Ch. 7–Foreign Information Technology Research and Development • 229

Figure 45.—Japanese Government Organization for Information TechnologyResearch and Development

SOURCE: “Science in Japan, ” Nature, Sept. 29, 19s3.

industries into three large groups. Some haveclaimed that MITI is on paper at least no moreinfluential than the Department of Commercein the United States.49

MITI was established in 1949 with thebroad charter of shaping the structure of Jap-anese industry, managing foreign trade andcommercial relations, ensuring adequate rawmaterials and energy supplies, and managingrelationships between particular business andtechnical industrial sectors and the govern-

‘g’’Science in Japan, ” Nature, vol. 305, Sept. 29, 1983.

ment. Despite this broad legal mandate, thepervasive MITI practice of “administrativeguidance” by which many policies are imple-mented depends on no statutory authority.Nevertheless, MITI does have the advantageof broad contacts across information technol-ogy industries and relies extensively on thisinformal practice to influence firms and wholeindustries in the direction it wants them totake.5o

‘OIra C. Magaziner and Thomas M. Hout, Japanese hfus-try PoL”cy (Berkeley, CA: Institute of International Studies,1980), pp. 40-41.


Within MITI’s bureaucratic structure, theIndustrial Policy Bureau has played the ma-jor role in guiding overall industrial develop-ment. This Bureau consults with representa-tives from all industrial sectors and throughthese informal and formal meetings sets Jap-anese industrial policy. The major MITI bu-reau involved with information technology isthe Machinery and Information Industries Bu-reau. In general, this Bureau oversees andcoordinates export, import, production, distri-bution and consumption of machinery and me-chanical apparatus. In addition to informationtechnology, aircraft, automobile, machinetools, as well as other industries, are withinthe Bureau’s responsibilities. Within the Bu-reau there are several divisions which dealdirectly with information technology.

The most significant division involved withinformation technology is the Electronics Pol-icy Division. Its responsibilities include: 1)planning comprehensive policies for electron-ics equipment industries; 2) the distributionof computers; 3) planning programs on the uti-lization of computers; 4) conducting surveyson the utilization of computers; 5) represent-ing the Japanese Government at internationalorganizations concerning information technol-ogy matters; and 6) overseeing the Data-Proc-essing Promotion Council.

The Industrial Electronics Division is re-sponsible for exports, imports, production, dis-tribution, promotion of consumption, and im-provement and adjustment of communica-tions products. These products include com-puters, laser application devices, radar, elec-tronic measuring instruments, telephone andtelegraph equipment, switchboards, facsimileequipment, broadcasting equipment, fixedmultiplex communication devices, and com-munication wire and cables.

Two other divisions, Data-Processing Pro-motion and Electrical Machinery and Consum-er Electronics, also are directly involved withinformation technology. As its name suggests,the Data-Processing Promotion Division re-sponsibilities include: 1) the examination andlicensing of data processing technicians; 2) the

cultivation and promotion of data processingservice industries; and 3) the promotion ofcomputer usage and applications programs de-velopment.

In addition to these major Bureaus and theirrespective divisions, MITI policies are influ-enced by several advisory councils, industryassociations, and research associations. Per-haps the most unique aspect of MITI policy-making, these advisory groups are where in-dustry and government officials develop aconsensus on goals and policies for technologi-cal development. In these councils and asso-ciations members from government, academia,and industry discuss technology trends, mar-ket potential, and policy. Problems, ideas, andproposals are discussed, and if a general con-sensus is obtained, it is reflected in govern-ment and industrial R&D policies and prac-tices. In addition to these formal channels,there are also a number of informal exchangesbetween government and industrial represent-atives.

Within MITI, the Agency for Industrial Sci-ence and Technology (AIST) is explicitlyoriented toward research and development oftechnology with industrial applications. In ad-dition to its responsibilities of planning andadministering policies and programs for re-search and development, AIST operates 16government laboratories, including the Elec-trotechnical Laboratory (ETL), the majorMITI laboratory for information technologyR&D. In conjunction with ETL, AIST alsooversees collaborative research with affiliatedlaboratories and private companies-particu-larly for MITI’s targeted national informationtechnology R&D programs. The AIST also ad-ministers the industrial standards programs.

Directly under MITI’s jurisdiction, the Elec-trotechnical Laboratory is the largest nationalresearch organization in Japan specializing inelectronics research. The ETL, with an annualbudget of $40 million, employs approximately730 researchers. ETL’s major areas of researchinclude solid state physics and materials, in-formation processing, energy, standards, andmeasurements. Similar to DOD facilities in the


United States, ETL has in addition to Its owninternal research, responsibility for advisingthe government on technology options andmonitoring industrial R&D programs. ETL,like other technology in-house research labora-tories, does not attempt to compete with in-dustry. As in many U.S. Government researchlabs, ETL concentrates on identifying researchprojects and directions (usually high risk)where it could supplement industrial researchactivities. ETL also oversees the industrial re-search efforts for MITI coordinated nationalR&D projects.51

As the Japanese Government began to moveinto more basic research activities and newstate-of-the-art technologies, many of its pol-icymakers felt that its institutions were ill--equipped (e.g., too constrained, rigid) for basic,pioneering research. Consequently, TsukubaScience City, which was begun in 1966, wasplanned and built by the government for thepurpose of centralizing research and educa-tional activities. In terms of its concentrationof high level personnel, it in many ways resem-bles Silicon Valley in the United States, al-though in terms of government organization,the North Carolina Research Triangle is per-haps a better comparison. Located withinTsukuba City are 30 of Japan’s 98 national re-search institutes. These 30 research institutesaccount for approximately 40 percent of thetotal research budget and 40 percent of thetotal number of researchers in Japan. In addi-tion to these 30 national research institutes,the Tsukuba Science City accommodates atotal of 46 research organizations, includingtwo national universities, six organizationsbelonging to government-funded special orga-nizations, and eight organizations affiliatedwith other administrative entities. Approx-imately 27 research-oriented private corpora-tions have also relocated to Tsukuba ScienceCity.”— —. — —

“George E. Lindamood, “The Rise of the Japanese ComputerIndustry, ” Scientific Bulletin, Department of the Navy, Officeof Naval Research Far-East, vol. 7, No. 4, October-December1982, p. 61, 62.

‘*For a detailed discussion of Tsukuba Science City see JustinL. Bloom and Shinsuke Asano, “Tskuba Science City: JapanTries Planned Innovation, ” Science, vol. 212, June 12, 1981,pp. 1239-47 and “Science City in Japan-Tskuba, ” Science andTechnology in Japan, January-March 1983, pp. 6-11.

Phofo cred(f Embassy of Japan

Tsukuba Science City

National Research and Development ProjectsAnother effort to stimulate basic long-term

research activities by coordinating industryand government research efforts was begunin 1966 with the initiation of National Re-search and Development Projects. Perhaps themost significant aspect that sets JapaneseR&D efforts apart from U.S. R&D, these na-tional projects are directed towards researchthat is in the Japanese national interest, long-term, high-risk, and precompetitive-researchthat is not directed towards any specific prod-uct, but technology that is useful for an en-tire industrial sector.

The initiation of a national R&D project isaccomplished through a series of steps. First,through meetings with government, academic,and industrial representatives, usually withinvarious councils and associations, MITI offi-cials derive a consensus on areas for nationalR&D attention. MITI’s “Vision of MITI Pol-icies in the 1980s ” is an example of the con-sensuaI decisions reached among the repre-sentatives. Reflected in these “Visions of the1980s” is Japan’s basic national economic phi-losophy which states that Japan should seekto ensure its economic survival by becominga technology-based nation and by makingmaximum use of brain power, which is itsgreatest resource to develop innovative tech-nology.53 More specifically, the report suggests

“’’The Vision of MITI Policies in the 1980s, ” provisionaltranslation, Ministry of International Trade and Industry,Tokyo, Japan, Mar. 17, 1980.

38-802 0 - 85 - 16


that Japan should encourage development ef-forts and a switch-over to “forward-engineer-ing” in the knowledge-intensive or informationtechnologies.

By targeting specific information technol-ogy areas for national priority, MITI identifiesthose areas to receive a combination of directand indirect project R&D support. This sup-port system for national projects, often termedseed money because of its relatively smallamount, initiates basic precompetitive re-search and leaves to industry detailed product-oriented decisions. Often this seed money isgiven to various information technology firmsin the form of 50-50 matching grants.

Lastly, MITI forms company groups or re-search associations to work on a specific na-tional project. Sometimes research associa-tions have actually overseen R&D activities(as in the VLSI project); however, these re-search associations generally coordinate eachmember’s separate research efforts. Staff ofthese research associations usually include em-ployees on detail from government and indus-try, as well as retired industry and govern-ment officials.

Between 1966 and 1979, the Japanese Gov-ernment contributed approximately $400 mil-lion to 16 different national research projects.54

A chronological history of Japanese Govern-ment support for national information tech-nology research and development projects ispresented in figure 46. Since the early projects,typical amounts committed to national re-search projects appear to be increasing and thescope of the projects is towards more basic re-search.

The VLSI development project exemplifiesone of the better known national informationtechnology R&D projects, largely because ofthe subsequent market success of the Japa-nese integrated circuit industry. Begun in1976, the VLSI project involved the formationof a new VLSI research association with sevenparticipating private companies in addition toNippon Telephone and Telegraph and the

——64 Leonard Lynn, “Japanese Technology: Successes and Strat-

egies,” Current History, November 1983, p. 370.

MITI Electrotechnical Laboratory. The proj-ect was jointly funded at $150 million fromgovernment and $200 million from industryover a 4 year period. The VLSI Research ASSO-ciation and MITI laboratory efforts werelargely generic and provided support for al-ready existing industry R&D efforts. The neteffect of these efforts was the worldwide in-troduction of the first 64 K RAM device. Theresultant successes of the Japanese informa-tion technology industry may signify that ef-forts across public (MITI), quasi-public (NTT),and private (major corporations) sectors inpursuit of a common national technologicalgoal is in fact one of the strengths of the Jap-anese national R&D projects system.55

The national information technology R&Dproject which has received the most attentionrecently is the Fifth-Generation ComputerSystems Project. Begun in 1979, the projecthas become an impetus to the initiation ofother major national information technologyR&D projects in the United Kingdom, France,and Europe.

In light of the Japanese Government’s goalof stimulating basic research efforts, the ob-jective of the fifth-generation computer proj-ect is to move Japan to a lead position in in-formation technology areas related to officeautomation, computer-aided design, computer-aided engineering, robotics, and computer-aided instruction. Moreover, the intent is todirect information technology development inJapan to specific societal needs. These include:coping with an aging society; increasing activ-ity in low productivity areas; increasing ener-gy savings; and assisting the transformationof society into one in which information playsa key role. The goal of the fifth-generationproject is to develop basic technology and pro-totype systems that can perform functionssuch as inference, association, and learning aswell as non-numeric processing of speech, text,graphics, and patterns.

“For an in-depth analysis of the VLSI projec~ see Kiyanori%ludcibara, “From Imitation to Innovation: The Very LargeScale Integrated (VIM) Semiconductor Project in Japan, ” Al-fred P. Sloan School of Management, Massachusetts Instituteof Technology, 1982-1983.


Figure 46.—Japanese Government Support for Information Technology

123.4

5

6.

789.

10

12.

Sponsors Large Scale MIT I All government funding through consignment payments.These are combined because IS a continuation of (1) Sponsors All government funding through consignment paymentsFONTAC was aimed at developing a large size computer with IBM systems. Corporate The was established by law 1970 to encourage the development of software by direct and operations are by M(TI Three longterm credit banks provide loans to software houses and data services through guarantee fund Total government support unclear, but totaled

for FY 1972.1980 (see text for additional material), of (2) Sponsors AIST and Corporate Toshiba, Sanyo.

Research and Hoya Glass aimed at developing a new series of computers competitive with IBM’s 370 series a 50 percent subsidy to three computer manufacturer groups

Corporate (produced M series), NEC.Toshiba (produced and (produced Sponsor 31 companies, 50/50 develop high efficiency Input-output and terminalsSponsor unclear, 50/50 (government/private).Sponsors Machinery and Information Bureau and Data Processing Division, Corporate large Japanese software companies to anI PA the Systems Development Corp , a number of unspecified smaller firms, to Increase the production and use of software pro-grams This constitutes most software development program to date Results unclear& 11, because 11 IS seen as a of (10), Sponsors Machinery & Information Bureau and Industrial Electronics Corporate two phases: Fujitsu, & Toshiba, OK I, Sharp, Matsushita; above, plus NTT and AlST’s Laboratory staff, formed Phase VLSI Research formed, Phase II Electronic Computer Basic Technology Research Association formed (July 1979). Funding (govern.

conditional loan, repayable profits are generated from technologies; Phase 1: Y30 from the government, Y42 from the sector Phase from the government; Y24.5 from the sector

Sponsors National Research and Development Program, Corporate Toshiba, Matsushita Electric, Research Association of Optoelectronics Systems (January 1981); Laboratory

ed by Optoelectronics Joint Research the Fujitsu Plant, all government funding through consignmentpayments.Sponsors National Research and Development Program, Corporate Fujitsu, Toshiba, Mitsubishi GovernmentLaboratory Ass/s Laboratory, the for the Development of Speed Computers(December 1981), Laboratory is also Involved, although majority of work be conducted at companies’ own research all government funding through consignment payments.

Next Generation Baste Technology Of 48 companies 3 areas; numbers ( ) Indicate number offirms, Area New Materials (33), Area Biotechnology (14), Area Ill: Semiconductor Elements (10), five formed, 3 forArea 1, 1 for Area 11, and 1 for Area Ill; all government funding through consignment paymentsSponsor” Machinery and Information Bureau, Corporate Part/c/pal/on Fujitsu, Toshiba, Government LaboratoryAss/stance. Laboratory, at preparatory stages; The Institute for New

Computer Technology, an endowed research foundation (April 1982), total funding yet to be determined

SOURCE Jimmy W Wheeler, Merit E Janow, Thomas Pepper, and Yamamoto, “Japanese Industrial Development in the 1980’s: Implications for U.S.Trade and Investment, ” Hudson Institute , 1982, for the Department of State.


The Japanese Government established theInstitute for New Generation Computer Tech-nology (ICOT) in April, 1982 as the centerorganization for coordinating the fifth-gen-eration computer project R&D activities.Although the government is funding the in-itial 3 year R&D stage, eight manufacturersdonated money to establish and run ICOT.The consortium of eight manufacturers whichequally support ICOT and share in the re-search results are: Fujitsu Ltd., Hitachi Ltd.,Matsushita Electrical Industrial Co., Mit-subishi Electric Co., NEC Corp., Oki ElectricIndustry Co., Sharp Co., and Toshiba Co.These companies, in addition to NTT andMITI’s Electrotechnical Laboratory, havesent 42 researchers to the ICOT researchcenter.

Beginning with a staff of 52 and a plannedbudget of $450 million over the first 5 years,the overall research program is scheduled tolast for 10 years. In addition to its relativelylong-term research, ICOT is unusual becauseit is a separate neutral organization with a cen-tralized research laboratory. This contrastswith the traditional Japanese approach in

which each of the participating research insti-tutions and companies conducts its own re-search work. In addition to its own internalresearch, I COT cooperates with two govern-ment laboratories, various Japanese univer-sities, and independent foreign researchers.ICOT also contracts with Japanese industriesto make and test prototype software and hard-ware. The specific universities and companiesinvolved vary with each individual researchproject and participation is not limited to theconsortium of eight major companies sponsor-ing ICOT. Figure 47 illustrates ICOT’s orga-nization for various research projects.

The research plans of the ICOT center focuson seven major areas:

●

●

●

●

●

●

●

basic application systems,basic software systems,distributed function architectures,new advanced architectures,VLSI technology,systematization technology, anddevelopment supporting technology.

Within these seven areas, 26 research projectsare to be conducted by teams of university,

Figure 47.—Cooperation Between Research Participants for the Fifth-GenerationComputer Systems Project

(lest-model making)

SOURCE: Trudy E. Bell, “Tomorrow’s Computers—The Quest,” IEIEE Spectrum, November 19S3.


industry, and government researchers. The 10year span for these 26 projects is divided intothree phases. Begun in 1982, the initial phaseinvolves reviewing and evaluating research onknowledge processing and developing basictechnology for the second phase. Hardwareand software subsystems such as simulators,prototypes for language processing, and ex-perimental natural language processing sys-tems are being constructed for several exper-imental systems. The intermediate phase willattempt to develop subsystems for hardwareand software as well as algorithms and basicarchitecture. The final stage will attempt tointegrate software subsystems, hardware sub-systems, and applications software in order todevelop the first fifth-generation computerprototypes. These three phases of research

Figure 48.—Conc

● Basic appl icat ion system. Bas ic so f tware sys tem

● New advanced arch itectul● Distributed function

architecture

● VLSI technology● S y s t e m a t i z a t i o n

t e c h n o l o g y

● D e v e l o p m e n t s u p p o r ts y s t e m s

within the seven different areas are illustratedin figure 48.56

Science and Technology Agency (STA)The Science and Technology Agency (STA)

is responsible for the overall coordination ofsocial needs-oriented science and technologypolicy and expenditure in Japan. It is respon-sible for the planning, formulating and promo-tion of basic policies pertaining to science andtechnology, and for coordination of these pol-icies and activities throughout the various

Richard Dolen, “Japan’s Fifth-Generation ComputerProject, ” Scientific U.S. Department of the Navy, vol.7, No. 3, July-September 1982, pp. 63-97, and “Research andDevelopment Plans for Fifth Generation Computer Systems, ”Japanese Embassy, April 1982.

ept Diagram Showing How Research and Development Are to Progress in the FifthGeneration Computer Systems Project

study of basic In termedia te theory and system systern on a trial basis

system (prototype)

re S a t s i c evaluation architect - a

● Trial production of ❁■

experimental machine ,GW UJ -

● ✤ ❉ ▲ ✤ ❉ ❅ Fifthn of network $yateql generationwe * . . * * .

●

-

..w ~ m a n dknowledge base

I N

WanCad afchitao- 1

I Sufq

SOURCE” Fifth Generation Computer Systems Conference, Japanese Ministry of Trade and Industry, Tokyo, Japan, Oct. 19-22, 1961


government ministries. STA has jurisdictionover councils, research institutes and develop-ment agencies which mainly concentrate ontechnological developments in nuclear energy,space and ocean development, aviation tech-nology, and laser technology. Practically noneof STA’S budget directly supports industrialR&D; therefore, information technology R&Dwhich is categorized as an industrial area, isnot widely influenced by STA. However, ap-proximately half of the STA budget indirectlysupports information technology R&Dthrough procurement of information and com-munication technology equipment and facil-ities for its agencies’ activities.

Information TechnologyPromotion Agency (IPA)

Another government organization aimed atdeveloping and disseminating information andcomputer systems is the Information Technol-ogy Promotion Agency (IPA), which was es-tablished in 1970 under the Information Tech-nology Promotion Agency law. Its goal is topromote the use of computers, encourage thedevelopment and use of programs, and helpsoftware firms. It is the only national orga-nization in the field of software promotion inJapan.

Financing for the I PA comes from govern-ment subsidies, private corporations, threelong-term credit banks (the Industrial Bankof Japan, the Japanese Development Bank,and the Long-term Credit Bank of Japan), andfrom revenues earned by the association itself.One of the more important of IPA’s activitiesis its credit guarantee programs. Informationprocessing firms and software houses are oftenin need of funds to develop software programs,but have limited property that can be used ascollateral. The I PA has a system for guaran-teeing such obligations, as long as they areregistered with the IPA.S7

57An IPA-style credit guarantee system is not uncommon inthe United States. However, the American credit guarantee sys-tems tend to be aimed at broad industries, such as housing,rather than at narrowly targeted sectors.

Japan Electronic Computer CO. (JECC)and Japan Robot Leasing Co. (JAROL)

Assistance is also provided by the JapanElectronic Computer Co. (JECC), which bor-rows money from the Japan DevelopmentBank (JDB) and also from private banks. Itis a jointly owned firm that purchases com-puters from participating manufacturers andleases them to customers. In 1980, the JapanDevelopment Bank provided $263 million tothe JECC, $218 million in 1981, and approx-imately $100 million in 1982. When the JECCwas first established, it provided major supportfor the Japanese computer industry; however,as the financial resources of these companiesincreases, the Japanese computer companieshave become less dependent on governmentsubsidy and are establishing their own leas-ing operations.

The Japanese Government also helped toes-tablish the Japan Robot Lease Co. (JAROL)which is made up of 24 members of the JapanIndustrial Robot Association and 10 insurancecompanies. JAROL’S objective is the encour-agement of the development and use of robotsin small and medium businesses. Like theJECC, JAROL buys robots from manufactur-ers and leases them at low prices to smallbusinessmen. JAROL also receives most of itsfunds from the JDB and is therefore able tolease its robots at low prices. Similar to theJECC, JAROL aims to create a mass marketfor robot technology while encouraging pro-duction.

Ministry of Posts andTelecommunications (MPT): NipponTelegraph and Telephone (NTT)

The Ministry of Posts and Telecommunica-tions (MPT) indirectly influences informationtechnology research and development becauseof its administrative guidance over NipponTelegraph and Telephone (NTT).S8 NTT’s

58There is Some contention over whether NTT should be clas-sified as a nongovernmental or governmental entity. BecauseNTT receives no direct funding from the Japanese Governmentsome argue that Nil’ is not a government entity. On the otherhand, the U.S. Government has encouraged NTT to open upits procurements to foreign suppliers on the grounds that NTTis a government entity and therefore is subject to the GATTgovernment code.

Ch. 7—Foreign Information Technology Research and Development Ž 237

budget, services, tariffs, and overall policies,as well as appointments of top officials, aresubject to MPT’s review and approval. How-ever, NTT has not received a government sub-sidy for over 30 years; in fact, over the last3 years NTT has returned to the JapaneseGovernment on request approximately $2 bil-lion. This contrasts sharply with the idea thatthe Japanese Government heavily subsidizesinformation technology R&D.

NTT is the domestic public telecommunica-tions monopoly in Japan, although it does nothave any manufacturing capability within theorganization. NTT, as the owner of virtuallyall the telephone lines in Japan, remains themost powerful single entity in Japanese tele-communications. NTT accounts for aboutthree-quarters of the Japanese market for tel-ecommunications and data communications,equipment, and services. As a result, NTTwith its demands for new equipment and serv-ices has a powerful influence on Japanese in-formation technology research and develop-ment—more so than MITI.

Typically, new communications productsdestined for NTT use are initiated in one ofits four Electrical Communications Laborator-ies (ECL). NTT spends approximately 2 per-cent of its revenue on R&D (which amountedto more than $350 million in 1980), mainly atECL labs. This system of labs corresponds toBell Labs although it is approximately one-fifth of the size of its U.S. counterpart. ECLtends to do more developmental R&D ratherthan the basic research for which Bell Labshas been so widely acclaimed. Much of the re-search carried out at NTT’s labs is devoted toachieving the extremely detailed and demand-ing specifications that the company requireswhen it issues R&D contracts to private firms.

More often, NTT launches research in col-laboration with one or more of the four majorJapanese electronics firms (NEC, Fujitsu, Hit-achi, and Oki Electric), which actually sendstaff to the NTT labs. The subsidized joint re-search normally results in NTI's appointmentof a preferred supplier from among the re-searchers when the time comes to purchase the

product. Because NTT is one of the largest in-formation technology and telecommunicationsmarkets in Japan, most electronics firms coop-erate with NTT. Competition often developsbetween these firms in efforts to be selectedas partners in new technologies and systemsdevelopments.

NTT divides its procurement procedurespractices into three tiers or “tracks” whichhave been agreed to in the General Agreementon Trade and Tariffs (GATT) and other bilat-eral trade agreements:59

Track 1 (competitive bidding) is applied toproducts to be procured based on the Gov-ernment Procurement Code agreed to by theGeneral Agreement on Trade and Tariffs(GATT). These procurements usually includeoff-the-shelf products such as PBXS, data ter-minals, modems, computers, peripherals, fac-simile machines, measurement instruments,etc.Track II is applied to equipment that is notavailable in the marketplace and which re-quires some research and development. Thistrack generally refers to products for whicha limited amount of collaboration betweenthe supplier and NTT is necessary to tailorapplications to NTT’s specifications. TrackII contracts are normally single-companycontracts.Track III is applied to equipment not avail-able in the marketplace and which requiresextensive R&D for NTT use. These are themost highly prized contracts and the mostdifficult for foreign or small companies topenetrate. NTT seeks a supplier with suffi-cient research ability to develop a product,or to develop one according to an NTT pro-totype.In addition, there are Tracks 11A and 111Athat allow for new producers to take over ex-pired Track II and III contracts.U.S. and other foreign telecommunications

equipment manufacturers have suggested thatNTY'S close collaborative R&D activities withseveral Japanese companies prevent themfrom penetrating the Japanese telecommuni-

5gJack Osborn, outgoing telegram from the U.S. Embassy inTokyo, Japm Oct. 21, 1983, 7 pages.

238 • Information Technology R&D: Critica/ Trends and /ssues

cations market. With a 20:1 trade deficit intelecommunications equipment and a substan-tial Japanese penetration of certain U.S. mar-ket niches, many U.S. companies would liketo balance some of these telecommunicationstrade deficits.6o Moreover, other motivationsfor insisting on participating in the Japanesemarket relate to the changing balance inAmerican and Japanese research and develop-ment. Just as the Japanese presence in theAmerican market has a dual purpose-exportsand the transfer of technology-so mightAmerican participation in Japanese marketsand R&D activities serve two functions.

The issue of NTT’s procurement policy canbe best viewed within the framework of theplans for liberalization of the entire Japanesetelecommunications monopoly. Because alarge amount of NTT'S profits (last year’s prof-it was $1.6 billion) goes to the Japanese Gov-ernment ($600 million) which needs revenue inthe face of its continuing deficits, NTT can-not make a large profit. Another fiscal con-straint is the rapidly decreasing rate of growthin the number of telephone users. More than90 percent of Japanese households now havetelephones, which means that the traditionalmarket (revenue) has leveled off. In addition,NTT is in the process of testing and eventuallyimplementing its ambitious 20-year all-digitalInformation Network System (INS) program,which is expected to cost up to $120 billionby 1990.

To solve these financial problems, PrimeMinister Nakasone is supporting a major fourstep reform plan recommended last year by aspecial study group commissioned to studythe Japanese Government. The plan, nowunder discussion, would first convert NTT toan incorporated government-owned entity.NTT would then be free, however, to set itsown management and personnel policies.Under the proposed plan, many operations,

‘In 1982 the Japanese exported more than $408 million intelecommunications equipment to the United States. Duringthis same period, Japan imported less than $88 million of thissame equipment from the United States. Peter J. Harm, “DataCommunications in Japan, ” Data Communications, August1983, p. 56.

such as data communications services, wouldbe delegated to spin-off companies operatingat a regional level-somewhat like the patternof the AT&T divestiture in the United States.

Hisashi Shinto, the new chairman of NTT,believes that transforming the massive bu-reaucratic Japanese telecommunications mon-opoly into a partly private company will bemore profitable, while encouraging greatercompetition and innovation in the telecommu-nications and information technologies indus-tries. It is believed that more Japanese com-panies will be encouraged to vie both forNTT’s business and for the private telecom-munications market previously controlled byNTT and its selected family of suppliers.

Kokusai Denshin Denwa Ltd. (KDD)The KDD operates Japan’s international tel-

ephone, telegraph, and other related commu-nications services.61 Divided from NTT in1953, it is 90 percent privately owned, with10 percent of its stock held by NTT. At thetime of the inauguration of KDD, the inter-national telecommunications research groupof the Electrical Communication Laboratoriesof NTT was transferred from NTT and reor-ganized as KDD’s Research Department. By1969, a development center was created andthe department was renamed Research andDevelopment Laboratories. The laboratories,located in Meguro, Tokyo, employ more than120 R&D personnel and are composed of 12special purpose laboratories and threedivisions.

A new laboratory being built in the Nerimasuburb of Tokyo, will be completed in 1985.This R&D reinforcement will permit furtherresearch and development concentration insuch fields as switching for integrated digitalnetworks, fiber-optic submarine cable trans-mission, satellite digital transmission, optical-memory disks, system conversion techniques,wideband video, etc. To help in this effort,KDD expects to add 50 more engineers to itsR&D staff by 1988.

8’Yasuo Makino, “Telecommunications in Japan: ChangingPolicies in a Changing World, ” Z’ekcommum”cati”ons, October1983, pp. 139-145.


Defense AgencyBecause of Japan’s small national market

for defense equipment, Japanese defense re-search and development activities are fairlylimited. Overall, Japanese procurement of de-fense equipment accounts for less than four-tenths of 1 percent of total Japanese indus-trial production.

Within the Defense Technical Research In-stitute there are very small-scale developmentprojects on electronics equipment (radar, etc.)with a budget of ¥ 1.3 billion ($5.6 million) infiscal year 1982, ¥ 0.6 billion ($2.6 million) infiscal year 1983, and ¥ 3.8 billion ($16.5 mil-lion) in fiscal year 1984. These projects aredone in cooperation with private firms. In ad-dition, the Defense Agency indirectly supportsinformation technology R&D by purchasinghardware for testing purposes from privateelectronics firms.

As a result of economic trends in both theUnited States and Japan, as well as a chang-ing international security environment, therehave been growing tensions over trade and de-fense issues. In general, some Americans be-lieve that the low level of Japanese militaryexpenditures frees funds for civilian researchand investment while requiring higher taxesand absorbing resources in the United Stateswhich in turn provides defense.62 As a result,the U.S. Government believes that Japanshould increase its military strength as wellas information technology R&D expendituresfor military applications. Moreover, U.S. com-panies argue that an inequality exists betweenUnited States and Japanese trade: no manu-factured U.S. civilian product (with the singleexception of airplanes) has captured as muchas 10 percent of the Japanese market, whileapproximately 14 percent of Japanese defenseequipment is purchased by the United States.63

— — —62David Denon, ‘‘Japan and the U.S.—The Security Agenda, ”

Current History, November 1983, p. 355,BqS~phen J. SOl~Z, “A Search for Balance,” Foreign Affm”rs,

p. 75.

Japan Development Bank (JDB)The Japan Development Bank (JDB) is anoth-

er government financial intermediary used totarget industrial development. The JapaneseGovernment’s Trust Fund Bureau (which isthe main organization in its Fiscal InvestmentLoan Program) provides JDB with its mainsource of capital, though it can also raise fundsby issuing certain types of bonds. JDB’s prin-cipal responsibility has been the extension oflong term, low interest loans for capital invest-ment in new industries. In the years immedi-ately after its formation, JDB concentrated onloans for the reconstruction of basic manufac-turing industries.

As a result of the consensus to increase thesupport for “knowledge-intensive’ industries,the JDB began to target support for what itterms “development of technology.” The fund-ing categories in area of development of tech-nology are illustrated in table 43. Most of thecomputer funds, as a matter of policy, havegone to the JECC, although some softwarefirms have also received funding. For the otherfunding areas for technology development,there are two general JDB loan programs.Both of these loan programs, which resembleMITI’s seed money grants, attempt to stim-ulate private investment in specific areas ofinformation technology R&D. The first loanprogram, set up under a 1978 law, amountedto ¥ 10 billion ($43.3 million) in 1981. Loansfrom this program must be directed towardspecific project areas designated by cabinet or-der. Should a designated project area be over-subscribed (as happened with semiconduc-tors), JDB can force larger firms that have bet-ter access to private financial markets to uti-lize those markets, while JDB loans arepreserved for the smaller firms.

The other technology development loan pro-gram was established by the bank itself andnot designated by specific laws, though it stillfalls within the broad policy guidelines of thegovernment. This part of the JDB budget to-taled ¥ 44 billion ($190.5 million) in 1981.


Table 43.—Japan Development Bank Loans for Development of Technology (in billions of yen)

Fiscal year Fiscal year Fiscal year Fiscal year1977 1978 1979 1980

New loans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ¥71.2 ¥129.0 ¥108.5 ¥96.4 $457”

Development of electronic computers. . . . . . . 38.2 55.3 47.1 55.4 262Domestically-manufactured computers . . . . 35.5 53.5 45.0 54.0 256Computer manufacturing plants . . . . . . . . . . 0.4 0.2 0.4 0.6 3Data processing systems. . . . . . . . . . . . . . . . 2.3 1.6 1.7 0.8 3

Use of high technology in certain electronicand machinery industries. . . . . . . . . . . . . . . . 8.3 7.8 10.2 14.5 69Electronic industry . . . . . . . . . . . . . . . . . . . . . 3.8 2.1 7.0 12.0 57Machinery industry . . . . . . . . . . . . . . . . . . . . . 4.5 5.7 3,2 2.5 12

Development of domestic technology . . . . . . . 24.7 65.9 51.2 26.5 126Development of new technology . . . . . . . . . 20.4 57.4 40.9 22.6 107Trial manufacturing for commercial use . . . 0.9 4.0 1.2 0.3 2Development of heavy machine. . . . . . . . . 3.4 4.5 9.1 3.6 17

aln millions of dollars.SOURCES: U.S. Facts arrd Figures About the Japan Deve/oprnent Bank, Japan Development Bank, 1981, p. 28; and Jimmy W. Wheeler, Merit E. Janow, Thomas Pepper,

and Midori Yamamato, “Japanese Industrial Development Policies in the 1980’s: Implications for U.S. Trade and Investment,” Hudson Institute Inc., 1982,for the U.S. Department of State.

These loans are devoted to new domestic tech-nologies and initial manufacturing efforts forcommercialization of these new technologies.Firms that believe that they have developeda process or technology falling within thebroad parameters established by the cabinetmust apply in order to be considered for loans;JDB does not solicit customers. The firm’sproposal is submitted to a council of scientificadvisors, which evaluates the proposal. If thetechnology is approved, the applicant thenfaces an evaluation of credit worthiness andof the financial characteristics of its loan ap-plication. If the applicant is a large companywith well established financial links, it mustconcurrently seek private financing, becauseJDB will provide only partial funding. If theapplicant is small, and has relatively weak fi-nancial links, or if the project is large-scale orviewed as a high priority for the nation, thenthe JDB may take a lead role in putting to-gether a consortium to finance the project.Finally, by general agreement, the JDB onlyfinances the first plant in a new area. Its roleis to help launch new technology, not to pro-vide low cost financing for the expansion ofindustry.

The small size of the JDB loans indicatesthat the importance of government’s financialmediary role “stems not from outright controlor from overall size, but rather from socializ-

ing risks, coordinating private investments,and processing information.”64

University

The Ministry of Education, Culture, and Sci-ence (MOE) funds research at three researchinstitutes, six National Institutes (for joint useby universities), and 93 national universities.

Among the various national universities inJapan, the big seven are the universities ofTokyo, Kyoto, Osaka, Tohoku, Hokkaido,Nagoya, and Kyushu. In the area of informa-tion technology, Tokyo University has activ-ities in several departments: informationscience (in the Faculty of Science), informationengineering and precision engineering (in theFaculty of Engineering), as well as a large cen-tral computer center. In most of the other uni-versities, there is just one department (usu-ally in the Faculty of Engineering), as well asa sizeable central computer center. Kyoto Uni-versity claims to have the oldest informationengineering department and is considered aclose rival to the University of Tokyo. Othernational universities that have significant

6tEi~uke S&&ibma, Robert Feldman, ~d YUZO H~ada!“The Japanese Financial System in Comparative Perspective,”a study prepared for the use of the Joint Economic Committee(Washington, DC: U.S. Congress, U.S. Government PrintingOffice, Mar. 12, 1982), p. 11.


computer science departments include Tsuku-ba University and the Tokyo Institute of Tech-nology. Among private institutions, only KeioUniversity and Waseda University are consid-ered to have sizeable computer scienceprograms.

Higher education has become an importantsocial investment in Japan. Between 1960 and1975, the numbers of students in higher edu-cation multiplied by more than three times (to2.2 million), including students in Japan’s jun-ior colleges (post high-school institutions con-cerned with teacher training, technical educa-tion, etc.). Of the 1.73 million undergraduatesenrolled at Japanese universities in 1981, ap-proximately 334,000 were enrolled in engineer-ing studies (exactly six times as many stu-dents in the natural sciences, includingmathematics). Many universities have noscience faculty, only engineering departments.These large numbers of engineering studentsindicate the heavy emphasis placed on engi-neering in Japan. Consequently, Japan contin-ues to maintain a large engineering manpowerbase. However, a small percentage of theseengineering students continue on to graduatestudy where they could contribute to basicuniversity research. For instance, in Japan,the proportion of graduate research studentsto the entire undergraduate student popula-tion is 3 percent. In the United Kingdom theproportion is over 19 percent; in France 22 per-cent; and in the United States 12 percent.65 Al-though Japan has more undergraduate stu-dents enrolled in electrical and electronicsengineering programs than the United States,Japan has approximately one-fourth as manyelectrical and electronic engineering graduatestudents as the United States. Because stu-dents at the graduate level make a large con-tribution to basic R&Din university environ-ments, the low number of Japanese graduateengineering and science students has beencited as one reason for such a small amountof university research in Japan. In this contextit is also worth noting that Japanese compa-nies, which prefer to train their own develop-ment engineers, like to recruit young inexpe-

“The Economist, Aug. 6, 1983, p. 65.

inexperienced persons-recruiting them beforethey go on to postgraduate work.

Recently, there has been concern that theenvironment for information technology basicresearch has generally not been adequate atJapanese universities. In addition to the in-adequate number of graduate students, twoother causes for the small amount of informa-tion technology basic research activities inJapanese universities have also been cited.

Some researchers believe that the heavy em-phasis on rote learning in Japanese schools hashelped to suppress creativity in the learningprocess. Moreover, the importance of severeuniversity entrance examinations has beenseen as a deterrence to specially talented orcreative students.

Researchers usually cite two indices of suc-cess in originality and successful basic re-search-the number of Nobel Prize winnersand the frequency with which scientists’ workis cited by other researchers. Japan has hadfour Nobel Prize winners. The citation indexdevised by the Institute of Scientific Informa-tion in the United States includes 19 Japaneseresearchers (and roughly half of those workedin American laboratories) among the 1,000international scientists accredited with themost frequent citations in the scientific lit-erature.66

A second major factor affecting basic re-search in university environments is the levelof government funding. More than 98.8 per-cent of R&D expenditures at national and pub-lic universities was funded by the governmentand only 1.2 percent by private industry.67 In1979, universities spent approximately $3.69milion for R&D: national and public universi-ties spent $2.459 million, and private univer-sities spent $1.15 million. The funds for re-search are distributed by the Ministry of

66The citation system however, does not take into accountthe fact that very few Western scientists are able to read theJapanese scientific literature that is published in Japanese. SeeMarch 1984 hearings held by the House Science and Technolo-gy Committee on Japanese science and technology information.

67Michiyuki Venohara, Nippon Electric Company, Ltd., “Jap-anese Social System for Technological Development-Its Meritsand Demerits, ” presented on Dec. 8, 1982 in Endohoven, Neth-erlands, p. 21.


Education in the form of formula support andproject research grants.

Through formula support the governmentprovides each Department Chairman withthree posts, two assistant professors and oneresearch assistant or vice versa. In each engi-neering or information science department orother related department, the chairman isgiven approximately ¥ 7 million ($28,000) tospend on research. Half of these funds maybekept by the university to cover administrationcosts. As a result, many departments are leftwith approximately ¥ 2 million ($8,000).

The research project grants, totaling ¥40,000 million ($160 million), also appear to beinsufficient because of their short term andsmall amount. A limitation is that the fundscannot be used for recruiting short-term assist-ance because most researchers (as well as mostJapanese workers) enjoy life-time employ-ment. The Japanese Government began in1982 to award 3 or 4-year support for researchprojects, each costing approximately $1 million.

Although funding levels for basic researchare sometimes perceived as being inadequate,the equipment in university laboratories hasbeen described as adequate by one recentAmerican visitor to the University of Tokyolabs:

In general, both in the industrial as well asin the university laboratories, the impressionI got was of a lot of equipment, some old,some new, mixed in a somewhat random fash-ion. In the university laboratories, in particu-lar, space seems to be at a premium. Theredoes not, however, appear to be any short-age of new equipment.G8

Research and Development LinksBetween Universities and Industry

Direct cooperation between universities andindustries in basic and applied R&D has beenrelatively limited. Japanese companies do not

‘Derek L. Lile, “Japanese Laboratory Visits,” %“entific Bul-letin, Department of the Navy, Office of Naval Research, FarEast, January-May 1983, p. 48.

encourage students to obtain work experiencein private industry. Furthermore, most Japa-nese firms do not generally look to universitiesfor innovative ideas; the larger, more impor-tant firms prefer to carry out their own basicresearch. Some university professors in Japanhave accused Japanese firms of suffering froma “not invented here syndrome” and this hascaused a great mistrust between industry anduniversity faculty. A further factor is that theMinistry of Education, which is the direct em-ployer of university staff, actively discouragesdirect links between academics and companies.Professors at state universities are forbiddento consult for private firms because it couldlead to nepotism in obtaining appointments.

The most effective channel for university-in-dustry collaboration in Japan is through in-formal personal links. In comparison withother countries Japanese graduates remain inclose contact with each other throughout theirprofessional careers. This leads to valuable co-operation between academics and industrial-ists, particularly in research. This is reinforcedby the role university professors play (as em-ployees of the Ministry of Education) in theestablishment and implementation of nationalresearch programs, such as the VLSI projectand the Fifth-Generation Computer SystemsProject. The presence of academics in the con-trolling bodies of these projects helps to ex-change research results between companiesand universities more effectively.

Other cases of informal collaboration includeexchanges of researchers between companiesand industries. For example, the ElectronicsDepartment at the University of Tokyo cur-rently has five visiting researchers from wellknown Japanese electronics companies for pe-riods of 1 or 2 years. Also the department re-ceives grants from at least 10 companies tobe used for purchasing equipment. The ar-rangement for visiting researchers from indus-try has some similarity to the U.K.’S TeachingCompany Scheme, but the emphasis seems tobe on the industrialist working in the univer-sities rather than on the academic working inindustry environments.

Ch. 7—Foreign /formation Technology Research and Development . 243

Collaboration between companies and uni-versity departments is also aided by the prac-tice of using universities as “shop windows”for new equipment that is often given or soldto them at very low prices. Through univer-sity use of new equipment, firms receive feed-back on the operation of new equipment andoften valuable suggestions for modificationsand extensions. Instances of this practice canbe found in the areas of computers (Fujitsu)and telecommunication receivers (NEC).

Another recent government incentive for in-dustry-university R&D collaboration concernspatents. As government employees, universitystaff in Japan were not allowed to profit fromthe commercial exploitation of their ideas.This has created a disincentive to patenting,while promoting dissemination of resultsthrough open publications. Currently aca-demic researchers are being encouraged toapply for patents and at least two universitieshave set up special offices to facilitate the pat-ent application process (Tokyo Institute ofTechnology and Tohoku University). Thiscould be to make research results more attrac-tive to private firms, which can in turn applyfor licenses in order to market the technology.

As the technological level of Japanese indus-tries approaches that of other advanced na-tions and the innovation of original technol-ogies is more widely demanded by domesticas well as foreign markets, the need has beenvoiced by Japanese industrial and governmentcircles for much closer cooperation between in-dustry and university researchers. As a result,the Japanese Government is also encouragingcloser research links between universities, in-dustrial firms, and government institutionsthrough the development of research parks,such as Tsukuba Science City.

Industry

To compete in global markets, Japanese in-formation technology industries, in general,are large-scale organizations in order to assuremaximum economies of scale and to sustainthe large amounts of capital necessary for con-

tinued innovation. Although there are manysmaller electronic firms in Japan, most of themsubcontract small-scale production or fill spe-cial niches (some in global markets) which re-quire custom or batch labor intensive produc-tion technologies. In general, however, majorinformation technology firms are more diver-sified and highly integrated than other com-petitor nations’ industries. For example, Jap-anese firms competing in the semiconductormarkets are significantly larger in total salesand assets than their U.S. counterparts-ap-proximately two to four times larger thanTexas Instruments and Motorola, and muchlarger than National Semiconductor, Fairchild,and Intel.69

In addition to large-scale operations, manyof the major Japanese information technologyfirms are vertically integrated. For example,most computer manufacturers produce semi-conductors and several of them are majorsemiconductor suppliers in the world market.This sharply contrasts with the United Stateswhere, although most computer manufactur-ers have at least some in-house semiconduc-tor development and production capability,only a few firms market both semiconductorchips and computers, and most of them do not

‘gGene Adrian Gregory, and Akio Etori, “Japanese TWhnol-ogy Today, the Electronic Revolution Continues, ” ScientificAmerican supplement, 1983, p. J, 22.

Photo credit: Overseas Public Affairs Office,Electronic Industries Association of Japan

Semiconductor research


sell the large range of products that are avail-able from Japanese companies.

Another type of structural integration thatis believed to give the Japanese a competitiveadvantage is in the area of consumer products.Compared to many American computer man-ufacturers that specialize in their productioncapabilities, Japanese firms which manufac-ture computers are also large producers of con-sumer electronics products. Because consumerelectronics products usually have a large mar-ket and generate large sales revenues, Japa-nese firms can utilize these large profits tofund research and development projects inother information technology areas.

As a result of structural integration, namely,production and development of computers inconjunction with telecommunications equip-ment, and consumer products coupled with in-tegrated semiconductor development and pro-duction capabilities, the Japanese informationtechnology firms may have a strong techno-logical base from which to continue their in-novation process. Moreover, because of theirvertical integration, Japanese informationtechnology firms can draw on cash flow gen-erated by consumer electronics sales to sus-tain large capital investments and researchand development costs of basic research.

Japanese information technology firms alsoparticipate in cooperative joint R&D pro-grams. Because there is relatively low technol-ogy transfer between Japanese firms (due tolifetime employment of engineers and fiercecompetition between firms), they often under-take identical research projects. The efforts oflimited numbers of research engineers in eachcompany are therefore spread out across manydifferent areas, making it difficult to concen-trate on basic research. In response to thesedifficulties, the Japanese Government createdresearch associations so that industries couldcollaborate on research projects centered ondifferent technology areas.70

70See the section on MITI for a description of the industrialparticipation in the research associations.

Because of the highly competitive nature ofthe electronics firms, industrial researcherswere at first extremely skeptical about theseresearch associations. However, the area of co-operation is limited to very basic research, anddevelopment efforts are still necessary to de-velop commercial products. The basic researchefforts generally leave many possible avenuesfor future product development and allowfirms to compete with one another. Becauseof their wide acceptance and heavy industryparticipation, these joint R&D programs havebeen useful to Japanese industry.

NEC Corp.NEC Corp. is Japan’s largest manufactur-

er of communications equipment with fiscal1982 sales totaling $4.98 billion and a pretaxprofit of $204 million. In 1982 sales to NTTand the Japanese Government agencies ac-counted for $888 million, or 19 percent ofNEC’S revenues, of which a significant portionwas for data communications. Foreign pur-chasers accounted for 30 percent of NEC’Ssales. NEC has a strong position in the worldmarket (including several plants in the UnitedStates), and it markets over 14,000 productsin over 100 countries worldwide. NEC has alsoproduced the distributed information-process-ing network architecture (DIMA) on which 60to 70 percent of all networks in Japan are nowbased. NEC’S success in part can be attributedto the fact that it is a vertically integrated sup-plier with strong shares of the Japanese com-puter, communications, and semiconductormarkets. Internationally, NEC is known forhaving installed over half of all satellite earthstations and for its microwave technology.NEC’S vertical integration activities are bestsummed up by Michiyuki Uenohara:

Combining computers and communica-tions (C + C) was conceived as a form of bus-iness best suited for our expansion into newareas, making the most of technological re-sources we have as a company that startedout as a communications equipment maker.It was in 1975 that the concept was put intopractical business programs, and in 1977, we


came up with a clear-cut expression of C + C.First of all, we achieved the ability to put onthe market a digital switching system whichcan serve as a link between communicationsand computer networks. Digital switchinghas made the combination of computers andcommunications not only possible, but nat-ural. 71

NEC’S central research laboratory, locatedjust outside Tokyo in the city of Kawasaki,houses approximately 700 scientists involvedin research on a wide variety of subjects re-lated to computers and communications. Specific research includes projects on GaAs cir-cuit development and semiconductor surfacetreatment.

Fujitsu Ltd.

Fujitsu is Japan’s largest mainframe ven-dor, and ranks second to NEC in telecommu-nications equipment sales. Fujitsu is nowranked sixth in the world, ahead of Honeywell(U.S.), CII-Honeywell Bull (France), ICL(U.K.), and Siemens (West Germany). With fis-cal 1982 sales of about $3.36 billion, telecom-munications equipment accounted for $638million of Fujitsu’s total sales. Sales to NTTaccounted for 37 percent and sales to the Jap-anese commercial market were 27 percent ofFujitsu’s total sales. Similar to NEC, Fujitsuis well established in the world informationtechnology market, with a solid history ofsales in the United States. Currently, Fujitsuis manufacturing and marketing digital PBXsin a joint venture with American Telecommu-nications Corp., and microcomputers and ter-minals in a joint venture with TRW, Inc.Other joint venture partners include Canadian,West German, and Spanish companies. Al-though Fujitsu ranks second to NEC inmicrowave equipment, carrier transmission,and automated office equipment, it is a leaderin Japan’s fiber optics and optoelectronics re-search.“Gene A~rian Gregory and Akio Etori, “Japanese Technol-ogy Today, The Electronic Revolution Continues, ” ScientificAmen”can, Special Supplement, 1984, p. J, 44.

The “tsu” in the title Fujitsu means com-munications and this was the basis of the Fu-jitsu laboratories. However, because of a per-ceived saturation of the communications field,Fujitsu decided to diversify and as a result,communications equipment now occupies aminor part of Fujitsu’s research operations.Approximately 60 percent of the research ac-tivities is currently devoted to computer andcomputer-related research. The Fujitsu Lab-oratories, also located in Kawasaki, employ ap-proximately 800 people. The semiconductordivision consists of 190 people divided intothree main subgroups: GaAs devices and cir-cuits, silicon integration, and materials. Be-cause its research is well advanced in the semi-conductor division, Fujitsu is expected to playan important role in the Japanese supercom-puter project.

Hitachi Ltd.Hitachi is Japan’s third largest company

(after Nippon Steel Corp. and Toyota Ltd.). Itentered the computer market in the late 1950sand was one of the first Japanese companiesto enter into a technology exchange agreementwith a U.S. computer manufacturer (RCA in1961). Since then, Hitachi has made agree-ments with Intel. During 1978-80, Hitachi alsocompleted marketing arrangements for itscomputer systems in Europe with BASF(West Germany), Olivetti (Italy), and St. Go-bain (France).

Hitachi generated more than $3.3 billion infiscal 1982 (out of total sales of $15 billion)from information and communications sys-tems and other electronics devices. AlthoughHitachi has in the past maintained a strongposition in the consumer electronics market,the company is placing more emphasis on in-dustrial electronic equipment. As a result,Hitachi is attempting to evolve into an in-tegrated office systems supplier in the UnitedStates. Hitachi is concentrating some of its ef-forts on the development and marketing ofPBXs.


Photo credit: Overseas Public Affairs Office,Electronlc Industries Association of Japan

A Hitachi engineer feeds a glass plate into an electronbeam lithography device which produces a photomaskby drawing LSI patterns on the plate with a sharplyfocused electron beam under high-precision computercontrol. The device is also capable of directly writing

patterns on silicon wafers.

Hitachi has rapidly increased research anddevelopment expenditures. Hitachi’s CentralResearch Laboratory at Kokubunji is one ofthe five Hitachi research labs and is devotedto the development of new materials and de-vices as well as new measurement equipment,medical engineering equipment, and commu-nications and information processing systems.The laboratory, established in 1942, employsapproximately 1,200 research and support per-sonnel. Their principal integrated circuit re-search project is aimed at developing a 1K bitstatic RAM for use in the central processingunit (CPU) and main memory of large comput-ers. Other projects that are also currently

underway include the development of GaAs,analog circuits for automobile telephones, andsemiconductor material research. There is alsoresearch taking place in lasers, light-emittingdiodes (LEDs) and light detectors, and somepreliminary work in electro-optic integration.

Oki Electric Industry Corp.

Oki Electric Industry Corp. was one of Ja-pan’s first entrants into the computer marketwith its transistorized OKITAC 6020 in 1959.Oki-Univac Company Ltd. was formed in Sep-tember 1963 to manufacture various Univac-based machines in Japan. Since that time, Okihas not engaged in the development of largecomputers. However, it remains strong inperipherals and terminals. In 1972, Oki estab-lished subsidiaries in the United States to produce and market communications equipmentand computer peripherals, terminals, and com-ponents, and has subsequently set up similaroperations in Germany and Brazil.

Oki Electric Industry Corp. had fiscal 1982sales of $1.03 billion, of which telecommunica-tions sales to NTT totaled $185 million andthose to government entities another $94 mil-lion. Oki also exported approximately $120million in equipment, mainly digital printers,to the United States. In July, the company an-nounced that it will build a plant in Atlanta,Ga., to produce mobile cellular radio tele-phones for a subsidiary of AT&T. The facil-ity is expected to produce 1.5 million units ayear eventually.

-France

Introduction reduced to between 7 and 8 percent.72 Infor-The historically agrarian society of France mation technology, particularly telecommuni-

has changed with tremendous speed since cations, has been a major component of thisWorld War II. In 1945 over half of the Frenchpopulation was dependent on agriculture for ‘zPierre Aigrain, “Seminar on High Technology in France, ”

Center for Strategic and International Studies, Georgetown Uni-its income; by 1970 that dependency had been versity, Feb. 9, 1983.


change. French efforts to expand advanced in-formation technology research and productionhave been directed toward two major goals:strengthening France’s international competi-tiveness and the development of an informa-tion technology-based infrastructure for thepreservation and continued development ofFrench culture and society. Coupled withhistoric French governmental involvement inindustry,73 the comparatively small size ofFrench participation in the world market forinformation technology, and the uniqueFrench social and political contexts for tech-nology, these goals have shaped French infor-mation technology research and developmentinto a pattern quite unlike that in the UnitedStates.

French information technology research anddevelopment activities occur in governmentlaboratories, in industrial settings, and in aca-demic environments. The structure, organiza-tion, and direction of R&D activities withinthese communities are all dissimilar, however,from the American experience. The pervasive-ness of government intervention in industrialand academic sectors makes it difficult to dif-ferentiate the three areas, but for ease of com-parison with the American experience, Frenchgovernment, university, and industry informa-tion technology research and development en-vironments are discussed separately below.Before discussing these environments, it is im-portant to understand the size of French par-ticipation in information technology, and thesocial and recent political environments for theconduct of information technology researchand development.

The Size of French Participationin Information Technology Markets

In 1982 the French estimated that they con-trolled about 5 percent of the world market ininformation technology. This compares with73The French Government has traditionally played a large rolein the coordination, funding, and direction of the French econ-omy since Jean Baptiste Colbert founded the Academy of Sci-ences in 1666. French Governments since have changed thescope and nature of that involvement, but the traditional mech-anisms used by government in industry, including those of thepresent French Government, have changed very little.

Photo credit’ Scientific Mission, Embassy of Franceand Ministere des PTT, French Government

Videotex terminal “Minitel”

a United States share of 48 percent in 1982.Moreover, if one were to compute the Frenchshare of the world market using informationtechnology goods and services produced ex-clusively by French-controlled corporations,the share would decline to 4 percent.74

Slightly over one-half of the French infor-mation market is served by foreign suppliers;the United States holds about 22 percent ofthe market, Japan and West Germany 7 per-cent each. The Netherlands holds 6 percent———. -.— - .

“French telecommunications and Electronics Council, TheElectronics Industry: U.S.A./France 1982, pp. 10-18.

38-802 0 - 85 - 17


and Italy, the United Kingdom, Austria andothers hold the remainder. Several major sub-sectors of the French market such as main-frame computers are nearly dominated by U.S.manufacturers. In areas of French strengthwithin the French market (e.g., in telephoneterminal and switching equipment, in whichthe French control about 91 percent of theirmarket), French participation in the UnitedStates and/or world markets is often verysmall. For example, France has less than one-tenth of 1 percent of the U.S. market in tele-

rminal and switching equipment. Thisphone tesituation may be more due to the structure ofthe U.S. telecommunications industry thanany French inadequacy. In addition, the futureFrench position in the U.S. telecommunica-tions market may improve as both CIT Alcateland Thomson CSF have recently establishedU.S. subsidiaries.”

The size of French participation in the mar-kets for information technology affects the lev-el of funds available for R&D. French indus-trial funding of information technologyresearch and development was reported as$2.2 billion in 1982, some significant portionof which was funded by the government. Inaddition to information technology researchand development in industrial settings, the ci-vilian French governmental funding of infor-mation technology conducted in public labora-tories was reported to be $0.6 billion in 1982.76

The Political Environment for FrenchInformation Technology Research

and Development

It would be difficult to describe French in-formation technology research and develop-ment activities without first considering thecontext of the present French Government’spolicy. The last French presidential election(1981) marked the first time science and tech-nology were used as apolitical issues.” Indeed,

751nterview with Mr. Chavance, CIT Alcatel director, June1983 and Telephony, July 25, 1983, p. 24.

7oA.F.P. Sciences, No. 325, Oct. 7, 1982, p. 30.77 Pierre Aigrain, “The French Experience in High Technol-

ogy, ” Center for Strategic and International Studies, George-town University, p. 2.

all candidates had some increased R&D fund-ing planks in their platforms. Before losing toMr. Mitterrand, Mr. d’Estaing had designeda plan for increasing real government R&Dfunding 8 percent per year for 5 years begin-ning in 1980. When Mr. Mitterrand waselected, he more than doubled that goal.

Mr. Mitterrand’s emphasis on increasingR&D spending was part of a larger industrialpolicy for France which included companionemployment and education policies as well asplanned market programs in several areas ofhigh technology.78 The overall government pol-icy, designed around the Socialists’ principlesof decentralization, democratization, human-ism, and volunteerism, included severalelements.

The first element of the French Govern-ment’s general policy was the declaration ofinformation technology development as a na-tional priority. Thus a major objective was tobring together universities, government lab-oratories, and industry to enhance national ef-forts in technological development.79 The sec-ond element was to convince the French peopleof the importance of industry, a particularlynecessary action in a country that has not yetcompletely integrated industrial activityamong its values and culture. The third ele-ment of the French Government’s policy wasto create conditions for people to accept morereadily changes in their work environment andsocial structure (caused by the introduction ofnew technologies), basically through a renewedsocial participation. This was viewed as a ne-cessity for the continuous introduction of newtechnology. The fourth element consisted ofthe introduction of mechanisms to increase in-dustrial investment. One of the major mech-anisms for attracting investment in major in-dustries has been the nationalization of majorFrench industrial firms.

‘a’’ French Technology Preparing for the 21st Century,” Sa”en-tific Ame~”can, November 1982, p. F3.

‘gRobert Chabbal, “The New Investment in Science and Tech-nology in France, “ in Thomas Langfitt, Sheldon Hackney, Al-fred Fishnay, Albert Glowaske (eds,), Partners in the ResearchEnter@”se, Um”versity-Corporate Relationsin Su”ence and Tech-nology (Philadelphia University of Pennsylvania Press), 1983,p. 138.


The first step taken by the Mitterand gov-ernment was to establish a ministerial depart-ment for research and technology. One yearlater, the three ministries of research and tech-nology, industry, and energy were combinedto create one very large ministry called theMinistere de la Recherche et de l’Industrie.This ministry was created in part to stimulateinteractions and exchanges between govern-ment and industry.

Another measure taken by the French Gov-ernment was the nationalization of major in-dustrial firms and most of the banking sector.Nationalism was viewed as a means of control-ling investment and ensuring that the govern-ment could exert economic leverage to achieveits goals of: expanding employment; trans-forming the workplace environment; enhanc-ing French productivity and competition;directing research and development into areasof government priority; and recapturing thedomestic market by replacing imports with do-mystically produced products.8o Moreover,government ownership was seen as means toenable those companies to receive ample fund-ing for innovative, relatively high risk researchand development.

8 Nationalization was alsoO

viewed as a mechanism to increase coopera-tion and technology transfer between indus-try and government.

The nationalization program was imple-mented in several stages over a period of 2years.

82 Following several delays, the nation-

alization program was approved on February11, 1984, giving the government control of 5industrial groups, 39 banks, and 2 financialorganizations. 83 In the information technolOgysector, almost every major company has beenreorganized to reflect a majority of govern-ment ownership. For example, the governmenttook over the central organizations (but notnecessarily all the subsidiaries) of the Com-

‘OMichael H. Harrison, “France Under the Socialists, ” Cur-rent History, April 1984, p. 155.

*“’Pitfalls in France’s Vast R&D Plan, ” Business Week, Nov.23, 1981, p. 94.

82Companies were actually nationalized on Feb. 11, 1982 (Law82-155); however, the reorganization plans were not in effectuntil January 1983,

‘3Michael H. Harrison, “France Under the Socialists, ” Cur-rent History, April 1984, p. 155.

pagnie Generale d’Electricity, Saint GobainPont-a-Mousson, Pechiney-Ugine-Kuhlman,Rhone-Poulnec, and Thomson-Brandt. Thegovernment also acquired majority shares inDassult and Matra, and later negotiated con-trolling or full ownership of three foreign-owned companies in France, Roussel-Uclaf,Cii-Honeywell-Bull, and I.T.T. France.

The actual effects of nationalization on theinformation technology industry and on theFrench economy as a whole are still uncertain.This uncertainty is created in part by the dom-inant role the French Government has playedin the French economy throughout all of thepostwar period, and in part by the confusionin the transition to Socialist industrial policy.Moreover, the effects of nationalization havealso been obscured by France’s economic de-cline in the first 2% years of the Socialist gov-ernment, which in turn has ultimately weak-ened the nationalization efforts to reshape orcontrol the activities of specific firms or cer-tain sectors of the economy.Nationalizationefforts have also been complicated by changesin three different ministers of the Ministerede la Recherche et de l’Industrie within a 2year period.

Most significant to the general Mitterandstrategy for the development of technologywas the enactment of the legislation, Law forProgramming and Orientation for Research,which established scientific and technologicalresearch and development as national priori-ties. Moreover, the law ensured funding forlong-term scientific efforts by stipulating bothquantitative and qualitative objectives for thefollowing 5 years. The law stated that between1981 and 1985 the percentage of the Gross Na-tional Product (GNP) devoted to research anddevelopment will increase from 1.8 to 2.5 per-cent, thereby increasing by 40 percent thespending for technological development in 5years.

85This process began in 1980 and there

“1bid.“Robert Chabbal, “The New Investment in Science and Tech-

nology in France, “ in Thomas Langfitt, Sheldon Hackney, Al-fred Fishnay, Albert Glowaske (eds.), Partners in the ResearchEnterprise, Um”versity–brporate Relations in Science andTechnology (Philadelphia: University of Pennsylvania Press),1983, p. 140,

250 “ Information Technology R&D: Critical Trends and Issues

have been 12 percent annual increases in eachof the three successive budgets.86 The FrenchGovernment’s contribution will be matched byindustry, which calls for industry to increaseits expenditure for research and developmentby 40 percent in the next 5 years.87

In addition to meeting broadly the scientificand development needs, these increased ex-penditures are also for financing nationalmobilizing programs that are focused on in-dustrial and government targeted priorities,such as information technology. Althoughthere are major programs in biotechnology,new sources of energy, machine tools, etc., theinformation technology program is perhapsthe most ambitious. La Filiere Electronique,implemented by the Mitterand government in1983, is designed to coordinate and stimulategovernment, university, and industry informa-tion technology research and development ef-forts in order to move France into the forefrontof advanced information technology R&D andproduction.88 Figure 49 illustrates some of thecoordination efforts between industry andgovernment for La Filiere Electronique pro-gram. The 5-year infusion of ƒ 140 billion (ap-proximately $18.7 billion) for R&D is expectedto accelerate the production of informationtechnology products by 3 to 9 percent eachyear, produce a surplus trade balance in infor-mation technology products, and create 80,000new jobs.

Fourteen national projects for research anddevelopment are outlined in La Filiere Elec-tronique program:

● large scientific and industrial Frenchcomputer,

• building blocks of mini- and micro-com-puting,

● consumer electronics systems,86This increase occurred after 10 years of decreasing spending

for scientific equipment and therefore, the need was extremelyacute.

sTRobert Chabbd, “The New Investment in Science ~d T~h-nology in France, ” in Thomas Langfitt, Sheldon Hackney,Alfred Fishnay, Albert Glowaske (eds.), Partners in the Re-search Enterprise, Um”versity and Corporate Relations inScience and !l’echnology (Philadelphia: University of Pennsyl-vania Press, 1983) p. 140.

“A “filiere” in France is a targeted industry grouping or othergoal around which a government plan for funding, production,investment, education and dissemination assistance has beendeveloped, There are currently six filieres in France today:

●

●

●

●

●

●

●

●

●

●

●

display technology,ergonomics of computerization,computer-assisted instruction,multiservice communications,widebank communications network,design and production assisted by verylarge scale integrated circuits,computer-assisted engineering design andproduction,voice-processing module,electrophotographic module,electronic editing, andcomputer-aided translation.

In addition to these projects and programs,the plan sets forth mechanisms for state/in-dustrial cooperation and outlines efforts whichare aimed at alleviating other constraints oninformation technology research and develop-ment not related to direct R&D funding. Theseinclude manpower programs to correct a per-ceived shortage of engineers and techniciansand government-sponsored market promotionefforts .89

Because of economic difficulties, there havebeen some funding problems for the Mitter-and government. Consequently, the plans foran overall increase of 4.5 percent funding forresearch activities have suffered significantly.For most areas of research, this reduction infunding means that the 1984 research spend-ing will remain at the same level as in 1983.However, the government has stressed thatit will maintain its commitment to increasefunds for high priority areas, including infor-mation technology.90 Other impediments to LaFiliere Electronique program and to Frenchinformation technology research and develop-

— . . —robotics, electronics, energy, biotechnology, work environments,and cooperation with developing countries.

89For example, the number of people with Level 1 qualifica-tions in information technology (a French Masters degree, ap-proximately equal to an American Ph. D.) is expected to fallshort of needs by 70,000 for the period 1981-1990 in France.In the French context, this number is quite large; in 1979 itwas estimated that 105,000 scientists and engineers were ac-tively involved in all aspects of French science (energy, phar-maceuticals, mechanics, aeronautics agriculture etc., as wellas information technology). Jean-Pierre Letouzey, ScientificMission, Embassy of Fran~ Sta&ment for the American ASSO

ciation for the Advancement of Sciences, Mar. 24, 1983, p. 9(unnumbered).

‘David Dickson, “Hard Times Force France to Cut Back Am-bitious Plans to Support Science, ” Chrom”cle of Higher Educa-tion, Apr. 21, 1984, p. 10.


ment may also stem from the proliferation ofnew projects which may in turn dilute avail-able funds to inconsequential levels.

Social Environment for FrenchInformation Technology Research

and Development

The French Government’s desire to pushFrance into a technologically based future hasencouraged information technology researchefforts in office automation, microcomputers,consumer electronics, and telephone terminalequipment. In addition, each of these researchareas has included a major effort in the humanfactors aspects of design and interaction, per-haps in recognition of the difficulty expectedwith the assimilation of these technologies intoFrench society. The emphasis placed upon hu-man factors engineering has given rise toFrench prototypes and products that are gen-erally esthetically pleasing and easy to use.Many French designs have been adapted elsewhere. For example, it was a French studythat suggested amber on black CRT screenswere the most pleasing and produced the leasteye strain.

Three French characteristics stand out asimportant with respect to information technol-ogy research and development. Risk taking inthe French industrial sector does not appearwith the frequency or at the level consideredcommonplace in the United States. One indica-tion of this is the small number of venturecapitalists in France and the unwillingness ofthe traditional banking industry to fund en-trepreneurs. However, there are efforts on thepart of government and industry to increasethe use of venture capital.

For the conduct of information technologyresearch and development, the French risk-avoidance characteristic may be translatedinto the general lack of leading-edge, oftenhigh-risk, technological research. Another pos-sible consequence for information technologyresearch and development is the prevalence oflarge organizational environments for the con-duct of R&D. It is difficult to judge what im-plications this has for French information

technology research and development. It canonly be contrasted to the fact that one of thestrong aspects of U.S. information technologyresearch and development has been consideredto be the infusion of small, innovative firmsinto the research community.

The second French characteristic which ap-pears to affect the conduct of information tech-nology research and development is the fre-quency of what might be termed reengineer-ing. Reengineering, or the production fromscratch, of French versions of an existing prod-uct is very common in French informationtechnology. Reengineering efforts maybe at-tributed to the French attempt to develop do-mestic production capabilities in order to mit-igate United States and Japanese dominancein some niches of the French information tech-nology market.

The French attention to reengineering isquite possibly related to a third French char-acteristic, strong national loyalty to French-made products. This adherence to Frenchproducts occasionally provides a serious hand-icap to French information technology re-search and development. Because the Frenchdo not manufacture all types of state-of-the-art instrumentation, French scientists andengineers (who in some cases maybe restrictedto purchasing French instrumentation) maybe limited in their research activities.

Government

There are a variety of French Governmentorganizations involved in information technol-ogy research and development. A few werenewly created with the advent of the Mitter-and government, but most have been oper-ating for decades. Although the nationaliza-tion of industries and the governmentprovision of research and development underMitterrand are often thought of as socialistgovernment actions, the link between Frenchpolitics and French research has been long-standing. Traditionally, France has closelyoverseen both basic and applied science re-search through government mechanisms.


Centre National de laRecherche Scientifique (CNRS)

The largest and oldest French Government-funded research organization is the Centre Na-tional de La Recherche Scientifique (CNRS),founded in 1939. CNRS supports basic re-search in chemistry, physics, earth, atmos-pheric and ocean sciences, life sciences, engi-neering, social sciences, mathematics, andhumanities. In 1984, CNRS had a budget ofƒ 7,735 million (almost 24 percent of the pub-lic civil research budget), and employed ap-proximately 25,000 people in 1,350 labora-tories or within universities, other governmentagencies, and industry. CNRS has seven labor-atories devoted to basic research in someaspect of information technology. The portionof the 1984 budget applicable to informationtechnology research and development isƒ 225.2 million, an increase of 13 percent overthe 1983 budget. This increase in the budgetfor the information technology research (incontrast to other CNRS research areas whichreceived less or no increases) is significant andis largely the result of the recognition of theimportance of information technology to theFrench economy and society.91

In line with France’s new efforts in strength-ening its information technology industry,CNRS information technology activities arecurrently coordinated with the Filiere Elec-tronique program and are becoming largely devoted to applied, industrial research. There areefforts in software development and program-ming techniques, speech recognition and syn-thesis, artificial intelligence, and robotics.CNRS draws heavily from the French univer-sity system for its personnel, unlike industryor most other government agencies, where thegrandes ecoles supply the researchers and ad-ministrators.92

In the past, CNRS ties with industry havebeen relatively weak. However, as CNRS in-creases its applied research activities, ties with

——“Discussion of CNRS based on “French Technology Prepar-

ing for the 21st Century, ” Scientific American, November 1982,pp. F4, F11.

gZ1ntimiew with Ckles Ga.rriques, President, Agence de l’ln-formatique, June 24, 1983.

industry have also grown. Closer ties betweenCNRS and industry seem to be important tothe Directeur General of CNRS who has re-cently established several agreements betweenCNRS and industry .93

CNRS has exchange programs and scientificaccords with 30 countries, including agree-ments with the National Science Foundation,the National Institute of Health, Massachu-setts Institute of Technology, the Universityof Chicago, and others.

Ministere des Postes, Telecommunications,et Telediffusion (PTT)

The French Ministere des Postes, Telecom-munications, et Telediffusion (PTT), throughthe Ministere de la Recherche et de l’Industrie,is responsible for the provision of all telecom-munications services and equipment, networkmaintenance, standards development, tele-communications policy, technical assistanceto former French colonies and other foreign en-tities, and research and development. ThePTT’s jurisdiction over telecommunicationspolicy has resulted in increased PTT involve-ment in information technology R&D. ThePTT often joins the Ministere de la Rechercheet de l’lndustrie and others in the funding ofprojects that cross the technological bound-aries between telecommunications and com-puters.94

The reach of the PTT through the Ministerede la Rechereche et de l’Industrie into the in-formation technology research and develop-ment communities is extensive. The PTTfunds the Centre National d’Etudes Telecom-munications (CNET) much in the same man-ner that AT&T funds Bell Labs. The PTT alsofunds the Institut National des Telecommu-nications (INT), a portion of the Agence del’Informatique (ADI), and all of the Ecole Na-tionale Superieure des Telecommunications

‘sFor example, agreements have been signed with Saint-Gobain, Renault, and Roussel-Uclaf. “French Technology Pre-paring for the 21st Century,” Scientific American, November1982, pp. F4-F1l.

“For example, le Project Pilote NADIR (exploration of newuses of Telecom 1 for data and voice transmission) is funded50/50 by the ITT and the Ministere de la Recherche et de l’ln-dustrie.


Photo credit: Scientific Mission, Embassy of Franceand Ministere des PTT, French Government

Videotex terminal “Minitel” with CCP card reader.

(ENST). The PTT’s terminal equipment, switch-ing and transmission needs are suppliedthrough contracts with CIT Alcatel, ThomsonCSF, the French Cable Co., and a host ofothers. The R&D for the products manufac-tured by these firms for PTT use is funded bythe PTT, generally through cooperative effortsbetween the industrial entity and the CNET.

The current government-sponsored push ininformation technology research and develop-ment follows upon a recent similar effort intelecommunications implemented by the PTT.By most accounts, the research, development,and implementation programs designed in1974 and 1975 to transform the French tele-communications system into a technologicallyadvanced network on par with the United

States has been a success.95 In 1974, Franceaveraged 12 main telephone lines per 100 in-habitants; in 1981 the figure was 33.96 Thenumber of lines in the national network grewfrom 7 million in 1975 to 20 million in 1982.In 1975 electronic exchanges were virtuallynonexistent in France; in 1981, 70 percent ofnewly installed switching capacity was elec-tronic. 97

As a result of the recent modernization ofthe French telephone network, the Frenchhave a greater percentage of digital equipmentthan any other country.ga This, in turn, hasspawned the provision of many sophisticatedservices such as Transpac (public data packetswitching network), Transmic (dedicated datatransmission), Teletel, and Videophone (videoconferencing). These technological possibilitieshave pushed PTT-sponsored telecommunica-tions and computer research and developmentin several directions. The three main directionshave been satellite technology (the Frenchlaunched Telecom 1 in the summer of 1983),an integrated services digital network (ISDN),and optical fibers (a wideband, multiservice op-tical subscriber network experiment is takingplace in 1,500 homes in Biarritz). The majorresearch efforts in all of these areas have takenplace at the Centre National d’Etudes des Tel-ecommunications (CNET).

Centre National d’EtudesTelecommunications (CNET)

The CNET was formed in 1944 to providescientific research and technical assistance tothe PTT.99 The CNET is active in appliedmathematics, computer science, solid-statephysics, and earth sciences. The total budgetin 1982 was about ƒ 1 billion (about $133 mil-lion); approximately 55 percent is spent on net-

““No Hang Ups for French Phones, ” (ZWecom France, June1982, p. 10.

‘Conseil Econornique et Social, La ‘1’elernatique et L ‘Amen-agement du Z’erritoim, Apr. 21, 1983, p. 35.

97P’IT Telecommunications, Bim”tz, The b“ghtwave Commu-nications World of the Future, p. 1.

‘81bid.‘gDiscussion based on the group of brochures and pamphlets

included in the CNET Dosw”er Presse.


work and service engineering, 25 percent oncomponents engineering and 20 percent on ba-sic research.

In 1982 2,600 scientists and engineersworked in six laboratories. Two of the groupsare located in Paris. Paris A performs long-term network planning and administers theother five centers. Paris B is the center of basicresearch for the CNET. Materials and geophy-sical research are the primary efforts. Appliedwork in components, transmission, marinecables, and satellites is also performed.

The two centers at Lannion work on localarea networks, ISDN, software, human-ma-chine interface, and acoustics (Lannion A) anddigital transmission, optical communications,and components (Lannion B). The center atGrenoble is devoted to microelectronics re-search, while the laboratory at Rennes isshared with the Centre Commun d’Etudes deTelevision et Telecommunications and studiesfuture telecommunications services and theirintegration with broadcasting technologies.

Since the advent of the Mitterrand govern-ment, the CNET has assumed a significantrole in the French national industrial develop-ment strategy. Under the plan for the electron-ics sector (La Filiere Electronique), the CNETlaboratory at Grenoble has undertaken proj-ects in CMOS technology for very large scaleintegrated circuits, gallium arsenide, comput-er-aided design, and artificial intelligence.

In conjunction with La Filiere Electroniqueresearch effort, CNET has introduced a pro-gram to improve its transfer of technology tothe industrial sector. The CNET owns over550 patents which it licenses to French andother companies. The licensing is done virtu-ally without regard to the royalty potential;CNET’S 550 patents provide approximatelyƒ 20 million in revenue per year.

L’Institut National de Rechercheen Informatique et Automatique (INRIA)

L’Institut National de Recherche en Infor-matique et Automatique (INRIA) is one of thenewest French Government information tech-nology research agency. It was formed in De-

cember 1979 under the d’Estaing Ministry ofIndustry. INRIA remains under the jurisdic-tion of the Ministere de la Reserche et de l’ln-dustrie in the Mitterrand government. Thename of this ministry has had several evolu-tions of late. With the advent of the Mitter-and government the Ministere de l’Industriewas changed to the Ministere de la Rechercheet de l’Industrie. Currently, the organizationis titled the Ministere de l’Industrie et de laRecherche. The first change effected by theMitterrand government was an effort to com-bine the mission of the Ministry of Industrywith that of the Ministry of Research andTechnology. The second change appears to beone of emphasis. INRIA is considered the lead-ing research institute in computer science inFrance. It has several locations. The main re-search center is in Rocquencourt (just outsideof Paris); another smaller center is located inSophia-Antipolis. INRIA shares facilities withthe CNET at Rennes and Grenoble and has asmall group in Toulouse. In 1982, INRIA’sbudget was ƒ 146 million (about $19.5 million)which funded 409 people, 225 of whom werescientists and engineers. In 1983, the budgetwas expected to be ƒ 200 million for fundingof INRIA contracts with industry.l00

INRIA has a three part mission: the conductof research on experimental computer sys-tems; international scientific relations; and thetransfer of technology. Each of the missionsis guided by industrial needs, at least insofaras those needs are articulated by the Ministerede la Recherche et de l’Industrie. Consequent-ly, the research performed by INRIA is ap-plied and the bulk of the work can be char-acterized as product development.

INRIA has eight research programs in areassuch as system architecture, languages, algo-rithms, automation, and man-machine inter-face. Each program conducts three to fiveprojects. In addition INRIA is responsible forfour of six pilot projects to be undertaken inconnection with La Filiere Electronique pro-gram. Those pilot projects are KAYAK (office

IOOINRIA, Dossier Presse.

.

256 “ information Technology R&D: Critical Trends and Issues

automation), NADIR (applications of Telecom1 satellite capability), SIRIUS (distributedsystems), and SOL (portable software).

Each pilot project has a different configura-tion of funding, personnel, and industry par-ticipation. For example, NADIR is financed50/50 by the PTT and the Ministere de la Re-cherche et de l’Industrie. The project admin-istrative responsibility is shared between theAgence de l’Informatique (ADI) and the Direc-tion des Affaires Industrielles et Interna-tionales, a division of the Direction Generaledes Telecommunications within the PTT. Theactual research is conducted at INRIA witha mixture of INRIA and CNET personnel. In-dustry personnel are not research team mem-bers, although several industrial representa-tives are participating through provision ofuser specifications.l0l

SIRIUS is an older project, and as such, itsoriginal funding sources and location havechanged. Currently, the project is adminis-tered by ADI. The research team includesINRIA, the University of Nancy, severalsmaller divisions of the Ministere de la Re-cherche et de l’Industrie and 15 industrialcompanies including CAP-Sogeti, Cii Honey-well Bull, and SNCF (the French nationalrailroad).1°2

The organization of INRIA is quite differentfrom the CNET and in many aspects providescomplementary research support. Unlike theCNET which draws heavily from the grandesecoles system for its researchers, INRIA re-cruits from the French university system.Many project directors at INRIA are Frenchuniversity professors who come to INRIA forthe duration of a project.

Also unlike the CNET, whose mission is tobe the research arm of the state telephone net-work, INRIA is not responsible to one central-ly defined set of research requirements. Rath-er, the institution generally must respond toa more diffuse set of requirements from indus-try. The coordination of research projects is

nominally the job of the Ministere de la Re-cherche et de l’Industrie, but most often, suchcoordination is effected by INRIA researchstaff members who take ideas forward to theMinistry for funding.

Agence de l’Informatique (ADI)Like INRIA, the Agence de l’Informatique

(ADI) was recently formed. Funded by theMinistere de la Recherche et de l’Industrie (75percent) and the PTT (25 percent), ADI’s 1981budget was F 300 million ($40 million). Fund-ing in 1982 was F 320 million as was the 1983budget. A 20 percent cut is expected for 1984.ADI was originally designed to be self sustain-ing though royalties from its research-derivedpatents and the sale of its published studies,but as yet, those items account for less than1 percent of ADI’s revenue. ADI has three ma-jor areas of activity: research and experimen-tation, application development and dissemi-nation, and training and education. It also hasthree support activities: regional programs, in-ternational affairs, and economic and legal stu-dies. Sixty professionals manage these pro-grams.103

The goal of ADI’s training and educationprogram is to produce more computer sciencegraduates and to improve the quality andavailability of science and engineering educa-tion. Included in the program is a project toput computers in the secondary schools. Theapplication development and disseminationprograms provide a forum for users and pro-ducers of information products to exchangeideas and develop computer applications tai-lored to the specific requirements of variousbusiness sectors.

ADI’s research program funds efforts incomputer science in a manner analogous tothat of the National Science Foundation; thatis, ADI funds but does not conduct the re-search. This program has approximatelyF 100 million ($13 million). In addition to thefour pilot projects ADI helps fund at INRIA,

l’JIThe pilot ~Oj~t NADIR English Language bulletin, May1983.

IOZINRIA, ~ggier Pregse Le Project Pilote SIRIUS.

‘OgAgence de l’Informatique, English language brochure, andinterview with Charles Garriques, President, ADI, June 24,1983.


partial funding is provided for the RHIN (sys-tem interconnection) and SURF (functional se-curity) projects. These six pilot projects wereoriginally to be undertaken as only a portionof ADI research mission. Longer-term proj-ects in new architectures, languages and pro-gramming, human-machine interface, designaids, automation, computer-aided translation,security, and translation are also planned. Re-cent budget cuts, however, have limited theseresearch endeavors.

Centre d’Etudes des Systemeset des Technologies Advancees (CESTA)

The Centre d’Etudes des Systemes et desTechnologies Advancees (CESTA) is the new-est French agency involved in informationtechnology. Founded in January 1982, it is oneof two completely new agencies formed by theMitterrand government in information tech-nology. Under the jurisdiction of the Minis-tere de la Recherche et de l’Industrie,CESTA’S 40 employees have two missions:technology forecasting and identification ofemployment impacts due to technological ad-vance. The technological scope of CESTA goesbeyond information technology (e.g., there areprograms in biotechnology) but the majorityof its work involves various aspects of infor-mation technology. CESTA’S forecasting re-sponsibility takes the form of evaluations ofmarket, cultural, and social acceptance of tech-nologically advanced products. When employ-ment impacts can be identified, retraining pro-grams are developed by CESTA personnel.CESTA’S major activities undertaken to fur-ther these goals are the conduct of seminarsand the commission of studies and papers ontopics of interest.l04

Although CESTA resides under the aus-pices of the Ministere de la Recherche et del’Industrie, the director described his agencyas independent, not administration-linked anddrew an analogy between CESTA and the Ten-nessee Valley Authority. He saw this inde-pendence as a necessary ingredient to his abil-

—.—I04CESTA, CESTA, December 1982, and interview with Yves

Stourdze, Directeur CESTA, June 20, 1983.

ity to gather groups representing divergentinterests.

Centre Mondial Informatiqueet Ressource Humaine

The Centre Mondial Informatique et Res-source Humaine is the other information tech-nology agency created by the Mitterand gov-ernment. The original mandate of the Centerwas threefold: research in those areas of infor-mation technology applicable to microcomput-ers; social experimentation in France; andThird World pilot projects designed to explorecomputer applications for the disseminationof medical and education information.1°5 How-ever, the Center’s role in information technol-ogy research and development has not alwaysbeen clear. Since its inception in November,1981, the Center has been racked with politi-cal struggles. Exactly what role the Centerplays or may play in French information tech-nology research and development, however,still remains uncertain.l06

University

Two parallel systems of higher education ex-ist in France; one is found in the universities,the other in the grandes ecoles. Both systemsproduce scientists, engineers and administra-tors with relatively little training in the ap-plied sciences and/or the business aspects ofresearch such as marketing, management fi-nance, or accounting. Beyond this similarity,the systems have few parallels.

In general, French university training doesnot usually prepare individuals for careers ingovernment or in the higher levels of indus-try. University trained scientists are occasion-ally found in industry and in government re-search laboratories in instances where the

Iobcentw Mond,i~ Informatique et ~ssource Hllmtie, ~’chesD’Information, Statuts et orgm”sation.

1%3ee for example, Dray and Menosky, “Computers and aNew World Order, ” Technology Review, May/June 1983, andWalsh, “Computer Expert Signs Off From World Center, ” Sci-ence, vol. 218, Dec. 3, 1982.


work is decidedly theoretical or, as in com-puter science, the discipline is so new that agrande ecole exclusively for the subject is non-existent. Although computer science is taughtin several of the grandes ecoles, computer sci-entists, particularly software engineers, inFrance have emerged from the university com-munity. As a result, several French univer-sities such as the university of Grenoble, haverecently become sites for information technol-ogy research and development activitiesaround which industry has begun to collect.This may eventually change the nature ofFrench research and development which tra-ditionally has been institutionally split intothree components. In general, the universities(and some government organizations) haveconducted France’s basic research, the govern-ment and the grandes ecoles have been thesites of applied research, and industry hasbeen the environment for development activ-ities. Although these various institutions havecertainly funded activities in each of the othercomponents, cooperation among them hasbeen minimal.

The highly respected grandes ecoles system,which produces the French cadre of govern-ment officials and industrial managers, wasformed by Napolean to develop an elite groupof French intellectuals who would be respon-sible for guiding France’s cultural, social, eco-nomic, and political futures. Today there areabout 150 grandes ecoles in France of which

ximately 10 produce state engineers. Theapproothers produce administrators, economists,sociologists, artists, and a host of other pro-fessionals in the sciences, liberal arts, andhumanities. All grandes ecoles have entry re-quirements. One must take a competitiveexam which requires between 2 and 3 yearsof study preparation. Typically, the Frenchhigh school graduate will prepare for this examat the university, within the high school set-ting, or in private preparatory institutions.Based on their test scores, students are gen-erally assigned to specific grandes ecoles.

The Ecole Polytechnique has traditionallyproduced the highest level of government offi-cials in France. As the name suggests, the cur-

iculum at Ecole Polytechnique is based on atheoretical education in a variety of scientificdisciplines. In addition, students who wish tospecialize in an area often enter one of theother grandes ecoles for additional trainingafter graduation.

The Ecole Nationale Superieure des Tele-communications (ENST) is the primary grandeecole for the production of researchers andengineers in information technology. TheENST is funded by the Ministere des Postes,Telecommunications, et Telediffusion (PTT),and its curriculum is overseen by the PTT. Theschool is considered to rank among the top tenof the 150 grandes ecoles in France. The schoolaccepts 70 first year students annually, basedon performance on the competitive exam.More students are added during the secondand third years of the school’s program. En-try into the second year at ENST can be ob-tained after completion of another grandesecoles education and/or after graduation witha maitrise (roughly equivalent to a U.S. bach-elor of science degree) from the university.These additions result in a total student pop-ulation of about 540 and the award of 220 de-grees annually. ’07

The first and second years of study at ENSTinvolve mathematics, physics, electronics,computer science, economics, foreign lan-guage, and humanities. During the third year,a student chooses from eight areas of specialtyto study for half of the year. Three months arespent in a work-study project in industry orgovernment or in a foreign study program.These work-study projects are a unique oppor-tunity for students to gain applied or indus-trial experience. As a result of ENST’S empha-sis on applied engineering, the proportion ofgraduates who work in industry is quite highin comparison to other grandes ecoles. Ap-proximately 60 percent of the graduates ofENST go to industry, and the remainder findjobs in the PTT.

The ENST has four laboratories where in-structional research is conducted: systems andcommunications, computer science, electronics

‘O’ Ministere des PTT, EA?ST.


and physics, image and sound, and life sci-ences. Occasionally research is conducted incollaboration with government research cen-ters and industry; however, the main purposeof the research is to provide student instruc-tion, not to further the state-of-the-art.108

The highly theoretical nature of thegrandes ecoles training does not always pre-pare government officials to direct and con-duct research for purposes of French industrialgrowth. Experience in the industrial commu-nity does not appear to be an alternative meth-od of developing such preparation for govern-ment officials, as the transfer of people fromcareers in industry to careers in government(or vise versa) is quite rare. However, in con-junction with its goal of strengthening thelinks between industrial and government re-search and development, the Mitterand gov-ernment has implemented several new pro-grams. These include increasing the numberof commercially oriented grandes ecoles, theteaching of marketing, accounting, and fi-nance throughout the grandes ecoles system,and encouraging more work-study programs.

Industry

The industrial component of French infor-mation technology research and development,just as in the United States, has many mem-bers. Similar to the situation in both Japanand the United Kingdom, a few large informa-tion technology firms are responsible for themajor proportion of research and developmentefforts. Cii Honeywell Bull, Thomson CSF,CIT Alcatel, and Sogitec, representatives ofthe spectrum of French industrial experiencein information technology research and devel-opment, are described below.

Cii Honeywell Bull

When nationalization became operational atCii Honeywell Bull in January 1983, the firmwas completely reorganized. It took on severaldivisions of Thomson and Alcatel and re-arranged its internal divisions into four

—. ———‘081bid.

groups: Bull Systems manufactures main-frames; Bull Sems (purchased from ThomsonCSF) manufactures minicomputers; Bull Per-ipheriques makes disks and printers; BullTransac (purchased from Alcatel) produces mi-crocomputers and office automation products.The collection of groups is now called Bull. Thetransfer of the company to state ownershiphas not only changed the structure of divi-sions, and their personnel, it has changed therelationships between management and work-er to reflect socialist principles. The work weekhas been shortened, salary differentials be-tween men and women have been eliminated,and the number of upper management person-nel and their salaries have been reduced.log

Research at Bull has been reorganized intosix working groups, each with 30 scientistsand engineers: advanced systems research; in-tegrated circuits research and technology; cus-tom design of integrated circuits; standard in-tegrated circuits; interfaces for technologyuse; and discrete components and subsystems.The integrated circuits research and technol-ogy group is mandated to “be a proponent oftechnology alternatives to our partners” andto [provide] updated competence for the bestchoices in integrated circuit technology at theBull group level. ” Similarly, the custom in-tegrated circuits group is to “establish know-how in design of custom VLSI” and “providesupport and expertise in our choices of newtechnologies available outside the company. ”In addition to the six research groups, addi-tional studies to determine the feasibility ofresearch in languages, artificial intelligence,vector processing, and distributed architec-ture are underway.ll0

Bull’s future position in information tech-nology research and development is unclear.It is important to note, however, that shouldit not improve, the French efforts in informa-tion technology research and development,and indeed in other advanced technology

“Wi.i Honey-well Bull, Bilan Social d’Entrepn”se, Exercise 1982and “Avis du Comite Central d’Entreprise sur le Bilan Social,1982, ”

““Bull, Corporate Z%chnology, June 17, 1983.


areas, may be diminished. Much of France’stechnological future is dependent on the avail-ability of a wide range of sophisticated com-puters and computer peripherals. As long asthe French rely on Bull for these products, avery important link in the information tech-nology research and development chain mayremain uncertain.

Thomson CSF

Thomson CSF produces a variety of aero-nautic electronics equipment, telecommunica-tions equipment for the public telephone net-work, medical devices, electronic components,and office automation products. Thomson CSFis partially controlled (40 percent) by Thom-son-Brandt, a producer of durable consumergoods, electromechanical capital goods, lampsand lighting fixtures, and engineering and fi-nancial services.111 Thomson CSF produces,sells, and distributes its products through anetwork of almost 60 domestic and foreignsubsidiaries and holding and associated com-panies. 112 Its revenues are dominated by thesale of electronic equipment followed by tele-communications and medical devices.1

13

In 1981, Thomson CSF spent over F 4 bil-lion ($530 million), approximately 10 percentof Thomson CSF and Thomson-Brandt com-bined revenues, on research and development.Seventy-five percent of the effort was inter-nally financed. Thomson CSF performs R&Dfor both Thomson CSF and Thomson-Brandt,and its spending represents 25 percent of allR&D spending in France. The vast majority(95 percent) of the research is product-relatedand takes place throughout the company’ssubsidiary structures. Basic research takesplace in the Laboratoire Central de Re-cherches. 114

Basic research at Thomson CSF includesprograms in gallium arsenide, molecular beamepitaxy, single mode fiber optics, and machine

I I lThom90n.Brmdt, l%nc~pales ~fi”df?s et ParticipationsFrancm”ses et Etrangers.

“’Ibid.“gThomson-CSF, Rapport Annual des Activities en 1981,‘]41bid.

level languages. The basic research effort,while small, is extremely important to Thom-son CSF. The basic research effort providesThomson with entree into the basic researchcommunity in France and abroad (Massachu-setts Institute of Technology, Stanford Uni-versity, CNRS and several French universitieswere mentioned as important research collab-orators). Moreover, a credible basic researcheffort helps with the recruitment of scientistsand engineers. Because of the excellent reputa-tion of Thomson CSF’S basic research functionand the small portion of corporate funds it rep-resents, nationalization is not expected tochange the nature of the activities and mayeven increase funding.

Like Bull, Thomson-Brandt, was national-ized with the passage of legislation in Febru-ary 1982. The effect on R&D activities atThomson CSF has been minimal in compari-son with the changes at Bull. Product-orientedresearch has been modified slightly to meetsome specifications of La Filiere Electroniqueprogram and state management has causedsome difficulties, but personnel and directionalchanges for the company have caused almostinsignificant disruption.

CIT AlcatelCIT Alcatel is a subsidiary of the Compag-

nie Generale d’Electricity (CGE), the fifthlargest company in France. CIT Alcatel(through its 8 French and foreign subsidiariesand affiliates) represents the public telecom-munications division of CGE. Another 11Alcatel Group members produce office auto-mation and professional electronics productsand provide computer services. Nationaliza-tion has not changed the personnel or the con-duct of research at CIT Alcatel. It is antici-pated that state ownership will improve therelationships between academia and industrialresearch and development, although resultsare not expected for another 10 years.lls

CIT Alcatel’s research and development ac-tivities are scattered throughout the com-pany’s subsidiaries. In addition, some research

I IbAlca~l, The Alcatel Group.


activities take place in association with CGE,CNET, and CNRS, and in conjunction withother French (e.g., Thomson CSF) and U.S.companies (e.g., SEMI Processes, Inc.). Thecompany’s director estimated that between 10and 12 percent of the employees (some 17,065in 1981) were involved in R&D116

Each product division within the CIT Alca-tel companies funds the R&D activities itdeems needed. Like Thomson CSF, the basicresearch at CIT Alcatel is performed at a cen-tral laboratory that is shared with all mem-bers of the CGE Group. Approximately 6 to8 percent of CIT Alcatel’s total research budg-et is devoted to this central laboratory.117

Sogitec, S.A.Unlike the majority of French information

technology firms, Sogitec is not a state-ownedcorporation and is representative of a small in-formation technology firm. Founded in 1964by its president, Christian Mons, Sogitec hasgrown to employ about 550 people in three lo-cations: Paris, Rennes, and Lakewood, Califor-nia. Sogitec also has sales offices in New Yorkand Washington, D.C. Its sales growth hasbeen impressive; 1978 revenues were doubledby 1979. Customers include General Electric,

“’Ibid.“’Ibid.

McDonnell Douglas, Ford, Daussalt, andothers in the aircraft, shipbuilding, and auto-mobile industries.118

Sogitec has two divisions that are designedto meet individual user needs. The data proc-essing services division provides softwarepackages and the related hardware for full textdocumentation, storage, and retrieval. Theother Sogitec division produces real-time simu-lators and simulation packages for aircraft,helicopters, land vehicles, and ships. TheFrench military has purchased Sogitec prod-ucts for combat pilot training in the FrenchMirage fighter bombers. In addition, Sogitec’ssimulation expertise is currently beingadapted for film production, television com-mercials, and animation.

Some notion of Sogitec’s research intensitycan be seen in its distribution of personnel.Over 250 of Sogitec’s 550 employees are scien-tists or engineers involved in research and de-velopment (50 are located in the UnitedStates). Eighty percent of the researcherscome from the grandes ecoles system, the re-mainder from French universities.ll9

“’Department of State, “WTDR on Sogitec (SIC) Data Sys-tems,” Oct. 10, 1983, p. 3.

“gIbid.

The United KingdomHistorically, the United Kingdom has heav-

ily relied upon industrial manufacturing for itseconomic well-being. As the post-industrialsociety arrives with the decline of manufactur-ing as a primary economic activity, the Brit-ish economy has suffered.120 The United King-dom’s share of world trade in manufacturedgoods in the decade 1963-73 fell from 15 to 9percent. For the first time in history theUnited Kingdom now appears to be approach-

‘z”Daniel Bell, The Coming of Post-Industrial Society (NewYork: BASIC Books, Inc., 1976).

ing a trade deficit in manufactured products.Moreover, the U.K. ’S needs for various infor-mation technology products are completely orsubstantially met by imports. As a result ofthe importance of exports to the U.K. econ-omy and the increasing importance of infor-mation technology to the world economy, boththe U.K. Government and industry have con-cluded:

Our basic economic situation dictates thatwe must become a net exporter of high tech-nology, high value-added products; informa-tion technology is a prime example of this.


Moreover, unless our [the United Kingdom]information technology industry achieves astrong world competitive position, then theefficiency of our other industries in the man-ufacturing services will suffer. Their capac-ity to be advanced users of information tech-nology has a close synergy with the level ofthe information technology industry itself.121

Traditionally, the United Kingdom has beenone of the world’s leaders in basic scientificresearch, particularly in the physical and bio-logical sciences. Britain holds the highest percapita ratio of Nobel Prize winners in the sci-ences which is more than double those for anyof the other industralized nations. Althoughthe United Kingdom has in the past had thelargest R&D expenditure percentage of grossnational product (GNP) outside of the UnitedStates, the United Kingdom has been tradi-tionally weak in converting and applying itsbasic research efforts to the production andmarketing of new products and services.

This difficulty of transferring scientificknowledge to commercial applications has con-sequently created serious problems in the in-dustrial sector and in the international com-petitiveness of British-made goods. Thisfailure to capitalize on basic research effortsis documented in a recent publication of theCentral Office of Information, London, “Brit-ish Achievements in Science and Technology”:

. . . there have been an array of ‘firsts’ thatapply to the information technology area, in-cluding radio navigation, computers, opticalfibers, liquid crystal displays, and flat screentelevisions, yet in none of these areas is aBritish manufacturer a principal supplier. ’22

Thus the current challenge in the United King-dom is to bridge the gap between the creativebasic research in information technologies andthe relatively weak state of industrial applica-tion of these technologies in order to create anenvironment which is more innovative in pur-suing both domestic and international mar-

‘*’The Department of Trade and Industry, “A Prograrnmefor Advanced Information Technology, ” the Report of theAlvey Committee (London: Her Majesty’s Stationary Office,1982), p. 14.

‘22Central Office of Information, London, British Achieve-ments in Science and Technology, April 1981, No. 11 l/RP/81.

kets. This is particularly important in the areaof information technology where the develop-ment of user applications of the technologiesfor the provision of new and innovative serv-ices is vital to the marketing of informationtechnology.

Reflecting the need to increase applied re-search activities, the current U.K. Govern-ment has taken measures in the opposite direc-tion of the traditional high level of governmentintervention in both industry and social pro-grams. The more conservative government’sefforts include privatizing the economy and re-storing entrepreneurial initiatives, thus at-tempting to make industry more independentof government and reducing government in-volvement in the marketplace. Governmentstrategies include privileged credit, deregula-tion measures, and tax benefits to encouragethe growth of small and medium businesses.

Privatization measures also entail exposingstate-run monopolies to outside competition.A major example of this introduction of com-petition is the termination of the monopoly ofBritish Telecom (BT) by granting a license toa major competitor to operate an alternativenational telecommunications network. It ishoped that this recently introduced competi-tion into the provision of telecommunicationsservices will stimulate demand for advancedtechnology transmission and exchange equip-ment. In conjunction with the privatizationmeasures in the field of information technol-ogy, the U.K. Government is also promotingthe marketing or applied aspects of researchand development by shifting some of its R&Dexpenditure from government research estab-lishments to the private sector. In Cambridgefor example, where a preference for basic re-search has long prevailed, anew view towardsresearch is emerging:

What has changed is the idea that moneyis dirty and that one must do pure science.Now, as in Cambridge, Mass., or in Califor-nia, people are not adverse to doing researchthat could be put to commercial use. Thischange in the work ethic has helped us.123

IZ3p~p Ld?cmnkr, “Is Britain Reviving?” World Press Review, September 1983, p. 31.


Because of radical changes in governmentpolicy over the past decades in the UnitedKingdom, the continuity of policy which has,for example, aided Japan in its post-war recov-ery, has not been there to support industry inits efforts to expand. This complicates anysimple characterization of the U.K. Govern-ment role in industrial policy and suggests adynamic situation currently regarding U.K.Government initiatives in information technol-ogy. Consequently, it is difficult at times todisassociate government, industry, and uni-versity roles in information technology R&D.However for purposes of comparison withUnited States, Japanese, and French R&D ef-forts, government, university, and industryR&D environments are described separatelybelow. Before discussing these environments,it is important to understand the size of theU.K. participation in information technologymarkets.

The Size of U.K. Participationin Information Technology Markets

While the United States and Japan are netinformation technology exporters, the UnitedKingdom is a net importer of information tech-nology products by F 300 million ($420 mil-lion) annually.’” In 1980, the U.K. industrycaptured approximately 50 percent of its ownF 2.1 billion ($2.94 billion) information tech-nology market. Moreover, the U.K. informa-tion technology industry captured 3.8 percentof the world information technology marketin1980.125

Because of the relatively small size of its na-tional markets, the U.K. information industryhas been somewhat inhibited. Moreover, it hashad difficulty in generating significant exportmarkets for its information technology prod-ucts to balance its heavy imports. For exam-ple, of the 19 top semiconductor companies inthe world, which account for approximately

IZ4AlI U.K. pund fi~res are converted into U.S. dollars ac-cording to foreign exchange rates as of Aug. 1, 1984, where L 1= $1.4.

‘25”A Strategy for Information Technology, ” National Enter-prise Board, 1981, p. 19.

75 percent of the world market, not one is fromthe United Kingdom.

Government

The U.K. Government officially recognizedthe importance of the development of informa-tion technology to the British society, indus-try, and economy by requesting the AdvisoryCouncil on Applied Research and Develop-ment (ACARD),126 which advises the PrimeMinister and the Cabinet Office on importantdevelopments in advanced technology, to ad-dress the following questions:

● Should development and application of in-formation technology in the United King-dom be stimulated?

● Are there constraints on British industrywhich supplies and applies informationtechnology equipment, software, andsystems?

The resultant 1980 ACARD report “Informa-tion Technology” has contributed to presentpolicy formulation in relation to informationtechnology, with an emphasis on the applica-tion of information technology as a key ele-ment in the future industrial and commercialsuccess of the United Kingdom as well as onthe potential significance of information tech-nology for both society and individuals.

The ACARD report recommended the fol-lowing:

1.

2.

3.

4.

— —

One minister and one government de-partment should be wholly responsiblefor information technology.One government department should beresponsible for the regulation of commu-nications and broadcasting.There should be a government commit-ment to information technology.The government should actively pro-mote and publicize British informationtechnology.— ——

‘z’Probably the nearest United States counterpart to ACARDis the Committee of Advisors, recently created by Dr. Keyworthin the Office of Science and Technology Policy in the WhiteHouse. ACARD, whose members are experts drawn from in-dustry, governmen~ and academia, offers specific recommen-dations for government action.

38-802 0 - 85 - 18


5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

Central government, local government,and the nationalized industries shouldapply information technology vigo-rously.The Post Office should provide aworldwide network.Education, training, and guidanceshould be accentuated in informationtechnology.Links between users and suppliers of in-formation technology equipment shouldbe improved.Strong British teams should participatein international fora on regulations andstandards.Legislation should be introduced to pro-vide for better protection of data, andother legal reforms will be required.The Post Office monopoly on use of itsservices should be ended.Public purchasing should be used to“pull through” development of equip-ment.The Science and Engineering ResearchCouncil and the Department of Indus-try should promote research and devel-opment in information technology.All publicly funded information technol-ogy research and development should becoordinated. This would involve Depart-ment of Trade and Industry, Ministryof Defense, the Post Office, and the Sci-ence and Engineering Research Council.

The United Kingdom has taken a wide-scopeapproach to encourage the development andapplication of information technology in Brit-ish industry and society. In response to theACARD report, Mrs. Thatcher’s governmenthas developed an overall strategy for informa-tion technology. The main objectives of theU.K. Government’s policy for informationtechnology are:

1.

2.

The development of a statutory regula-tory framework favoring the growth of in-formation technology products andservices.The development of new products andtechniques through direct research and

34

4.

development support and enlightenedpublic purchasing.Action to make individuals more awareof what information technology offers andso enable them to take advantage of thenew information technology products andservices.The provision of a national telecommuni-cations network capable of stimulating,and meeting, demands for new services.

For implementation of these goals, 1982 wasdesignated as IT Year in the United Kingdom.A wide range of promotional aids was used toincrease the awareness of the general public,industry, and schools, and main procurementagencies. The major force in IT Year, in addi-tion to the information technology awarenesscampaign, was a major and intensive govern-ment-industry initiative to encourage re-search, development, and application of infor-mation technology in order to help strengthenthe overall U.K. economy.

Also during 1982, the British Governmentacknowledged the need to address the field ofinformation technology in a coherent mannerand subsequently made significant changes inits policymaking structures. A minister withspecial responsibilities for information tech-nology was appointed, the first such appoint-ment in any nation. The Minister for Indus-try and Information Technology hasresponsibility, under the Secretary of State forTrade and Industry, for all the Departmentof Trade and Industry’s activities concernedwith information technology, including thoserelated to research and development. Morespecifically, the Minister has responsibilitiesfor information technology, telecommunica-tions, computer systems, microelectronics,electronics applications, robotics, and space.He oversees British Telecom and the Post Of-fice public purchasing, research and develop-ment (including the industrial research estab-lishments) and the British Technology Group.He is also responsible for sponsorship of thechemical, mechanical and electrical engineer-ing, and paper industries; and for distributionand service trade industries; newspapers,


printing and publishing; and for standards andquality assurance and firms. Although the re-sponsibilities of the Minister of Industry andInformation Technology are central to the pro-motion of information technology in theUnited Kingdom, at times the press coverageof ministry activity tends to be exaggerated.

Department of Trade and Industry (DTI)As a result of the 1980 ACARD recommen-

dations which suggested increasing coordina-tion between industry and government, theDepartment of Trade and Industry (DTI) hasbecome the major focus for U.K. initiatives inareas related to information technology. In re-cent years there has been a shift in U.K. Gov-ernment support from its own research estab-lishments (which still account for a substantialpart of the nation’s scientific resource) to theprivate sector. The DTI total expenditure in1982-83 was approximately F 230 million($322 million), of which approximately two-thirds was spent in the private sector. DTI al-located approximately F 80 million ($112 mil-lion) for information technology R&D in 1979-80.’27 This trend is greatly supported by theU.K. Government because it locates researchcloser to the point of application (which somefeel is vital in the area of information technol-ogy) and encourages private sector initiatives.

The DTI’s decisions on funding priorities forresearch and development rely on the adviceof five Research Requirements Boards (RRBs).Each RRB is chaired by a Senior Industrialistand consists of industrialists, scientists, andgovernment representatives. The RRB’s areseen as an effective system for monitoring anddeveloping strategies to ensure that researchsupport priorities match the demands ofchanging technologies and future industrialneeds. The DTI, therefore, works closely withindustry through the Research RequirementBoards to ensure that research in the nationallaboratories is directed towards industrialneeds. Three of the Research RequirementsBoards, Computers, Systems, and Electronics;

‘*’John K. Thompson, “ ‘IT’ in Britain, ” speech given at theBritish Embassy to the Potomac Chapter of the American Soci-ety of Information Scientists, June 9, 1982.

Mechanical and Engineering; and Metrologyand Standards, are involved with informationtechnology R&D support. The Computers,Systems, and Electronics RRB is the majorsupporter of information technology R&Dwith 1980-81 expenditures of F 6 million ($8.4million). 128

In keeping with the current U.K. Govern-ment policy, which aims to achieve a profita-ble, competitive, and adaptable private sector,the DTI has recently implemented a varietyof programs that are intended to supplementdirect R&D support. Because of the recent re-cession that has inhibited U.K. companiesfrom investing large resources in R&D activ-ities, these programs are intended to encour-age the private sector to research, develop, anduse information technology. The DTI’s sup-port programs generally comprise threeelements:

1.

2.

3.

An awareness program to stimulate inter-est in the potential of the new technology.Consultancy to explain how a particulartechnology can be applied to a particularcompany’s needs.Support for ensuing projects.

These DTI R&D support programs, catego-rized by the various applications of informa-tion technology, are presented in table 44.

THE ALVEY PROGRAMME FOR ADVANCED1NFORMATION TECHNOLOGY

In addition to industrial R&D supportschemes, the DTI recently initiated a collab-orative national information technology R&Dprogram. One of the catalysts to the forma-tion of the Alvey Programme was the an-nouncement of the plans for Japan’s Fifth-Generation Computer Systems Project and Ja-pan’s invitation to other countries, includingthe United Kingdom, to discuss participationin the program. The scale and cohesiveness ofthis and other Japanese programs were seenby the U.K. representatives to the Japaneseconference as a major competitive threat. TheBritish also believed that the U.S. industry’s

‘z8Statement of Kenneth Baker, before the U.K. House ofCommons, Dec. 21, 1982.


Table 44.—Department of Trade and Industry R&D Programs

CAD/CAM

CAD/MAT

FMS

FOS

MAP

MISP

ROBOTICS

SPS

— The Computer-aided Design and Computer-aidedManufacture program is designed to promote andaccelerate the acceptance and application of CAD/CAMprimarily in the mechanical and electrical engineeringindustries.

— The Computer-aided Design Manufacture and TestProgram is designed to encourage the use of CAD/MATfor design testing and production in the electronicsindustries.

— The Flexible Manufacturing Systems Scheme is designedto encourage fiirms to install flexible manufacturingsystems.

— The Fiber Optics and Opto-Electronics Scheme isdesigned to encourage the development, production,and application of fiber optics and opto-electronics.

— The Microelectronics Application Project is designed toencourage the application of microelectronics inproducts and processes in manufacturing industries.

— The Microelectronics Industry Support Program isdesigned to promote the microelectronics componentsindustry, particularly for the manufacture of siliconintegrated circuits.

— The Industrial Robotics Scheme is designed toencourage the development and application of robotsin manufacturing industries.

— The Software Products Scheme is designed to encouragethe development and application of software productsand packages.

Awareness programs, demonstration projects and other promotionalactivities

- 6 million ($8.4 million) (over 3 years from1981)

- 9 million ($12.6 million) (over 3 years from1982)

- 60 million ($84 million) (over 3 years from1982)

-25 million ($35 million) (over 5 years from1981)

-55 million ($77 million) (commenced 1978)

-55 million ($77 million) (commenced 1978)

+ 10 million ($14 million) (over 3 years from1981)

+80 million ($1 12 million) (over 4 yearscommencing 1981)


reaction to the Japanese Fifth GenerationComputer Systems Project could also causean equal if not greater degree of competitionfor the U.K. industry.

In the light of these factors, the U.K. delega-tion called for an urgent study into the feasi-bility for a collaborative R&D program gearedto particular strengths and requirements. Thisstudy, completed by the Alvey Committee,outlined plans for a national information tech-nology R&D effort to improve the UnitedKingdom’s competitive position in world in-formation technology markets.129 John Alvey,Chairman of the Committee, comments on thecoordination aspects of the program:

This is the first time in our history that weshall be embarking on a collaborative re-search project on anything like this scale. In-dustry, academic researchers, and govern-

—— —-—-—‘z’See “A Pmgr amme for Advanced Information Technology,

The Report of the Alvey Committee, ” Department of Indus-try, Her Majesty’s Stationary Office, 1982.

ment will be coming together to achievemajor advances in technology which nonecould achieve on their own. The involvementof industry will ensure that the results asthey emerge are fully exploited here in Brit-ain to the advantage of our economy. Infor-mation technology is one of the most impor-tant industries of the future and thereforeone upon which hundreds of thousands ofjobs in the future will depend. Collaborationwill ensure that the results of the research arewidely disseminated particularly to smallerfirms which have such an important contri-bution to make to the industry. No one canguarantee success, but the government isconvinced that this program will ensure forBritish industry secure access to the newtechnology and to the products and processeson which our future prosperity depends.The Alvey Program for Advanced Informa-

tion Technology is a 5 year program fundedby three government ministries and industry.Total funding for the program will be approx-imately $525 million over a 5 year period. The


Department of Education and Science (DES)through the Science and Education ResearchCouncil (SERC) will fund approximately 50million ($70 million) for promoting advancedresearch in academic institutions and thetraining of necessary manpower. The Minis-try of Defense (MOD) will fund approximately%40 million ($56 million) for research believedto be important to the defense industry andwill contribute its experience in the field of in-tegrated circuits. The Department of Tradeand Industry will provide the major portionof the government’s funds, approximatelyL 110 million ($154 million), and will have theoverall responsibility for management of theprogram. Industry will fund the remainingL 150 ($210 million) in the form of 50 percentmatching funds for each R&D project.

The Alvey Program R&D projects are con-centrated in four technical areas, known as“enabling technologies. ” These enabling tech-nologies, seen as crucial to the developmentand application of information technology inthe United Kingdom, include:

● very large scale Integration (VLSI) siliconintegrated circuits,

● software engineering,● intelligent knowledge based systems

(IKBS), and● man/machine interface.

The research projects in these four technicalareas will be managed by the Alvey Director-ate consisting of staff from industry, DTI,MOD, and SERC. Each of the four technolo-gy areas has its own director in addition to adirector in charge of networks and communi-cations among the various R&D projects.Each of the research teams will generally beorganized in small consortia—e.g., two infor-mation technology firms, together with a gov-ernment research establishment team, and auniversity team. Unlike the Japanese ap-proach of creating a center for research, re-search teams will rely on a data network andelectronic mailbox service that will allow in-teractive communication among the R&D pro-gram participants.

The United Kingdom will also be a majorparticipant in the European Strategic Pro-gram for Research in Information Technology(ESPRIT). Currently, U.K. companies are in-volved in more than half of ESPRIT’s pilotprojects. The Alvey program is also designedto complement the ESPRIT program with in-terl.inking communications networks and com-mon parallel research strategies.

THE DEPARTMENT OF TRADE ANDINDUSTRY’S NATIONAL LABORATORY

FACILITIESThe U.K. national research establishments

involved with information technology R&Dare the National Physical Laboratory, the Na-tional Engineering Laboratory, and the Com-puted-Aided Design Centre.130

The National Physical Laboratory (NPL)has a long distinguished history in computertechnology. In 1981-82, NPL R&D expendi-tures were approximately L 22 million ($30million). Currently, R&D projects are focusedon data networks, data security, special inputdevices, and microprocessor applications. TheLaboratory also is developing standard net-work protocols and evaluating cryptographicmethods for data protection. NPL also devel-ops software for solving engineering problems.

The National Engineering Laboratory(NEL) carries out research, development, de-sign, consulting, and testing in automatedmanufacturing. The 1982 Information Tech-nology Year campaign highlighted the Labora-tory’s involvement in robotics and automatedproduction systems. NEL R&D expenditureswere approximately L 16 million ($22.4 mil-lion) in 1981-82. Key areas of NEL researchinclude automated assembly, control and op-timization of production systems, the devel-opment of an advanced turning cell, and otherflexible manufacturing systems.

Computer-Aided Design Centre (CADCENTRE)is the primary center in the United Kingdom

l30Discussion of the U.K. National Research Labs, based on“Research and Development Report, 1981 -82,” Department ofIndustry, 1982.


for the development of computer techniquesin design and engineering. In 1981-82 theCADCENTRE R&D expenditures were approx-imately L 4 million ($5.6 million). The centeroffers approximately 30 computer softwarepackages in CAD/CAM and provides a com-prehensive range of services including soft-ware development, consultancy, productionservices and the provision of hands-on experi-ence in CAD/CAM techniques. Because pri-vate sector initiatives were encouraged in DTIsponsored R&D, the DTI agreed to sell theCambridge-based CADCENTRE to a U.K.consortium led by the U.K. computer firm In-ternational Computers, Ltd. (ICL) for approx-imately L 1 million ($1.4 million). The newlyprivatized CADCENTRE employs ICL staffand DTI management staff. In addition, theDTI has agreed to provide some financial sup-port in order to ease the transition from a gov-ernment research establishment to a commer-cially run company. The DTI will be entitledto a royalty based on the CADCENTRE’Sturnover.131

The Ministry of Defense (MOD)The Ministry of Defense (MOD) is also a ma-

jor supporter of applied R&D. Almost half ofthe United Kingdom’s R&D budget has beendevoted to defense. In the 1970s, defense R&Dremained relatively constant although allother areas of R&D decreased by almost 50percent. In 1978, the Ministry of Defense ex-penditures for electronics research and devel-opment were approximately $900 million.132

Beyond the direct support from industrialresearch funds, the defense sector has poten-tially the greatest possibility for contributingto civil information technology. There hasbeen a recent shift in emphasis within theMOD away from aircraft and towards elec-tronics research and development. Because the

‘g’ Under the royal~ arrangements, the U.K. Government willbe repaid and could receive a further net amount of L 4.5 mill-ion over a 10-year period, assuming forecast revenue levels areachieved by the company.

‘s’J. Thyme, “Information Technology in the U. K.: Gover-nment Policy, ” in G.P. SweeIWy (cd.), Informah”on and the 7hns-formation of Society, North-Holland Publishing Co., 1982, p.261.

overseas defense electronics market is fairlystrong, it is difficult to predict what the long-term consequences of this shift maybe. If themarket grows for defense electronics, therecould be a positive effect on the U.K. supplierindustry. On the other hand, because the spin-off effect of U.K. military R&D to commercialproducts has not been particularly significant,any increase in attention to military needs bythe limited U.K. electronics industry mighthave the effect of further reducing their civil-ian-oriented work and reducing their marketcompetitiveness.

Department of Education and Science (DES)

Under the advice of the Advisory Board forResearch Councils (ABRC), the Departmentof Education and Science (DES) allocates thescience budget to five Research Councils. TheScience and Engineering Research Council(SERC) has the primary responsibility for in-formation technology R&D. The SERC budg-et allocated to information technology R&Dwas L 5 million ($7 million) in 1979-80, L 8.5million ($11.9 million) in 1980-81, and L 11million ($15.4 million) in 1981-82.*33

SERC, analogous to the U.S. National Sc -ence Foundation, accepts competitive bidsfrom universities for special projects. UnlikeNSF, SERC provides only half the funds forsponsored projects. The rest of the fundingmust come from industry, charitable founda-tions, or other government departments.SERC also provides research establishmentswith central computing support. The Engi-neering Board of SERC is most involved withinformation technology, with funding areas in:

● device-related research,● skilled manpower training,• software technology, database utilization,

and system reliability,● distributed computing systems.

SERC operates the Rutherford Appleton Lab-oratory (RAL). The lab itself has a consider-able research program in information technol-

‘S9John Thompson, “ ‘IT’ in Britain,” speech given at the Brit-ish Embassy to the Potomac Chapter of the American Societyof Information Scientists, June 9, 1982.


ogy, with efforts in distributed computingsystems, industrial robotics, computing appli-cations in engineering, electron-beam lithog-raphy, and image processing.

British Technology Group (BTG)

The British Technology Group is an inde-pendent public corporation established to pro-mote the development and application of newtechnology. The British Technology Groupwas formed in 1981 and includes the formerNational Research Development Council(NRDC) and the National Enterprise Board(NEB).

BTG provides funds for technological inno-vation through:

●

●

●

joint venture finance under which BTGcan provide 50 percent of the funds re-quired for the business in return for a levyon sales of the resulting product orprocess;recirculating loans, which are a form ofworking capital loan, through which BTGcan help a company to meet specific or-ders for innovative products; andequity and loan finance on venture capi-tal terms where a company is set up forthe purpose of developing and marketingan invention or new technology; equitymay also be provided in the form of re-deemable preference shares.

Projects from all sectors of industry areeligible for consideration. The primary consid-eration for BTG support is that the proposalproject must be based on a new invention ora genuine technical innovation.

University

Basic research is supported primarily at theuniversity level (approximately 22 percent ofgovernment R&D spending) both by disciplineoriented committees of the Science and Engi-neering Research Council (SERC) and the Uni-versity Grants Committee (UGC). Recent cutsin university funding have been substantial—15 percent over the last 4 years. In addition,lowered enrollment has caused the universities

to become top-heavy with senior, relatively ex-pensive faculty. The steadily rising cost of re-search equipment has also affected the avail-able funds for university research. Together,these three factors have resulted in substan-tially less spending for university R&D.134

Although university research funding hasbeen decreasing over the last few years therehave been marked changes in the distributionof funds—away from basic research towardsengineering and applied R&D. This trendtowards greater emphasis on industrial ap-plication of research results is exemplified inseveral recently initiated schemes. Theseschemes, designed to promote high quality re-search in fields of applied science, are cospon-sored by SERC and DTI. In one of the moresuccessful joint SERC/DTI initiatives, theTeaching Company Scheme, DTI pays for en-gineers to do postgraduate work in industry.Students work on product development, de-sign, and manufacturing processes, but withclose attention and support of academic staffwho supervise the innovative aspects of thegraduates’ work. The program so far has at-tracted great interest from industry, andmany firms such as General Electric Co., Brit-ish Aerospace, Ferranti, and IBM are partici-pating.

Another program aimed at promoting indus-trial training is the introduction of approxi-mately 150 information technology centers(ITEC centers) throughout the United King-dom. Funded by DTI, Manpower ServicesCommission, and industry grants, these cen-ters collaborate with local industries to pro-vide computer training for unemployed high-school age youths. Approximately 70 percentof the students upon leaving the centers findemployment in a computer-related field. Otherprograms that offer technical training includeComputers in Schools, through which DTIfunds 50 percent of the cost of up to two micro-computers in each U.K. school, and a Manpow-

194 Robin B. Nicholson, “Science and Technology Policy in theUnited Kingdom, ” address given at the Nineteenth AnnualMeeting of the National Academy of Engineering, Nov. 2-3,1983.


er Services Commission Program that fundsinformation technology training in computer-related subjects. Approximately half of all theprogr ammers training in the United Kingdomis provided by these programs.

In addition to industrial training programs,the U.K. Government has encouraged greaterindustrial participation in R&D as well ascloser industry-university ties through the ini-tiation of science parks. Similar to the Japa-nese and U.S. science parks, the U.K. scienceparks provide an opportunity for industry tolocate in the proximity of major researchuniversities.

Industry

Since the late 1960s, the U.K. Governmenthas played an important role in supporting theU.K. computer industry. For example, in 1968the government encouraged a series of merg-ers that led to the establishment of Interna-tional Computers Ltd. (ICL) and then providedfunds amounting to approximately $12 millionannually until 1976. Moreover, the govern-ment encouraged ICL’S growth through pref-erential procurement policies that guaranteedalmost all large central government contractsto ICL. ICL became the largest European com-puter company, with a wide customer base inthe United Kingdom as well as overseas, in-cluding the United States. ICL became so suc-cessful that government assistance was with-drawn in 1976, and in late 1979 the govern-ment sold its 25 percent share in the company.However, by 1981, ICL was heavily in debt,and the government (although it encouragedprivate funding and joint research programs)arranged L 270 million ($378 million) in loansfor ICL.

In an effort to become profitable again, ICLis currently exploring joint activities with non-U.K. firms. In 1982, ICL and Fujitsu reacheda collaborative agreement. The arrangementprovides for special access to Fujitsu’s ad-vanced microelectronics technology, and pro-vides for purchases of semiconductor chips de-veloped by Fujitsu. ICL will in turn marketlarge Fujitsu computers in Western Europe,

thus broadening the ICL product line intomore powerful computers. It is hoped to ex-tend this collaboration at a later date to othertechnology areas, including communicationstechnology. ICL also recently reached anagreement with a small U.S. company, ThreeRivers, to manufacture and market worldwidea microcomputer designed by Three Rivers.

In 1979, the U.K. Government also invested$100 million for the creation of a national semi-conductor firm, Inmos. Operating as an inde-pendent producer, Inmos was established tomanufacture a limited range of products (prin-cipally high-capacity semiconductor memorychips) to sell to large electrical goods manu-facturers. Also, to assure indigenous produc-tion of semiconductors, the U.K. Governmenthas convinced several U.S. semiconductorcompanies to setup production in the UnitedKingdom by offering them grants up to 33%percent for R&D costs, under the Support forInnovation Program (SFI).

Currently, more than half of the U.K. indus-trial electronics R&D expenditure is funded bygovernment. The funding may be in the formof cost-sharing grants, procurements, or anyof the multitude of funding schemes. More-over, the pervasive influence of the U.K. Gov-ernment is exemplified in the original Alveyproposal which called for the government tofund 90 percent of industrial R&D activities.The proposal, however, was finally amendedto the current commitment of 50 percent onlyafter considerable debate, on the grounds thatindustry would not be committed if it only wasrequired to invest 10 percent of its own re-sources.

The historically low industrial funding forinformation technology R&D has been attrib-uted to the U.K. information technology in-dustry’s focus on highly specialized markets.Because the U.K. information technology in-dustry in some instances fills small but impor-tant market niches, it usually does not capturemass markets; consequently, the U.K. infor-mation technology industry has not alwayshad adequate resources for R&D funding.

A case in point is the British semiconduc-tor chip industry. The leading U.K. informa-


tion technology firms have concentrated onproducing specialized chips known as “spe-cials” where the demand is low and the mar-ket relatively small. For example, world salesin 1980 of custom-built special chips came toonly L 100,000 ($140,000) almost one-twen-tieth of the comparable figure for standardchips.135 Although the leading British electron-ic firms’ demand for these standard chips areextremely high, U.K. industry has left themanufacturing of these types of chips mainlyto companies from the United States and Ja-pan, which together capture approximately 75percent of the world market. However, theUnited Kingdom’s leading firms—Ferranti,Plessey, and GEC—all plan to expand theirmanufacturing capability of standard chips fordomestic use. This pattern has also been re-peated in other information technologies; forexample, few British companies produce hard-ware that directly serves the semiconductorand electronics industry (i.e., for testing andproduction needs).

There is some indication that U.K. industrymay be reassessing its investment policies.Current estimates suggest, for example, thatthe volume of R&D has been maintainedthrough the recession on a selective basis withsubstantial advances in areas such as micro-electronics and corresponding decreases inmetals and traditional engineering. For exam-ple, industrial electronics R&D expendituresincreased form L 279 million ($390.6 million)in 1975 to L 442 million ($618.8 million) in1978.136

Similar to the situation in France and Ja-pan, a few large firms are responsible for alarge proportion of information technologyR&D expenditure in the United Kingdom.Several of these major firms are describedbelow.

‘35 Peter Marsh, “Britain Faces Up to Information Technol-ogy, ” New Scientist, Dec. 23, 1982, p. 637.

“’Robin Nicholsou “Science and Technology Policy in theUnited Kingdom, ” address given at the Nineteenth AnnualMeeting of National Academy of Engineering, Nov. 2-3, 1983.

Plessey

In 1983, Plessey invested approximately$225 million in R&D, 15 percent of its sales.Approximately $22 million of the R&D ex-penditures were allocated to basic researchactivities. Of the 320 researchers in the mainPlessey laboratory working on microelectron-ics, 170 people are working on gallium arsenideand related materials and 150 are working onsilicon. Plessey’s silicon research is gearedtowards speciality or custom circuits.

Plessey relies heavily on outside contracts,and approximately 45 percent of its researchis done for the MOD and British Telecom. Ofthe remaining 55 percent, half is conducted forthe Plessey operating divisions and half is con-ducted for the head office. Currently, Plesseyis attempting to reduce its reliance on outsideR&D funding in favor of more operating divi-sion work.137

General Electric Co. (GEC)

In 1983, the General Electric Co. (GEC) in-vested approximately $900 million in researchand development, approximately 10 percentof its sales. Unlike Plessey, only 25 percent ofGEC’s research is for the MOD and BritishTelecom. Fifty percent of the research is forGEC’s 120 operating companies, and 25 per-cent is for basic, speculative research. Becausea large percentage of its R&D is supported byits operating companies, GEC’s research activ-ities are more oriented towards commercialproduct development. GEC’s key areas of re-search include microelectronics, fiber optic de-vices, software engineering, and custom chipdesign.138

British Telecom (BT)

Previously a public monopoly, the U.K. Gov-ernment has recently privitized British Tele-com (BT). Moreover, the U.K. Government haspermitted the licensing of other competitors

“’’’Profile: GEC and Plessey: Two Approaches to R& D,” TheEconom”st, Nov. 20, 1982.

‘3’’ ’Ibid.


to provide telecommunications services. It ishoped that the introduction of competitioninto the provision of telecommunications serv-ices will stimulate new demand and provisionof a vast array of new services. Consequently,U.K. companies with advanced products areexpected to be well placed to take advantageof market development and to use the UnitedKingdom as a springboard for capturing European and other world information technologymarkets.

The privatization efforts, however, havecaused some speculation about the future ofBT laboratories in much the same way thatdivestiture has caused speculation about thefuture of Bell Labs. Currently, BT Labs re-

search expenditure is approximately L 100million ($140 million) annually. BT has tradi-tionally had strong research programs in fourmajor areas:

●

●

●

●

advanced technology (e.g., CAD in largescale integrated circuit design, opticalcommunications systems, gallium arse-nide, high speed logic);transmission (e.g., digital transmission inLANs, small earth station satellites);customer service/apparatus (e.g., Presteland Viewdata);advanced systems (e.g., microprocessorsoftware development ISDN local con-nection).

European Strategic Program for Research inInformation Technology (ESPRIT)

In an attempt to reverse the decline of Eur-ope’s competitiveness, to ensure a strongertechnological base, and ultimately to ensureeconomic and political independence, the Com-mission of European Communities proposedthe European Strategic Program for Researchin Information Technology (ESPRIT).139 Al-though EEC members hope that ESPRIT willsucceed in strengthening Europe’s technologi-cal base and international competitiveness,there are concerns as to whether research canbe effectively undertaken and shared amongpotential international competitors.

ESPRIT is designed to address three ma-jor difficulties that currently face the Euro-pean information technology industry as it at-tempts to develop new state-of-the-art tech-nologies: the problem of raising adequatefunds for long-term research and developmentduring a period of economic recession and fall-. — — —

‘sgIn 1975, the European Community had a trade surplus ininformation technology products; however, by 1980, the tradedeficit in information technology products reached $5 billionand reached approximately $10 billion in 1982. At present, Eur-ope represents one-third of the world information technologymarket but accounts for only 10 p-cent of world informationtechnology production. For example, in the European domes-tic market, 2 out of every 5 information technology productssold are European; 8 out of 10 personal computers sold in Eur-ope are imported from the United States; 9 out of 10 videotaperecorders sold in Europe come from Japan. ESPRIT Proposal,COM(83) 258 final, Commission of the European Communities,June 2, 1983, p. 8.

ing sales; a domestic market which, unlikethose of the United States and Japan, is frag-mented into a number of relatively small na-tional units; and reluctance by some within in-dividual countries to subsidize other nationswho have historically been economic and po-litical rivals.140 Consequently, the ESPRITprogram seeks to: increase the size of researchteams, optimize the use of human and finan-cial resources, and initiate definition and adop-tion of European standards for informationtechnology products.141

ESPRIT attempts to address the problemof the fragmentation of the European market,as well as the divided efforts of the individ-ual member nations’ national R&D supportprograms, by linking a significant proportionof key European engineers and scientists fromgovernment, industry, and universities. In thisrespect, ESPRIT is similar to other nationalresearch and development programs such asJapan’s Fifth-Generation Computer Project,the United Kingdom’s Programme for Ad-vanced Information Technology, and the U.S.Semiconductor Research Corporation (SRC),which are also partnerships among variouscompanies, academic research laboratories,“=David Dickson, “Europe Seeks Joint Computer ResearchEffort,” Science, Jan. 6, 1984, p. 28.

“’ESPRIT Proposal, COM(83) 258 find Commission of theEuropean Communities, June 2, 1983, p. 9,


and government agencies. However, unlikethese other research and development pro-grams, ESPRIT also represents an interna-tional institutional arrangement.

Unlike the MITI program in Japan, theESPRIT program is limited to precompetitiveresearch; it does not intend to develop a com-mercial product as does the Japanese fifth-generation computer effort. At any point inthe research, however, the participating com-panies are free to take the technical resultsgained from the ESPRIT projects and developcommercial products on their own. Therefore,ESPRIT is not seen to be in competition withnational research and development programsor individual companies, but as a reinforce-ment to make them more effective.

Funding for ESPRIT is approximately $1.3billion for the next 5 years and approximately2,000 researchers will take part in researchactivities. 142 The funding be equally sharedamong 12 principle corporate partners andother participating companies with the Com-mission of European Communities. The 12main corporations participating in ESPRIT in-clude: General Electric Co. (United Kingdom),Plessey plc. (United Kingdom), InternationalComputers Ltd. (United Kingdom), Compag-nie General de L’Electricity (France), CIT-Alcatel (France), Cii Honeywell Bull, (France),Thomson-Brandt (France), AEG-TelefunkenAG (West Germany), Nixdorf Computer AG(West Germany), NmbH Phillips Gloeilampen-fabrieken (Netherlands), Olivetti SPA (Italy),and Societa Torinese Esercizi Telefoncini(Italy).

The ESPRIT program consists of five ma-jor research areas: advanced microelectronics,software technology, advanced informationprocessing, office automation, and computerintegrated manufacturing. The EEC has cho-sen these five areas of information technology————

‘42Actual funding for ESPRIT is 1,500 million European cur-rency units (ECU). Approximately 1 AU equals 1 U.S. dollar.Of the total European Community countries’ research and de-velopment expenditures, 1.7 percent is for EEC research anddevelopment activities. Nine percent of the EEC R&D budgetis reserved for industry of which 20 percent is for ESPRITfunding.

because of their perceived importance for fu-ture European industrial competitiveness. Thefirst three of these research areas were selectedin part to develop better enabling or core tech-nologies. Office automation and computer in-tegration were selected as specific applicationsareas where information technology is ex-pected to have a large economic and social im-pact–automation of the office and the factory.

The advanced microelectronics project’s ma-jor goal is to develop smaller, more reliable,and more powerful integrated circuit technol-ogy so that devices can perform more func-tions or operations than circuits availabletoday. More specifically, the goal of the ad-vanced microelectronics project is to improvethe current state-of-the-art process, which isbased on three-to-five micrometer structures,to processes that are based on structuressmaller than one micrometer. The Europeansare hopeful that this advanced microelec-tronics project will improve Europe’s currentintegrated circuit trade deficit.143

The major emphasis of the advanced infor-mation processing project is on informationand knowledge engineering, information stor-age and usage, signal processing, and exter-nal interfaces. The research project’s overallgoal is to develop technological capabilitiesthat underlie machine intelligence. Advancesin the new types of information processing willalso entail breakthroughs in advanced comput-er architecture, further miniaturization inmicroelectronics, and higher reliability.

The goal of the software technology projectis to improve software engineering techniques.More specifically, the project’s goals includeestablishing standardized software interfaces,automating the software engineering process,and disseminating and centralizing softwareresearch results in a common database so thatindividual modules of software programs canbe reused where similar functions are re-quired.

‘43Currently, Europe absorbs 20 percent of the world’s inte-grated circuit market, although Europe produces only 6 per-cent of the world’s integrated circuits.

274 • /formation Technology R&D: Critical Trends and Issues

The office automation project, one of the twospecific applications projects, is directed at de-veloping a multimedia interface for all officecommunication needs, and at developing effi-cient electronic filing systems for unstructuredinformation (text, voice, graphics, images,etc.). The project will also examine the cross-cultural interaction of human factors, educa-tional, sociological, and industrial effects of of-fice automation systems. Moreover, researchon machine translation (which is of great im-portance to the European Community) and as-pects of machine-user interfaces, such as in-tegrated image test speech communication,document creation, and distribution will alsobe conducted.

The goal of the computer integrated manu-facturing project is to develop improved sys-tems for automated factories. These systemswill integrate in a common data base comput-er-aided design, computer-aided manufactur-ing, computer-aided testing and repair, andassembly. Such integration will require furtherdevelopments in integrated systems architec-ture, advanced components, real-time basedimaging, and integrated control subsystemsmounted on semiconductor chips.

Experts from the 12 main industry partnersas well as outside consultants have developeda work plan of specific projects in the five proj-ect areas. As these specific projects have beendefined, proposals are solicited. Proposals maybe submitted by research teams from any in-dustry or university, although the team mustbe composed of nationals from two or moreEEC countries. This arrangement is meant toencourage international cooperation betweennations and industries, and therefore preventduplication of research efforts and make op-timal use of limited financial and human re-sources. The main criteria for evaluating pro-posals include: technical soundness; contri-bution to industrial strategy in light ofESPRIT objectives; European Communityusefulness; technical, scientific, and mana-gerial capability to undertake the proposedproject; and proposed activities that will fa-cilitate the dissemination of research results.

The Commission and the advisory board,which consists of industry representatives(mainly from the 12 contributing industries),review research proposals and approve grants.Two broad types of proposals are consideredfor the different technical projects. Type A,which represents the strategic long-term re-search activities of ESPRIT, involves large re-search establishments and large commitmentsof resources, both human and financial, aswell as clear long-term strategic plans toensure continuity of research and long-termbenefits. Type A projects receive 50 percentfunding from the European Community, andthe research participant is expected to providethe remaining funds. Type B proposals requirerelatively smaller resources and account for asignificant share of the overall efforts underESPRIT. Type B projects could range fromvery long-term, very speculative R&D toshorter-term and more specific R&D. Type Bprojects receive at least 50 percent of theirR&D funds from the European Community,or more if the applicant is from an academicinstitution or smaller business with limitedavailable finance.

One of the first ESPRIT research proposalsto be funded is a project to develop advancedinterconnection between very-large-scale inte-grated circuits. The research project is a jointeffort between Plessy and GEC in the UnitedKingdom, Thomson CSF in France, and Tele-funken in West Germany. Another initial proj-ect, jointly shared between the Polytechnic inLondon and the University of Amsterdam, isfocusing on the development of 11 differentaspects of tools and methods for developingmachine intelligence.

The Future of ESPRIT

The initial response to the ESPRIT pilotphase has been favorable; the 1 year pilotphase of ESPRIT, launched in mid-1983 witha budget of $20 million and funded 50 percentby the European Community and 50 percentby industry, attracted over 200 research pro-posals. However, only 36 could be selected toreceive EEC matching funds. EEC officials


were quite surprised not only at the scale ofresponse, but also at the apparent willingnessof companies to permit their scientists to worktogether with few restrictions.144

Currently, however, there is still some doubtas to whether ESPRIT will further achieve itsstated goals. In the past, the EEC has hadsome difficulties with other joint projects. Inthe 1960s, Pierre Aigrain, then President Pom-pidou’s chief scientist, proposed a Europeanproject to strengthen the technological baseof the computer and communications indus-try. However, the project was never launchedbecause of resistance from the European tele-communications monopolies to the suggestionthat their own research was insufficient andthe lack of a plan to suggest how a coopera-tive research program supported by estab-lished companies and nationalized industriescould be beneficial to all the participants.145

Moreover, in the early 1970s, Dutch, German,and French computer manufacturers formeda joint venture, Unidata. Each company senttheir top engineers to a joint research and de-velopment facility. However, the project hadmany difficulties and some of the participantseventually withdrew from the project.146

As in past joint European research efforts,the future of ESPRIT depends as much on theresults of political struggles around the re-structuring of Europe’s economic and indus-trial base as it does on any judgment of itstechnical and scientific merits.147 This is illus-trated by the first few unsuccessful attemptsin early 1984 for EEC endorsement of ESPRIT,which became intermingled with broader eco-nomic issues ranging from the efficiency ofFrench farming practices to the EEC’s bud-get procedures. Moreover, it is still unclear asto how the national information technology re-search programs will mesh with ESPRIT proj-

——-— . —“’David Dickson, “Europe Seeks Joint Computer Research

Effort, ” Science, Jan. 6, 1984, p. 28.“’’’What Hope for ESPRIT?” Nature, Feb. 16, 1984, p. 582.“’Beth Karlin and George Anders, “Europe Looks Abroad

for High Technology It Lags in Developing, ” The Wall StreetJourzud, Oct. 5, 1983, p. 1.

“’David Dickson, “Europe Seeks Joint Research Effort, ” Sci-ence, Jan. 6, 1984, p. 28.

ects. Consequently, there are debates in eachcountry over whether the results of informa-tion technology research and development, asan important key to future political and eco-nomic strength, should be shared with poten-tial competitors.

International rivalry is also a large indus-try concern. The lack of an established legalframework in which companies will be allowedto collaborate may cause difficulties for Euro-pean information technology industries. Be-cause ESPRIT is concerned primarily withlong-term precompetitive research, there maynot be any conflict with the EEC antitrustlaws which are intended to apply primarily tomarketing strategies, rather than product de-velopment. However, some companies believethat as the gap between scientific discoveryor basic research and commercial applicationnarrows, collaborative precompetitive re-search will be extremely difficult.148 It is forthis reason that three of Europe’s largestmainframe computer manufacturers (Siemens,ICL, and Bull) have established the EuropeanComputer Industry Research Center. The Re-search Center discourages open collaborationby initially excluding participation by othercompanies because:

If you recognize the fact that you are in acompetitive market, in which companies arefighting against each other, then you mustaccept that it is not in the interest to offerall research results to everyone who might beinterested in them, and that at least someprojects will be of character that will forbidthe open publication of research results fromthe beginning. . . . We do abstract researchon an international basis where the balanceof cooperation and competition should bedetermined by the rules of international com-merce, and market-like research on a nationalbasis, where individual companies can adoptthe most appropriate strategies for their do-mestic and political environment. 149

— — . . . —148This problem maybe particularly acute in research projects

such as office automation and computer-aided manufacturing.“’Statement by Mr. Heimann of Siemens, David Dickson,

“Europe Seeks Joint Computer Research Effort, ” Science, Jan.6, 1984, p. 29.


Despite the preliminary difficulties in achiev- ments. If ESPRIT does succeed, it is likely toing agreement on the funding for ESPRIT, the be used as a model for similar cooperativeestablishment of the controversial Siemens- European Community projects in other fields,ICL-Bull Research Center, and other under- such as telecommunications and biotech-lying political and economic rivalry, ESPRIT nology.may succeed in overcoming these impedi-

Chapter 8

Information Technology R&Din the Context of U.S. Science and

Technology Policy

Contents

Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Part I: Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .General U.S. Science and Technology Policies . . . . . . . . . . . . . . . . . .Information Technology R&D Policies . . . . . . . . . . . . . . . . . . . . . . . . .

Part II: Key Issue Areas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Issue A: Organizationof Government . . . . . . . . . . . . . . . . . . . . . . . . .

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Dimensions of the Issue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Options for Addressing the Issue . . . . . . . . . . . . . . . . . . . . . . . . . . .

Issue B: Military/Civilian Balance . . . . . . . . . . . . . . . . . . . . . . . . . . . .Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Dimensions of the Issue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Options for Addressing the Issue . . . . . . . . . . . . . . . . . . . . . . . . . . .

Issue C: International Competitiveness . . . . . . . . . . . . . . . . . . . . . . . .Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Dimensions of the Issue . . . . . . . . . . . . . . . . . . . . . ...!..... . . . . .Options for Addressing the Issue . . . . . . . . . . . . . . . . . . . . . . . . . . .

Concluding Thoughts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

TablesTable No.45. Policy Tools, Actors, and Goals of Science Policy and Technology46. Tenets of Science and Technology Policy, 1960-84. . . . . . . . . . . . . .47. Forces Affecting Science and Technology Policy Since 1960 . . . . .

279

279

280280286

290290290291293294294295297298298298300301

Policy . . . . . . . .. . . . . . . . . . . . . .. . . . . . . . . . . . . .

48. Federal Government Policies Toward Information Technology R&D . . . . . . . . . . .49. Department of Defense Funding for Basic Research by Discipline,

Fiscal Years 1982, 1983, and 1984... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figures

281285287288

296

Figure No. Page50. Federal R&D Budget Authority for Defense and Nondefense Activities . . . . . . . . . 29551. Federal R&D Budget Authority for Defense Activities by Character of Work . . . . 295

Chapter 8

Information Technology R&Din the Context of U.S. Science and

Technology Policy

Findings

● Information technology is one of the mostdynamic and controversial areas of U.S.science and technology due to its rapidpace of change, the emphasis placed onthis technology for economic growth andfor national security, and the pervasiveor “core” nature of the technology and itseffects.

● In this area there is a growing conflict be-tween policies that emphasize basic re-search and policies that focus on interna-tional competitiveness and applications.These issues are prominent in informationtechnology R&D because the lines be-tween basic and applied research are souncertain.

● Interest in coordinating Federal policy forinformation technology is intensifying, inpart because of foreign government poli-cies, growing costs for R&D, and grow-ing concern for international competitive-ness. Although coordination of variousaspects of Federal policy has been de-bated for decades, it is a particularly sa-lient issue in information technology: manyGovernment agencies are involved, butnone devote high-level policy attention to

●

●

this area. The advantages of centraliza-tion or coordination are that it could savemoney and more effectively focus R&Din critical areas; the possible disadvan-tages include the establishment of a cum-bersome bureaucracy and the loss of agen-cy autonomy and flexibility.The dominance of the Department of De-fense in information technology R&D hasraised questions: Is military work siphon-ing off too much talent from civilian ap-plications? Is the military work changingthe direction of research in informationtechnology in ways that are disadvanta-geous for the commercial sector or for thepublic? And are existing efforts to trans-fer technology from military to commer-cial applications adequate? Evidence cur-rently suggests that there are growingproblems in this area.Current policies and practices toward in-formation technology R&D conflict withthe realities of increasing internationalcompetition. The situation may call for amore sophisticated Government role inmonitoring and support of industry andresearch.

IntroductionEarlier chapters documented the rapid themselves continue to change in cost, power,

changes occurring in information technology and the variety of functions they can perform.research and development. The technologies At the same time, institutional structures for

279

38-802 0 - 85 - 19


R&D are quickly evolving, international com-petition is intensifying, and the technology’simpact on a wide range of social issues andproblems is increasingly prominent. Becauseinformation technology is pervasive, its effectscascade through many aspects of society—from science itself to education, business, anddefense–and at each point create seeminglyindependent changes and conflicts.

These changes are bringing increased atten-tion to U.S. policy toward information tech-nology R&D. In particular, Japanese and Eu-ropean policies, as noted in chapter 7, havebrought increasing demands for U.S. policy responses.

Because the effects of information technol-ogy are so wide-ranging, any policy to respondto this technology must consider not only ac-tions within specific issue areas such as man-power, but also broader issues in science andtechnology policy, such as the organization of

Government and the roles of different agen-cies in R&D. The purpose of this chapter is toexamine these more general frameworks forpolicy toward information technology R&D.

The chapter is divided into two major sec-tions. The first begins with some brief back-ground on science and technology policy in theUnited States, and the forces that have af-fected this policy as it has unfolded over thelast few decades. Then, the chapter shows howthese broad policies and forces set the contextfor and are closely tied to policy toward infor-mation technology. In the second section ofthe chapter, OTA discusses three key areasthat are central to the science and technologypolicy issues raised by information technologyR&D. The areas are the organization of Gov-ernment, the balance of military and civilianroles, and policy measures to enhance inter-national competitiveness.

Part I: BackgroundGeneral U.S. Science and

Technology Policies

Historically, science policy has been theterm used to describe the actions of Govern-ment that affect the funding, organization,performance, and use of science.’ The term hasincluded policy for technology and engineer-ing as well as for science. More recently, how-ever, as technology has played a more promi-nent role in society and industry, manyexperts view “science policy” as inadequatein addressing concerns of technology.2 Theterm “technology policy” has been used in-creasingly to refer to policy measures muchmore directly related to development and use

IScience Policy Research Division, Congressional ResearchService, Sa”ence Policy, A Working Glossary (Washington, DC:U.S. Government Printing Office, 1976).

‘See, for example, J. J. Baruch, “The Cultures of Science andTechnology,” Science, Apr. 6, 1984. Baruch argues that lump-ing science and technology policy together is unwise, since thetwo enterprises have quite different approaches, goals, andneeds.

of technologies, particularly as they relate tointernational competitiveness. In some recentdiscussions of industrial policy, technologypolicy has sometimes been considered an ele-ment of, or even a synonym for, industrialpolicy.

Table 45 sketches some of the actors andpolicy tools involved in both science and tech-nology policy. The two types of policies havedifferent, yet overlapping, constituencies andgoals. In an area such as information technol-ogy, where “science” and “technology” areoften commingled, the boundary between sci-ence policy and technology policy is vague. Re-cent statements of science policy (box A) il-lustrate the priorities of various policymakers,and show how science and technology are of-ten mentioned together and blurred in the for-mation of policy. Note that although executivebranch statements of science policy may bethe most visible, other actors in the sciencepolicymaking arena-particularly Congress

Ch. 8—information Technology R&D in the Context of U.S. Science and Technology Policy ● 281

Table 45.—Policy Tools, Actors, and Goals of Science Policy and Technology Policy

Science Policya Technology Policya

Primary policy tools: Primary policy toolsFunding of basic research Mission-oriented R&D fundingScientific manpower and education measures Engineering manpower and education measuresScience information dissemination Technology transfer mechanismsInternational exchange programs Limits on international flow of technology and

informationR&D tax creditsStandards and patent policiesUniversity/industry research collaboration

Primary Actors: Primary Actors:Office of Science and Technology Policy Office of Science and Technology Policy

(The White House)National Science Foundation, National Science Foundation

National Science BoardNational Academy of Sciences National Academy of EngineeringAgencies conducting basic research, e.g., Department of Mission agencies— e.g., Departments of Defense, Energy,

Defense, National Institutes of Health the National Aeronautics and Space AdministrationCongress CongressUniversity science community Industrial R&D communityAmerican Association for the Advancement of Science Industry associations— e.g., Information Industry

AssociationProfessional societies—e.g., American Medical Professional societies—e.g., Institute for Electrical and

Association, American Chemical Society Electronics Engineers, Association for ComputingMachinery

Department of Commerce—International TradeAdministration, National Bureau of Standards, NationalTelecommunications and Information Administration

Social goals: Social goals:Quality of life Economic well-beingKnowledge for knowledge’s sake National securityEquity, education Technological leadershipa!,sCl~nC~ ~OllCY,, and ,~teChnOIOgY ~OllCY>> are often difficult to sepa~ate, The policy tools, actors, and goals listed under each category are those that tend to be associated

with science policy or with technology policy However, in many practical situations, the issues and actors are intertwined.

SOURCE” Office of Technology Assessment

and the scientists and engineers themselves—have a strong (some would say dominant) in-fluence over actual policy.

A brief history of U.S. science policy (which,as noted above, has usually been defined to in-clude technology policy) is helpful in order toprovide a context for the gradual unfolding ofpolicy toward information technology.3 U.S.

3This chapter’s analysis of science and technology policy isa synthesis of published books, articles, statements and legis-lation; in addition, OTA and its contractor (J. F. Coates, Inc.)conducted interviews with several dozen science policy experts.OTA is indebted to this group (see acknowledgments at thefront of this volume) for their insights and assistance, althoughOTA takes full responsibility for the content of this report. Fora fuller elaboration of history and issues in science policy,readers should consult, for example: Harvey Brooks, The Gov-ernment of Science (Cambridge, MA: MIT Press, 1968); W.Henry Larnbright, Goverm”ng Science and Technology (NewYork: Oxford University Press, 1976); Daniel S. Greenberg, ThePofitics of Pure Science (New York: New American Library,1971); A, Hunter Dupree, Science in the Federal Government(Cambridge, MA: Harvard University Press, 1953); U.S. Gen-eral Accounting Office, Major Science and Technology Issues

science policy has evolved since the 1940s outof tension between two fundamental premises:

1. that research should be supported in or-der to push ahead the frontiers of humanunderstanding (“science for the sake ofscience”), lay the groundwork for techno-logical advances, and train future scien-tists and engineers; and

(Washington, DC: GAO, Jan. 30, 1981); Congressional BudgetOffice, Federal Support for R&l) and Innovation (Washington,DC: CBO, April 1984); Science Policy Research Division, Con-gressional Research Service, National Sa”ence Board: SciencePolicy and Management for the National Science Foundation,1988-1980, January 1983; Frank Press, “Science and Technol-ogy in the White House, 1977 to 1980, ” Science, Jan. 9, 1981,pp. 139-145, and Jan, 16, 1981, pp. 249-256; and the annual se-ries of reports on R&D from the American Association for theAdvancement of Science. For a comprehensive view of the roleof one key player in science policymaking, see Toward the End-less Frontier: History of the Committee on Science and Tech-nology, 1959-79, House Science and Technology CommitteePrint (Washington, DC: U.S. Government Printing Office,1980).

Ch. 8—information Technology R&D in the Context of U.S. Science and Technology Policy . 283

. .

From President Carter’s 1979 message to Congress on Science and Technology:Yet despite the centrality of science and technology in our lives, the Federal government has

rarely articulated a science and technology policy, This message sets forth that policy. The thesisis that new technologies can aid in the solution of many of our Nation’s problems, These technol-ogies in turn depend upon a fund of knowledge derived from bade research. The Federal govern-ment should therefore increase its support both for basic research and, where appropriate, forthe application of new technologies. . .

The Federal government’s support of research and development is criticaI to the overall ad-vance of science and technology. Federal responsibility lies in three major categories:

1. The largest fraction of Federal investment serves the government’s direct needs such as de-fense, space, and air traffic control. . .

2. The Federal government undertakes research and development where there is a national needto accelerate the rate of development of new technologies in the private sector. . . . when therisk is great or the costs are inordinately high. , . .

3. The Federal government supports basic research to meet broad economic and social needs. . . .

The majority of Federal support for basic research is in the mission agencies.

From the White House Office of Science and Technology PoIicy’s 1982 Annual Science and Tech-nology Report to the Congress:

The U.S. science policy is● to enhance the contribution of science to the two most pressing long-term needs of the United

States: national defense and the international competitiveness of U.S. industry.● To maximize the return on national R&D investments; and● To ensure the long-term vitality of the U.S. science and technology base.The strategy to implement U.S. science pOLiCY

Emphasizes excellence–in research results and in people:Stresses the importance of scientific relevance to national needs, and more clearly defines theappropriate roles of the Government and the private sector in supporting R&D;Facilitates cooperation in scientific research among Government, industry, and academia;Seeks to support sufficient basic and long-term applied research to ensure that the United Statesmaintains the world’s strongest science and technology enterprise;Emphasizes the importance of having the leading research universities in the world and oftrainin g the highest quality scientists and engineers to ensure continued U.S. qualitative leader-Ship; and . . . .Allocates Federal R&D resources to support this strategy.. -

The U.S. technology policy is to ensure that U.S. scientific leadership results in economic anddefense leadership. .. ~+ . ~ . *The strategy to implement U.S. technology policy’ ‘ r -

Provides tax and other incentives to the private sector for commercial R&D;Continues to emphasize the different private sector and government roles in developing newtechnologies, products, and processes so as not to discourage private sector initiative withthe threat of Government intervention and competition; -Improves the climate of cooperation so that maximum Government,cross-stimuli occur among industry, and academia; . .=Improves the ability of Federal laboratories to contribute to U.S, industry, and also takes advan-tage of foreign research results;Encourages the change in industry’s outlook to emphasize long-term viability rather than onlyshort-term gain;Recognizes that the service sector in the U.S. economy is gaining in importance, and focusesemphasis on R&D accordingly; andRecognizes the effect of economic and regulatory policies on U.S. science and technology and,ultimately, on U.S. economic competitiveness.


2. that the investment of public resourcesin research should be moderated and di-rected at specified high-priority nationalneeds, to which the private sector is un-willing or unable to respond.

The first principle was embodied in the cre-ation of the National Science Foundation andthe National Institutes of Health to managethe distribution of public funds to supportbasic research. The primary mechanism forthis support is research grants, which aremade in response to requests by recognizedscientists and validated by the judgment oftheir peers.

The roots of the second principle, whichunderlies all “mission-oriented” Government-funded research, go far back into our nationalhistory. The second principle is evident in thescience policy statements of box A, particu-larly those from the Reagan administration,which emphasize the payoffs of science andtechnology for the economy and defense.

The accountability and focus indicated bythe second premise is often at odds with basicresearch, which sets its own directions andoften leads investigators down blind alleys ortoward ends that may have no immediate orforeseeable practical applications.

Until the 1940s, most federally supportedresearch was closely related to well-establishedGovernment responsibilities such as defenseor exploration and development in the West,or to areas basic to the national economy (agri-culture, water, and public health). After WorldWar II, leaders such as Vannevar Bush–real-izing that we had entered an era of rapid ad-vancement in scientific knowledge that couldcreate new technologies and industries-force-fully led the Nation to accept increased, sys-tematic, and continuing support for sciencethrough funding of basic research and scienceeducation.4

4See Vannevar Bush, “Science, the Endless Frontier: A Re-port to the President on a program for Postwar Scientific Re-search, ” originally pubIished in 1945, reprinted by the NationalScience Foundation, Washington, DC, 1960.

The clearest landmark event in the post-warera was the Soviet launch of Sputnik in 1957,which had two major kinds of effects. The re-sponse to Sputnik-the technological ventureto put a man on the Moon, and bring him back,by the end of the decade of the 1960s–wasunique for a non-war effort in having a singu-lar clarity of mission and unequivocal criteriafor success. This mission galvanized a largeportion of the scientific and technologicalenterprise to a single clear goal. The second,more diffuse consequence was to redirect theNation’s attention, albeit for only a briefperiod, to science, science education and newscientific opportunities.

The premises of science policy, as describedabove, have gradually evolved into a set of rel-atively consistent basic tenets or assumptionsthat guide Government’s actions. OTA de-rived the science policy statements in table 46primarily from the practices and behavior ofU.S. policymakers and institutions over thepast 25 years, as well as from published state-ments and policies and interviews with sciencepolicy analysts. Although these principleshave been relatively stable, they may contra-dict one another and come into conflict in spe-cial cases, and exceptions could certainly befound for each item in the list. Furthermore,they have rarely been stated explicitly; in-stead, they are embodied in a diverse collec-tion of decisions, practices, and legislation.While table 46 is in no way a complete set ofthe principles which guide U.S. science policy,the essential tenets relevant to informationtechnology are included.

Each of the science policy statements dis-played in table 46 has varied in importanceand salience in driving programs, projects, andorganizational relationships. Often, the proc-esses by which a policy issue is resolved re-sult in an overcorrection of some situationwhich, in turn, later leads to the recognitionthat the pendulum has swung too far. For ex-ample, the advent of Sputnik was perceivedto indicate that support for basic science hadbeen too weak. On the other hand, the Mans-


Table 46.–Tenets of Science and Technology Policy, 1960-84

Basic Research1. Basic research is a Federal mission.2. The best model for conduct of the basic scientific enter-

prise is physical science, and in particular, physics.3. Peer review will be the primary means for selecting topics

for basic research. Management concerns will play a rolein more mission oriented research.

4. Manpower for the scientific enterprise will be producedprimarily as derivative of, and as an intimate part of, basicresearch at universities.

5. Social sciences will flourish under the traditional (physi-cal science) model of scientific research. Social and in-terdisciplinary research are keys to the more effective ap-plication of knowledge to many classes of societalproblems.

Mission Agency R&D6. There is a useful and significant distinction between basic

and applied research and between research, engineering,and technological applications. These distinctions are ofprimary value in defining the role of Government in rela-tion to the general economy and the role of Governmentagencies in relation to their missions and to each other.

7. Mission agencies will define their knowledge needs whichmay be satisfied through R&D and present their casethrough the budget process.

8. Federal agencies are expected to undertake research insupport of the commercial, business, and private sectorinsofar as support of that research will yield substantialpublic benefit, especially to the clients and constituentsof that agency. Support is encouraged only in those caseswhere research to satisfy nongovernmental needs is un-likely to be adequately sustained by private initiative.

Defense R&D9. Defense research, although a major part of U.S. R&D ex-

penditures, will be treated as an isolated, separate casewith the expectation that side benefits will accrue to thelarger scientific and industrial community.

10. DOD will have a restricted and limited role in support ofbasic and social research at universities. This policy,manifested in the Mansfield amendment under the re-newed pressures of the Cold War, has been relaxed.

Organization of Government11. Voluntary coordination, rather than legislative or cen-

tralized control and coercion, will be the primary instru-ment by which programs in and among agencies will beintegrated, Coordination will be a primary mechanism forassuring completeness of coverage of essential fields andthe primary instrument for reducing overlap and redun-dant budgets and programs.

12. At the Executive level, the Office of Management and Bud-get will exercise its statutory role in assuring that mis-sion needs are met and that research and developmentprograms are reasonable and realistic. There is also a rolefor a White House science policy advice mechanism.

13. In the Congress, oversight, both general and budgetary,will be the primary technique by which quality, complete-ness, and fullness will be assured.

14. Planning for science and setting the agenda for scienceand technology are best handled by the mission agenciesor the specific disciplines.

15. Public and stakeholder participation in science and tech-nology decisionmaking is appropriate, desirable, and en-couraged.

Special Federal Roles16. The Federal Government will help assure the strength of

the research system by collecting, analyzing, and dis-seminating information on subjects such as science,scientific and engineering manpower, and technologicalinnovation.

17. National laboratories are general assets to the nation, well

18.

19.

20.

21.

22.

beyond the particular missions for which they were estab-lished.

R&D FundingThe scientific community may operate on the assump-tion that there is a firm long-term implicit commitmentto incremental funding increases.To avoid disturbances in the established pattern of sup-port for science, the identification of new problems,issues, and options will be handled primarily by budgetaugmentation, rather than by reprogramming of existingprograms. The primary instrument for effective infusionof money in large quantities for new scientific enterpriseswill be the establishment of an office, a bureau, an agency,or a division.In most fields, the most appropriate method of supportwill be funding individual projects by individual in-vestigators.On large expensive basic science projects, the Federalrole is to provide large block funding and long-term sup-port. It will stand clear of the programmatic side of thoseactivities.

Nonfederal R&DBoth basic and applied research in the commercial sec-tor is best and most effectively handled by individual cor-porations and will best prosper under competition. To fa-cilitate that development certain public strategies, suchas patents, copyrights, tax write-offs, and a variety of othermeasures are appropriate for Government. American com-mercial research requires no particular Government in-tervention, attention, or assistance, since it can cope withany foreign competition.

23. Applied research ”applicable to the private and nonfederalpublic sector requires little attention. It will take care ofitself.

24. Good relationships between universities and industry arebeneficial to both institutions, and Government will actto support such relationships, but not directly intervene.

25

26

Utilization of ResearchThe free and open dissemination of research results, ex-cept those of a commercial proprietary sort or affectingnational security, is the best guarantee of the effectiveuse of new knowledge in the service of the nation.With regard to basic research, technology transfer, thatis, the practical use of research results, is best handledby the delivery of scientific information through journalsand monographs. Commercial use best occurs throughscientific channels, through the employment of univer-sity scientists as consultants, and through private sec-


Table 46.—Tenents of Science and Technology Policy, 1960-84—continued

tor organizations assuming responsibility for remaining be placed upon trade and exchange where appropriate,alert to developments in their own interests. primarily at the discretion and behest of the national secu-

27. Mission agencies have the primary responsibility for get- rity establishment.ting the scentific and technical results of research to po-tential users within their mission areas. Technology

28. It is in the national interest to deprive Iron Curtain coun- 29. Policy toward technology is unnecessary or will be treated

tries of the benefits of Western, that is American, science as an adjunct to science policy. This policy has, of course,

and technology. Consequently, systematic restraints will come under some challenge in the last few years.

SOURCE: OTA analysis, interviews with science policy experts, and synthesis of published materials. See footnote 3. Note that these are statements of underlyingpolicies, and there are exceptions and contradictions to each statement

field amendment6 was a rebuke to DOD forobscuring the distinction between basic andmission-oriented research, but it also swungGovernment away from the Sputnik-inducedchanges and back toward accountability andstrictly defined applied research.

Nevertheless, most of the principles high-lighted in table 46 have remained effective andfunctional over the past quarter-century. Asgeneral science policies they influence researchand development in the area of informationtechnology. While many of these influences aresubtle or indirect, the principles shown in table46 can be seen in the current situation. For ex-ample, the U.S. Government’s position towardthe global market, reflecting the propositionsabove, is that industrial competition will dealeffectively with issues of international com-petition and no special Government policy isrequired. This is in sharp contrast, of course,to Japanese and European strategies, as notedin chapter 7, and this contrast has intensifiedthe debate about appropriate Federal roles in

—. —.. .—‘As described in W. C. Boesman, “U.S. Civilian and Defense

Research and Development Funding,” Congressional ResearchService, Science Policy Research Division, Aug. 29, 1983, p. 23:

Even after the establishment of the National ScienceFoundation, whose mission is the support of basic andapplied research and education in the sciences, DOD con-tinued to fund a significant amount of basic research un-til, in 1969, the Congress passed the “Mansfield amend-ment” to the fiscal year 1970 military procurementauthorization which prohibited funds authorized by thatlaw from being used to conduct R&D not having “a di-rect and apparent relationship to a specific military func-tion or operation. ”

The following year, the Congress passed the “modifiedMansfield amendment” to the fiscal year 1971 militaryprocurement authorization which prohibited funds au-thorized by that “or any other Act” from being used toconduct R&D unless the Secretary of Defense determinesthe existence “of a potential relationship to a militaryfunction or operation. ”

international competitiveness. While these arelargely issues of trade policy, there are strongconnections between trade and science pol-icies, especially in information technology.

Other aspects of policy toward informationtechnology R&D that stem directly from thesegeneral science policies include the separationof military and civilian research in informationtechnology R&D, and the implicit belief thatthe market process will take care of the down-stream social effects of information technol-ogy. The next section of this chapter will setforth policies toward information technologyR&D in more detail.

Since 1960, several important trends andforces have affected both general science pol-icy and policies toward information technol-ogy R&D in particular. Table 47 highlightssome of these forces. As is evident from thesecond column of table 47, many of the forcesaffecting science policy generally have beenparticularly prominent in information technol-ogy R&D. In some respects, policy toward in-formation technology R&D is the leading edgeof issues in science and technology policy. Thisis in large part due to the rapid pace of changein information technology, the emphasis placedon information technology for economic growthand for national security, and the pervasiveor “core” nature of the technology.

Information Technology R&D Policies

Like general science and technology policies,policies toward information technology R&Dare not often explicit or coordinated. Insteadthere have been many decisions, actions, statements, and organizations, which taken togeth-


Table 47.— Forces Affecting Science and Technology Policy Since 1960

Trend or force Implications for R&D in in format ion technology

1. Growing pervasiveness of science and technology insociety. This is accompanied by rapid blurring oftraditional distinctions between basic and appliedresearch, and between science and technology.

2. /integration of the global economy. This isaccompanied by an increase in internationalcompetition, a challenge to U.S. supremacy in certainresearch areas, and a growing consensus that U.S.industries are not invincible and may need help.

3. The shifting role of the Department of Defense. D O Dsponsorship of R&D was dominant in the post warera, then was shifted away from basic research andother agencies played stronger roles. Now DOD isonce again dominant in most areas of R&D funding,although the funding is much more directed than itwas after World War Il.

4. The side effects of technology. The public seems tohave grown increasingly wary of technology,particularly in the ’60s and ‘70s. At the same time,science and technology are viewed as a way out ofour economic malaise.

5. Big budgets for R&D. Demand for accountability hasgrown as R&D budgets have swelled and agencieshave undertaken major projects (e.g., accelerators,weapons systems).

6. Internal upheaval in the science enterprise. Thedecade-long search for a more effective sciencepolicy apparatus has bounced around government,focusing at various times on agencies such as NSFand OSTP. None have been conspicuously effective ina broad-scale science policy role.

Information technology is one of the most vivid examplesof this growing pervasiveness. The intertwined natureof basic and applied work, and of information scienceand technology, raises questions about appropriateFederal roles, which have traditionally been based onthose distinctions.

Information technology is an area in which thesechallenges have become quite intense: while we stilllead in most areas of R&D, our lead is narrowing, andour ability to use our technological leadership inapplications for economic gain is in question. Themargin of error for actions in information technologyR&D has been dramatically reduced because ofinternational competition.

DOD was an early and strong supporter of many areas ofinformation technology. 70-80 percent of Federalfunding for R&D in information technology now comesfrom DOD. In certain areas (e.g., artificial intelligence,software engineering), DOD continues to be a verystrong influence on the directions for R&D.

Though there are concerns about privacy and equityissues, use of information technology seems to beviewed as inevitable, and, in many cases, desirable.R&D in information technology, as the basis forinnovations, is viewed as essential to support aneconomy heavily oriented toward high technology.

The demand for accountability has just begun ininformation technology R&D, particularly in majorsoftware projects, or use of supercomputers.Universities in particular are squeezed by rapidly risingcosts for this research.

Information technology R&D is acutely affected by themultiplicity of agencies and roles in setting policy.Because of the technology’s pervasiveness, more thana dozen agencies set policy for R&D in informationtechnology. None of them have devoted high-levelpolicy attention to information technology; it tends tobe viewed as a tool.

—.

SOURCE OTA analysls, !ntervlews with science and technology policy experts, and synthesis of published material. See footnote 3

er comprise the de facto U.S. information tech- judgment that they are necessarily appropri-nology R&D policy. Using techniques similar ate at present or in the future.to those used to develop table 46—i.e., analy-

—1. The Federal Government has operatedsis of published material and the actions of

policymakers, and discussions with policy ex- under the assumption that the UnitedStates should not be dependent on foreignperts—OTA produced a list of policies specif-

ically related to information technology R&D, information technologies to ensure its na-tional security.which are displayed in table 48. Each of these,

elaborated below, can be seen to stem rela- A primary mission area for information tech-tively directly from one or more of the tenets nology R&D is national security. Defenseoutlined in table 46. Note that the following spending dominates Federal support in thisare statements of the effective principles that area of R&D. The U.S. position as leader ofappear to underlie Government’s actions over the Western military alliance has led to a com-the past quarter-century. As in the statements mitment to keep the United States at the fore-of general science policies, OTA has made no front of information technology developments.


Table 48.-Federal Government Policies Towardinformation Technology R&D

1. The Federal Government has operated under the assump-tion that the United States should not be dependent onforeign technologies to ensure its national security.

2. Information technology R&D has been funded separatelyfor civilian and military applications.

3. R&D priorities have been set in Government by missionagencies, and in commercial application areas by theprivate sector.

4. The Federal Government has assumed that it should pro-mote continuous innovation in information technology.

5. The market has been assumed to be the best mechanismto bring the civilian benefits of R&D in information tech-nology to society.

6. The market has been the primary means to attend to theconsequences and effects of information technologies.

7. Where necessary, the Government has used traditionalmeans for regulating the behavior of firms in informationtechnology industries.

8. The short- and long-term manpower needs of informationtechnology R&D have been addressed through traditionalmeans.

9. Government has followed industry’s lead in setting stand-ards except where Government is a dominant purchaser.

10. The Federal Government has assumed that free tradepolicies benefit the United States in the long term.

11. U.S. Government has restricted the export of sensitivetechnical information, as well as advanced informationtechnology itself, to Eastern Bloc nations.

12. The primary international role for the U.S. Government ininformation technology has been to promote equitableuse of common global resources.

SOURCE: See text.

2. Information technology R&D has beenfunded separately for civilian and militaryapplications.

This policy is not unique to information tech-nology R&D. It assumes that a useful distinc-tion can be drawn between civilian and mili-tary uses of information technology. It alsoassumes that there is little overlap betweenthe civilian and military uses in this area, andthat where such overlap exists, as in weatherforecasting, the results of military R&D willfind their way into commercial uses. In a fewcases, there are small transfers of funds frommilitary to civilian agencies performing R&D.

3. R&D priorities have been set in Govern-ment by mission agencies, and in commer-cial application areas by the private sector.

Government sees information technology asa tool. Therefore, information technology R&Dis decentralized. Each agency sets its own

R&D priorities—the National Weather Serv-ice, U.S. Postal Service, the Department of theTreasury, the Federal Aviation Administra-tion, the Federal Reserve Board, and so on.This area of R&D has not received high levelor coordinated policy attention.

4. The Federal Government has assumedthat it should promote continuous growthand change in information technology.

Innovation in information technology is viewedas overwhelmingly beneficial to society. Onthe military side, Government seeks continualinnovation in order to keep ahead or abreastof potential adversaries. On the civil side, thecontributions of information technology toproductivity argue for continued advances tokeep the U.S. economy prosperous. On theother hand, the concentration on innovationtends to shift attention, especially within Fed-eral R&D, to new and glamorous technologiesand away from improving or reducing thecosts of existing technologies.6

5. The market has been assumed to be thebest mechanism to bring the civilian bene-fits of R&Din information technology tosociety.

For the most part Government policy hasassumed that the market will identify andmeet the needs of society for information tech-nology and that the market will make the ap-propriate investments in R&D to meet thoseneeds. Similarly, Government policy has as-sumed that industry will develop the support-ing technologies and infrastructure such assoftware quality control processes as part ofmeeting the market’s needs.

Government frequently encourages innova-tion in the private sector through indirectmeasures such as procurement and tax allow-ances. In cases where developments are impor-tant to the national interest, such as nationalelectronic mail, Government has used R&D

‘See L. Thurow, “The relationship between defense-relatedand civilian-oriented research and development priorities, ” inPrion”ties and Effia.encyin Fderal Research and Development,a compendium of papers submitted to the Subcommittee on Pri-orities and Economy in Government of the Joint EconomicCommittee of the Congress, Oct. 29, 1976.


contracts to promote innovation while limitingthe risks to industry.

6. The market has been the primary meansto attend to the consequences and effectsof information technologies.

This policy assumes positive impacts will re-sult in new markets, while producers will re-duce adverse impacts in order to remove im-pediments to present and future applications.Protection of individual rights is the major ex-ception to the reliance on the market. The Con-gress, the executive branch, and the courtshave all attempted to cope with the issues ofindividual privacy, intellectual property, andfreedom of access. For the most part, the as-sumption underlying their deliberations andactions is that traditional legal, regulatory, ororganizational mechanisms can handle theseissues.

7. Where necessary, the Government hasused traditional means for regulating thebehavior of firms in information technol-ogy industries.

Historically, the Government seems to haveassumed that there is nothing special aboutinformation technology industries. Govern-ment has not seen antitrust, patent, tax andother regulatory policies as major impedi-ments to innovation. For the past half-century,Government has assumed that: 1) regulatedmonopolies such as AT&T are effective per-formers of R&D; and 2) developments in reg-ulated and unregulated areas of telecommu-nications and computers are not in conflict.

8. The short- and long-term manpower needsof information technology R&D have beenaddressed through traditional means.

As noted in chapter 5, the Government hasrelied on the universities to meet the needs ofthe market for the trained scientists and engi-neers necessary for innovation in informationtechnology. Support of research in informationtechnology at universities is the primary meth-od by which the Federal Government supportsmanpower development in the field. Govern-ment has also assumed that the universities,assisted by various subsidies, will make the

necessary investments in equipment to pro-vide the appropriate training for these futureinformation scientists and engineers.

9. Government has followed industry’s leadin setting standards except where Govern-ment is a dominant purchaser.

Government treats information technologylike any other industrial product in terms ofstandards, relying mostly on voluntary indus-try standards. When Government does get in-volved in standards-setting, it is usually at therequest of industry. In the computer field, theInstitute for Computer Sciences and Technol-ogy (ICST) at the National Bureau of Stand-ards is responsible for developing standardsfor the Federal Government, and it also par-ticipates in and coordinates a variety of indus-try standards efforts. In certain cases such ascomputer networking standards, ICST hastaken a firm leadership role, both domesticallyand internationally.7

10. The Federal Government has assumedthat free trade policies benefit the UnitedStates in the long term.

From transistor radios to microchips, the U.S.Government has maintained a position of freetrade in the area of information technology.The assumptions underlying this policy havebeen: 1) free trade will open up foreign mar-kets for U.S. products, 2) the U.S. lead in in-formation technology is largely unassailable,3) the marginal benefits of lower costs to theconsumer outweigh the threat to U.S. indus-try, and 4) competition promotes innovation.

This openness has included access to R&Dthrough the published literature, licensing,joint ventures, and other commercial routes.

11. The U.S. Government has restricted theexport of sensitive technical information,as well as advanced information technol-ogy itself, to Eastern bloc nations.

The importance of information technologies toU.S. national security has led to Governmentactions to preserve the superiority of U.S. in-

7J. H. Young, “Effects of Standards on Information Tech-nology R& D,” paper prepared for OTA, Nov. 25, 1983.


formation technology. One way is to preventits adversaries from getting access to, for ex-ample, certain research publications, advancedchip designs or cryptography software. Thispolicy assumes that: 1) the Government caneffectively control the international operationsof U.S. researchers and firms, 2) these controlswill not unduly harm the viability of the U.S.information technology industry, and 3) theadvantage of such controls outweighs harmdone to the U.S. R&D enterprise through re-striction of information flow among re-searchers.

12. The primary international role for the U.S.Government in information technologyhas been to promote equitable use of com-mon global resources.

U.S. actions regarding international use of in-formation technologies have focused on issuessuch as spectrum allocation, the use of the geo-synchronous orbit, the international use ofcommunications satellites, and access to datafrom weather and other Earth applicationssatellites.

Part II: Key Issue AreasThe previous section provided background

on the nature of policies related to informationtechnology R&D, and on the connections be-tween those policies and broader science andtechnology policies. A central conclusion isthat information technology R&D has been in-fluenced to a substantial degree by policiesapplicable across many areas of science andtechnology. However, factors such as the re-liance on information technology for economicgrowth and national security, and the perva-sive, core nature of the technology, are increas-ingly stressing policy toward information tech-nology R&D, focusing attention upon it, andsetting it apart from policy for other areas ofscience and technology.

This section attempts to build on that foun-dation by examining three particular areas ofpolicy toward information technology R&Dthat may be ripe for change or improvement.OTA selected the three issue areas becausethey are key problems for the future develop-ment of information technology R&D. Through-out this report, scores of issues have been iden-tified which, in themselves, merit attention.The three issues discussed in this chapter areoverarching, in that they subsume many of theearlier issues. Addressing these three areascould help set the direction for many of themore detailed issues. The first, and most fun-damental, topic is the organization of Govern-

ment to deal with information technologyR&D. The second issue area is the balance ofcivilian and military funding in this area ofR&D. The final area is international com-petitiveness.

ISSUE A: Organization of Government

Demands for coherent Federal policy to-wards information technology R&D conflictwith the traditional system of pluralist deci-sionmaking by various agencies and the pri-vate sector.

Introduction

The search for coherence and effectivenessin science policy of all kinds has been the ob-ject of many commissions, proposals, legisla-tive initiatives, and reorganization plans. Ahundred years ago the first plan for develop-ing a Department of Science and Technologywas introduced in the Congress.8 Several timesin the 1970s and again within the last year,the idea of such a central department has re-emerged.9 The fundamental issue coming outof all such proposals is whether science is bet-ter managed through a central organization

— —‘For a discussion see A. H. Dupree, Science in the Federal

Government, Harvard University Press, 1953, p. xi.‘See, for example, H.R. 481, The National Technology Foun-

dation Act, introduced Jan. 6, 1983, by Rep. George Brown,et al.

Ch. 8—Information Technology R&D in the Context of U.S. Science and Technology Policy . 291

or as an adjunct to the missions of the Fed-eral Government as reflected by the missionagencies.

While the proposals for a central Depart-ment of Science and Technology have con-sistently failed, the Nation has reorganized itsscience and technology apparatus severaltimes to meet new and emerging needs. TheDepartment of Energy, the EnvironmentalProtection Agency (EPA), and the NationalOceanic and Atmospheric Administration(NOAA) all resulted from the coalescence orrearrangement of diverse scientific and tech-nical functions.

Nevertheless, the traditional Governmentorganization for science in general is decen-tralized, pluralistic, and only loosely coordi-nated. The strong centralizing force on sciencecomes from the annual budget review by theOffice of Management and Budget. Eventhere, concern for scientific issues is for themost part split up along agency lines; for ex-ample, NASA is in one area, NOAA in an-other, and Defense science in still another.This pluralistic system of Federal policymak-ing has the advantages of allowing missionagencies to tailor R&D to their own needs,which is a basic tenet of science policy as dis-cussed earlier. However, several trends areputting stress upon the ability of the currentdecentralized system to cope with new andemerging problems:

● Increasing international competition, andthe presence of coordinated technologypolicies among our trading partners, arehighlighting our absence of coordinationand causing many to call for reexamina-tion and change; and

● The costs of R&D of all kinds have risen.At the same time, there is increased pres-sure on the Federal budget from entitle-ments, defense, and the deficit. Some ar-gue that a more coordinated and coherentFederal science and technology apparatuscould be more cost-effective and ac-countable.

There is a broad spectrum of possibilities forcoordination of Federal activities in science

and technology areas, ranging from completedecentralization and pluralism to a centralagency which handles the bulk of R&D fund-ing and science policymaking. Despite the ap-peal of coordination in principle, it has coststhat include decreased flexibility of missionagencies, creation of cumbersome bureaucra-cies, and potential loss of multiple fundingavenues—and hence multiple approaches—forresearchers. One report notes:

Coordination is like motherhood; everyoneagrees it must be done but it lacks an opera-tional definition.

Coordination is not a homogeneous activity;but rather an umbrella which encompassesmany different activities performed by dif-ferent people, for similar effect.

Coordination requires significant effort atall levels of management and, therefore, bothhorizontal and vertical structures need to beconsidered.

A certain amount of coordination is goodfor the health of Government, but like exer-cise, too much will cripple or kill.10

Consequently, most science policy experts ar-gue that some combination of centralized deci-sionmaking, ad hoc coordination, and missionautonomy is appropriate.ll

Dimensions of the Issue

Current responsibilities for informationtechnology are dispersed all over Govern-ment—from the Department of Defense to theGeneral Services Administration, from the De-partment of Justice to the Federal Communi-cations Commission. The ad hoc nature of pol-icy in this area is even more evident than mostother types of science policy because agenciestend to see information technology as a tool,not as something warranting significant pol-icy attention in itself. In addition, there aresimply more agencies involved because of thepervasive nature of information technology.

10W. A. H~n, D. S. Alberts, and J. Lovelace, “InteragencyCoordination: Workshop Report,” in The Management of Fed-erai Research and Development: An Analysis of Major Issuesand Processes (McLean, VA: The Mitre Corp., 1977), pp. 93-97.

*’See, for example, Harvey Brooks, The Government ofScience (Cambridge, MA: MIT Press, 1968).


More than a dozen agencies fund or affect rele-vant R&D.12

There is, however, some coordination in Fed-eral policy. To the extent that DOD dominatesR&D funding, it is the de facto lead agencyand informal or formal coordination point.And agencies often coordinate their work onan ad hoc basis. For example, the Defense Ad-vanced Research Projects Agency (DARPA)and NSF brief each other on computer scienceresearch programs. 13 And ICST at the N a t i o n -al Bureau of Standards has a group of seniorofficials from major agencies who advise theInstitute on its programs.

Several specific factors focus attention onthe degree of coordination and coherence inpolicy toward information technology R&D.14

Among them:

In general, there is a lack of high-level pol-icy commitment in this area, which hasseveral kinds of effects. One is that therole of mission agencies in informationtechnology is shifting, uncertain and asdivergent as the roles of those agenciesthemselves. Coordination would be usefulin such subjects as database collection,R&D research topics, and compatibilityof technology and information.Many have argued that there are substan-tial shortages of manpower. However, asdiscussed in chapter 5, the evidence forsuch shortages is inconclusive, except in

‘These include the Department of Defense (itself divided intothe Defense Advanced Research Projects Agency, the Officeof Naval Research, the Air Force Office of Scientific Researchand various other units in the Pentagon and the three services),the National Science Foundation, the Department of Commerce(largely through the National Bureau of Standards, the NationalOceanic and Atmospheric Administration, and the NationalTelecommunications and Information Administration), the Na-tional Institutes of Health, the Department of Energy, the Na-tional Aeronautics and Space Administration, the Federal Avia-tion Administration, the Patent and Trademark Office, and theDepartment of Justice (both in their jurisdiction over antitrust,and in R&D for law enforcement systems).

‘SElias Schutzman, National Science Foundation, personalcommunication, Mar. 2, 1984.

14For fither elaboration of some of these factors ~d discus-

sion of institutional options, see “Institutional Options for Ad-dressing Information Policy Issues: A preliminary Frameworkfor Analyzing the Choices, ” a staff memorandum prepared bythe Communication and Information Technologies Program ofOTA, NOV. 29, 1983.

●

●

●

certain very specific areas. The lack ofreliable assessments of manpower and theassociated uncertainties hinder policy-making in all areas of the Governmentthat work with information technology.The Federal Government has an exten-sive network of national laboratories, al-though the quality and relevance of someof these facilities has periodically been inquestion. 15 Researchers at various na-tional laboratories constitute the largestconcentration of expertise in use of super-computers. The question of how best touse the national labs in this and otherfields of information technology R&Dcuts across a variety of agencies, in par-ticular the Departments of Defense andEnergy.Related to use of the national labs, chap-ter 3 pointed out that researchers are in-creasingly requiring advanced computersor “supercomputers” to perform a widevariety of research. The question of whereto house such machines and how to pro-vide access is of concern to a wide vari-ety of agencies involved with informationtechnology R&D. Committees of the Fed-eral Coordinating Council for Science,Engineering and Technology (FCCSET)at OSTP have attempted to address thisissue.As discussed in chapters 2 and 3, theemerging shared wisdom in the industryis that software is a key problem in costand effectiveness of computing systems.Many believe that American industry hasfailed to give adequate research attentionto software problems, preferring for a va-riety of reasons to emphasize hardware.The reliability and maintainability of soft-ware will become an increasingly largeissue, raising questions of quality control,standards, manpower, and education pol-icy for many agencies in the Federal Gov-ernment with large information systems.

“See the Report of the Wlu”te House Su”ence Council FederalLaboratory Review Panel, May 1983, sponsored by the Officeof Science and Technology Policy. (Also called “The PackardReport, ” after its chairman, David Packard.)


● In the area of regulation, forcing newtechnology into old categories is nearlyuniversal because the new is not alwaysseen as new, or the new is seen as a prob-lem, not an opportunity. For example,cable television throughout the countryis being treated as a local utility occupy-ing some status resembling, perhaps, elec-tric power. Each local government treatswhat could be an integrated national in-formation utility on a short-term andsomewhat parochial basis. In sharp con-trast, Canada and France have adoptedpolicies toward cable that aim to developa national utility.

● Uncertainties about funding levels for in-formation technology R&D contrast withthe needs for stability in budgets or sup-port as a base for long-range research inuniversities and industry. Examples ofsuch uncertainty include some recent va-cillations in funding of certain informa-tion technology areas by DOD agencies,particularly DARPA; and the currentuncertainties concerning supercomputerresearch support between DARPA andNSF. (See ch. 3).

● Information technology industries arenow combining technologies developedunder regulation (as in radio, telephone,and television) with computer technolo-gies basically developed in the marketsystem. The convergence of these twotypes of technologies creates new regula-tory issues dealing with ownership, pub-lic versus private control, privacy, andaccess.

Options for Addressing the Issue

As a core technology, information technol-ogy is used by everyone but is not clearly theresponsibility of anyone. Yet, given its valueto the balance of trade and productivity ofU.S. industry, the demands for new, morecoherent action have become increasinglystrong. l6

“%, for example, Science Policy Research Division, Congres-sional Research Service, The Information Science and Technol-ogy Act of 1981, June 1982.

Option 1: Maintain the Status Quo. Thestrongest argument for no major change inFederal activity is that the technologies andtheir effects are highly fluid, and it maybe tooearly to devise appropriate policy or Govern-ment organizations. In addition, some mayview coherent, coordinated Federal policytoward information technology R&D as unnec-essary or infeasible. The ad hoc coordinationcurrently used has been relatively effective,and attempting more formal coordination ormore elaborate national policy could be cumber-some. In addition, pluralistic research fund-ing has the advantage of funding more thanone approach to a research topic or problem.

The disadvantage of maintaining the statusquo is that we may reduce opportunities to en-hance our competitiveness and to use ourR&D resources in a more socially productivemanner.

Option 2: Improve Monitoring and Coordina-tion. A first step toward coordination of Fed-eral roles would be to provide new mechanismsfor the various agencies involved in this areato communicate in a systematic way. Such co-ordination mechanisms would at least raisethe level of attention to information technol-ogy R&D issues and provide a forum whichcould facilitate a common understanding ofareas of strength and weakness in Federal sup-port. Congress could designate a formal coor-dination group with representatives from ap-propriate agencies involved in informationtechnology R&D. In fact, the first priority ofa coordination group could be a report to theCongress, and subsequent hearings, on thoseareas of strength and weakness. Though DODwould be a major player in such a coordina-tion effort, it is important that it not domi-nate; the status, needs, and objectives of thecivil sector should have an adequate platform.

Other coordination and monitoring stepsmay also be desirable. To the extent thatStates play a stronger role in promoting in-formation technology R&D centers, and inusing information technology for delivery ofservices, it may be useful to establish mecha-nisms whereby States and the Federal Gov-ernment can cooperate in setting priorities for

—


information technology R&D. Such mecha-nisms could include a national conference onthe intergovernmental research needs in infor-mation technology R&D, hearings on Stateand local information technology R&D needs,and commissioning studies of the needs ofState and local governments for improved in-formation technology.

Option 3: Set New National Policy. A morecomprehensive alternative is to make a high-level policy commitment to information tech-nology R&D. This could be accomplished byreestablishing an office such as the Office forTelecommunications Policy in the ExecutiveOffice of the President, or elevating the re-sponsibilities, status and visibility of the Na-tional Telecommunications and InformationAdministration and the Institute for Com-puter Science and Technology (ICST) in theDepartment of Commerce. To complementthis action, Congress could establish a leadagency for information technology R&D poli-cy that could devote a substantial amount ofhigh-level attention to the issue. However, theestablishment of a lead agency has the disad-vantage that that agency’s mission may bepursued at the expense of others.17 Note thatin the last several years Congress has beenconsidering various proposals for centralizedoversight bodies for information technologypolicy-la Con=ess may also wish to considerrestructuring the basic oversight mechanismsfor information technology R&D in the Con-gress and/or the executive branch.

Finally, it maybe an opportune time to takeaction on several more detailed issues. The onethat is most prominent is software, as dis-cussed in the chapter 3 case study of softwareengineering. Federal standards for supporting,using, testing, updating and documentingsoftware could add much reliability to Gov-ernment information systems and consistencyto relations between the Government and in-dustry. One mechanism for dealing with theseissues is to work through ICST.

“See Brooks, op. cit.‘Wee “Institutional Options for Addressing Information Pol-

icy Issues, ” op. cit.

Option 4: Establish a New Federal Organiza-tion. Congress could create a new organization,transferring to it much of the current dis-persed responsibility for information technol-ogy and adding new functions. These newfunctions could include compiling and inter-preting information on Federal procurementof information technology, civilian vs. militarypriorities in R&D, regulatory actions with di-rect or indirect effects on the technology, theU.S. position in domestic and internationalmarkets, social impacts of information tech-nology, high priority issues to be resolved, andrecommendations for congressional action.The advantage of such a new organizationwould be that it would assure that the tech-nology would be visible and explicitly ad-dressed; on the other hand, it could diminishthe effectiveness of other organizations thatpursue information technology R&D as partof their mission. A new organization could bepart of a new “National Technology Founda-tion, ” or it could be a freestanding “Instituteof Information and Communication. ”

ISSUE B: Military/Civilian Balance

Relying primarily on DOD for funding ofinformation technology R&D may conflictwith the pressing demands of internationalcompetitiveness and productivity.

Introduction

The Department of Defense (DOD) has beenby far the largest supporter of informationtechnology R&D among Federal agencies.With increasing budgets for R&D in the re-cent past, DOD is sponsoring a higher propor-tion of many fields of R&D activity. The dom-inance of DOD in information technology isperhaps the most striking of all, however; esti-mates of the proportion of DOD funding rangefrom 70 to 80 percent or more of all Federalfunding.19 In some parts of the field, DOD hassponsored pioneering work which established

‘These estimates are based on W. C. Boesman, “U.S. Civil-ian and Defense Research and Development Funding, ” SciencePolicy Research Division, Congressional Research Service, Aug.29, 1983. Also see ch. 2 for further discussion.


foundations for both commercial and militaryapplications. However, with tightening budg-ets and growing concern over internationalcompetitiveness, the wisdom of DOD’s contin-ued dominance of information technologyR&D funding is coming into question. Specif-ically, three questions have surfaced in thecourse of OTA’s study:

1. Is the military work siphoning off toomuch talent from civilian applications?

2. Is the military work changing the direc-tion of research in information technologyin ways that are disadvantageous for thecommercial sector or for the public?

3. Are existing efforts to transfer technol-ogy from military to commercial applica-tions adequate?


DOD and civilian agency funding of R&Dhave varied in relative emphasis and roles overthe past decades. The issue of DOD vs. civilfunding of R&D has received little emphasissince the late 1960s, when DOD R&D wasdrastically reduced because of a perceptionthat the agency had overstepped its mandate,and because of social concerns about the DODbudget. As shown in figure 50, in the past dec-ade (and particularly during the Reagan ad-ministration) DOD funding for R&D of allkinds has risen dramatically faster than civil-ian agency funding, which has actually droppedin real terms. It can be misleading to use thecombined term, R&D, in this discussion; as fig-ure 51 shows, for all fields combined, thedramatic increase has been almost exclusivelyin development, rather than in basic or appliedresearch.

More specifically, DOD support for work ininformation technology, particularly throughthe Defense Advanced Research ProjectsAgency (DARPA), has remained strong andhas grown dramatically. As shown in table 49,support for basic research in mathematics and

Figure 50.—Federal R&D Budget Authority forDefense and Nondefense Activities

ars

1975 1977 1979 1981 1983 1985Fiscal year (est.)

SOURCE: Division of Science Resources Studies, National Science Foundation“Federal R&D Funding: The 1975-85 Decade, ” March 1984

Figure 51 .—Federal R&D Budget Authority forDefense Activities by Character of Work

1975 ’77 ’79 ’81 ’83 ’85

c

6

3 Applied research

L0

Basic research

1975 ’77 ’79 ’81 ’83

Fiscal year (est.) Fisca l year (es t . )

SOURCE: Division of Science Resources Studies, National Science Foundation,“Federal R&D Funding: The 1975-85 Decade,” March 1984

38-802 0 - 85 - 20

— . ”

—


Table 49.—Department of Defense Funding for Basic Research by Discipline,Fiscal Years 1982, 1983, and 1984 (budget authority In millions)

Fiscal year Fiscal year Fiscal year Fiscal year1982 1983 1984 1985

Physics, radiation science, astronomyand astrophysics . . . . . . . . . . . . . . . . . . . . 76.0 80.9 87.2 96.6

Mechanics, aeronautics, andenergy conversion . . . . . . . . . . . . . . . . . . . 73.1 79.5 86.3 92.2

Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71.5 81.0 82.8 87.5Electronics . . . . . . . . . . . . . . . . . . . . . . . . . . . 90.0 90.5 97.9 93.7Oceanography. . . . . . . . . . . . . . . . . . . . . . . . . 51.1 50.2 53.4 57.5Biology and medical sciences . . . . . . . . . . . 64.9 66.3 79.8 86.7Chemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53.1 58.9 62.0 66.3Mathematics and computer sciences. . . . . 83.3 98.8 111.7 124.9Terrestrial sciences, geophysical

research . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24.3 29.0 30.8 33.9Atmospheric sciences . . . . . . . . . . . . . . . . . . 20.8 21.8 25.0 28.2Behavioral sciences, human resources . . . 33.9 33.6 35.2 36.3Special studies . . . . . . . . . . . . . . . . . . . . . . . . — 2.0 —University instrumentation . . . . . . . . . . . . . . — 30.2 30.0 -300In-house laboratory independent

research , . . . . . . . . . . . . . . . . . . . . . . . . . . . 54.1 57.5 58.4 67.0

Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 696.1 780.0 839.3 899.9SOURCE: Leo Young, Department of Defense, presentation to AAAS Colloquium on R&D Policy, Mar. 29, 1984.

computer science grew from $83.3 million infiscal year 1982 to a planned $124.9 millionin fiscal year 1985. In applied research and de-velopment, several major projects at DODhave pumped many hundreds of millions ofdollars into information technology. Theseprojects include Very High Speed IntegratedCircuits (VHSIC); Command, Communica-tions, Control, and Intelligence (C31); and morerecently the Strategic Defense Initiative (SDI)or “Star Wars” program, and the StrategicComputing program.

Science policy experts interviewed by OTAwere almost universally concerned about thisresurgence of DOD funding for R&D, and forinformation technology R&D in particular.Comparing the current situation to the post-War era when DOD research funding was alsodominant, they point out that current researchis generally much more mission-oriented and,consequently, less productive for nonmilitaryuses. Some argue that we are endangering ourinternational competitiveness in the long termby monopolizing the information technologyR&D community with defense-related proj-ects. Others point out that it is unwise to havea monolithic source of funding for any area—e.g., certain technical approaches may tend tobe ignored —and argue that the current situ-

ation desperately calls for a civilian balanceto DOD’s funding. Despite these strong warn-ings, however, there is inconclusive evidencethat these negative results of DOD’s fundingare occurring.

For example, in artificial intelligence (AI)the pool of researchers is very small andalmost all receive DOD funding. As noted inchapter 3, DARPA and ONR have been almostthe exclusive funders of artificial intelligencefrom the start. Yet, some AI researchers notedduring OTA’s case study that relatively basicresearch which could lead to nonmilitary ap-plications-such as intelligent libraries-is be-ing neglected. The assumption that AI R&Dfunded by DOD is equally applicable to bothmilitary and civilian applications is, therefore,under question, although more than anecdotalevidence is needed to assess the problem.

Other controversial topics for the sciencecommunity in general are export and publica-tion restrictions on scientific and technical in-formation. The dominance of DOD funding ofinformation technology R&D raises the dan-ger that the research will be classified too earlyto allow nonmilitary users to benefit. This dan-ger is particularly prominent in large scale de-fense initiatives such as the “Strategic Com-

Ch. 8—Information Technology R&D in the Context of U.S. Science and Technology Policy . 297

puting” program. For academic scientists inparticular, free information flow is viewed asessential to productivity and to the ethos ofscience as an international enterprise. Hence,the tension has produced some strong rhetor-ic. A university association president recent-ly told a gathering of the American Associa-tion for the Advancement of Science: “Thesepeople feel that any delay or inconvenience forthe Russians is worth whatever it costs tous . . . What we are seeing now is not disagreements among reasonable people; it is ideolo-gy without restraint, and it is dangerous tous all. ”2°

Again, however, the extent to which restric-tions on information flow have actually beenonerous or counter-productive has not beencarefully examined. More broadly, except forscience policy analysts, most people—at uni-versities, in Congress, or in associations or re-search groups-have not raised DOD fundingas an issue. They maybe comfortable with thecurrent situation, or they may be uncomfort-able alienating a powerful source of funding.

Part of the reason for infrequent question-ing of DOD’s dominance in this area is thatdefense applications for computer-related de-vices are fascinating problems. One computercolumnist noted that state-of-the-art equip-ment, challenging problems, and the mystiqueof “secrecy” are powerful lures for computerscientists. 21

—. —‘“R. Rose~weig, president, Association of American Univer-

sities, address to American Association for the Advancementof Science colloquium on R&D policy, Washington, DC, Mar.30, 1984. This comment was made when the Department of Defense was considering placing restrictions on the publicationof “unclassified but sensitive’ research. At the time, thepresidents of Stanford, California Institute of Technology, andMassachusetts Institute of Technology “warned the ReaganAdministration that their institutions may be forced to stopconducting unclassified research for the Pentagon if they arerequired to give military reviewers the right to restrict pub-lication of some findings. ” (Kim McDonald, “3 UniversitiesWarn Pentagon on Censorship,” The Chronicle of Higher Edu-cation, Apr. 4, 1984, p. 1.) In part in response to this outcry,the Pentagon rescinded its plan for restrictions. (“White HouseDecides to Cool Campus Secrecy Issue, ” Science and Govern-ment Report, June 15, 1984, p. 1), See also Albert H, Teich andJill P. Weinberg, American Association for the Advancementof Science, “Issues in Scientific and TechnicaJ Information Pol-icy, prepared for Office of Special Projects, National ScienceFoundation, Dec. 28, 1982.

2*D. Clapp, “While Japan Builds Computers, We’re MakingMissiles, ” Infoworld, June 27, 1983.


Option 1: Maintain the Status Quo. While itmay be desirable to have a stronger civiliangovernment presence in information technol-ogy R&D, some would argue that our nationalsecurity requirements mandate the currentlevel of DOD involvement, and that it is im-practical or unnecessary for civilian agenciesto have a balancing involvement.

Potential negative consequences of main-taining the status quo are that we may be com-promising international competitiveness, andhence national security, in the long term. Theconcerns surfacing about DOD’s funding ofR&D may be early signals of a serious prob-lem, or they could be insubstantial worries.Currently we do not have reliable informationto tell the difference.

Option 2: Increase Monitoring and Analysis.The clearest need in addressing the impact ofDOD priorities is that it be explicitly addressedand more monitoring and analysis be done.Specific topics in need of monitoring and anal-ysis include:

●

●

●

●

●

Effectiveness and effects of national secu-rity restrictions on access to informationtechnology research and devices—espe-cially the exchange of ideas among lead-ing researchers and the ability to use for-eign graduate assistants on DOD-relatedprojects.Effects of DOD support on the researchpriorities of leading researchers in thefield.Transferability of information technologydeveloped for DOD–the ease of transfer,the time lag—compared with primarilycommercial development.Use of limited manpower in certain fieldssuch as artificial intelligence and softwareengineering.Relation of DOD’s requirements for infor-mation technology R&D to commercialrequirements, and more broadly, thetradeoffs between national security andinternational competitiveness in this area.

One factor working against explicit consid-eration of military vs. civilian priorities in


R&D is the fact that Congress is ill-equippedto balance the funding of research among dif-ferent agencies because the agencies’ budgetsare independent and handled by different con-gressional committees. The executive branchis theoretically capable of such considerations,but in practice, as noted above, it also handlesagencies’ budgets as discrete units. Hence,there may be a need for a new mechanism toweigh R&D goals in information technologyfrom a multiagency perspective. Such mech-anisms could include joint congressional hear-ings, activities of the interagency coordinationbody discussed in Issue A, and/or a joint studyby DOD and a prominent civilian group suchas the National Science Board on the relation-ship of DOD spending to R&D priorities. Tobe most effective, such a study should prob-ably be tied to subsequent congressional hear-ings on the issue.

Option 2 does not preclude either of theother options, and hence may be a wise courseof action in any case.

Option 3: Bolster Civilian R&D funding. Con-gress could act to provide a stronger civilianbalance to DOD’s information technologyR&D funding, on the basis of the suggestiveevidence of problems, or on the assumptionthat domination of information technologyR&D by one mission agency is unwise. Thoughsuch a move would require budgetary in-creases of several million dollars, many in-dustry and policy experts suggest that the ul-timate payoffs in innovation and productivitywould be substantial.

Such funding may go beyond some policy-makers’ notion of appropriate roles for Gov-ernment. There does seem to be room, how-ever, for more civilian agency funding of“fundamental” (in the sense of being widelyapplicable and long-term) if not “basic” (in thesense of being disinterested in applications) re-search in information technology. The fund-ing agency involved would have to be carefulthat the research community had sufficientmanpower to absorb such funds. Some expertshave called for a civilian research effort thatwould mobilize the research community in a

way similar to that of the Apollo program—it could be a 5- or 10-year effort toward spe-cific objectives such as uses of computers foreducation, to aid the handicapped or poor, orother social goals.

ISSUE C: International Competitiveness

U.S. policies and practices are based on anassumption of unassailable U.S. dominancein information technology R&D, which is in-creasingly inaccurate.

Introduction

Information technology is an important ele-ment of global high-technology trade. As dis-cussed in chapter 7, the efforts of the French,Japanese, and other governments to target in-formation technologies as tickets to future in-ternational prosperity attest to that fact. Sincethe advent of information technologies, theUnited States has had the lead in developmentand in global market share. That situation ischanging as other advanced nations are in-creasing their patents in international com-merce, as the U.S. balance of trade in infor-mation technology begins to weaken, and asthe industry becomes more global in charac-ter and thus less amenable to traditional meth-ods of governmental control.


Stresses of global integration and foreigncompetition on U.S. information technologyR&D policies emerge in several areas. One isthe effect on policies promoting developmentof the R&D base of U.S. industry-manpower,facilities, R&D information and R&D behav-ior. The issues in this area include:

● Foreign versus domestic high-tech man-power. A large percentage of the graduatestudents in science and engineering areforeign nationals. It has been a matter ofsignificant national pride that the worldcomes to the United States for trainingin science; on the other hand, some haveargued that we are investing resources inthese foreign students which those that


leave then take away to their home coun-try. The value we have placed on scienceas an international enterprise is understress as international competition inten-sifies.

● A related issue is that in those areas thatinvolve national security or commercialsecrecy, there are increasing pressures torestrict access of foreign scientists andgraduate students. Such restrictions couldbe a major dislocation in the ethos of theuniversity.

● Restricting the access of foreign scientiststo U.S. research may isolate the U.S. re-search community. Such isolation couldreduce the infusion of new approaches andideas into the R&D process, and thus hin-der R&D in the United States.

● Our traditional linguistic chauvinism con-flicts with a recognition of the need totranslate literature from other countriesin this area. Little of the Japanese tech-nical literature, and only somewhat moreEuropean literature, is routinely availablein English translations.

● Internationalization of the informationtechnology industry is leading to the glo-balization of R&D. Countries such asGreat Britain, the United States, France,and Italy are competing for the locationof research centers of the major multina-tional firms in information technology.Scotland’s Silicon Glen is an example. Inaddition to the competition for multina-tional R&D facilities, there is growing in-terest in joint ventures among firms in ad-vanced nations such as Japan, Germany,the United States and France.

Issues also emerge concerning U.S. policiesto maintain or improve the current U.S. leader-ship in information technology R&D and mar-keting. These include issues related to thestructure of the industry, the role of DOD, andthe effects of regulation on the industry:

● The individual American corporation inthe information technology market mayconfront foreign government-coordinated,sustained, and supported consortia orconsortia of private companies enjoying

●

●

●

●

●

subsidies and generic research input fromtheir governments.The openness of U.S. markets to foreigncompetition is not met by symmetricalU.S. access to foreign markets. U.S. man-ufacturers still primarily focus on the U.S.market-the world’s largest for informa-tion technology. These firms may not begiving adequate attention to developingnations’ markets, leaving them largely toother nations. For example, the Japaneseare now vigorously pursuing countertradewith the Chinese to exchange mineral re-sources for Japanese high technology.One element of the dominance of DOD ininformation technology R&D is the poten-tial diversion of talent and resources awayfrom nonmilitary science and technology.The rigid specifications and limited appli-cability of many DOD-sponsored technol-ogies could skew the development of U.S.information technologies away from thoseproducts that are of most use in foreignmarkets-especially markets in develop-ing nations.There is a conflict between the need forfree trade and the need to protect sensi-tive science and technology. The issue iswhether U.S. export controls on informa-tion technology are unnecessarily ex-cluding U.S. companies from effectivelycompeting for large foreign markets.With the development of foreign marketsfor information technology, concerns ariseover the access of small and medium-sizedfirms to foreign markets. Some feel thatwhere U.S. firms have penetrated foreignmarkets, the larger firms have dominatedtrade, to the exclusion of the smallerfirms. This is not unusual, since a largemajority of manufacturing exports comefrom large firms. However, with the glo-bal integration of the industry, the UnitedStates may wish to encourage smaller, in-novative firms to seek out foreign trade.Promotion of a rational world system formanaging the use of information technol-ogies is also important to the long-termleadership of the United States. Thoughthis has long been an area of recognized


importance for U.S. policy, a number ofimportant global standardization issuesremain. For example, an increasing issuewill be global compatibility of communi-cations systems. The fact that foreignsystems are generally run by nationalgovernments, while the U.S. system is amarket-based system, tends to put theUnited States in a different, often disad-vantageous, position from all the othercontenders in international negotiations.


The central issue facing information tech-nology industries is how best to enhance theirability to compete on equal footing with com-panies from other nations, especially wherethose companies are strongly supported bygovernment. This problem affects many otherindustries-from steel to shoes. Continuing de-bate over the need for an industrial policyflows directly from this issue. As noted inchapter 7, the essential question is not how toimitate the policy strategies of Japan or othercountries which appear successful, but to comeup with a response that could build on uniqueU.S. strengths.

Option 1: Maintain the Status Quo. Somewould argue that the Federal Government isill-equipped to become more involved in a fast-paced area such as information technology.The current scheme of activities to promoteinternational competitiveness works well insome respects; the prime role of the Govern-ment should be to provide a healthy macro-economic business climate.

The disadvantage of the status quo is theincreasing evidence that the traditional pat-tern of policies related to international com-petitiveness does not allow our industries tocompete on “a level playing field” with com-panies from other countries. At present, thereis little basis for deciding among options fordealing with foreign competition and its ef-fects on information technology R&D. The rel-ative newness of the threat and the rapid ideo-logical polarization of the industrial policy

debate have left the Nation long on conjectureand short on facts.

Option 2: Monitor and Support InternationalTrade in Information Technology, and RelatedEfforts in R&D. Various measures have beenproposed for the support of internationaltrade, and it is beyond the scope of this reportto discuss them in detail.22 A key aspect oftrade support is ensuring that foreign marketsare open to U.S. industry, and helping U.S.companies to actively seek developing mar-kets for the technology. This could involveincreasing the commitment and attention ofthe U.S. Special Trade Representative, the In-ternational Trade Administration, and theForeign Commercial Service to the needs ofinformation technology firms.

Options more specifically related to R&D in-clude promotion of generic information tech-nology R&D centers in the United States, andclose monitoring and evaluation of alternativeinstitutional models-both domestic and for-eign-for cooperative research in informationtechnology.

Further, support for international competi-tiveness in R&D could include establishingmechanisms to monitor foreign technical lit-erature and disseminate translations to Amer-ican scientists and technologists. Such sup-port could also provide funding for our researchpersonnel to travel overseas for conferencesand consultations, and for American studentsor professors to study overseas. Congress maywish to beef up scientific bilateral agreementsand exchange programs.

In addition, it would be appropriate toanalyze:

● The amount of foreign purchasing by theBell Operating Companies that wereformerly part of AT&T. As a vast mar-ket for information technologies, the be-havior of these companies will be criticalto the future of the U.S. industry.

— . — —22See the recent OTA report, International Competitiveness

in Ek+ctroru”cs.


●

●

●

●

The extent and type of foreign ownershipof U.S. information technology firms andthe effects of such ownership on the trans-fer of information science and technology.The career paths of foreign computer andelectrical engineering graduate students.The extent to which major U.S. informa-tion technology firms undertake R&D inforeign countries versus in the UnitedStates.The effects of joint ventures, countertradeagreements, licensing, and other arrange-ments on the transfer of U.S. informationscience and technology.

These information gathering activities wouldbe helpful regardless of the path Congresschooses to take in addressing this issue area.

Option 3: Set National Policy. While there ismuch we need to know, one alternative is tobegin setting a long-term policy on the role of

information technology in U.S. trade, and tocontinue the debate on how the United Statesmight restructure its trade policies to respondto those of Japan and other nations. In addi-tion, the United States could assist the inter-national competitiveness of U.S. firms by de-veloping a national position in internationalstandardization which would take into accountthe needs of Government, the private sector,and the consumers as well as balance short-term needs with the long-term developmentof foreign markets.

The United States could establish a morecoherent policy on the flow of scientific infor-mation, and could establish a review and ap-peal mechanism for DOD’s restrictions on theflow of information and technology.

Options 2 and 3 are not mutually exclusive,and probably make best sense in concert witheach other.

Concluding ThoughtsAs part of the preparation for this chapter,

interviews and workshops were conductedwith several dozen experts in science policyand information technology R&D. Box B is asample of their responses to the question,“What single message would you like to getacross to the Congress concerning informationtechnology R&D?” The diversity of these re-sponses indicates the multifaceted nature ofissues related to information technology R&D.Their responses are reprinted in box B in or-der to illustrate the wide-ranging priorities ofa group of well-informed specialists, and toprovide a different perspective on some of theissues discussed earlier in the chapter.

A common theme in these comments, andin many other discussions of policy in thisarea, is a drive for perspective: for a long-rangeview of technological and social changes, andfor policies that work together effectively ina wide range of areas.

Indeed, U.S. policy toward information tech-nology R&D, as in many other areas, is par-tial and incremental. This lack of long-termperspective may in part be inherent in thepolicymaking machinery; in other cases, pol-icymakers have explicitly assumed that theGovernment will be most effective when itresponds to a mature issue—an issue that hasreached a level of public concern where actionis clearly called for, and the background of theissue is well understood.

However, the nature of a “core” technology,facilitating major and pervasive social changes,raises questions about the utility of partial andincremental policies. Many of the issues evolv-ing from a core technology are likely to evolvelate rather than early, and are likely to bestructural-that is, deeply built into the so-ciety, and hence very disruptive and traumaticto correct.


to do so.

ons Of this country re-en given the latitude

t legislate on matters of R&D except where you

In response to some of these concerns, vari-ous interested parties have called for somekind of prestigious national body which couldhelp sort out issues and lay the groundworkfor a long-term perspective.23 Congress maywish to consider such an option. Although onecould be skeptical about creating another com-mission, other countries have tried variationson this theme with some apparent success indeveloping long-term perspectives.24 In theUnited States, this area of research is ratheranemic. For example, the National ScienceFoundation recently reorganized its Policy Re-search and Analysis Division to address theshort-term needs of the executive branch

“see, for example, J. L. Kirkley, “Backing into the Future, ”Ikkrmtion, February 1982, p. 31; M. R. Wessel and J. L.Kirkley, “For a National Information Committee,” DAuna-tion, 1982, p. 234; David Burnharn, The Rise of the ComputerState (New York: Random House, 1983). For a discussion alsosee OTA, “Institutional Options for Addressing InformationPolicy Issues, ” op. cit.

24% Telecom Australia, Telecom 2000: An Exploration ofthe Long-Term Development of Telecommunications in Aus-tralia (Melbourne, Australia: Australian Government PrintingUnit, 1975); and Nora, S. and A. Mine, The Computerizationof Society (Cambridge, MA: MIT Press, 1980).

—

-

Take “T

rous action now to enhance science educationfor peo e at aU levels and all education. Provide %~~”:strong support for it.

“ .-WXIRC13: UI’A wdshaw ad intervkws.

rather than long-term research on social im-pacts of science and technology .25

Given that such little effort is now beingundertaken to understand the long-term ef-fects of information technology, it is difficultto say whether development of such a perspec-tive would be possible in the United States.However, such efforts probably entail littlerisk, in that any insights derived could helpinform policymakers on information technol-ogy R&D, and on use of the technology itself.An examination and anticipation of the socialand cultural impacts of information technol-ogy could at a minimum suggest avenues toexplore and monitor, possible options to con-sider promoting, and identification of poten-tial developments that one might wish tothwart or prevent.

Z6SW Nation~ Science Foundation, Division of policyResearch and Analysis, “Program Announcement and Solicita-tion, ” January 1984.

- -

Chapter 9

Technology and Industry

— —

ContentsPage

Importance of Information Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307

Composition of Information Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308Functions .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308Examples of System Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309

Characteristics of theU.S. Information Technology Industry . . . . . . . . . . . . . . . . . . . . . 316Employment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317International Trade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317Industry Structure . . . . . . . . . ..., . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . 320Small Entrepreneurial Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......321

Where the Technology is Heading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323Microelectronics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324Silicon Integrated Circuits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324Gallium Arsenide Integrated Circuits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328The Convergence of Bio-technology and Information Technology . . . . . . . . . . . . . . . . 330Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331The Changing Role of Software in Information Systems . . . . . . . . . . . . . . . . . . . . . . . 332The Software Bottleneck . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333Software and Complexity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335

TablesTable No. Page50. Functions, Applications, and Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30951. Characteristics of Human Workers, Robots,

and Nonprogrammable Automation Devices. . . . . . . . . . . , . . . . . . , . . . . . . . . . , . 31252. Comparison of the U.S. Information Technology Industry

With Composite Industry Performance, 1978-82 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31753. Employment Levels in the U.S. Information Technology Industries . . . . . . . . . . . . 31854.U.S. Home Computer Software Sales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32055.U.S. Office Computer Software Sales. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32156. Percentage of Total industry Sales and Employment

by Size of Small Businesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32257. Venture Capital Funding and Information Technology Firms, 1983 . . . . . . . . . . . . . 32358. ICs Made With Lithographic Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32659. Software Sales by Use , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332

FiguresFigure No. Page52. Direct Broadcasting Satellite System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31153. The Changing Structure and Growth of the

U.S. Information Technology Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31854. Balance of Trade, Information Technology Manufacturing Industry . . . . . . . . . . . . 31955. U.S. Trade Balance with Selected Nations

for R&D Intensive Manufactured Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31956. Historical Development of the Complexity of Integrated Circuits . . . . . . . . . . . . . . 32457. Read-Only Semiconductor Memory Component Prices by Type . . . . . . . . . . . . . . . . 32558. Median Microprocessor Price v. Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32559. Minimum Linewidth for Semiconductor Microlithography . . . . . . . . . . . . . . . . . . . . . 32660. A Schematic of an Electron Beam Lithography System, and Its Use to Make

LSI Logic Circuits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32761. Trends in Device Speed–Stage Delay

per Gate Circuit for Different Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32862.The Relative Costs of Software and Hardware. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332

Chapter 9

Technology and Industry

Importance of Information Technology

Information technology and the industriesthat advance and use its products are becom-ing increasingly important to America’s eco-nomic strength. Information technology is acore technology, contributing broadly to theNation’s trade balance, employment, and na-tional security. The information sector alreadyaccounts for between 18 and 25 percent of thegross national products of seven of the mem-ber nations of the Organization for EconomicCooperation and Development (OECD) and be-tween 27 and 41 percent of employment int h e s e c o u n t r i e s .12 2

An important characteristic of this technol-ogy is the rapidity with which research resultsare translated into “advanced” products andfrom there to the mainstream consumer market-place. This rapid transfer has been particularlyevident in semiconductor research leading tocheaper and more powerful microprocessors .

G i v e n t h e a c c e l e r a t e d m o v e m e n t f r o m t h eresearch laboratories to a worldwide market-place , i t i s natural that the at tent ion of thet r a d i n g n a t i o n s h a s f o c u s e d o n i n f o r m a t i o ntechnology research and development (R&D).Two basic building blocks of information tech-nology are:

● t h e m i c r o e l e c t r o n i c c h i p - t h e l a r g e s c a l eintegrated c ircuit that permits the stor-age , rapid retr ieval and manipulat ion ofvast amounts of information, and

● s o f t w a r e - t h e s e t s o f i n s t r u c t i o n t h a t d i -rect the computer in i ts tasks .3

As noted in chapter 2 , there i s no consen-sus on what composes the “ informat ion tech-

‘Information Activities, Electrom”cs and Tekwomtnum”cationsTechnologies: Impacts on Employment, Growth and Trade(OECD, Paris, 1981), pp. 22 and 24.

‘M. R. Rubin, Information Econom”cs and Policy in theUnited States, Libraries Unlimited, Littleton, CO, 1983,pp. 32-44.

3Software may also refer to the stored information.

nology industry. ” Because of the varyingdefinitions and the differing statistical anal-yses arising from them, it is impossible to pindown the size of the industry. A relatively con-servative estimate, discussed later in thischapter, places the annual sales of the U.S. in-dustry at over two hundred billion annuallyand growing. From another perspective: overone-half million jobs depend on computer in-dustry exports .4

Findings

In examining the composition of informa-tion technology, the characteristics of the U.S.industry, and where the technology is heading,several trends become apparent:1.

2.

3.

The U. S. information technology industryis large and growing rapidly; there is a worldmarket for its products and services as thedeveloped countries increasingly move intothe information age.The technology is pervasive, making pos-sible major productivity improvements inother fields, creating new industries, andenlarging the range of services available tothe public. The technology diffuses throughmost facets of life, including business, engi-neering and science, and government func-tions.With each level of technological advance inbasic information technology, the level ofcomplexity and cost increases for R&D, asdoes the demand for additional technicaltraining for R&D personnel.

‘Robert G, Atkins, The Computer Industry and InternationalTrade; A Summary of the U.S. Role, Information ProcessesGroup, Institute for Computer Sciences and Technology, Na-tional Bureau of Standards, December 1983, p. IX-1. Draft ofa report under joint development by ICST and the InternationalTrade Administration, as of February 1985.

307

308 • information Technology R&D: Critical Trends and Issues

4. Domestic and foreign competition for world 5.markets is both intensive and escalating.The stakes for the United States are greatin terms of corporate sales and competitiveness in world markets, and employment.

There will be continued rapid advances inthe capabilities of the industry’s underlyingbasic building blocks, microelectronics andsoftware.

Composition of Information TechnologyIndividual information technologies have

grown explosively in technical sophisticationand, at the same time, diminished in cost. Thatrare combination has propelled the new tech-nologies into the mass marketplace. As thetwo basic building blocks continue their ad-vances, the parade of new information tech-nology capabilities and applications will go on.

There is a recurrent pattern in which onetechnological advance makes possible stillother advances, often in a “bootstrap” fash-ion. For example, sophisticated computer-aided design (CAD) equipment is a tool usedin development of state-of-the-art random ac-cess memories and microprocessors; develop-ment of the next generation of super-comput-ers depends on use of today’s most advancedcomputers; “expert” systems— still in theirtechnological infancy-are already employedin configuring complex computer systems andcustom-designing integrated circuits.

Functions

For purposes of this report, the term “infor-mation technology” has been used to refer tothe cluster of technologies that provide the fol-lowing automated capabilities:

1.

2.

Data Collection. Examples of automateddata collection systems range from large-scale satellite remote-sensing systemssuch as weather satellites to medical ap-plications such as CAT-scans and elec-trocardiograms.Data Input. Input devices include the fa-miliar keyboard, optical character read-ers, video cameras, and so on. They arethe means by which data are inserted andstored, communicated, or processed.

3.

4.

5.

6.

Information Storage. The storage mediaassociated with the information industryare the electronic-based devices whichstore data in a form which can be read bya computer. They include film, magnetictape, floppy and hard disks, semiconduc-tor memories, and so on. The ability tostore increasingly vast amounts of datahas been essential to the information tech-nology revolution.Information Processing. Information proc-essing is the primary function of a com-puter. The information stored by acomputer can be numeric (used for com-putations), symbolic (rules of logic usedfor applications such as “expert” sys-tems), or image (pictorial representationsused in applications such as remote map-ping). The stored information–in what-ever form—is manipulated, or processed,in response to specific instructions (usu-ally encoded in the software). The increas-ing speed of information processing hasbeen another essential factor in the infor-mation technology revolution.Communications. Electronic communica-tions utilizes a variety of media-the airwaves (for broadcast radio and television),coaxial cable, paired copper wire (used,among other things, for traditional tele-phony), digital radio, optical fibers, andcommunications satellites. Communica-tions systems play a major role in broaden-ing the use of other facets of informationtechnology and make possible distributedcomputing, remote delivery of services,and electronic navigation systems, amongmany other applications.Information Presentation. Once the infor-mation has been sent, it must be “pre-

‘ ~

Ch, 9–Technology and Industry ● 309

sented” if it is to be useful. This can beaccomplished through a variety of outputdevices. The most common display tech-nology is the cathode ray tube or videodisplay terminal. Hard-copy output de-vices include the most commonly used im-pact printers as well as those using non-impact technologies such as ink-jet andxerography. There are also audio systemsthat permit the computer to “speak”-exemplified by the automobiles that ad-monish you to fasten your seat belts.

Table 50 shows some of the technologies andapplication areas that depend on these functions.

Examples of System Applications

A few examples of the many informationtechnology applications are provided below.They have been chosen to illustrate the diver-sity of applications, and in most cases reflectcapabilities that have only become feasible ona significant scale in recent years. The ex-amples include: cellular mobile radio commu-

table 50.—Functions, Applications, and Technologies”

RepresentativeFunction Typical application areaa information technology

Data collection Weather prediction

Medical diagnosis

Data input Word processingFactory automationMail sorting

Storage ArchivesAccounting systemsScientific computation

Information processing

Ecological mappingLibrariesSocial Security payments

Traffic controlDistributed inventory control

Medical diagnosesEngineering designScientific computation

Communications

Ecological mappingFactory automation

Office systems

TeleconferencingRescue vehicle dispatchInternational financial

transactions

Data output and presentation Word processing

Management informationPedestrian traffic control

Radar, infra-red object detection equipment,radiometers

CAT-scanners, ultrasonic cameras

Keyboards, touch-screensVoice recognizes (particularly for quality control)Optical character readers

Magnetic bubble devices, magnetic tapeFloppy disksWafer-scale semiconductors (still in research phase),

very-high-speed magnetic coresCharge-coupled semiconductor devices, video disksHard disksGeneral purpose “mainframe” computers, COBOL

programsMinicomputersMulti-user super-micros, application software

packages“Expert” systemsSpreadsheet application packages, microcomputersSupercomputers: multiple instruction-multiple data

(MI MD) processors, vector processors, data drivenprocessors, FORTRAN programs

Array processors, associative processorsRobotics, artificial intelligenceLocal area networks, private branch exchanges (PBX),

editor applications packagesCommunications satellites, fiber opticsCellular mobile radiosTransport protocols, data encryption, Integrated

Services Digital Networks (ISDN)

Personal computers, printers (impact, ink jet,xerographic)

Cathode ray tubes, computer graphicsVoice synthesizers

aThi~ list is not ~xhau~tive; any given technology may also be used for some of the other applications mentioned.

SOURCE Office of Technology Assessment.


nications, direct broadcast satellites, robotics,computer simulation, a biomedical applicationin heart pacemakers, and financial services.

Cellular Mobile Radio CommunicationsFor decades, use of mobile telephone serv-

ice has been limited by over-crowding of theelectromagnetic spectrum; major cities havehad only some few hundred subscribers be-cause of spectrum constraints. Demand for ad-ditional mobile voice service has been grow-ing at a 12 percent rate annually for the past20 years. A new technology-cellular mobileradio communications—holds promise forvastly increasing the number of potentialsubscribers (to more than 100,000 in particu-lar geographic areas),’ while providing serv-ice of improved quality compared to that availa-ble previously. What is most significantabout the cellular concept is that it permitsconservation of the electromagnetic spectrum—a limited natural resource.

The older technology provided service froma single antenna to the area served. The cel-lular concept is based on dividing the servicearea into a number of geometric shapes, orcells, each served with its own antenna. Thecell antennas are interconnected with leasedtelephone lines. When a vehicle leaves a cell,that cell passes control to the next cell, andso on. The number of subscribers is muchgreater with the cellular concept because thetransmission frequencies can be reused re-peatedly in nonadjacent cells. Thus the impactof increasing demands on the spectrum is min-imized.

According to some estimates, cellular radiowill be a $4 billion U.S. industry by 1990 andcould reach $6 billion by the mid-1990s. Cel-lular mobile telephones recently cost about$3,000 but prices are likely to fall rapidly be-cause of competition among vendors.

Other countries are also finding considerabledemand for mobile radio services. In Japan,

60peration of cellular mobile radio systems in dozens of U.S.cities has been approved by the Federal Communications Com-mission.

mobile telephone services are now available inTokyo. In Spain, some 20,000 subscribers areanticipated by 1990. The Netherlands reachedapproximately 50,000 inhabitants by 1984.Saudi Arabia introduced cellular mobile radiocommunications in 1981 in three cities; by1982, there were 19,000 subscribers, and theSaudis have now extended coverage to 32cities.

Direct Broadcast SatellitesThe use of direct broadcast satellites (DBS)

has recently become feasible. With DBS, over-the-air signals (e.g., TV, radio) can be receiveddirectly by small rooftop antennas (see fig. 52).These new, commercial, geostationary sys-tems may have widespread use in the UnitedStates and in other countries, primarily pro-viding entertainment, but also with potentialfor education, advertising and other businessuses. As of early 1985, nine applications had

Photo credit: Motorola

Mobile cellular radiotelephone

Ch. 9—Technology and Industry ● 311

Figure 52.—Direct Broadcasting Satellite System

, Indoor unit

been approved by the Federal Communica-tions Commission (FCC) for subscription serv-ices to the public, some of which may becomeoperational.

DBS services will compete with other mediasuch as multiple distribution systems, STV(subscription TV), and cable TV (CATV) forserving television viewers. DBS services mayprove especially valuable for providing serv-ice to sparsely populated geographical areasthat are not economical to reach by othermeans, as well as to densely populated citieswhere new cable installations are prohibitivelyexpensive.

Among the technical improvements thatmake DBS systems feasible are higher powerand more efficient on-board power amplifiers;solar power generation equipment; more ac-curate satellite station-keeping control; im-proved ground receiver sensitivity; the use ofhigher frequencies; and better transmitterdownlink beam control. Higher power andhigher frequencies also make practical the useof small (about 1 meter or less) rooftop an-tennas. Some of the early U.S. experimentalsatellite communications systems, along withparallel advances in solid-state electronics andnew materials processing techniques, havecontributed significantly to the improvements

in components and subsystems availabletoday.

DBS systems had their genesis when NASA’sATS-6 satellite was launched in 1974—a sys-tem having a multiplicity of payloads, two ofwhich included TV broadcast capabilities. Anumber of countries have DBS systems inplace or in planning stages. Included amongthese are: Canada’s medium-powered ANIKC-2, which is serving Canadian audiences aswell as providing five channels of televisionentertainment to the Northeastern UnitedStates; Japan’s modified BS-2 satellite, launchedin early 1984; and France’s TDF-1 and WestGermany’s TV-SAT, which are expected to beoperational in 1986.

RoboticsRobots are mechanical manipulators which

can be programmed to move workplaces ortools along various paths. They are one of thefour tools employed in computer-aided manu-facturing (i.e., robots, numerically controlledmachine tools, flexible manufacturing sys-tems, and automated materials handlingsystems).6

Robotics emerged as a distinct disciplinewhen the century-old industrial engineeringautomation technologies converged with themore recent disciplines of computer scienceand artificial intelligence. The convergenceproduced the growing field of robotics, whichhas had wide factory applications in areas thatinclude (but are not limited to) materials han-dling, machine loading/unloading, spray paint-ing, welding, machining, and assembling. Therobots are most often used in performing par-ticularly hazardous and monotonous jobswhile offering enough flexibility to be easilyadapted to changes in product models.

Neither today’s robots, nor those likely tobe available in the next decade, look likehumans nor do they have more than a fractionof the dexterity, flexibility, or intelligence ofhumans. A simple “pick and place” machine

‘See Computerized Manufacturing Automati”on: Employment,Education, and the Workplace (Washington, DC: U.S. Congress,Office of Technology Assessment, OTA-CIT-235, April 1984.

38-8o2 0 - 85 - 21


with two or three degrees of freedom may costroughly between $5,000 and $30,000, whilemore complex progr ammable models, oftenequipped with microcomputers, begin at about$25,000 and may exceed $90,000.

Human workers, robots, and nonprogram-mable automation devices each have certaincharacteristics which offer advantages for themanufacturing process. Table 51 comparessome of the salient characteristics.

Table 51 .—Characteristics of Human Workers,Robots, and Nonprogrammable Automation Devicesa

Non programmableHuman Automation

Characteristic Worker Robot Device

Flexibility . . . . . . . . . . 1Consistency. . . . . . . . 3Endurance . . . . . . . . . 3Ability to tolerate

hostileenvironments. . . . . . 3

cost . . . . . . . . . . . . . . n/aSpeed . . . . . . . . . . . . . 2-3Intelligence/

programmability . . 1Sensing . . . . . . . . . . . 1Judgment. . . . . . . . . . 1Ability to adapt

to change. . . . . . . . 1Dexterity. . . . . . . . . . . 1

22

1-2

1-22

2-3

222

22

31

1-2

1-211

333

33

aNumerical ranking indicates relative advantage, With “l” Indicating the greatestadvantage.

SOURCE: Computerized Manufacturing Automation: Employment, Education,and the Workp/ace, (Washington, DC: U.S. Congress, Office ofTechnology Assessment, 1964). This table was derived from the fore-going report

At the end of 1983 Japan had about 43,000operating robot installations-by far the larg-est number in any country. The United States,where the original patents for robots were ob-tained, had about 9,400 installations (v. about13,000 in early 1985 and 6,300 in 1982), fol-lowed by West Germany (4,800), and France(3,600). Some projections indicate that by 1990there will bean installed base of nearly 90,000robots in the United States. *

As reported by OTA6a, significant roboticsresearch is being conducted in over a dozenuniversities, about three dozen industrial firmsand independent laboratories, and in FederalGovernment labs at the National Bureau ofStandards (NBS), NASA, and DOD. The re-port further noted that the research was in-tensively addressing several problem a r e a s :improved positioning and accuracy for the ro-bot’s arms; increased grace, dexterity, andspeed; sensors, including vision, touch, andforce; model-based control systems; software;mobility; voice recognition; artificial intelli-gence; and interface standards. The interfacearea is critical to widespread use of automatedmanufacturing. NBS has created an Auto-mated Manufacturing Research facility in partto perform research on the interfaces betweendifferent computerized devices in a factory.

Computer Simulation

“Computer simulation” is a process thatemploys a computerized model of certain sig-nificant features of some physical or logicalsystem which is undergoing dynamic change.We therefore find computers used in applica-tions such as economic modeling and wargames where “what if” scenarios can be playedout to test suggested social policies or militarystrategies. As relatively inexpensive computermemories grow in size and integrated circuitsgrow in speed, increasingly complex opera-tions are being modeled. Software advanceshave also played a large part in the growinguse of computer simulations.

*worldwide Robotics Survey ~d D&toW, 1984, to ~ pub-lished, Robotics Industries Association, Dearborn, MI. Thisdata is based on the RIA’s definition for robots which excludesthe simpler, nonreprogr amrnable machines.

@Compute~~ M~ufactufl-ng Automation: Employment,Education and the Workplace, op. cit.

Ch. 9–Technology and Industry ● 313

When combined with other informationtechnologies, some startling simulations arepossible. In flight simulators, for example,computer-generated imagery techniques areused to create background, images, sounds,and sensations, and to position discrete ele-ments in a scene—elements which have beenobtained from video images of real trees,buildings, clouds, aircraft, and other charac-teristics of the environment to be simulated.The flight simulator is placed in the dynam-ically changing recreated environment and thepilot then “flies through” in the simulator,which duplicates the performance of the actualairplane. Realistic simulations are made pos-

sible by the rapid calculation and implemen-tation of flight characteristics related to theknown aerodynamic and engine properties ofthe craft, air speed, altitude, attitude, and lift.Some systems use 10,000 variables associatedwith a single-engine jet fighter. In order toachieve visual accuracy, simulator systemscalculate the aircraft’s position between 25and 60 times per second.

The Federal Aviation Administration per-mits 100 percent simulation training for ex-perienced pilots who are upgrading their flightcertifications to more advanced aircraft. Simu-


lated dogfight training at Williams Air ForceBase is so realistic that some F-4 pilots havebecome airsick.7

Computer simulation is proving effective forimproving the quality of training in a varietyof jobs—e.g., operating a locomotive, control-ling a supertanker, operating a space shut-tle—while reducing the cost, time, and riskassociated with conventional training.

Heart Pacemakers

Microelectronics and software systems arefinding a variety of applications in medicaldiagnoses and treatment. The heart pace-makers now being used by more than 500,000patients in the United States are illustrativeof the widespread use of these technologies formedical purposes.

The heart muscle contracts in response toelectrical impulses generated by apart of theheart itself. The contractions force the bloodthrough the arteries to all parts of the body;but if there is something wrong with the elec-trical impulses, blood flow is impaired and theheart muscle becomes injured.

——-——. -“’The Technology of Illusion,” Forbes, Feb. 27, 1984, pp.

158-162.

Photo credit: Medtronlc, Inc.

Heart Pacemaker

Pacemakers are medical devices that can besurgically implanted under a patient’s skinwith electrode leads fed through the veins intothe heart. The two main components of a pace-maker are the pulse generator and the leads.Pulse generators contain a power source andthe electrical circuitry for sensing, pulsing, andprogr amming. Some new pacemakers are alsocapable of telemetry functions, which transmitcertain critical measurements of performanceof the pacemaker and the patient’s condition.Programming changes and telemetry func-tions require external microprocessors thatcommunicate with the implanted devices in or-der to monitor performance and to modify thepacemaker’s operation.

The first implantable pacemaker was devel-oped in 1958, and the first implantation in ahuman was performed in 1959. Early pace-makers sent electrical impulses to the heartat fixed intervals. By 1970, the technology for“demand” pacemakers was developed. Thesepacemakers sense when the heart is not work-ing properly and, when necessary, send outelectrical impulses to trigger the contractionof the cardiac muscle. Programmable pace-makers-which are continually being updatedwith advanced technology— can be repro- grammed without additional surgery, in re-sponse to changes in the patient’s physiology.This capability reduces the risks of additionalsurgery for cardiac patients and decreases thecosts of their care.8

The latest generation of pacemakers is de-signed to be flexible, enabling updating of thesoftware and external keyboard to accom-modate newer technology. The flexibility ex-tends to enabling pacemaker parameter set-tings to be customized for each patient’sspecial needs, and to the use of the same tech-nology for other medical applications, such asdrug dispensers and pain controllers.9

%pecisl Committee on Aging, U. S. Senate, Fraud, Waste,and Abuse in the Medz”care Pacemaker Industry, (Washington,DC: U.S. Govemxnent printing Office, 98-116. September 1982).

‘Richard M. Powell, “A New Programm er for Implanted PulseGenerators,” pp. 678-687, technical paper presented at the 15thHawaii International Conference on Systems Sciences, Janu-ary 1982.

—

Ch. 9—Technology and Industry . 315

Financial ServicesThe financial service industry would not pro-

vide the level of service it does without infor-mation technologies. 10 The numbers of checks(over 37 billion annually), credit card drafts(over 3.5 billion annually), and securities trades(over 30 billion shares traded annually) aresimply too much for any manual system tohandle. In the 1960s, for example, before thefinancial service industry was substantiallyautomated, there were days when the NewYork Stock Exchange suspended operationsbecause the broker/dealers were unable to han-dle the workload.

Both users and providers of financial serv-ices have made use of virtually every categoryof informat ion technology . Banks and otherfinancial service providers have been longtimeu s e r s o f c o m p u t e r s . H o w e v e r , a n i n c r e a s i n gn u m b e r o f r e t a i l e r s - l a r g e a n d s m a l l - a r e i n -s ta l l ing the hardware and us ing the te lecom-munications networks to verify checks and toauthorize and complete credit transact ions .

Applications software packages for financialservices can process market data and gener-a t e i n f o r m a t i o n u s e d f o r p o r t f o l i o m a n a g e -ment. They can also generate and analyze loaninformat ion, and process bank card appl ica-t ions . The spread of remote terminals inter-connected with central computers i s becom-ing ubiquitous, and expanding the number ofservices avai lable to the publ ic .

T h e v i d e o - r e l a t e d t e c h n o l o g i e s - v i d e o t e x ( atwo-way information service) and te letext (ao n e - w a y i n f o r m a t i o n s e r v i c e ) — a r e n o t y e twidely used in the delivery of financial serv-ices, but experiments are under way and it isl ike ly that they ul t imate ly wi l l gain accept-a n c e . O n e v i d e o t e x s y s t e m , c u r r e n t l y b e i n gmarketed in Flor ida, uses an AT&T terminalattached to a television set to link the finan-cial service provider with the customer.

— . —‘°For detailed discussion, see Effects of Information Tech-

nology on Financial Service Systems, (Washington, DC: U.S.Congress, Office of Technology Assessment, OTA-CIT-202, Sep-tember 1984).

There are also several card technologies. Theembossed plastic card with its strip of mag-netic tape provides the primary means foraccessing credit and debit services that are de-livered through both paper-based and elec-tronic systems. Laser cards-not yet used bythe financial service industry–can store dig-ital data signifying the bearer’s fingerprints.Electron cards combine three encoding tech-nologies-the banking industry’s magnetictape, the retai l industry’s opt ical characterrecogni t ion , and the UPC bar code .

Document and currency readers have hadsome acceptance, part icularly for the readingof checks encoded with magnetic ink. Systemsthat process credit and debit card transactionsalready truncate the paper f low at the earl i -est pract ical t ime. The data are recorded onmagnetic media and transferred electronicallyfor processing by the card issuer.

Together, these technologies have resultedin a financial service industry that is offeringan increasing number of services such as auto-mated redemption of money market funds. Atthe same t ime these services are being pro-v ided in an increas ing number of ways , in-c luding the use of automated te l ler machinesand telephone bill payers.

With all this, there is the possibility ofredistribution of functions among traditionalsuppliers as well as potential new entrants.Whereas in the past the payment system hasbeen reserved largely to banks, because theyhad access to facilities for clearing and settle-ment, movement of funds electronically makesit possible to avoid the traditional paymentsystem and to settle directly between tradingpartners. Alternative means of distributing in-formation could diminish the role of brokersfor such products as securities, real estate, andinsurance. Cash-oriented businesses, such asgas stations and supermarkets, already use on-site automated teller machines to relieve therequirement to cash checks while minimizingthe amount of currency that is held at eachstore location.


Characteristics of the U.S. InformationTechnology Industry

The information technology industry is acomposite of several industries both in man-ufacturing and services. There is no single,generally accepted definition of the industry.The definitions in use by the Information In-dustry Association, for example, are too broadfor the purposes of this report, while defini-tions used by the Computer and BusinessEquipment Manufacturer’s Association andby the Association of Data Processing Serv-ice Organizations are too restricted.

Despite the ambiguities, this section at-tempts to characterize the industry in termsof size and structure, growth, employment, in-vestment in R&D, and other areas of specialconcern. The part of the information technol-ogy industry portrayed here includes primar-ily manufacturers of: electronics; informationprocessing equipment, including computers,office equipment, and peripherals and services;semiconductors; and telecommunicationsequipment.

Table 52 is a summary of the key indicatorsfor the composite 30 industry categories cov-ered by Business Weekll for the period 1978to 1982, compared with aggregated data forpart of the information technology industry.”A point of contrast worth noting is that infor-mation technology firms’ business perform-ance during the period outpaced the compos-ite industry average significantly as measuredby the following criteria:

●

●

growth in sales revenues: 40 percent forthe composite industry groups vs. 66 per-cent for the information technology sector.growth in profits: 6.4 percent for the com-posite vs. 36.4 percent for the informationtechnology sector.

llBu~jne~~ W=k, R&D Scoreboard, June 20, 1983, PP. 122-

153. See also Business Week, July 9, 1984, pp. 64-77, whichshows a continuation of the trends through 1983.

‘zData shown by Business Week for national composite in-dustry R&D expenditures correlates closely with NSF data inScience Resources Stud-es “Highlights,” June 11, 1982.

●

●

●

●

profits/sales ratios ranging between 4.2to 5.7 percent for the composite vs. 7.9to 9.7 percent for the information technol-ogy sector.growth in the number of employees: a de-crease of 7.8 percent for the composite vs.an increase of 11.8 percent for the infor-mation technology industry.growth in R&D expenditures: 81 percentfor the composite vs. 111 percent for theinformation technology sector. The infor-mation technology sector’s R&D expend-itures per employee were higher than thecomposite by about 20 percent annuallyfor the period, and both groups increasedtheir R&D investments in spite of a reces-sion.13

investments in R&D as a percentage ofsales: averaging 2.1 for the composite vs.4.2 for the information technology sector;and, as a percentage of profits, 42 vs. 48,respectively.

The above statistics are incomplete becausefirms whose primary business is not informa-tion technology do not appear as part of theindustry—despite the fact that their informa-tion technology activities may be significant.Among the firms not represented are GeneralElectric, Rockwell International, and West-inghouse.

The above data for the information technol-ogy manufacturing industry, when combinedwith related sectors such as the software andcomputer services, and telephone and tele-graph services sectors of the economy, hadtotal revenues of $229 billion for 1982, up from$180 billion in 1980, for a 27 percent increase.Figure 53 illustrates the increasing portion of

‘Klf the four sectors-industry, Federal Governrnent, collegesand universities, and other nonprofit organizations-only in-dustry increased its funding for R&D in constant 1972 dollarsduring the 1978-82 period. Probable Levels of R&D Expench”-tures in 1984: Forecast and Analysis, Battelle, December 1983.

Ch. 9—Technology and Industry . 317

Table 52.—Comparison of the U.S. Information Technology Industry withComposite Industry Performance, 1978-82

Percent1978 1979 1980 1981 1982 change— —.

C o m p o s i t e

Infotech

C o m p o s i t e

Infotech

C o m p o s i t e

Infotech

C o m p o s i t e

Infotech

C o m p o s i t e

I n f o t e c h

C o m p o s i t e

Infotech

C o m p o s i t e

Infotech

C o m p o s i t e

Infotech

Sales (mill Ions of dollars) . . . . . 1,085,291

131,872

1,277,764149,783

1,421,551

174,449

1,586,510

193,921

1,520,313218,862

4066

P r o f i t s ( m i l l i o n s o f d o l l a r s ) . . 59,57812,780

72,50513,821

73,49315,474

81,75716,056

63,36517,436

6.436.4

Profits/sales (percent) . . . . . . . . . . 5.59.7

5.79.2

5.28.9

5.18.3

4.27.9

Employees (thousands) . . . . . . . . . . 15,1332,952

15,5423,099

15,4983,226

15,0453,252

13,9593,301

– 7.811.8

R & D ( m i l l i o n s o f d o l l a r s ) . . . . 20,6104,961

24,6745,885

28,9847,221

33,2858,531

37,17910,473

81111

R&D $/sales (percent) . . . . . . . . . . . 1.93.8

1.93.9

2.04.1

2.14.4

2.54.8

R&D $/profits (percent) . . . . . . . . . . 34.638.8

34.042.6

39.446.7

40.753.1

59.080.1

R&D $/employee ... . . . . . . .

R&D expendi tures per employee

lnfotech/Composite (percent) . . .

1,3621,680

1,5881,899

1,8702,238

2,2122,623

2,6673,173

120123 120 121 119Business Week “Scoreboard” Numbers Notes:

A This IS a sample of R&D spending in information technology by U.S. corporations, It is based on total R&D expenditures for those companies that are publiclyheld, have annual revenues over $35 million, and R&D expenses of $1 million or 1 percent of revenue. Only that spending by companies whose primary businessis Information technology (electronics, computers, office equipment, computer services and peripherals, semiconductors, and telecommunications) is included.

B Sales, R&D spending, and R&D spending per employee figures have been adjusted to reflect the numbers from Western Electric and other AT&T subsidiariesthat are not included in the “Scoreboard” numbers, This adjustment involves:1 addition of revenues received from Western Electric to the total operating revenues figures in the AT&T Annual Reports for the years covered,2 use of the total AT&T spending figures for R&D which include spending by Western Electric and other AT&T subsidiaries as provided in Business Week for

the years 1980-82 and as estimated from a chart in the 1983 AT&T Annual Report for the years 1978 and 1979.3 use of total AT&T employment figures provided in Forbes each May for the years 1978-82.

C Employment numbers for all sectors have been calculated from the R&D spending per employee and the R&D spending figures provided in Business Week andmay reflect rounding errors,

SOURCE. Data obtained, or calculated from Business Week, Scoreboard, June 30, 1983 the U.S. Commerce Department; Forbes, May 1979 throuah 1983; 10K formsfiled by AT&T and Western Electric Corp

s e r v i c e s c l o s e l y m a t c h e s t h a t o f c o m p u t e rm a n u f a c t u r i n g , b o t h g r o w i n g b y a b o u t 1 4 0percent between 1972 and 1982 .

International Trade

Technology-intensive products have impor-tant implications for the balance of trade.Studies on U.S. trade and the influence of tech-nology have generally concluded that technol-o g y s e r v e s a s a n i m p o r t a n t d e t e r m i n a n t o fcomparative advantage in manufactured goodst r a d e . 14 15

“C. Mi~el Aho and Howard F. Rosen, “Trends in Technol-ogy-Intensive Trade: With Special Reference to U.S. Competi-tiveness, ” Office of Foreign Economic Research, Bureau of In-ternational Labor Affairs, U.S. Department of Labor, p. 11.

“T. C, Lowinger, “Human Capital and Technological Deter-minants of U.S. Industries Revealed Comparative Advantage, ”Quarterly Review of Econozzu”cs and Business (1974), winter1977, pp. 91-102, as reported in Aho and Rosen, pp. 11, 12.

total sales due to services, which are approach-ing 50 percent of total industry revenues.

Employment

Employment in information technologymanufacturing increased significantly between1972 and 1982 in most segments of the indus-try (table 53), experiencing employment growthranging between 9 and 142 percent. Only theconsumer products (radio and TV sets) showeda decline (28 percent). Total employment in in-formation technology manufacturing grew by51 percent , in spi te of economic recess ions .

Employment in information technologyservices i s about equal to that of the manu-facturing segment , with some 1 .6 mi l l ion em-ployees . The employment level in computing


Figure 53.—The Changing Structure and Growth of the U.S. Information

300

200

-

100

0

Technology IndustryInformation industry revenues; 1980-83 (dollars in billions)

Software 180na n d c o m p u t e r . servicesl’2

Telephoneand telegraph 34.3%

service2

Informationtechnology

manufacturing

229

258

34.7%

1980 1981 1982 1983

YearsSOURCES: IADAfJSO; 2U.S. Industrial Outlook, 1983, 1984.

Table 53.—Employment Levels in the U.S. Information Technology IndustriesEmployees (in thousands)

Percent change1972 1982 1972-1982

Manufacturing a

Computers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 351 + 142Office equipment. . . . . . . . . . . . . . . . . . . . . . . 34 51 +50Radio and television receiving sets . . . . . . . 87 63 –28Telephone and telegraph equipment . . . . . . 134 146 + 9Radio and television communications

equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . 319 454 +42Electronic components. . . . . . . . . . . . . . . . . . 336 528 +57

Totals, manufacturing . . . . . . . . . . . . . . . . . 1,055 1,593

ServicesTelephone and telegraph . . . . . . . . . . . . . . . . 949 1,131 +11Computing b. . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 360 + 141Radio and television broadcast . . . . . . . . . . 68 81 +19Cable television . . . . . . . . . . . . . . . . . . . . . . . 40 52 +30

Totals, services . . . . . . . . . . . . . . . . . . . . . . . . . . 1,206 1,624aEstimates provided by the U..S. Department of Commerce, Bureau of Industrial Economics.bF\gures are for 1974 and 1983. Source: U.S. Industrial Outlook, 1984.CFigures are for 1979 and 1983, Source: Federal Communication Commission in telephone interview with OTA staff, May 1984,

dF\gures are for 1981 and 1982. Ibid. (FCC).

—


The U.S. merchandise trade balance hasbeen negative in recent years: deficits of $42.7billion in 1982; $69.4 billion in 1983; and over$100 billion in 1984. In 1980, advanced tech-nology products showed a positive trade bal-ance of $31 billion, compared with a deficit ofm o r e t h a n $ 5 0 b i l l i o n f o r a l l m a n u f a c t u r e dgoods.”

Information technology manufactured prod-ucts17 usually made positive contributions tothe U.S. balance of trade between 1972 and1982 (fig. 54), led by sales of computer equip-m e n t . H o w e v e r , 1 9 8 3 a n d 1 9 8 4 s a w t r a d edeficits of $0.8 and 2.3 billion according to De-p a r t m e n t o f C o m m e r c e e s t i m a t e s .18 T h e d e f i -cits would have been much greater withoutabout $6 billion in exports of computer equip-ment in each of those years (see f ig . 55) .— . — —

“International Competitiveness in Advanced Technology:Decisions for Aznen”ca, National Research Council (Washing-ton, DC: National Academy Press, 1983, pp. 23-24.

“The products include: computer equipment, SIC 3573; of-fice equipment, SIC 3579; radio and TV sets, SIC 3651;telephone and telegraph equipment, SIC 3661; radio and TVcommunications equipment, SIC 3662; and electronic comp~nents, SIC 367.

‘aU.S. Industn”d Outlook, 1978 through 1984.

Figure 54. —Balance of Trade, InformationTechnology Manufacturing Industry

(dollars in billions)

6 -

equipment (SIC 3573)4

2

o

,-6 Radio and TV Communications

equipment,

1972 1974 1976 1978 1980 1982 1984Years

SOURCE: U.S. Industrial Outlook for years 1978 through 1984.

est

Figure 55.— U.S. Trade Balancea With SelectedNations for R&D Intensive Manufactured Products

$30

28

26

24

22

20

18

16

14

12

4

2

0

-2

-4

-6

●✍

Developing nations

1966 67 68 69 70 71 72 73 74 75 76 77 78 79 80Years

aExports less imports.blncludes West Germany,

SOURCE: Science Indicators— 1982, National Science Board, Na-tional Science Foundation, 1983.

Although the United States holds the ma-jor market share of the industrialized coun-t r i e s ’ e x p o r t s o f h i g h - t e c h n o l o g y p r o d u c t s ,t h a t s h a r e h a s d e c l i n e d f r o m 3 0 p e r c e n t i n1962 to 22 percent in 1978, and has increasedonly marginal ly s ince . In absolute terms, theU.S. positive trade balance in high-technologyproducts increased over e ightfold from 1962to 1980. During that same period West Ger-m a n y , a n d e s p e c i a l l y J a p a n , s t a r t i n g f r o msmaller bases , have had impress ive gains int h e i r s u r p l u s e s o f e x p o r t s o v e r i m p o r t s — aninefold increase for West Germany and two-hundredfold increase for Japan.1 9

19U+S. Dep~ment of Commerce, International made Admin-istration, as reported in “An Assessment of U.S. Competitiveness in High-Technology Industries, ” prepared for the Work-ing Group on High Technology Industries of the CabinetCouncil on Commerce and Trade, final draft, May 19,1982.

—— —


Industry Structure

Information technology companies in theUnited States are heterogeneous in dollarvalue of sales, breadth of product lines, num-ber of employees, and degree of vertical in-tegration. Competing for those markets arethousands of firms ranging in size from thevertically integrated multinational companiessuch as IBM, ITT, and the still-large, post-divestiture AT&T to the more typical, smallercompanies that produce a narrow cluster ofproducts with a large technology content.While the majority of sales dollars are concen-trated in the large companies, the vast ma-jority of companies are relatively small.

The diversity in the sizes of the businessesthat populate the information technology fieldis important for reasons that include the bene-fits of competition; the impetus to innovationdue to firms’ willingness to take risks and totry new directions; and the tendency to fillniche market demand for specialized products.The U.S. telecommunications, electronics, andsoftware industries provide three illustrationsof the contrasts in the breadth of that di-versity.

The telecommunications services industryis simultaneously both a mature industrydating back to the 19th century and a grow-ing industry marked by limited competitionin some markets and diversity in others. Whilelocal public telecommunications services arelargely dominated by regulated monopolies,long-distance services are now offered in acompetitive market. New services, such aspaging, cellular mobile radio services, bypassservices and direct broadcast satellite services,are not dominated by the established carriers,but are being offered by a variety of firms.

Table 54.—U.S. Home

The telecommunications equipment provid-ers are exceptionally diverse and competitive.There are hundreds of firms in this field, rang-ing from a few U.S. and foreign multinationalcompanies manufacturing a full range ofequipment to dozens of medium size, andmany hundreds of small companies concen-trating on a more limited range of products—e.g., speech compression products, multiplex-ers, modems, data terminals and local areanetworks.

The electronics industry is even more di-verse. The thousands of electronic systemsand equipment providers, as well as consultingfirms, range in size from those with billionsof dollars in annual sales to small, start-upventures such as the Apple Computer’s garageoperation of a few years ago. For those seg-ments of the industry undergoing rapid tech-nological change, diversity is often acceleratedby spin-offs from rapidly growing companies,and other new entries into the field.

The U.S. microcomputer software industryis an example of rapid change, growth, anddiversity. Future Computing, Inc. estimatesthat the U.S. market for home computer soft-ware grew by 168 percent from 1982 to 1983,and projects an 85 percent growth from 1983to 1984, with the growth rate declining to 26percent by 1988 in a projected $5 billion an-nual market by then (table 54). Software unitsales for office use are projected to enjoy com-parable growth. Table 55 indicates growthrates of 87 percent between 1982 and 1983,and 58 percent from 1983 to 1984, graduallytapering off to 24 percent in 1988 in a $6.7 bil-lion annual market.20 Like the electronics in-dustry, software firms number in the thou-

2 0 F u t u r e c o m p u t i n g , ~ r g on ~ c o m m u n i c a t i o n , Mmch 1984.

Computer Software Salesa

1981 1982 1983 1984 1985 1986 1987 1988Software units (millions) . . . . . . . . . . 1.1 6.9 19.0 37.0 58.0 82.0 112.0 144.0Percent growth (units) . . . . . . . . . . . . 486 168 85 52 42 32 26Revenue (millions of dollars). . . . . . . 48 282 757 1,400 2,100 3,000 4,000 5,000aData for 19u.M are estimates by Future cOrnPUllW.SOURCE: Future Computing, Richardson, TX, by OTA staff telephone interview, March 1964,

Ch. 9—Technology and Industry . 3 2 1

Table 55.—U.S. Office Computer Software Salesa

1981 1982 1983 1984 1985 1986 1987 1988

Software units (millions) . . . . . . . . . . 3.2 6.1 11.0 17.0 25.0 35.0 46.0 59.0Percent growth (units) . . . . . . . . . . . . 111 87 58 44 36 30 24Revenue (millions of dollars) . . . . . . . 343 724 1,351 2,100 3,100 4,200 5,400 6,700aData for lgsd.aa are estifnates by Future Computln9

SOURCE Future Computing, Richardson, TX, by OTA staff telephone interview, March 1984

sands. Although no hard data are available,it is the impression of industry analysts thatthe rapid pace of startups in the software in-dustry seen in recent years now appears to beslackening.

An important trend in industry structure istoward concentration in the software and dataservices industries, where larger firms have ac-quired hundreds of smaller ones. Among theselarger firms are Automatic Data Processing,Electronic Data Systems, and General Electric.

Larger firms are also integrating their prod-uct lines-some moving into subsystems andcomponents, and others, such as semiconduc-tor manufacturers, moving into subsystemand computer manufacturing. AT&T, TexasInstruments, Intel, and Northern Telecom-munications Corp. are examples of companiesrecently expanding into computers and relatedproducts.

There is also an active trend toward affilia-tions between information producers and dis-tributors. Among these are publishers of news-papers, books, and magazines and broadcast,cable TV, and interactive computer networkcompanies.

Geographic diversification is taking place ata rapid pace as foreign-primarily European—companies acquire U.S. firms. Among theEuropean firms are: Olivetti, CAP Gemini,Racal Electronics, Schlumberger, Thomson-CSF, and Agfa-Gevaert.

A number of U.S. firms are expanding theirmarkets by establishing joint ventures or byacquiring foreign outlets—e.g., IBM, AT&T,and Datapoint.

Small Entrepreneurial Firms

Small businesses play a central role in U.S.industry in general, accounting for:

● thirty-eight percent of the Nation’s grossnational product;

● two-thirds of all new jobs;● two-and-a-half times as many innovations

per employee as large firms;21andŽ a tendency to produce more ‘‘leapfrog’

creations compared to large companiesand to introduce new products morequickly than large firms.22

The contributions of small businesses to theinformation technology industry are signifi-cant. Small businesses with less than 500 em-ployees account for 33 percent of sales in elec-tronic components and 34 percent of theNation’s employment in this sector. In thecomputer services sector, small businesses ac-count for 69 percent of the industry’s sales and67 percent of its employees. These firms alsomake contributions in sectors where largecompanies dominate, such as in office com-puting equipment, consumer electronics, andcommunications equipment (table 56).

Opportunities for innovation based on ad-vances in information technology have beennumerous. Consequently, hundreds of marketniche applications have been created that

“Advocacy: A Voice for Small Businesses, Office of Ad-vocacy, U.S. Small Business Administration, 1984. p. 1.

22Stroemann, 1977. Karl A. Stroetmann “Innovation in Smalland Medium Sized Firms.” Working paper presented at InstitutFur Systemchnik und Innovations forschung, Karlsruhe, WestGermany, August, 1977. See Vesper (Entrepreneurship).


Table 56.—Percentage of Total Industry Sales andEmployment by Size of Small Businesses

Companies with Companies withunder 100 employees under 500 employees

Sales Employment Sales Employment

Office computing machinesand computer auxiliaryequipment . . . . . . . . . . . . . . . 2.6 2.6 6.0 6.3

Consumer electronics . . . . . . . 5.4 5.9 9.3 10.1Communications equipment. . 4.7 5.2 9.9 10.8Electronic components . . . . . . 18.3 17.2 32.7 33.7Computer services . . . . . . . . . . 51.2 47.7 68.6 67.1SOURCE: Small Business Administration, 19S4.

small firms are quick to fill-niches too nar-row or specialized to attract large firms.23 Thedevelopment of the computer in the 1940s andthe subsequent invention of the transistor ledto an explosion of new applications and to op-portunities for then small companies whosenames are now familiar in the informationtechnology industry: IBM, Fairchild, Intel,Texas Instruments, DEC, Data General,Wang, and Computervision. One market re-search company’s listing of the fastest-grow-ing niche markets shows that about 90 percentof the top 50 entries are in information tech-nology,24 and many of these technologies arerelatively new.

As the successful firms in this industrygrow, they tend to generate spin-off companiesthat often specialize initially in a narrow rangeof innovative products or processes. Fairchildis such a company, having begun with onlyeight employees and now accounting for over80 spin-off enterprises.25

Venture capital funding is extremely impor-tant to the small entrepreneurial firms. In-.— — ———

231n a July 1981 study, the General Accounting Office foundthat in concentrated industries, such as the information tech-nology industry, small businesses are likely to perform special-ized innovative functions and develop products or processes tobe used or marketed by other, usually larger firms in that in-dustry, For example, in the semiconductor industry there aremany small companies that manufacture diffusion furnaces, ionimplantation machines, epitaxia.1 growth systems, and mask-making systems—all part of a necessary supporb structure forthe semiconductor industry.

‘421st Century Research, Supergrowth Technology U. S. A.,newsletter as published in EDP News Service, Inc., ComputerAge-EDP Weekly, Feb. 28, 1983, p. 9.

“Carl H. Vesper, Entrepreneursiu”p and National Policy,Heller Institute for Small Business Policy Papers, p. 27.

vestments through organized venture capitalinvestment businesses in 1983 amounted to$2.8 billion, up 55 percent from $1.8 billion in1982.26 It is improbable that this growth ratewill be sustained for very long.27

The supply of U.S. venture capital financ-ing has grown at least eightfold since the mid-1970s, helping to fund the more than 5,500U.S. small startup and expanding firms. Thesefirms tend to operate in innovative marketniches where the risks are high and the poten-tial payoffs much higher.

Information technology firms supported byventure capital funds numbered in excess of1,000 in 1983, and received about $2.1 billionfrom organized venture capital investmentbusinesses. These funds were used for bothstartups and expansion of existing companies.The distribution of funds is shown in table 57.

These firms account for a significant num-ber of advances in the information technologyfield. Among the now-well known informationtechnology firms started or aided by venturecapital are: Intel, Apple Computer, Visicalc,DEC, and Data General. R&D limited partner-ships also contribute to this process (see ch. 2).

The existence of so many entrepreneurialfirms, the current abundance of venture capi-tal to fund their growth, and the strengthen-ing university-industry R&D relationshipsserve as a powerful source of strength for in-novation in the United States. For example,

“Data provided by Venture Economics, Inc., from its data-base covering organized venture capital investment companies.

“’’Scramble for Capital at Almost-Public Companies, ” For-tune June 25, 1984, p. 91.

Ch. 9–Technology and Industry . 323

Table 57.—Venture Capital Funding and Information Technology Firms, 1983

Distribution of Distribution ofavailable available

Percentage of 1,000 venture capital venture capitalventure capital firms (percent) (millions of dollars)

Computer hardware and systems. . . . . . . . . . . . . . 28 39 819Software and services . . . . . . . . . . . . . . . . . . . . . . . 12 7 309Telecommunications and data communications . 9 11 231Other electronics. . . . . . . . . . . . . . . . . . . . . . . . . . . 10 10 210

Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 61 1,569alncl”des semiconductor fabrication and test equipment, instrumentation, fiber Optics, laser-related devices, etc

SOURCE Venture Economics Inc , prowded In telephone interwew with OTA staff, May 22, 1984

over 30 new computer chip firms have startedoperation since the late 1970s. The fastest-growing niche markets noted earlier are heavi-ly populated with products that are new, orradically different from those of only a decadeago. Few of our major competitors now havea comparable combination of factors to spawnthe next Silicon Valley. In fact, our West Ger-man,28 French, and Japanese trading partnersare now actively seeking ways to emulate theenvironment that fosters this type of entre-prenuership. 29

Small Businesses and Joint R&D

Congress has passed, and the Presidentsigned, legislation that would ease the restric-tions of antitrust laws on joint R&D ventures.— — — —

28’’ The Technological Challenge: Tasks for Economic, Social,Educational, and European Policy in the Years Ahead, ” speechby Hans-Dieter Genscher, (West German) Federal Minister forForeign Affairs, Bonn-Bad Godesberg, Federal Republic of Ger-many, Dec. 13, 1983.

*g”In This New Age of Entrepreneurs, We’re Number OneAgain, ” Joel Kotkin, Wdu”ngton Post, Apr. 29, 1984, p. B1.

Small business may find both new opportun-ities and impediments. Although small busi-nesses are currently permitted to undertakejoint research activities among themselves,the new joint venture legislation may en-courage increased opportunities in joint ven-tures between small and large firms.

This type of cooperation between large com-panies and small companies has already begunto occur. For example, Control Data Corp. hasmade its advanced computer design toolsavailable to two small companies, Star Tech-nologies, Inc. and ETA Systems, Inc. (the lat-ter a CDC spin-off). IBM also has a market-ing agreement with Floating Point Systems.If the joint venture legislation can stimulatetransfers of marketing and technological re-sources of major U.S. companies to small in-formation technology firms, small companiescan realize definite benefits from the en-couragement of these new types of arrange-ments. A potential impediment could occur iftechnology developed through R&D by largefirms is not made available to small firmsthrough licensing or other means.

Where the Technology is HeadingThis section looks ahead to prospects for fur- tant tools for coping with increasing complex-

ther advances in two building blocks of infor- ity through the use of computer-aided designmation technology-microelectronics and soft- (CAD) and computer-aided engineering (CAE)ware. As discussed later, these technologies systems, and are driving a wide range of ap-are being integrated and are providing impor- placations.


Advances in microelectronics and softwareare dependent on entirely different under-pinnings-e.g., physics, chemistry, materialsscience and electrical engineering for theformer, and computer and information science,psychology, and software engineering, amongothers, for the latter. Prospects for continua-tion of the explosive growth of informationtechnology during the past two decades willbe paced by advances in microelectronics andsoftware.

Microelectronics30

The rate of technological improvement insemiconductors has been phenomenal duringthe past quarter century, and has sparked newapplications and economic advances world-wide. This section reviews the growth ofintegrated circuit capabilities and seeks to pro-vide insight into future progress, its depend-encies and foreseeable limitations, and thelikely timeframe of that progress.

Probable trends include:●

●

●

Silicon is likely to remain the most impor-tant semiconductor material for the nextdecade and beyond, while gallium arse-nide is likely to come into wide use in cer-tain applications where its special prop-erties provide specific advantages. Theimpact of other innovative materials islikely to be marginal until after the turnof the century, when biotechnology maybegin to converge with information tech-nology.Prices for the main output of microelec-tronics technology, logic and memory, canbe expected to continue to decline on a perfunction basis, following the general pat-tern of the past two decades.Continued advances during the next twodecades in large-scale integration of sili-con and gallium arsenide will require the

——— .—.—.30A ~i~ficmt mount of data in this section is t~en from,

or based om 1) the Sixth Mountbatten Lecture, M2”c.melectmm”csProgress Prospects, delivered by Ian M. ROSS, President, AT&TBell Laboratories, Nov. 10, 1983, London, England; and 2) J.D. Meindl, Theoretical, Practical, and Analogical b“zm”ts toULLU, Technical Digest of the International Electron DevicesMeeting, December 1983, pp. 8-13.

●

continued advance of other sciences andtechnologies, particularly physics, chem-istry and materials science, and manufac-turing and software engineering.The R&D costs for incremental advancesin microelectronics technology are ex-pected by industry experts to increase asthe complexity of circuits increases andas the theoretical physical limits of micro-electronics are approached.

Silicon Integrated Circuits

Transistor Size and Density

Between 1972 and 1981, the number of tran-sistors that could be packed on a chip doubledeach year (11,000 in 1972 and 600,000 in 1981)(see fig. 56). Today, integrated circuit technol-

Figure 56.—Historical Development of theComplexity of Integrated Circuits

1 05

. ->

1 04

1 03

1970 1975 1980 1985

Introduction year

SOURCE: Arthur D. Little, Inc. estimatea.


ogy is approaching the capability of packinga million components (mostly transistors) ona single silicon chip.31

Along with these improvements in density,costs per function have dropped dramaticallyuntil now some memory components are lessthan 0.01 cent per bit, and many microproc-essors are under $10 per chip (see figs. 57 and58). This is the result of a thousand-fold de-crease in the cost of manufacturing comparedwith 20 years ago.

The substantial increases in transistor den-sity have been made possible by reductions inintegrated circuit feature size, which in turnhave depended on advances in photolithog-raphy, the technique by which integrated cir-cuits are manufactured. Minimum line widthshave been reduced by a factor of two every 6or 7 years. Using photolithography with visi-ble light, line widths have been reduced from25 microns in 1972, or l/1,000th of an inch, to

$lwhile other technologies (gallium arsenide, iridium ~-timonide, cryogenic superconductors, photonic devices, and POtentially biotechnology) with higher switching speeds, lowerheat dissipation, or other qualities, will compete in certainspecial applications, current expert opinion is that silicon willhave the major role for the next decade or more in meeting mostneeds for digital memory and logic.

Figure 57.—Read-Only Semiconductor MemoryComponent Prices (per bit) by Type

H P R O M , a n d E P R O M .

Fixed. ..L— . Mask-ROM

SOURCE: Arthur D. Little, Inc., estimates.

Figure 58.—Median Microprocessor Price v. Time(1,000 unit purchase)

100

10 i

1

Uncerta inty

,10 1 1 11977 1982 1987 1992 1997

Years

o

4-Bit

SOURCE: Arthur D Little, lnc , estimates

about 1.5 microns currently, or less than 1/10,000th of an inch.

Optical, or visible light, lithography, whichcurrently remains the leading integrated cir-cuit manufacturing method, is expected toreach the practical limits of its capacity forsmall feature sizes in the range of 1.0 to 0.5microns (the wavelength range of visible light).At the same historical rate of progress, theselimits will be reached in the 1990-94 period (seefig. 59.).

Most experts believe that advanced litho-graphic techniques using light of shorter wave-lengths, in the X-ray range, and electron-beam(see fig. 60) and ion implantation machinesthat can directly draw lines and features onsemiconductor substrates32 will be needed in

gZ1n Cwent co~erci~ lithographic techniques, the semicon-ductor substrate is coated with a light-sensitive emulsion, orphotoresist, and a mask, much like a stencil, is used to exposea pattern in the substrate that is then etched away with a chem-ical, creating lines and features. Electron beams can be focusedand directed to create patterns in the emulsion without a mask,and ion beams can implant materials on the substrate withoutthe need for photoresist or etching.


Photo credit: AT& T Bell Laboratories

Microprocessor WE 32100

order to achieve further advances.33 A num-ber of such machines are in use for experi-mental purposes and for production of circuitsto be used in-house at major companies suchas IBM and AT&T.

Electron beam lithography and ion implan-tation34 may become the ultimate semiconduc-tor manufacturing techniques, with minimumline widths possible in the range of 0.1 to 0.01micron. The latter widths are comparable tofeature spacing on the order of 20 to 200——————

~~~me en~nWr9 have begun to question the premise thatmore densely packed circuits are necessary. See “Solid State, ”IEEE Spectrum, January 1984, p. 61.

WI’he Defense Department’s Very High Speed Integrated Cir-cuits (VHSIC) program supports research on electron-beam andion-beam manufacturing techniques. Discussion of both of thesecan be found in IEEE Spectrum, January 1984, pp. 61-63.

Figure 59.— Minimum Linewidth for SemiconductorMicrolithography

10

1

0.11977 1982 1987 1992 1997

Years


atoms. The major drawback of such systemsis that they are much slower than mask lithog-raphy; they are now used only in producingcustom chips of low production volume. Elec-tron-beam and ion implantation technologyare expected to be in use on commercial in-tegrated circuit production lines by 199435 (seetable 58).

Research is under way in the United Statesand Japan that is striving to achieve moredensely packed circuit components by stack-ing chips into three dimensional structures. Anumber of problems need to be solved to real-ize 3-D circuits, among them is the difficultyof making connections with elements buriedwithin layers of semiconductor material. Ion-beam equipment is seen as critical to produc-ing the complex microscopic structures re-quired for such 3-D chips.36

—. .—.—351EEE Swtnn.n, January 1984, p. 63.‘6’’3-D Clups, ” The Econonu”st, Feb. 12, 1983, p. 84.

Table 58.—ICs Made with Lithographic Techniques(percent)

Technique 1983 1985 1987

optical scanner. . . . . . . . . . . . . 85 58 38Optical stepper . . . . . . . . . . . . . 15 40 52Electron beam . . . . . . . . . . . . . . 0 1 8X-ray . . . . . . . . . . . . . . . . . . . . . . 0 1 2Ion beam . . . . . . . . . . . . . . . . . . 0 0 0SOURCE: IEEE Spectrum, January 1984, p. 80,

Ch. 9–Technology and Industry ● 327

Figure 60.—A Schematic of an Electron Beam Lithography System,and its Uses to Make LSI Logic Circuits

Compu te r

(G)

Beam on-off control I(F) I I I ICI I a~e

. . . . . , I. .

analog circuits

Electrostatic deflection(p)

’ on a w x o r

SOURCE Mosaic, SS//MS//LVLSMJLSJ,S/, VOI 15, No. 1, p 15.

In devices with such subminiature dimen-sions, other limitations become significant,e.g., the dielectric strength, electronic isola-tion, and heat dissipation of the silicon mate-rial-some of which may represent more se-vere limitations than the theoretical limitsimposed by the most advanced manufactur-ing techniques. Transistors with critical di-mensions of 0.1 micron have been demonstratedto operate by Bell Telephone Laboratories.These dimensions would theoretically lead tosome billion components on a square centi-meter of silicon. A more realistic, practicallimit for silicon may be 100 million compo-nents per square centimeter.37

Increasing the physical dimensions of thechip may also contribute to increasing thecomponent count per chip. Chip sizes are—

“ROSS, op. cit., p. 6.

determined by manufacturing control–theability to minimize the number of impuritiesand defects in the semiconductor substratematerial and in finished circuits. Today’stypical chip size is about 1 square centimeter.There are development efforts under waytoward wafer-scale integration38 which couldprovide a factor of 100 increase in integratedcircuit chip size, to 100 square centimeters orlarger. This factor, coupled with componentdensities of 100 million transistors per squarecentimeter made possible with the advancedmanufacturing techniques noted above, couldresult in component counts of 1 billion tran-sistors per integrated circuit.39

. —SsChips me cut from 3 to 6-inch-diameter wafers which are

currently manufactured with some 100 separate integrated cir-cuits per wafer.


38-802 0 - 85 - 22


Integrated Circuit Speed and Reliability

Concurrent with, and largely the result of,decreasing circuit feature size and increasingdensity, circuit switching speeds have in-creased significantly. A little more than 20years ago, switching delays were crossing themillionth of a second threshold; by 1977 thebillionth of a second threshold had been passed;by 1990, delays of 1 trillionth of a second (apicosecond, or 10-12 second) may be possible(see fig. 61). Practical limits are foreseen, how-ever, that would reduce actual maximum per-formance to one-tenth that speed, or 10 pico-seconds .40

As well as reducing circuit delays, the con-tinued shrinking in the size of individualcomponents and the concentration of morefunctions in each succeeding generation ofintegrated circuits have reduced computer sys-tem chip counts and thus the time used up incomputers when signals travel between chips.

— —.—‘“Ibid.

Figure 61 .—Trends in Device Speed—Stage Delayper Gate Circuit for Different Technologies

1977 1982 1987 1992 1997

Years


Chip reliability has also improved. It is esti-mated to have increased by at least a factorof 100, while the scale of integration has in-creased by a factor of 1,000.41

Gallium Arsenide Integrated Circuits

Another compound, gallium arsenide (GaAs),should find an increasing number of applicationsas a semiconductor substrate because of itsspecial characteristics. These include high fre-quency operation, low power consumption, ra-diation resistance and optoelectronic and pho-tonic properties.

Analog Devices

Currently, GaAs is used mainly in discretecomponent (non-integrated) analog devicessuch as microwave and millimeter wave radar,telecommunications and electronic warfaretransmitters. Electronic signals at these highfrequencies (10 to 100 Gigahertz or billioncycles per second) can be generated at muchlower power with GaAs than is possible withconventional transmitter technologies. Inte-grated circuit, or monolithic, microwave andmillimeter wave GaAs circuits have recentlybecome available. Integration promises im-proved performance and lower prices in militarysystems and private sector microwave com-munications (local loop bypass and other)applications. Cellular radio is expected to pro-vide a large market for GaAs transmitters,receivers and amplifiers.

Another application that offers a large po-tential consumer market for analog GaAs in-tegrated circuits is direct broadcast satelliteservice (DBS). The principal French laboratoryof N.V. Philips Co., based in The Netherlands,recently announced development of a GaAschip that includes all of the functions neces-sary for the reception of DBS signals.

Optoelectronic and Laser DevicesGallium arsenide is provoking excitement in

another telecommunications technology, fiberoptics. It can be fashioned into tiny lasers to



function as light sources for fiber optic trans-mission, into optoelectronic photodetectors toreceive optical signals and translate them intoelectrical signals, and into signal processingcomponents for fiber optic system amplifiersand repeaters (see Fiber Optics Case Study).Recent work at Bell Labs has developed large-scale integrated (LSI) GaAs signal processingcircuits that are scheduled for production.42

Heretofore, separate devices have been re-quired for fiber optic light generation, recep-tion and signal processing. Communicationamong these separate devices introduces de-lays in fiber optic transmission systems.Lasers have been particularly difficult to inte-grate with the other circuit components onsingle chips because of their complex structureand the large amount of heat that is generatedby their operation.

Fujitsu, the Japanese electronics giant, hasconducted GaAs R&D funded by that govern-ment as part of a $74 million, 7-year projectamong a number of Japanese companies. In1983, it announced development of a processthat integrates a laser light source and thesignal processing transistors necessary for fi-ber optic transmission, all on a single GaAschip that operates at room temperature. Asimilar integrated optoelectronic photoreceiv-er is in the works.

American researchers at Honeywell andRockwell International have also announcedthe development of integrated GaAs opticaltransmitters. 43 The Rockwell device in particu-lar is reported to be adaptable to well-knownLSI circuit manufacturing techniques, promis-ing low production costs.”

— .— .-.— ——‘zl??~ectrom”cs, Oct. 6, 1983, p. 154.“Honeywell announced an integrated optical transmitter in

1982, but it originally required special cooling. The Honeywelldevice has since been improved. See C. Cohen, “Optoelectronicchip integrates laser and pair of FETs, ” Electrom”cs, June 30,1983, pp. 89-90. See also The Econonu”st, Feb. 5, 1983, pp. 82-83.

441, Waler, ‘lG~s Optical IC Integrates Laser, Drive Tra-nsistors, ” Electronics, Oct. 6, 1983, pp. 51-52.

Digital Devices

Interest has begun to intensify in the useof GaAs as a replacement for silicon in digitallogic and memory devices. High switchingspeeds and low power consumption make thissemiconductor attractive in certain applica-tions. For instance, Cray Research has an-nounced its intention to develop and use GaAschips in the Cray-3 supercomputer which isscheduled to be available in 1986 (see Ad-vanced Computer Architecture Case Study).But commercial semiconductor firms havethus far been reluctant to move GaAs logicand memory into production. Only low levelsof integration (fewer devices per chip thansilicon) and low yields (few good chips in abatch) are currently attainable with GaAs. Al-though 42 U.S. companies conduct laboratorywork on GaAs, only two firms have announcedplans to sell digital GaAs chips.45

The Department of Defense is interested indigital GaAs because, as well as having speedand power consumption advantages, GaAscircuits are highly resistant to radiation ef-fects, making them attractive for military andspace applications. Since 1975, DOD, throughDARPA, has funded a total of $6 million to$7 million in R&D of digital GaAs integratedcircuits. DARPA has announced plans tospend an additional $25 million to set up pilotproduction lines for GaAs 64K RAM and6,000 to 10,000 element gate array (program-mable logic) chips. The contract has beenawarded to a joint venture between RockwellInternational, who will develop the gate arraychip, and Honeywell, who will design the RAMchip. Each of the two companies will developproduction capacity for both the logic andmemory chips to provide second sources foreach.46

“W. R. Iversen, “Pentagon Campaigns for GaAs Chips, ”Electronics, July 28, 1983, pp. 97-98.

*J. Robertson, “Say Rockwell, Honeywell Get $25M DARPAIC Pact, ” Ekctrom”c News, Feb. 30, 1984, p. 4.


The DARPA program has received mixedreviews from the industry. On one hand, Rock-well believes that the DARPA-funded facil-ities will help make digital GaAs profitable forthe entire industry, since the technology is ex-pected to be made available to other defensecontractors, 47 but neither of the two compa-nies that have announced plans to developdigital GaAs for the merchant semiconductormarket. The companies-Gigabit Logic, whichrecently entered into a private R&D ventureto develop GaAs RAM chips,48 and HarrisMicrowave Semiconductors-submitted bidsfor the DARPA contract. Both cited theDARPA emphasis on low power for portablemilitary systems at the expense of higherspeed as the major factor in their decision toforego bidding on the contract.49 This trade-off is seen as forfeiting the major advantageof GaAs in commercial competition with sil-icon digital memory and logic devices.

Optical Logic

Another area of applied research in whichgallium arsenide is playing a critical role is op-tical logic. The replacement of computer proc-essors based on electronics with ones based onphotonics, which would theoretically be capa-ble of vastly higher speeds, has been consid-ered a possibility for some time. Such opticalcomputers might be capable of trillions ofoperations per second.5o With the spread of fi-ber optic communications technology, opticalprocessing is even more attractive. The largecapacity of light-wave transmission demandshigher processing speeds to fully exploit theavailable fiber optic bandwidth. Conversionsbetween optical and electronic signals intro-duce delays into current fiber optic systems.

Only recently, with the use of GaAs andanother new semiconductor material, iridiumantimonide, has the technology been available

471 bid., see also Iversen, op. cit., p. 97.‘a’’ Gigabit in $6.lM Ga& static RAM R&D Venture,” Elec-

trom”c News, Jan. 9, 1984, p. 66.4eIversen, op. cit., pp. 97-98.50E. Abraham, C. T. Seaton, and S. D. Smith, “The Optical

Computer,” Scientific American, February 1983, pp. 85-88.

to begin to develop optical devices that func-tion like the logic elements that make upmicroprocessors. Hughes Aircraft ResearchLaboratories, which is performing R&D workon optical logic under contract from the Na-tional Security Agency, has announced the de-velopment of a prototype device that is ex-pected to be capable, in about 3 years, of morethan 10 times the speed of current electroniclogic devices. The device is to be applied indata encryption systems.51

The major drawback seen with current ap-proaches to optical computing is that the logicelements are much larger than comparableelectronic ones. Barring fundamental theoreti-cal and practical advances, especially in physicsand in materials sciences, optical logic systemswill probably not see widespread use beforethe end of the century.

The Convergence of Bio-technology andInformation Technology

Advances in the biological sciences andbreakthroughs in genetic engineeering haveencouraged speculation and some preliminaryresearch on the possibility of designing bio-computers. Devices constructed from bio-chemical molecules potentially offer greaterlogic and memory densities, and the possibil-ity of totally novel information processingmethods, including the emulation of brainfunctioning. The first interdisciplinary scien-tific conference on this subject, sponsored bythe National Science Foundation and the Uni-versity of California, Los Angeles, was heldin October 1983.62 The meeting brought to-gether biologists, chemists, physicists, math-ematicians, computer scientists and electricalengineers to discuss the prospects for thistechnology and to generate questions to guideresearch.

The field is just beginning to adopt theoreti-cal concepts on how such biochemical-basedmachines might operate, and on the more dif-

SIL. Wdler, “components for Optical Logic Stat to Click, ”Electronics, Dec. 29, 1982, pp. 31-32,

sZInt,ernation~ Conference on Chemically Based Computing,Santa Monica, CA.

Ch. 9—Technology and Industry • 331

ficult questions of how one might constructa bio-computer and make it work in concertwith existing information systems. LewisMayfield, head of the Chemical and ProcessEngineering Division at NSF concluded, “Theconsensus of the Santa Monica Conferencewas that a great deal more work will be neededat the fundamental level before developmentcan begin. So although we will accept grantproposals in this area, we have no plans toorganize a program to promote it. ”53

Software

Software and hardware are primary ele-ments in information technology systems. Thehardware provides the set of generic functionsthat information systems may perform (input,storage, processing, communications and out-put), and the software combines and organizesthese functions to accomplish specific func-tions. The two elements are responsible for theburgeoning range of capabilities of informa-tion technology, and, increasingly, for ad-vances in the technology itself.

Software is the part of information technol-ogy that is most readily modifiable. A majorattraction of computer-based systems is thatthey are programmable-machine functionsare fixed, but their sequence may be modifiedto achieve different ends at different times orto respond to changing conditions. The con-tent of information systems maybe changed,expanded and contracted, within limits, to ac-commodate the changing needs of users. Soft-ware is the vehicle for providing the flexibilityof information systems to the users of thetechnology; software aids users in communi-cating and managing large bodies of informa-tion and in coping with the complexity of largeor specialized systems.

Traditionally, the term “software” narrowlyreferred to programs for large mainframe com-puters and was of concern only to computerprofessionals. But as computer power hasspread to wider segments of society, the char-

——“J. B. Tucker, “Pioneering Meeting Puts Biochips on the

Map, ” High Technology, February 1984, p. 43.

acteristics of software have been modified bythe uses to which computer systems are put,and the meaning of software has broadened.Now, software is both the instructions that di-rect the operations of computer-based sys-tems, and the information content, or data,that computer systems manipulate.

The focus of software production and use ischanging from concerns of computer profes-sionals to the concerns of a broad range of endusers of information technology -i.e., thosewith no particular interest in the technology,but only its capabilities and results. Becauseof this and the increasing availability of com-puter and telecommunications capabilities,future limits on the utility of information tech-nology will depend less on hardware capabil-ity and more on the difficulties of defining andaccommodating the needs of users.54

Possible implications of an expanding num-ber of users of information technology, and ashift in the focus of concern toward the needsof that expanding set of users include:

●

●

●

●

●

� ✎ ✍ �

The relative importance of software issuesin the research and development of infor-mation technologies will likely increase.Small companies will continue to be im-portant players in software developmentbecause of their ability to respond to thespecialized needs of small segments of theuser population.There will be movement toward standard-ization of software for common applica-tions, such as in accounting, banking, in-surance and government.There will be continuing pressure to in-troduce systematic engineering practicesinto the production of software to lowersoftware development costs and to assurethe reliability of software in critical ap-plications. (See Software EngineeringCase Study.)There will be a demand for higher level,nonprocedural computer languages forthe development of new and unique appli-

“An analogy can be drawn between computer users and automobile drivers: all want the benefits of the technology but mostare indifferent to how it works.


cations; such languages will allow the userto focus on the characteristics of the tasksto be performed and not on how the hard-ware performs them.

● The search will intensify for powerful andsimplifying concepts for the organizationof information and knowledge in manyfields.

● Computerization will require the users ofinformation systems to more clearly de-fine and organize their habits and opera-tions related to information use.

● There will be more concern among usersof information technology with the usesof information and the content of infor-mation systems, and less concern withhow the technology works.

The production of software has become amajor industry in the United States, with esti-mated expenditures of $40 billion in 198255 forsoftware products and for in-house develop-ment of programs. Sales of software in theUnited States doubled between 1982 and 1983from $5 billion to more than $10 billion, andare expected to increase to more than $24 bil-lion in 1987 (see table 59).

As information technology applications pro-liferate in the office, factory and home, and inthe military, computer software is required tofulfill many new and complex functions. Thegrowing demand for software is causing pres-

5%. Olsen, “Pathways of Choice,” Mosaic, July/August 1983,p. 3.

sure to make software more useful, reliableand cost effective, and to make software de-velopment more productive. Not unexpectedly,software comprises higher and higher propor-tions of the cost of new information systems.Estimates of the relative cost of software inlarge systems range to above 80 percent (seefig. 62).

The Changing Role of Software inInformation Systems

Computer software is traditionally classifiedas two general types: applications softwarethat is designed to apply computer power to

Figure 62.—The Relative Costs of Software andHardware

1955 1970 1985

Years

SOURCE: Steve Olson, Pathways of Choice, Mosaic, July/August 1963, p. 6.

Table 59.—Software Sales by Use

1982 1983 1984 1985(millions of dollars)a

Software, total , . . . . . . . . . . . . . . . . . . . . . . . 5,001 10,309 15,017 24,677Application programs, total . . . . . . . . . . . . . 1,997 3,448 5,455 11,100Computer-aided design, manufacturing

and engineering . . . . . . . . . . . . . . . . . . . . . 776 1,126 1,636 3,200Other applications . . . . . . . . . . . . . . . . . . . . . 1,221 2,320 3,819 7,900Systems software, total . . . . . . . . . . . . . . . . 3,004 6,861 9,562 13,577

Compilers, interpreters, and assemblers . . 541 610 700 932Data-base management systems . . . . . . . . . 1,100 1,430 1,888 3,500Diagnostic and performance monitoring . . 493 645 710Operating system . . . . . . . . . . . . . . . . . . . . . . 870 4,176 6,264 8 , %aAll figures in current U.S. dollars.

SOURCE: Electronics, Jan. 12, 1964, p. 128,

Ch. 9—Technology and Industry • 333

a specific task or tasks, such as computer-aided design of automobiles or inventory man-agement in a retail store; and systems soft-ware that is used to manage the operations ofinformation systems themselves, such as com-puter operating systems or database manage-ment systems.

Applications software has come to includea user interface that allows persons other thancomputer professionals to work with the com-puter system, while systems software is de-signed to help computer professionals writeand execute programs, make efficient use ofsystem components, diagnose and correctfaults, and manage and audit system resources,contents and usage. This distinction, thoughuseful, is beginning to break down with theadvent of personal computers, in which oper-ating systems are important tools to helpusers control the workings of their machinesand perform such mundane tasks as copyingfiles. However, the ends to which the softwareis applied still remain distinct-applicationssoftware solves user problems, systems soft-ware solves computer problems.

In general, systems software is an integralcomponent of the hardware because its job isto control the hardware and to schedule andaccommodate the development and executionof applications programs. The trend has beenfor the manufacturers of hardware or special-ized software vendors to write systems pro-grams, and for more of the systems programsto be embedded in hardware, or ROM (ReadOnly Memory), making the end user incapableof altering them. There is also a trend towardthe standardization of operating systems toincrease the portability of applications soft-ware, or the ability to use identical applica-tions programs on different machines.

The expansion of the computer user popula-tion is being fueled by the expanding contentof information systems, and is leading to anexpanding definition of the term “software.”Information systems are increasingly encom-passing and integrating many collections ofinformation formerly considered to be separateand segregated, such as the files of corporate

divisions or government departments. Thetechnology that has heretofore embodied suchlarge information bases, for the most part pa-per and print, was not integrated; thus the in-formation contained in those files has not beenaccessible to single individuals. Now, rela-tively cheap computer terminals and telephonelinks are making the enormous amounts of in-formation being stored in digital electronicform (software) accessible by people in theirhomes and offices.56 Through such services asvideotex, newspapers, periodicals and even en-tire libraries may eventually be made available“on-line.

Integration of information management onsuch a large scale will require sophisticatedcontrol software to help people find the infor-mation that they want; advances in informa-tion science and contributions from such fieldsas artificial intelligence and psychology willbe needed for this to occur. The integration ofinformation management will also have pro-found effects on the way people view andhandle information, and on how organizationsconduct their business; the social sciences (so-ciology, political science, anthropology) will in-creasingly be looked to for understanding ofthe inevitable organizational and human con-flicts that will flow from the changes in infor-mation access and use made possible by infor-mation technology.

The Software Bottleneck

Demand for computer applications is out-stripping the ability of the traditional largedata processing organizations (e.g., banks, in-surance companies, government agencies) tosupply the programs these organizations needin a timely and cost-effective manner. The lag

~eThe massive integration of information SYSteIXIS rtiSeS a

number of public policy issues including questions of the gov-ernment role in protecting intellectual property, pn”vacy andother civil h“berties assuring eqw”ty of access to information,and the effect of information technology on the processes ofgovernment itself. These and related issues are the subject oftwo OTA assessments now in progress, “Intel.lectusl PropertyRights in an Age of Electronics and Information” and “Fed-eral Government Information Technology: Administrative Proc-ess and Civil Liberties. ”


time between identification of an applicationand completion of the software in the averagedata processing department is 2½ years andis increasing, probably discouraging the pro-posal of many new applications.57 There areanumber of factors that contribute to this “soft-ware bottleneck, ” chief among them that ap-proximately 50 percent of the data processingbudgets are spent maintainingg old programs—making changes in existing codes because oferrors in the design or coding of programs, orbecause applications requirements were inadequately specified.58

Another factor that slows the developmentof new programs is that software develop-ment, as it is practiced now, is project-ori-ented. Each new program is built up fromscratch with little systematic reuse of partsof older programs, even though identical func-tions may be implemented. The knowledge andexperience gained in one project are not sys-tematically preserved and transferred to otherprojects. Innovations in programming tech-niques tend to be adopted piecemeal as proj-ect budgets allow. As a result of this projectorientation and lack of standardization, largeprograms that must rely on the efforts of anumber of programmers tend to be patch-work of pieces that reflect different program-ming styles, and are often nearly incomprehen-sible to people other than the original authors.These problems are exacerbated by a turnoverrate among computer professionals estimatedat 15 percent per year.59

—.. ——‘7A recent survey of IBM mainfrarnebased data processing

departments found that the average time lag for the initiationof applications programming projects is 2 years and this lagis growing by 3 months per year. Once under development, theaverage application requires 8 months to complete. At the sametime, software representing 10 months of programming effortis discarded yearly by the average data-processing departmentbecause of obsolescence. A certain backlog of work is desirableto keep data-processing staff busy, but it is clear that manyorganizations are falling behind. Application Development inPractice, Tech Tran User Survey, Xephon Technology Trans-fer Ltd., 1983., p. 2.

‘“W. Rauch-Hindin, “Some Answers to the Software Prob-lems of the 1980s, ” Data Commum”cations, May 1981, p. 58.See also the Case Study on Software Engineering.

‘This figure reflects a recent decline in turnover attributableto the last recession and may well be on the rise. Also, it isbelieved by experts that the greatest turnover is among thoseprofessionals writing code, so that this figure understates theeffect of turnover on program development productivity.

These difficulties combine to produce thepresent software development situation inwhich time is wasted in writing essentiallyidentical parts, testing is laborious and expen-sive, and maintenance programmers spendmost of their time trying to understand pro-grams rather fixing them. The difficulties havegrown out of the historical context in com-puting in which hardware resources were ex-pensive, and the creative aspect of program-ming was devoted mostly to the efficient useof these resources through the clever compres-sion of code. Now, with the precipitous declinein hardware costs, progr amming productivityrather than code efficiency is widely recog-nized as the limiting factor in the use of com-puting power.

The software development resources now inuse appear to be inadequate and too frag-mented to keep up with the demand for com-puting power. These resources include trainedmanpower and tools (software developmenttools include workstations, computer-aided-design, coding and testing programs, and doc-umentation) to aid designers, programmersand managers in creating and maintainingcost-effective, reliable software.

Computer manufacturers and software de-velopers, and government, university and in-dustry experts are recognizing the problemsassociated with the software bottleneck andsee the potential for sizable productivity gainsby changing the nature of software develop-ment. For example, some avenues to alleviat-ing the productivity problems of software de-velopment and maintenance lie in the use andreuse of standard function software modules,and the development of software to produce,test, and maintain other software. Researchand development efforts in software engineer-ing are seen as essential to this change (seeSoftware Engineering Case Study).

Software and Complexity

Computer software is an important factorin making information technologies useful ina widening array of complex applications—from space vehicles to telecommunications


networks to weapons systems. Software isused to produce simulations of complex sys-tems and events, so that scientists and engi-neers can experiment and analyze withoutbuilding expensive prototypes, and to controlthe operations of the most complex of man’screations, such as nuclear reactors and thespace shuttle.

The need to manage a variety of increasinglycomplex systems is a major source of the pres-sure to improve software capability and reli-ability. For example, a large air traffic controlprogram can consist of more than half a mil-lion instructions that must mesh perfectly andbe essentially free of errors.60 The efforts ofhundreds of programmers over several yearsmay be required to produce a software systemof this size. The complexity of the softwaremakes it virtually impossible to assure thatno errors are present in a completed program.61

Software written for critical applicationssuch as air traffic control, manned spacevehicles, nuclear reactors, weapons systems,or electronic funds transfer, although impos-sible to guarantee as error-free when com-pleted, must be constructed in such a way thatat least the parts of programs and their inter-connections are rigorously designed and ade-quately tested as they are built up. The com-plexity of software development is likely toincrease dramatically with the advent ofmassively parallel computing systems and theconstruction of large knowledge-based systems.

. —~OolSen, Op. cit., p. 3.GIEvery twochoice brnch point in a program (where decision

is made on what procedure will follow depending on the condi-tions that are present) doubles the number of possible pathsthat the program can take, so that a program with 10 branchpoints will have 1,024 possible paths. A program for a largeair traffic control system may have more than 39,000 branchpoints. If a computer could test a trillion paths per second (acapability far beyond toda ‘s most powerful supercomputers)

?’it would require over 1011’ 31 years (that’s the number one fol-lowed by 11,731 zeros) to test all possible paths in such a pro-gram. See P. E. Olsen, and R. W. Adrion, M. A. DennisBranstad, and J. C. Chemiavsky, ACM Computing Surveys,June 1982, p. 184.

Conclusion

The microelectronic and software buildingblocks for information technology are likelyto advance dramatically in coming decades.The path of progress for microelectronics isreasonably clear into the 1990s as practicaland theoretical limits are approached forsilicon and as other technologies come intouse. For software, the path is less clear, butmost likely will include far more standardiza-tion, reuse of software modules, and develop-ment of software tools to produce, test, andmaintain other software. Most importantly,the integration of the concepts behind ad-vances in these two building block technol-ogies will shape future capabilities.

Information technology capabilities andtheir efficient utilization will depend upon theintegration of ideas from a variety of dis-ciplines. For example, the acceptance and ef-ficient use of emerging computer and telecom-munications resources in organizations requireunderstanding of behavioral and social vari-ables, as well as understanding of computertechnology. Similarly, computer-aided-designrequires both new software and new hardwarefor the display and manipulation of complexdesigns. The integration of hardware and soft-ware in new microelectronics and software de-sign tools leads to further advances in infor-mation technology itself.

This integrative interaction constitutes apowerful technological engine which is fuelingthe advance of information technology inmany applications—including computers forscience, engineering and business, robotics andcomputer-integrated manufacturing, telecom-munications, and artificial intelligence, amongmany others. These applications in turn willinfluence opportunities for economic growthand create alternatives for change in society,and thus will have major impacts on some ofthe policy issues that will become preeminentfor legislators.

Index

Alvey, John, 106American Council on Education, 150American Electronic Association, 147AT&T, 14, 16, 78, 111-134

Babbitt, Governor Bruce, Arizona, 172Battelle Memorial Institute, 32Bell Laboratories, 11, 14, 16, 74, 113, 120, 153British Alvey Program, 18Brown, Charles L., president, AT&T, 128Buchsbaum, Solomon J., 123Bureau of Labor Statistics (BLS), 142, 146

California Institute of Technology, 17case studies, 9

advanced computer architecture, 9, 55changing computer architecture, 56computer architecture R&D, 57critical areas of research, 63international efforts, 65manpower, 65

artificial intelligence, 13, 87content and conduct, 96foreign national efforts, 105R&D environments, 90

fiber optics, 12, 67commercialization trends, 69directions of U.S. research, 73research in Japan, 74United States R&D, 71

software engineering, 11, 75content and conduct of software

engineering R&D, 79international efforts in software

engineering R&D, 85software R&D environments, 76

Chynoweth, Alan G., 129Computer Technology Corp., 36Congress:

House Committee on Energy and Commerce,3

House Committee on Science andTechnology, 3

Joint Economic Committee, 172Senate Commerce Committee, 128

Corning Glass Works, 73

Defense Advanced Research Projects Agency(DARPA), 9, 29, 58, 93, 95, 219

Department of commerce, 32Department of Defense (DOD), 5, 11, 13, 20,

29, 41, 58, 75, 78, 89, 93, 279, 294, 329Department of Energy, 4, 10, 20, 58

Department of Justice, 33, 111, 116deregulation and divestiture on research,

effects of, 111-135antitrust laws, deregulation and divestiture,

113divestiture, 116

factors affecting research, 123allocation of research and development

expenditures, 125availability of research results, 130basic research, 128Bell Communications Research, Inc., role

of, 129stability of earnings, 123

policy implications, 131research at AT&T, management of, 120

Bell Labs after divestiture, 122modified final judgment and Bell Labs, 121

education and human resources for researchand development, 139-166

concern about manpower, 139elementary and secondary education, 154Federal role in manpower development, 157identifying particular manpower problems

and solutions, 141problems in higher education, 149range of manpower predictions, 145relationship between manpower development

and economic growth, 140societal context for determining Federal

manpower policy, 161supply of manpower, 149

environment for R&D in informationtechnology, 25-52

concepts for R&D, 27roles of the participants, 28

conflicts in perspectives, goals, andpolicies, 42

Federal Government, 28Government funding, 28industrial R&D, 34other Federal policies, 31

antitrust policy, 33industrial policy, 33patent policy, 31R&D limited partnership, 33tax credits, 32technology transfer, 32

pattern of Government funding, 30universities’ role in R&D, 37

measures of the health of U.S. R&D ininformation technology, 43, 51

339


information industry profile, 43U.S. patent activity, 45

European Community (EC), 66European Economic Community (EEC), 216European Program for Research and

Development in Information Technology,(ESPRIT), 18, 202, 216, 219

Federal Communications Commission, 31, 111,114, 115

Federal Trade Commission, 33foreign information technology R&D, 201-276

conclusions, 221international trends, 201

adapting technology, 202international technology agreements, 208multinational corporations, 206patents, 210science and engineering students, 212scientific and technical literature, 211technology exchange agreement, 207trade, 201U.S. computer trade, 203

Japan, 222government, 226

Defense Agency, 239Information Technology Promotion

Agency (IPA), 236Japan Electronic Computer Co., 236Japan Development Bank, 239Japan Robot Leasing Co., 236Kokusai Denshin Denwa Ltd., 238Ministry of International Trade and

Industry, 228Ministry of Posts and

Telecommunications: Nippon Telegraphand Telephone, 236

national R&D projects, 231Science and Technology Agency,

industry, 243Fujitsu Ltd., 245Hitachi Ltd., 245NEC Corp., 244Oki Electric Industry Corp., 246

size of Japanese participation ininformation technology markets,

university, 240France, 246

government, 252Agence de l’Informatique, 256Centre d’Etudes des Systemes et

Technologies Advances, 257Centre Mondial Informatique et

Ressource Humaine, 257

235

225

des

Centre National d’EtudesTelecommunications, 254

Centre National de la RechercheScientifique, 253

L’Institut National de Recherche enInformatique et en Automatique, 255

Ministere des Postes,Telecommunications, et Telediffusion,253

Industry, 259CII Honeywell Bull, 259CIT Alcatel, 260Sogitec, S. A., 261Thomson CSF, 260

political environment, 248size of participation in information

technology markets, 247social environment, 252university, 257

United Kingdom, 261ESPRIT, 272government, 263

British Technology Group, 269Department of Education and Science,

268Department of Trade and Industry, 265Ministry of Defense, 268

industry, 270British Telecom, 271General Electric Co., 271Plessey, 271

size of participation in informationtechnology markets, 263

university, 269U.S. information technology R&D policies,

implications for, 214government/industry/university

institutional arrangements, 217government role, 216industry participation, 219science and technology policy goals, 215

France, 66, 70, 87, 246French La Filiere Electronique program, 18

General Accounting Office (GAO), 39, 78, 155General Electric, 181Georgia Institute of Technology, 30goals for Federal R&D policy, 3Great Britain, 66, 71, 86, 261Greene, Judge Harold H., U.S. District Court

for the District of Columbia, 116

Heckler, Representative Margaret, 139

Index ● 341

IBM, 12, 36, 53information technology R&D, nature of, 6

close boundary between theory andapplication, 8

complexity, 8multidisciplinary nature, 7software as technology, 7

Inman, Admiral Bobby, 194Institute of Electrical and Electronic

Engineers, 148issues and strategies, 14

changing roles of universities, 17foreign programs, 18science policy, 19scientific and technological manpower, 16telecommunications deregulation, impacts of,

14

Japan, 65, 70, 74, 85, 222Japanese Fifth Generation Computer Project,

13, 18, 65, 105Japan Ministry of International Trade and

Industry (MITI), 65, 75, 105, 219, 228

Lawrence Livermore National Lab, 58legislation:

Economic Recovery Tax Act, 32House Authorization Act of 1984, 41Merrill Act of 1862, 158, 159National Cooperative Research Act of 1984,

33National Defense Education Act of 1958, 159Public Law 96-517, 31Sherman Antitrust Act, 113Stevenson-Wydler Technology Innovation

Act of 1980, 32, 215Tax Equity and Fiscal Responsibility Act of

1982, 32Lewis, David, 148Library of Congress, 133Los Alamos National Lab, 58Lotus Development Corp., 202

Massachusetts Institute of Technology (MIT),17, 177

McCarthy, John, 90Microelectronics Center of North Carolina, 36,

184, 188Microelectronics & Computer Technology

Corp., 36, 120, 172, 193Microsoft, Inc., 204

National Aeronautics and SpaceAdministration (NASA), 4, 10, 30, 31

National Bureau of Standards, 37, 85National Center for Education Statistics, 34National Oceanic and Atmospheric

Administration (NOAA), 4National Science Foundation (NSF), 8, 9, 27,

29, 30, 40, 41, 58, 94, 142, 147, 159National Science Teachers Association, 155National Telecommunications and Information

Administration, 133Nelson, Richard R., 129new roles for universities in information

technology R&D, 169-195forces driving new relationships, 171impacts of new university arrangements, 173

expected benefits, 174potential costs, 175

new roles, 176Microelectronics Center of North Carolina,

184, 188Microelectronics & Computer Technology

Corp., 193Microelectronics and Information Sciences

Center, University of Minnesota, 178MIT Microsystems Industrial Group, 177Rensselaer Polytechnic Institute for

Industrial Innovation, 180Semiconductor Research Corp., 189

Stanford University Center for IntegratedSystems, 182, 185

North Carolina State University, 30

Ohio State University, 30

Penzias, Arno, vice president, Bell Labs, 130Perot, H. Ross, 194principal findings, 5

Rensselaer Polytechnic Institute, 30, 73, 180Research Triangle Park, NC, 173, 231

Semiconductor Research Corp., 36, 120, 189Stanford University, 182, 185

technology and industry, 307characteristics of the U.S. information

industry, 316employment, 317industry structure, 320international trade, 317smd entrepreneurkd firms, 321

composition of information technology, 308examples of system applications, 309

cellular mobile radio communications, 310computer simulation, 312


direct broadcast satellites, 310financial services, 315heart pacemakers, 314robotics, 311

importance of information technology, 307where the technology is heading, 323

convergence of biotechnology andinformation technology, 330

gallium arsenide integrated in circuits, 328microelectronics, 324silicon integrated circuits, 324software, 331-334

Texas A&M University, 194top 15 U.S. companies in R&D spending, 44Trilogy Ltd., 33Tsukuba Science City, Japan, 231Turing, Alan, 90

University of Arizona, 74University of Minnesota, 178University of Rhode Island, 30

University of Rochester, 73University of Texas, 194U.S. patent activity, 45U.S. Patent and Trademark Office, 47U.S. science and technology policy, 279

background, 280general policies, 280science policy statements, 282tenets of science and technology policy,

285information technology R&D policies, 286

key issue areas, 290international competitiveness, 298military/civilian balance of R&D funding,

294organization of government, 290

Vetter, Betty M., 147von Neumann, John, 90

West Germany, 66, 70

0