DOCUMENT RESUME

ED 132 354 CE 008 864

AUTHOR Kriesberg, Harriet M.; And OthersTITLE Methodological Approaches for Determining Health

Manpower Supply and Requirements. Volume I.Analytical Perspective.

INSTITUTION Nathan (Robert R.) Associates, Inc., Washington,D.C.

SPONS AGENCY National Health Planning Information Center(DHEW/PHS), Rockville, Md.

REPORT NO DHEW-HRA-76-14511; HRP-000-7377PUB DATE 76CONTRACT HRA-230-75-0067; HRA-230-75-00073NOTE 83p.; For a related document see CE 008 865

EDRS PRICEDESCRIPTORS

MF-$0.83 HC-$4.67 Plus Postage.*Employment Projections; Health Occupations; *HealthPersonnel; Health Services; Labor Market; *laborSupply; *Manpower Needs; Manuals; *ResearchMethodology; Statistical Analysis

ABSTRACTDesigned to contribute to planr understanding of

the state of the art and to the improvement of 41.-11th manpowerplanning, this monograph describes and evaluate_ various methods usedto determine present and future health manpower supply andrequirements. The methodologies presented, chosen after a review ofthe documents identified in the Inventory of Health Manpower PlanningActivities of 1973 in the Bureau of Health Manpower, are consideredto be practical with regard to the resources available to state andlocal health planners. This first volume provides an analyticalperspective, the definition of terms, the factors that determinesupply and requirements, the concepts that underline eachmethodological approach, and the uses and limitations of healthmanpower statistics within the context of the subject area. A seLondvolume is a manual that describes the methodology step-by-step.(WL)

***********************************************************************Documents acquired by Er-C include many informal unpublished

* materials not available fic other sources. ERIC makes every effort ** to obtain the best copy avai_ Inle. Nevertheless, items of marginal *

* reproducibility are often encountered and this affects the quality *

* of the micrcfiche and hardcopy reproductions ERIC makes available* via the ERIC Document Reproduction Service (EDRS). EDRS is not* responsible for the quality ot the original document. Reproductions ** supplied by EDRS are the best that can be made from the orijinal.***********************************************************************

MethodologicalApproaches

For DeterminingHealth Manpower

Supply andRequirements

VOLUME IAnalytical Perspective

U S DEPARTMENT OF HEALTH.EDUCATION & WELFARENATIONAL INSTITUTE OF

EDUCATIONDOCUMENT HAS BEEN RE0,F0

DUCED YACTLY AS RECEIVED FROMTHE PERSON OR ORGANIZATION ORIGINATING IT POINT:, OF VIEW OR OPINIONSSTATED DO NOT NECESSARILY REPRESENT OFFICIAL NATIO%Al INSTITUTEEDUCATION POSITION C. POLICY

HRP-00C A77

U.S. Department ofHealth, Education, and WelfarePublic Health Serviceaureau of Health Planningand Resources DevelopmentDivision of Planning Methods and TechnologyNational Health Planning Information CenterRockville, Maryland 20850

Prepared By

Harriet M. Kriesberg, John Wu,Edward D. Hollander, Joan Bow,Robert R. Nathan Associates, Inc.Washington, D.C.Under Contract No. HRA 230-75-0067

Edited By

Aspen Systems CorporationGermantown, Md. 20767Under Contract No, HRA 230-75-00073

DHEW Publication No. (HRA) 76-14511

2

Foreword The purpose of this monograph is to describeand evaluate various methods used to determinepresent and future health manpower supply andrequirements. The methodologies presented werechosen after a review of the documents identifiedin the Inventory of Health Manpower PlanningActivities of 1973 in the Bureau of Health Man-power. Documents selected for their methodolog-ical content were then analyzed in detail. Theauthors also provided supplementary material byresearching other references and informationsources.

This monograph does not address the full rangeof methodologies that may be used in overallhealth manpower planning; rather, it is limited tothe specific subject of estimating manpowersupply and requirements. It deals only peripher-ally with related manpower issues that affectsupply and requirements, such as labor produc-tivity, task delegation, and geographic andspecialty distribution.

The methodologies presented are practical andcan be implemented with the resources availableto state and local health planners. Theoretical ormore complex methodologies commonly used atthe national level -have been deliberatelyexcluded.

3

The monograph is presented in two volumes.The first volume provides an analytical perspec-tive, the definition of terms, the factors that deter-mine supply and requirements, the co, icepts thatunderlie each methodological approach, and theuses and limitafions of health ihanpower statis-tics within the context of the subject area Thesecond %.-llume is a practical manual that de-scribes the methodology step-by-step, includingquestionnaire samples and selected tables forestimating health manpower :-:upply and require-ments.

It is hoped that this monograph win contributeto planners' understandirg of We state of the artand to the improvement of health manpowerplanning.

r

Z. Erik Farag, Ph.D.Director, Division of

Planning Methods and Technology

Bureau of Health Planningand Resources Development

4

lv

Preface The National Health Planning InformationCenter, a component of the Division of PlanningMethods and Techno!ogy, Bureau of HealthPlanning and Resources Development, is cur-rently publishing two series of monographs onhealth planning.

The first series, of which this two-volume mono-graph is a part, involves health planning methodsand technology. Future monographs will alsoinclude detailed descriptions of pragmaticmethods developed and used by practicing healthplanners in a variety of locations and situations.The overall purpose is to provide health plannerswith suggestions for methods and proceduresthat can be adapted to their own health planningenvironment and experience.

The second series includes monographs onhealth planning information, as well as generaland topical bibliographies of the health planningliterature and reference works to currentprograms, to individuals, and to institutionsinvolved in health planning.

Copies of all monographs published by theNational Health Planning Information Center canbe purchased from:

The National Technical Information Service5285 Port Royal RoadSpringfield, Virginia 22161

This second monograph, MethodologicalApproaches for Determining Health ManpowerSupply and Requirements, was prepared byRobert R. Nathan Associates, Inc., of Washington,D.C. The monograph is presented in two volumes,Analytical Perspective (Volume I) and 'PracticalPlanning Manual (Volume II).

The report on which this monograph is basedwas originally entitled "Critical Review of Meth-odological Approaches Used To Determine HealthManpower Supply and Requirements." The reportwas funded under HRA Contract No. 230-75-0067.

The project team for Robert R. NathanAssociates was directed by Harriet M. Kriesberg.The report was written by Ms. Kriesberg, John Wu,Edward D. Hollander, and Joan Bow. Advice in thepreparation of the report and review of the finalmanuscript were provided by an outside panel ofpracticing health planners. The panel consisted ofIrene H. Butter of the School of Public Health,University of Michigan; Lewis Dars of the NewJersey Department of Higher Education; andThomas L. Hall of the School of Public Health,University of North Carolina.

Mary C. McGuire of the Bureau of Health Plan-ning and Resources Development, Health

6

Acknowledgments

vi

Resources Administration, DHEW, served asproject officer for the contract. The entire projectwas under the direction of Frank A. Morrone, Jr.,of the Division of Planning Methods and Tech-nology, DHEW. Jack Lass of Aspen Systemsedited and revised the final manuscript prior topublication.

The authors express their deep appreciationand gratitude to all the above-mentioned personsand especially to Hyman K. Schonfeld of theNational Research Council. Mr. Schonfeld kindlyvolunteered to review the sections of the mono-graph that dealt with the health needs approach.While acknowledging their indebtedness to all ofthe contributors who made this monographpossible, the authors wish to assume full respon-sibility for any deficiencies in the contents.

7

CONTENTS

ForewordPreface ivAcknowledgments

I. INTRODUCTION 1

Background 1

DEFINITION OF TERMS 5

Definitions Affect Estimates 5Concepts of Requirements for

Manpower 7Importance of the Assumptions in

Planning Outcomes 14

FACTORS AFFECTING DEMAND ANDSUPPLY FOR HEALTHMANPOWER 17

Determinants of Demand 17Relative Importance of Determinants

of Demand 23Determinants of Supply of Health

Manpower 24Relative Importance of Determinants

of Supply 29The Interaction of Demand and

Supply 30

vii

viii

IV. METHODOLOGICAL APPROACHES TOESTIMATING MANPOWERREQUIREMENTS 32

Introduction 32

The Manpower/Population RatioMethod 33

The Service Targets Approach .... 34

The Health Needs Approach 37

The Economic (Effective) DemandApproach 41

Future Requirements 42Strengths and Weaknesses

of AlternativeMethodologies 45

V. METHODOLOGICAL APPROACHESTO ESTIMATING SUPPLY 47

Current Supply 47

Current Supply Methodologies 50

Future Supply 51

Estimating Inflows 53

Estimating Outflows 55

Strengths and Weaknesses ofSupply Methodologies 56

VI. HEALTH MANPOWER ISSUES: USESAND LIMITATIONS OFSTATISTICS 60

Introduction 60

Problems in Data Gathering 61

Distribution of HealthManpower 64

Effective ManpowerUtilization 65

Productivity: The "PhysicianExtenders" 67

Qualification of Manpower 69Technological Innovation and New

Modes of Delivery 70

I. Introduction

The passage of the rational Health Planning Backgroundand Resources Development Act of 1974 began anew era for health resource planning. Congressdecreed that

... each health system agency shall have asits primary responsibility the provision of ef-fective health planning for its health servicearea of health services, manpower, and facili-ties which meet identified needs, reducedocumented inefficiencies, and implementthe health plans of the agency.....In providing health planning and resourcesdevelopment for its health service area, ahealth systems agency shall perform thefollowing functions:1. The agency shall assemble and analyzedata concerning

a. the status (and its determinants) of thehealth of the residents of its health servicearea

b. the status of the health care deliverysystem in the area and the use of thatsystem by the residents of the area

c. the effect the area's health care de-livery system has on the health of the resi-dents of the area

d. the number, type, and location of thearea's health resources, including healthservices, manpower, and facilities

e. the patterns of utilization of the area'shenIth resources, and

the environmental and occupationalxposure factors aflecting immediate and

long-'erm health conditions.

1 93-641, Sec. 1513.1 0

1

2

The specific purpose of this monograph is tohelp local health planners perform a criticalfunction: the measurement of their area's man-power supply and requirements. It is intended foran audience which has no special training or ex-petrtise in statistical methods, but which has theresponsibilities enumerated in the new law fornealth resources planning. Our objective is to pre-sent the wide range of methodologies used byhealth planners in assessing health manpowersupply and requirements, ranging from the mostsimple to the more complex, but omitting thehighly technical approaches that are not feasiblefor area planners at this time.

The manpower issues before health systemsagencies today differ from those faced by plan-ning agencies in the 1960's. These differencesreflect the dramatic changes that have occurredin the system for the delivery of care in theUnited States in the last decade and the excitingdevelopments on the horizon. Indicative of thesechanges is the rise in the nation's health bill from$26 billion in 1960 to $69 billion in 1970, as wellas the growth of heaRh services employmentfrom 2.6 million in 1960 to 4.1 million in 1970. By1974, 7.7 percent of our gross national product,$104 billion, was devoted to health care. Projec-ted employment in 1975 is 5.3 million, and healthservices spending is expected to reach $126bi!lion.2

The burgeoning demand for health care in thelast decade, whetted by Medicare, Medicaid, andthe spread of pfivate insurance coverage, pro-duced a manpower ,;hortage of crisis proportionsin the late 1960's the health industry's seem-

2. U.S. Bureau of the Census. Statistical Abstract of theUnited States: 1974. Washington, D.C.: U.S. Government Print-ing Of fice, 1974; Worthington, Nancy L. "National healthexpenditures, 1929-74," Social Security Bulletin, 38 (2): Feb-ruary 1975; U.S. Department of Commerce, U.S. Industrial Out-look: 19 75 With Projections to 1980. Washington, D.C.: U.S.Government Printing Office, 1974.

11

3

ingly insatiable demand for more manpower raninto the barrier of a limited supply. The shortagepersisted despite a tremendous increase inhealth personnel. From 1960 to 1970, while U.S.population increased 18 percent; the number ofdoctors grew 21 perce:it; registered nurses, 33percent; radiologic technicians Fnd technolo-gists, 32 percent; nurses' aides, orderlies andattendants, 78 percent; and dental hygienists, 173percent.3

By the early 1970's there were signs that man-power needs were beginning to stabilize andeven to decline. Although shortages continue toexist in some areas, there are areas andoccupations with a surplus of manpower, andsome health graduates have experienced diffi-culty finding jobs. The Committee on EconomicDevelopment in its 1973 proposal for a nationalhealth care system pointed out that:

Poor distribution, together with inadequateutilization, training, and organization, haveaggravated the shortages of manpower insome areas while causing surpluses inothers. Beyond some crude and increasinglydoubtful ratios of professionals to popula-tion, it is not even known how many peopleare now needed, let alone how many wouldbe needed under a better-organized system.'

Different diarnoses call for different remedies.If (as is apparently true) the manpower problemswe face today are closely tied to maldistributionand malutilization, we must try prescriptions tocounteract these difficulties that differ fromthe treatment we would use to ovei come generalshortages. The need for precise, accurate andtimely manpower data to assess correctly thesituation has never been more pressing, and the

3. Goldstein, Harold M. and Horowitz, Morris A. Rese 17 andDevelopment in the Utilization of Medical Manpower. Boston:Center for Medical Manpower Studies, Northeastern Univer.sity, 1974.

4. Committee on Economic Development, "Building a na-tional health care system," April 1973, p. 33, quoted in Gold.stein and Horowitz, op. cit.

12

4

Importance of methodologies for .nakIng reliableprojections to plan properly for the future hasnever been more .t!te.

The data base t'ir aduquate health manpowerplanning does ric.t c. t and needs to be built.Some of the put.' d data on supply andrequirements are bs,oi described as "guessti-mates" and are of questionable value. The Inad-equacies of current statistics are compounded byshortcomings of data related to region and local-ity, occupation and specialty, length of requirededucation or training, type of employment situa-tion, demographic characteristics, salary, and soon.

In addition to the problems generated by an in-sufficiency of data, difficulties arise In the selec-tion of an appropriate methodology for determin-ing present estimates and future projections. TheIssue being addressed and the underlying con-cepts dictate both the data requirements andestimating methodology. How an estimate Ismade depends on the available data, and theavailable data determine the choice of method-ology. No methodological approach Is universallyapplicable; each planner must weigh the alter-natives and select the methodology or mix ofmethodologies appropriate for hls purposes andsuitable for hls data.

A planner's choice Is not made In a vacuum,but in the particular political and organizationalsetting In which he operates. The environment ofeach planner Is sul gonerls. In each locale, how-ever, disparate forces, varying from health pro-viders and educational Institutions, professionalassociations and labor unions to courts and In-surance companies, are present and Impact onthe production, distribution and utilization ofhealth manpower. The planner's goal In preparingestimates of supply and requirements Is tomake a rational, effective contribution to policydecisions. For hls findings to Influence policy,the methodology he uses should be comprehen-sIble and persuasive to policy-makers In hls area.

1.3

However we approach the methodology fordetermining "health manpower supply and re-quirements," we must begin with precise defini-tions of every term in the phrase, since we mustknow exactly what we are proposing to measure,as well as how we propose to measure it

"Health manpower" in its most br^ad definitionapplies to more than six million persons in thehealth care industries, engaged in any capacity inactivities devoted to providing health care: em-ployees and self-employed; the public and privatesectors; those engaged in patient care, in ancillaryhealth services, and in administrative and sup-portive activities. Although such a broad defini-tion is useful in measuring and planning the totalimpact of the health care sector in the allocationof national resources, many occupations are in-cluded in it which are beyond the concern ofhealth resources planning (for example, in main-tenance, laundry, food preparation, computeroperation). On tho other hand, other definitionsmay be too narrow for our purpose; they may in-clude only the private sector, to the exclusion ofthe public; or only employees, to the exclusion ofthe selfemployed. Thus, in using any statistics of"health manpower," it is necessary to take carefulnote of the definitions and coverage, down to thofine print and the footnotes, since those willgreatly affect the results of the analysis.

1. 4

II. Definitionof Terms

Definitions AffectEstimates

5

Information needs are defined by the planningpurposes to be served; and for the purposes ofhealth resources planning, the most useful infor-mation is that concerning occupations or clustersof related occupations, of persons qualified toperform professionaI or technical services in pro-viding health care directly or indirectly to indi-viduals or groups of people. The occupation is thecommon unit for organization, delivery of ser-viues, training, credentialing, employment, com-pensation, etc., and therefore an appropriate unitfor manpower planning.

We must also be precise in defiffing "occupa-tions" or "occupation clusters." "Occupations"are typically defined by a set of functions, re-quiring a specific combination of knowledge andabilities, and are usually associated with a specif-ic title (for example, "registered nurse" or "respi-ratory therapist"). Title alone is no a sufficientidentification of an occupation, since the sameset of functions may be called by different titles(e.g., "radiologic technician" or "X-ray techni-cian"); and the same title may be applied toseveral distinct occupations (e.g., "nurses" maybe registered nurses, licensed practical nurses, ornonlicensed practical nurses). Thus, unless occu-pations are defined specifically, estimates of re-quirements and supply are misleading.

Sometimes the focus of the manpower planningis on an alleged or prospective "shortage," andhere again we must take care that the term is pre-cisely defined. "Shortage," of course, exists when"supply" is less than "demand," but there havebeen many instances of reported "shortages"coexisting with unemployment in the same occu-pation either because the supply was narrowlydefined to exclude qualified persons who mightbe available, or because the demand was broadlydefined to include requirements for health man-power In excess of the means to pay personnel tofill them. Thus, the definitions of manpower "re-quirements," "demand" and "supply" are centralto manpower planning.

15

As used here, the term "manpower require-ments" is defined as the manpower necessary toprovide health services to a. population. Themagnitude and composition of the requirementsare specific not only to the population in question,but also to the level, intensity and quality of thehealth services, and to the delivery system and themodes of manpower utilization prevailing orassumed to prevail at a given time and place.Therefore, manpower requirements are derivedfrom these characteristics of the health caresystem. This does not mean that we must alwaysestimate the use of health care services before wecan estimate manpower requirements; as we shallsee, short cuts may ue used at times to estimatemanpower needs directly. In view, however, of thederived nature of manpower requirements, wemust explicitly take into account any'hing thataffects the levol and of the services to beprovided (e.g., health insurance), since it will alsoaffect manpower requirements.

"Demand" is an economic term. It refers to thequantities of goods or sec/0es which users arewilling to buy at various prices. Demand for man-power (labor) refers to the amount of manpower ofspecific kinds (in specific occupations) whichconsumers or employers will use (employ) at vari-ous rates of compensation (wages, salaries, orfees). In general, the lower the price, the greaterthe demand and vice versa. Thus, the demand formanpower is expressed in the employment (orself-employment) for pay of personnel who pro-duce goods or provide services which people wishto buy at specific prices. Demand for health man-power is derived from the demand for health careservices and from consumers' willingness andability to pay for those services; it is expressed inthe employment and compensation of personnelto provide the services. This is what economistscall "effective demand" the use of money topurchase goods and services to satisfy needs or

1 6

Concepts ofRequirementsfor ManpowerDerivation ofManpower Requirements

Conc,otn nf Demand andDemand Elasticity

7

8

wants. The money, and the willingness to use itfor these purposes, make demand "effective."

Needs can be converted to effective demand inmany ways. When people have more moneyincome, they spend more on health services. '

(This is what we mean when we say that "thedemand for health care is income-elastic.") Simi-larly, when the price paid directly by individualconsumers for health services declines (for

xample, through health insurance), people at anyn income level tend to use more health ser-3 than they would at higher prices; and the

total expenditures for services are greater. 2

Finally, needs can be converted to demand by theintervention of a third party (government or philan-thropy) to pay (in whole or in part) the provider ofthe needed services on behalf of the consumer.

For ttie past 25 years, and particularly for theE.F1 decade, all of these influences have been at

vr.)1(.. rising family incomes; the spread of health,...ance, including Medicare; and the various

punc subsidies for health care to individuals,such as Medicaid, and for health care facilities,principally hospitals. The result has been an ex-plosive increase in expenditures for health careand a narrowing of the gap between needs andeffective demand. These trends must be explicitlytaken into account in projections of demand forhealth care services and the derived demand forhealth manpower.

1 Resell, R. and Huang, Lien-fu. The &loot of Health Insur.ance on the Demand tor Medical Care. Rochester, N.Y.: Uni-versity of Rochester, 1971. See, for example, table 2.

2. Ibid. Also, U.S. Department of Health, Education, andWelfare, Social Security Administration. A Report on theResults of the Study ol Methods of Reimbursement for Physi.cians Services Undor Medicare. SS Pub. No. 92-73 (10-73).Washington, D.C.: U.S. Department of Health, Education, andWelfare, 1973, Vol. II, part II, table 21. Note that the increase Indemand for Insured services occurs when consumers Inavidually pay less directly, even though they are paying more Inthe rising costs of the insurance.

17

9

When we speak of "need" we refer to some Concepts of Need

level of health services that ought to be consumedduring a period of time in order to attain a desiredhealth status, determined according to some pro-fessional standard in the light of existing medicalknowledge as well as the health manpower to pro-vide these services. Since need, both for servicesand manpower, is determined without regard toprice and effective demand, its use in health re-sources planning emphasizes the gaps in healthservices and facilities for which resources shouldbe sought in order to bring the health care for aspecified population to the standard consideredto be minimum or optimum. In this respect, theconcept of need is associated with an earlierstage of planning than the concept of effectivedemand, which presupposes facilities and ser-vices in being (or in prospect, with fundingassured), and is concerned with planning theiroperations, including their manpower resources.

Needs may be formulated in terms of observedgaps in available health services or in terms ofstandards of health care that are not being met. Ifsick people are not being adequately cared for, orif defects of eyes, ears or teeth among childrenare not being detected or corrected, obviouslythere are needs for health services and for person-nel to provide them. In these cases, there are bydefinition "shortages" of health services. Medicalprofessionals may determine needs on the basisof their judgments of what is required to provide asatisfactory level of health care, given the state-of-the-art. But just as hungry people or nutritionists'findings do not create market demand for fooduntil there is money to buy it, so poor health doesnot create a demand for health services (and thederived demand for health manpower) until thereis money for the operation of facilities and theemployment of personnel. As one study com-mented:

Perhaps more in health than in any other ser-vice Industry, "need" and "demand" areclouded and confused by interest groups'

18

10

Concepts of Supply andSupply Elasticity

biases. A given area might have a well-docu-mented need for, say, five physical thera-pists, but as long as the money is not avail-able to fund these positions, there Is nodemand.DEMAND. The number of Jobs that can befinanced with current or future funds.NEED. The number of persons in a field whowill be required to produce a given level oramount of service judged to be desirable.

The distinction is between social ideals(what people feel ought to be done) andeconomic realities (what people are able topay for). 3

The definitions of Supply of manpower in anoccupation are even more numerous than the defi-nitions of demand. "Supply" may be variouslydefined as:

1. The number of persons in the specifiedarea qualified to practice the occupation,whether actively practicing or not

2. The number available to practice the oc-cupation at a given time, including those em-ployed (or self-employed) and those unem-ployed (available for work but not employed)

3. The number of unemployed

4. The number of entrants the newly quali-fied practitioners entering the market, andthe previously qualified practitioners re-

3. Connecticut Institute for Health Manpower Resources,Not by Numbers Alone: Study of Educational Programs andEmployment Opportunities in Health in Connecticut and theNortheast. Hartford, Conn.: The Connecticut Institute forHealth Manpower Resources, January 1974. For further discussion of the Implications of the dif ferent concepts, see Dars,Lewis. An Analysis of the Concepts of Demand and Need forMedical Care and Their Implications for Manpower Planning,Health Manpower Planning Series, Monograph 1. Trenton, N.J.:Department of Higher Education, 1973. See also, Jeffers,James R., Bognanno, Mario F. and Bartlett, John C., "On thedemand versus need for medical services and the concept of'shortages'," American Journal of Public Health, 46.63,

January 1971,

1 9

11

entering the market after a period of non-availability5. The numbers who would be in the marketif the pay (or other form of compensation)were raised by specified amounts abovespecified prevailing levels.

All of these definitions are "correct," and eachhas its uses in connection with a specific plan-ning operation and time frame. The importantpoints are that the definition chosen should beexplicitly suited to the operational situation towhich it is applied, and that it should be compati-ble with the definition of demand used in thesame situation. In addition, because part-timework is an accepted practice in many healthfields, the number of persons referred to in thedefinition should relate to full-time equivalentpersonnel.

The last definition (item 5 above), which de-scribes the relationship between supply and com-pensation, is the most rigorous economic con-cept of supply and the most difficult to measureand apply. It recognizes that supply is not a fixedquantity or a static element in the labor market butresponds to economic (and other) incentives,more or less depending on the occupation, thetraining facilities and training time, and thecharacteristics of the potential supply. The moreresponsive the supply to the various incentives,the greater the "supply elasticity" is said to be.For example, a substantial increase in salariesoffered to nurses may produce an almost imme-diate inflow of nurses already traine6 but inactive;an increase in compensation to physicians, how-ever, will have a delayed effect In inducing physi-cians to move from one area to another, or an evenlonger delayed effect (as long as a decade) whilemedical education institutions are expanded totrain additional physicians. The supply of nursesIs much more elastic than the supply of physi-cians, and the supplies of allied health profes-sionals requiring one or two years of training aremore elastic still.

2 0

The use of other definitions of sunply dependson the planning problem at hand. If the opening ofa new hospital is being considered, it is necessaryto know whether the unemployed reserve plus thenew entrants and re-entrants in the various oc-cupations will be sufficient to fill the budgetedslots. If the choice is between establishing a train-ing program for allied health occupations orundertaking a sustained recruitment campaign, itis necessary to measure the potential supply ofqualified persons and the incentives which wouldbe required to activate them.

Thus, the relevant concept of supply in large

part determines the choice of methodology.Whether information is sought from employers,from labor force statistics, from licensing or reg-istry rosters, or from education and training insti-tutions depends on the aspect and definition ofsupply that is relevant to the planning problems.

Concept A geographic labor market may be definedof the Relevant simply as "the area within which workers compete

Labor Market for jobs and employers compete for workers." 4

Typically, labor markets are local, encompassingan area in which workers can change jobs withouthaving to change residence. However, Americansare traditionally mobile, and moving from place toplace In search of job opportunities Is common;also, for some occupations, labor markets may be

regional or national. Especially in the professions,young men and women often move about as theyobtain their professional education and training;they are, therefore, at least at the time they com-plete their training, relatively unsettled and free todecide where to settle for the most promisingopportunities. Nevertheless, for manpower plan-ning purposes, occupational labor markets are

4. U.S. Department of Labor, Bureau of Labor Statistics,Occupational Supply: Concepts and Sources of Data for Manpower Analysis, BLS Bulletin 1816 (Washington, D.C.: U.S.Government Printing Office, 1974),

2 1

13

essentially local, limited by the distance overwhich commuting is possible.

The larger and denser the population of thelocal area, the more likely it is that there will bedemand, supply, and an active market for most ofthe health care occupations. This is particularlytrue of the 265 metropolitan areas in which nearlythreefourths of Americans live and work. The sizeof these areas generates effective demand formost kinds of health care services; this demand inturn offers opportunities for livelihood for a widevariety of generalists and specialists in medical,dental, nursing and allied health occupations.:::onsequently, metropolitan areas have attracteda more than proportionate number of health pro-fessionals. Thus, in most cases, metropolitanareas can be considered relevant labor marketareas for most purposes of health manpower plan-ling. Moreover, many of the kinds of statisticsuseful in planning are available for "standardmetropolitan statistical areas."

Since the relevant labor market encompassesthe area from which resources are drawn, thegeographic boundaries for non-metropolitan labormarket areas are set by the search for qualifiedpersonnel or by the location of educational insti-tutions that train them. Non-metropolitan areasmay not, in fact, contain supplies of health carepersonnel sufficient even for their limited healthfacilities. Consequently, the salaries offered mayhave to be adjusted upward, depending on theelasticity of supply, to attract manpower intothese areas.

The National Health Planning and Resources Concept ofDevelopment Act of 1974 (Pl. 93-641) requires Appropriate Healthestablishment of health service areas, which it Service Areadefines as:

a geographic region appropriate for theeffective planning and development of healthservices, determined on the basis of factorsincluding population and the availability ofresources to provide all necessary health ser-

2 2

14

vices for residents of the area .... To theextent practicable, the area shall include atleast one center for the provision of highlyspecialized health services. 5

The Act provides further that the area (except inunusual circumstances) shall have a populationbetween 500,000 and thrue million, with empha-sis on SMSA's but with due regard to the differ-ences in health planning and services betweenmetropolitan and non-metropolitan areas. Theserequirements of law explicitly recognize the na,ture of health service areas and their relation tolabor market areas and economic areas generally.Like the health services from which it is derived,the demand for labor tends to be more diversifiedqnd more specialized as we progress from thesparsely populated areas to the most populousand richest communities, while by the same tokenthe labor supplies in the larger populations aremore plentiful, diverse, and specialized, and thetraining facilities more accessible. Labor marketareas and health service areas will tend to becompatible as sources of information and as unitsof analysis for health manpower planning.

Importance of the Projections are measurably influenced by theAssumptions in underlying assumptions. The validity of the future

Planning Outcomes estimates depends upon the extent to which theirunderlying assumptions realistically anticipatethe ch _riges that the future will bring.

Different estimatec follow from different con-cepts of what is being measured and from differ-ent ideas of what changes and rate of change willoccur in the future. With regard to the supply ofhealth occupations, for example, governmentsupport for academic institutions, especiallymedical and allied health programs, in the yearsahead is vitally important. The assumptions madeabout the level of government support are key

5. U.S., National Health Planning and Resources Develop.merit Act of 1974 (P.L, 93-641), Sec. 1511 (a).

23

15

determinants of the size of the graduating classand thus of the estimate of future increments ofsupply. For example, the Federal Government'sprojections of health manpower supply arepremised on:

... one critical set of assumptions ... thatthe health manpower production system willreceive the future support from Federal orother sources that is needed for the main-tenance and operation of the system at itsassumed 1974-75 capacity level. Implicit inthe assumption is that the effect on thehealth manpower production system of thelate-1972 levels of Federal biologicalresearch and support programs will notchange significantly in the years ahead, andthat the effects of military and Public HealthService recruitments and needs will remainroughly as they are today.

With regard to requirements, the key detern.1-nants are the number of people requiring care andthe amount of money available to pay for it. As-sumptions about population characteristics, thegeneral economy, income levels, consumer ex-penditure patterns, and government policy regard-ing health care financing underlie the projectionsof the demand for health care. The manpower re-quired to provide the projected level of care isconditioned by the form of the delivery systems,the state of technology, labor productivity and soon.

Since a predominant proportion of health work-ers are women, assumptions about labor forceparticipation rates and tha lifetime career activitypattern of women are important to the futuresupply and requirement estimates.

As one moves further in time from the present,one's assumptions are inevitably subject to moreuncertainty. For this reason, It is advisable to

6. U.S. Department of Health, Education, and Welfare, Bureauof Health Resources and Development, The Supply of HealthManpower: 1970 Profiles and Projections to 1990, DHEW Pub.No. (HRA) 75-38. Washington, D.C.: U.S. Government PrintingOf fice, 1974.

2 4

6

prepare more than one projectiun based on differ-ent sets of assumptions, ranging from the mostconservative to the most conjectural amongplausible alternatives. The use of different sets ofassumptions provides test of the sensitivity ofthe estimates. If a -,o in assumptions doesnot affect the final estin c.,;e substantially, one canhave more confidonce in the projection.

2 5

The demand for health manpower is determinedby (1) the quantity and quality of health servicesdemanded, which are conditioned by such factorsas the pupulation served, prices and incomes, andthe supply of health services available; and (2) theproductivity (output of services per person) ofhealth manpower, which is specific to the state cfmedical technology and to the health servicesdelivery system. Demand for manpower is thusinfluenced by many factors.

The size, the growth, and the characteristics ofthe population are basic determinants of demandfor health services and therefore of health man-power. Health resources planning must takeaccount not only of the numbers of people and the,ates of increase or decrease, but of the birth anddeath rates, the proportions of the very youngand the very old, the ethnic composition (sincesome diseases are endemic to certain ethnicgroups), and the urban-rural composition. Forexample:

The health industry was one of the first to beaffected by the decline in births . . . . Thisdecline undoubtedly permitted a shift of hos-pital paramedical staff to other urgent and

26

17

1

Ill. FactorsAffecting Demandand Supply forHealth Manpower

Determinantsof Demand

The PopulationServed

8

Economic Factors

growing needs and may have already influ-enced some doctors and nurses to choosespecialties other than obstetrics or pediat-rics .... Assuming maintenance of the two-child family average and the continued valid-ity of the doctor-patient ratio currently esti-mated by the Department of Health, Educa-tion, and Welfare, the need for physicians inthe year 2000 will be fewer than the numberneeded to service a population increasing atthe higher growth rate (the three-child norm).Demand for other health personnel will besimilarly affected.'

Demand is strongly influenced by consumers'incomes (expenditures for health services risefaster than income rises) and by the orices ofhealth services (the lower the price, the greaterthe utilization). The income elasticity and priceelasticity of demand have been the principal fac-tors in the sharp increase in demand over thepast decade, as family incomes have risen andhealth insurance has lowered the effective pricesof health care services, principally those of hos-pitals and physicians. Estimates and projectionsof demand for health care and for the manpowerto provide it must take explicit account of the in-comes of the population for whom the servicesare intended and of the Insurance coverage of theservices among that pcpulation: for example,families with young children are likely to havemore income and !:.'ss insurance coverage thanthe elderly; in 1ustr;a1 workers are likely to havehigher incomes and more insurance than farmworkers; etc.

Productivity Given the demand for health care services on agiven scale, the question arises: How manypeople in what occupations are required to fill thedemand? The answer obviously depends on the

1. U.S. Department of Labor, Manpower Administra-tion Manpower Report of the President, 1973. Washing-ton, D.C.: U.S. Government Printing Office, p. 78, 1973.

27

19

output per person or productivity, as it iscalled. 2 For our purposes here, output is definedas units of service (visits to the doctor, days ofroutine hospital care, surgical procedures, dentalprocedures, laboratory tests, X-rays, etc.). Num-bers of personnel (full-time equivalent) of variouskinds required to perform specified numbers ofthese service "outputs" are the "inputs" whichdetermine measures of productivity: X-rays perday (or month, or year) of radiologic techniciantime; "well-baby" clinic visits per pediatrician-day; patient-day of routine nursing service per RN-day; etc. The variety and complexity of health careservices are such that it is often difficult andsometimes impossible to identify homogenousclasses of service outputs with defined occupa-tional personnel inputs. Yet it is unmistakable thatthere are significant differences among healthcare facilities depending on their modes of de-livery (e.g., inpatient vs. outpatient), types of ser-vice (e.g., acute care hospitals vs. skilled nursinghomes), standards of service, etc., and that thesedifferences are valid and useful in planning healthmanpower resources. Similarly, among facilitiesof the same types, productivity differs between

2. The problem of measuring productivity in health servicesis complicated by theoretical ambiguity of concept and defini-tion. If the "output" is maintenance of a given state of health ofa population, then health services are "inputs." So far, satis-factory measures of "a state of health" have not been de-veloped. On the other hand, "outputs" can be defined andmeasured as the services performed by health care personnel,and the numbers and mix of personnel are "inputs." The latterdefinition is used here because it is compatible with other ele-ments of the system for planning the development and alloca-tion of health resources.

Still another concept is "... a synthesis of the two alreadydiscussed. This is a definition in terms of episodes of illnessappropriately cared for ... using the hospital 'case rather thanpatient-day as the unit of output ... relating service to somehealth objective ...." ("Issues in the measurement of pro-ductivity," (in) Donabedian, Avedis. Aspects of Medical CareAdministration: Specifying Requirements for Health Care.Cambridge, Maw: : Harvard University Press, 1973.)

2 8

20

the more as compared with the less effi-ciently organized and managed.

A line of research in "task analysis" in recentyears has begun to prlvide systematic informa-tion on the tasks which are the basic component"building blocks" of health care services. 3 Thisinformation points the way to the organization oftasks and personnel for the most efficient assign-ment of work and the delegation of tasks ac-cording to the knowledge, skill and training theyrequire so that highly skilled personnel are notused for tasks that can be satisfactorily performedby those less highly trained. In this way progress

can be made toward defining standards of pro-ductivity; toward utilizing most efficiently scarce,skilled personnel; toward achieving better service;and toward effecting lower costs.

The State-of-the-Art The technology of medicine and dentistryof Health Care affects both the demand for health care services

and the manpower required to provide them.Obviously, technology determines what services

can be offered: the development of diagnostic andtherapeutic procedures creates a demand forthem. One has only to reflect on the limited scopeof medicine and dentistry at the beginning of the20th century to realize how scientific discoveriesand technologies have enlarged demand. Theprocess is apparently without end. Technologyinfluences the tasks that manpower performs andthe skills that must be acquired.

The advance of technology also increases theefficiency and productivity of health care person-nel and permits services to be provided by fewer,and often more specialized, personnel. For

3. For an overview of this approach, see U.S. Department of

Labor, Manpower Administration, Job Analysis tor HumanResource Management: A Review of Selected Research and

Development, Manpower Research Monograph No. 36. Pre-

pared by Michael Wilson. Washington, D.C.: U.S. Government

Printing Office, 1974; especially pp. 18-19, 26 and 29-34, 39,

and appendices A and B.

29

21

example, while biochemistry has greatly in-creased the types of laboratory tests and theiruses in diagnosis, the development of suchsophisticated equipment as the autoanalyzer hasmade it possible for a technician to perform manytests on a blood sample more speedily and at rel-atively lower cost than formerly. Thus technologyon the one hand increases the demand for healthservices and the requirements for health man-power, and on the other hand, raises productivityand thereby reduces the number of persons re-quired. Technological change, by altering thedemand for services, the tasks performed and theoutput per worker, creates or eliminates healthjobs, increasing or decreasing manpower require-ments.

The system for the delivery of health care ser- The Characteristics ofvices is composed of the facilities, the personnel, the Delivery System

and their relationships to patients, in the organiza-tion, administration and financing of health careservices. Delivery of physicians' services on afee-for-service basis through the offices of indi-vidual physicians differs in important ways fromdelivery through the outpatient department of ahospital or the clinic of a prepaid group or otherHMO, and these differences influence not onlythe demand for services of various kinds but thenumber and occupational mix of the personnelproviding them.4 Shifts in the utilization of ser-vices between, for example, inpatient and out-patient care, or between treatment and preventionof illness, produce corresponding shifts in de-mand for health service personnel. In fact, the

4. For examples of ef fects of various national health insur-ance plans on manpower demands, see U.S. Department ofHealth, Education and Welfare, Assessment and Evaluation ofthe Impact of Archetypal National Health Insurance Plans onU.S. Health Manpower Requirements. Prepared by Huang,Lien-fu and Shomo, Elwood (Robert R. Nathan Associates).DHEW Pub. No. (HRA) 75-1. Rockville, Md.: U.S. Department ofHealth, Education, and Welfare, 1974.

3 0

22

explosion of demand and cost since 1965, and theprospects for further extension of health insur-ance have stimulated variations in delivery sys-tems in the search for cost-containment mea-sures. Consequently, health resources planningshould take account explicitly of the changes inthe delivery sy: tem impending or prospective atthe time the planning is done.

The Supply of Health care is one of the sectors of the econ-Health Services omy in which supply is an important determinant

of demand. Effective demand presupposes asupply of facilities and personnel; demand can beeffective only if there are services to be purchasedand providers from whom to purchase them. More-over, the availability of health services as the con-sequence of the discovery and application of newmedical and dental technology in itself generatesa demand, as we have seen again and again. Thisis particularly true with the proliferation of spe-cialized services. In the face of the consumer'simperfect knowledge, the physician or dentistbecomes the professional adviser or even thedecisionmaker: "Once the consumer has placedhimself in the physician's hands, the physicianplays a major role in deciding the amount of ser-vice consumed." 5

The influence of supply on demand is nowhereclearer than in health resources planning. When-ever services are limited by the unavailability offacilities or personnel (e.g., inadequate hospitalcapacity or lack of the full range of medical/dentalcapabilities), as may be the case in rural or remoteareas villages in Alaska are extreme examples

the demand is truncated accordingly. Planningthe expansion of the supply of services must takeaccount explicitly of the effective demand for ser-

5. Report of the National Advisory Commission on HealthManpower. Washington, D.C.: U.S. Government Printing Of-fice. Vol. II, 238, November 1967.

31

23

vices and personnel which will be activated by thesupply.

Although ail demand for health manpower is Relative Importancederived from demand for health services, the of Determinantsseveral determinants of demand affect different of Demandoccupations differently. Some occupations (e.g.,physicians and dentists) are linked primarily tosize and distribution of population; otrs (e.g.,pediatricians) are linked to birth rates. Thedemand for laboratory technicians, on the otherhand, is strongly influenced by the developmentof biochemical technology and equipment. Forpurposes of manpower resources planning, it isimportant to recognize the specific dominant in-fluences on the demand for manpower at differentpoints in the delivery system.

While the demands for "free-standing," auton-omous professionals like physicians anddentists may derive in the first instance fromexogenous variables of population and income,demands in other health care occupations arestrongly affected by the number of physicians anddentists and other aspects of the delivery system.The demand for nurses, for example, derives fromthe facilities in which nurses are employed, suchas physicians' offices, hospitals, clinics andnursing homes; and from such secondary determi-nants as school health programs and the numberof children in school, and nurse training programsand the requirements for nurse instructors. Thedemand for pharmacists, on the other hand, isdetermined by the number of prescriptionswritten

which itself is affected by the development ofdrug therapy (the drugs available for treating pa-tients and the physicians' propensity to use them)

and the productivity of pharmacists (largely aresult of the extent of pre-compounding in manu-facture). 6

6. Maryland Council for Higher Education, A Projection ofMaryland's Health Manpower Needs Through the 1980's. Balti-more, Md.: Maryland Council for Higher Education, 1969.

32

24

Medical, dental, nursing and allied healthoccupations seldom occur in isolation, unrelatedto one another. Depending on the relationship ofdemand and supply and on the delivery system,tasks may be reallocated or delegated betweenoccupations functionally related in the deliveryprocess though of different skill levels; e.g., MDand RN, or RN and LPN. Thus, the demand forRN's may be determined in part by a deficiency inthe supply of MD's. The reallocation of tasks maygo the other way, as well: an abundance of RN'scan limit the demand for LPN's.

In sum, there are no neat formulas for derivingdemand for health care occupations mechanisti-cally from the various determinants. Not onlyevery community, but every occupation as well,must be anaiyzed with intelligent regard to thepeculiar circumstances of the times, arising asthey do from determinants as diverse as birthrates, health insurance, facility construction pro-grams, or technological developments.

Determinants of Over the long range of time and place, supply ofSupply of health care manpower will be determined, like

Health Manpower supplies in other occupations, by underlying eco-nomic and institutional factors: rates of pay andearnings expectations in relation to time and costof training; working conditions, rewards of statusand other intangibles; opportunities for trainingand for employment; geographic and occupa-tional mobility; institutional incentives and con-straints that encourage, discourage, or circum-scribe occupational practice. Because of thepublic interest in health care, health service occu-pations are subject to these long-term influencesmore than most: training is relatively prolongedand exacting; opportunities are institutionallydetermined; credentiallny as a prerequisite toemployment is widespread; and occupational andclass distinctions are more persistent than Inmost fields. On the other hand, earnings in manyhealth service occupations are good or better thangood, prestige is high, and the prospect of a

33

career in the medical and healing arts is persis-tently attractive to young men and women.

In many medical, dental, nursing and allied Education and Training

health occupations the flow of qualified person- Programs

nel into the labor supply is governed by the train-ing system: the capacity of training facilities andprograms, the length of training and its costs, thenumber of applicants, the standards for admis-sion, the number of students in each class, thedropout rate during training, and the number ofgraduates. The health resources planner needs topay close attention to the admissions policiesand practices of local and regional health educa-tion and training institutions: How many areadmitted? What are the qualifications for admis-sion? What are attrition rates between admissionand graduation?

In the case of physicians, medical education isprolonged and costly, and the number of places inmedical schools is limited and requires years (anda great deal of money) to expand. Because medi-cal school facilities expanded so little for so manyyears, the supply of American-trained physicianshas been inadequate to national requirements,Nith the result that the U.S. supply has been (andis being) augmented by large numbers of foreign-3Orn, foreign-trained physicians. Thus, the addi-:ions to supply are governed by the capacity ofnedical schools and the admission of immigrant3hysicians. The supply of dentists is similarlygoverned, though the time and cost of training areess. But the shortage of dentists is masked byiental neglect, because effective demand (whatonsumers will pay for) falls far short of the dental

'weds of the population in the absence of insur-ince coverage.

The planner must assess the extent to whichthose who are graduates of local and regionaltraining institutions become part of the local andvgional supply. Nurses and allied health workers,

3 4

25

Locational and OccupationalDecisions of QualifiedPersonnel

26

for example, are more likely to remain in the areaIn which they are trained than are doctors. Ingeneral, the more highly trained the occupation,the wider the relevant labor market and the greaterthe geographic mobility among its practitioners.Physicians tend to move, sometimes over con-siderable distances, in search of opportunities topursue their specialties, to have access to well-equipped, well-staffed hospitals, and to enlargetheir earning possibilities. The markets for radio-logic and laboratory technicians, on the otherhand, tend to be more strictly local.

The movement of personnel at all professionallevels both in and out of the area reflects, in part,the fact that Americans are mobile people; in anygiven year about one American in 16 moves to adifferent state. In the case of health service pro-fessionals, the movement has been unbalancedas physicians, dentists, nurses, and other highlytrained personnel have gravitated to larger com-munities and especially to metropolitan areas,drawn by opportunities for more specialized prac-tice, higher earnings, and an atmosphere con-ducive to professional development. The obverseis that many smaller communities and rural areasare undermanned. This s another example of theperverse economics of the health services sector:the concentration of health manpower supplies inmetropolitan areas has not restrained the pricesof health services, nor has the paucity of supply Insmall communities served to raise them.

The number of graduates of local and regionaltraining programs, therefore, is not synonymouswith the numbev available to the local or regionalsupply. Not only do graduates leave the area (toreturn to their horres elsewhere or to seek morefavorable opportunities), but some even leave theoccupation, perhaps because they find they canearn more in some other field. In any event, toestimate local supply it is necessary to discountthe number of graduates on the basis of pastexperience, especially in the occupations withrelatively short training where the personal invest-ment in time and money is not so great.

35

27

Although the output of education and training Labor Force Partici-

institutions is the principal determinant of the P ation Rates

supply of new practitioners in health serviceoccupations, much the largest part of the supplyat any time is composed of previously trained per-sonnel, either those employed, unemployed, orout of the active labor market for personal rea-sons. Because rates of participation by women inthe active labor force 'vary from time to time ac-cording to their personal circumstances and pref-erences, "potential supply" is likely to be signifi-cant in occupations heavily populated by women.For example, women 18 to 25 and women 35 to 55years old have higher labor force participationrates than those 25 to 35. Women with pre-schoolchildren are only two-thirds as likely to be in thelabor force as those with children of school age.Women are more than five times as likely as mento be working in part-time jobs. These participa-tion rates all reflect ways in which women adaptto the competing demands on them which arisefrom their several roles. They are particularly ap-plicable to health service occupations becausewomen are so prominent a part of the labor forcein these occupations, especially in nursing,which is almost entirely staffed by women. It isnot unusual for young women to work until thebirth of the first child and to leave the labor forceor to work part-time for a few years, returning tofull-time employment in their middle or latethirties. These withdrawals and re-entries mustbe reckoned in estimating supply. They reflect apattern of lifetime career 'activity that is under-going dhange as women's role in society is being

redefined.

For many occupations some form of credential- Licensing and CertifIcition

ing Is practiced either licensing by the state asa prerequisite to practice or certification by a pro-fessional organization attesting to the completionof accrectited education or training necessary toestablish the qualifications of individuals. Cre-dentialing requirements may constrain the supply

36

28

Rates of Pay and Earn-ings Expectation

Rate of Deaths andRetirements

of manpower in any locale by limiting the numbereligible for employment and by preventing moreefficient utilization of manpower. A significantsource of future supply may develop from the useof proficiency examinations as a basis for creden-tialing, as mandated by the 1972 Social SecurityAmendments, as an alternative to the completionof prescribed, accredited education.

Because physicians and dentists are among thebest paid occupations, qualified applicantsexceed the limited number of places in profes-sional schools, despite the long, difficult andexpensive training. The high returns have at-tracted relatively large numbers of foreign medi-cal graduates, who (as noted above) in recentyears have numbered nearly as many annually asU.S. graduates. There is no doubt that many moreAmerican men and women will enter the medicalprofessions as places are made for them.

There is a wide gap in earnings between physi-cians and dentists and all other health serviceoccupations. Nurses in particular have beenunderpaid relative to comparable professionals inother fields, and this fact has been reflected inshortages from time to time. The increased de-mand for hospital nurses resulting from hospitalinsurance brought about improvements in .pay andwork conditions for nurses and an increasedsupply of nursing students and graduates. In addi-tion to long-run responses of this kind, it is oftenpossible in the short run to draw nurses from thepotential supply by offering higher pay as well ashours of work compatible with their preferences.

Death and/or permanent retirement of part ofthe labor force each year are factors which alsolower supply. It Is estimated that on the average,death and retirement annually remove about twopercent of the male labor force and about fivepercent of the female labor force. Estimates ofannual rates of separation for all causes by

37

29

individual occupation for each state arepublished by the U.S. Department of Labor andshould be explicitly used in projecting netchanges in the supply.'

The relative importance of the several determi- Relative Importance ofnants of supply is specific to the various occupa- Determinants of Supplytions, the sources of supply, and the means ofgenerating supply. Among health service occupa-tions requiring significant training, the capacity ofthe education/training facilities is the overall long-range constraint. The supply of applicants isample and in most occupations exceeds thenumber of places in the schools. The capacity ofthe schools is not a static constraint, since placescan be made for more students; but the lags be-tween new capacity and output of health profes-sionals may range from several to 11 years (thelatter in the case of physicians).

For the supply of physicians the rate of immi-gration of foreign-trained doctors can be almostas important as the capacity of U.S. medicalschools, as well as the only possible short-runadjustment to supply, especially for residentstaffs of hospitals.

For RN's also, immigration has been a signifi-cant factor when there was an apparent shortageof nurses. The expansion of nurse training capa-city, however, has lessened the dependence onforeign-trained nurses. Among nurses, neverthe-less, the qualified but inactive segment provides asource of potential supply, part of which may beresponsive to recruitment offering terms of em-ployment, pay and working conditions calculatedto attract nurses back into active status.°

7. See the detailed description of the use of labor force sepa .ration rates for estimating outflows due to death and retire-ment in Part II of this monograph.

8. A major metropolitan hospital recently disclosed that it.was having trouble recruiting staff nurses because it did notknow that it was paying below the market.

38

30

Since health services are essentially local,questions of supply of personnel (or services) willordinarily focus on the local (or regional) labormarket and on the factors which determine thesupplies which are (or will be) available to thatmarket. In major metropolitan areas, supplies inmost health service occupations are likely to bedetermined locally for the most part by the out-puts of local training institutions, adjusted by theflows of persons to areas of greater opportunity .and away from areas of less opportunity. Theseflows are not automatic, however, and manpowerplanning may be itself a determinant of supplyinsofar as it stimulates recruitment efforts. Lesspopulous areas, on the other hand, may encountershortages even when national supplies are ade-quate and may have to compete vigorously withmetropolitan centers for the output of trainees.

For physicians and dentists, both U.S. andforeign trained, the relevant labor market is wider.Immigrants are often recruited for resident staffby the less well-known hospitals, and if theyare able to meet licensing requirements andremain in the United States may be moremobile and available to smaller communities.Those trained n the United States have a choiceof practicing in their home communities, in theplaces where they irit&n, or in markets whichoffer attractive prospects of earnings, specializa-tion and professional opportunities. Even in largecommunities it is difficult to anticipate whichdeterminants of supply will dominate; in smallercommunities it is much more difficult.

'The Interaction of We have noted earlier the tendency in healthDemand and Supply services for supply to generate demand, for con-

sumers to avail themselves of services that ap-pear likely to be conducive to health and comfort.This is most clearly illustrated by the resort tonew modes of surgery and chemotherapy, whichfind a ready or even avio market when they be-come available. From this we may conclude thatin a relatively affluent society with a large com-ponent of income, both private and public, to be

39

31

used for discretionary spending (that is, for con-sumption other than food and shelter), there islatent demand for more health care than is avail-able, and that the appearance of a doctor or ahospital or a mode of treatment where there wasnone before activates this demand. Technical andbiochemical advances a new drug or a newpiece of equipment thus change the demandboth for health care and for the numers andqualifications of the manpower to deliver it. Thechanges in the manpower demands, in turn, affectthe institutions and programs for education andtraining.

There is also an interaction of demand and sup-ply in health services manpower. A half centuryago, when the effective demand for health carewas limited by the relaUvely low level of personalincomes, the supply of physicians was greaterthan the demand could employ efficiently, withthe result that physicians were underutilized andperformed many health service tasks now per-formed by allied health personnel. Under condi-tions of shortage, on the other hand, tasks aredelegated from scarcer to more plentiful classesof manpower, e.g., from physicians to nurses. Inthe same way, an inadequate supply of RN's stim-ulates a demand for practical nurses.

Though they may originate as adaptations toimmediate demand-supply situations, thesechanges, once they are initiated, may becomeembedded in conditions of productivity, pay andoccupatioral scope. The use of heaith service per-sonnel at their highest skills increases their pro-ductivity and lays an economic basis for increasesin pay. As these conditions become more wide-spread, they are reflected in tables of organizationin hospitals and in the programs of training insti-tutions. The continuous chain of interactions be-tween demand and supply should be explicitlyrecogrind, In manpower resources planning, andits effects should be anticipated in demand andsupply projections.

4 0

32

IV. MethodologicalApproaches to

EstimatingManpower

Requirements

Introduction Alternative methodologies for estimating re-quirerrents are available to the local planner, butno one method is universally applicable and errorfree. Experience has shown that differentmethods produce different estimates. A 1970investigation of six earlier projections of 1975physician requirements based on differentmethodologies found that the 6stimates varysignificantly, ranging from 305,000 to 425,000, andthat the findings led to opposing conclusions con-cerning the adequacy of the projected supply ofphysicians.'

All methodologies are tools in the hands of theresearcher, and the quality of the product is asmuch, if not more, a reflection of the skill of thecraftsman as of the excellence of the instrument.At every step in preparing an estimate, judgmentand assessment are necessary, and no mechanis-tic technique is a substitute for insight.

In this chapter, we will present the alternativemethodological approaches to estimating require-ments In a general, conceptual framework. A de-tailed, stepby-step, how-to-doit description andan in-depth analysis of the pros and cons of each

1, Hanson, Leo W. "An appraisal of physician manpoworproloct Ions," Inquiry, March 1970,

4 1

33

method are given in Chapter III of Volume II.Most studies are not pure examples of any onemethodology but a mix or blend, adapted to theparticular circumstances of the area planner. Inthe social, economic, and political setting inwhich he operates, the planner pinpoints thespecific problem he is addressing and in the lightof the resources available to him selects themethodological approach or mix of methodolo-gies best suited to his needs. His alternatives areapproaches based on the manpower/populationratio, the service targets, health need, and eco-nomic (effective) demand. 2

By far the most popular and frequently used The Manpower/methodological approach relates manpower to Population Ratiopopulation. 3 A ratio of the number of health per- Methodsonnel to the total population served is selected.To calculate manpower requirements, the ratio isapplied to the target year population.

Manpower Estimated manpowerX target population =

Population requirements

The numerator in the ratio, manpower, may referto discrete occupations (e.g., radiological tech-nologist) or to generic categories (e.g., alliedhealth occupations). The term may be limited tohealth personnel providing service in a particularsetting (e.g., nursing homes) or to a particular typeof care (e.g., pediatric), or it may be all-inclusive,encompassing the totality of workers in the healthindustry.

The denominator of the ratio, population, maybe defined in different terms depending upon the

2. The typology of methodological approaches was adoptedfrom Hall, Thomas L. "Estimating requirements and supply:Where do we stand?" (in) the Pan American Conference onHealth Planning, 10.17 September 1973. Washington, D.C.: PanAmerican Health Organization, Scientific Publication No. 279,1974.

3. Some researchers invert the ratio, relating population tomanpower, e.g., 3,500 population per x.ray technologist.

4 2

34

planner's concerns: total community population ifthe problem deals with environmental health;citizens 60 years of age and older if the concern isto staff nursing homes and treat chronic illness;the residents of a geographically defined servicearea if the issue is adequate ambulatory care.

The kingpin of this method is the ratio and itsappropriateness. The validity of the estimatedepends upon the fit of the ratio. Experts are fre-quently asked to choose the ratio on the basis ofrequirements for opVmal care or for someminimum standard. Alternatively, planners mayadopt the ratio found to prevail in the nation as awhole, or in their region or in a comparable geo-graphic area.

The attributes of the manpower-populatjon ratiomethod its simple data requirements, low cost,ease of understanding and application explainits appeal and popularity. However, one should beaware of its basic limitations. To assume thatpopulation size explains manpower requirementsignores important influences that do not operatethrough population size. Such variables as popu-lation density, the organization of the deliverysystem, the money available to pay for health ser-vices, and the productivity of manpower areexamples of factors that are overlooked in theratio method. In using the ratio method, the plan-ner assumes, explicitly or implicitly in the choiceof the ratio, that these variables operate in hissituation in the same way as in the situation fromwhich the ratio is selected, or that he can makeexplicit allowances for differences. Despite thesecaveats, it should be pointed out that other, moresophisticated methodologies may use the man-power-population ratio as an Input.

The Service Targets Recognizing that the relevant measure of man-Approach power requirements is not tho number of people

but the volume of service, the service targets

4. The Delphi method may be used to achieve a consensus. ItIs described In Chapter I of Volume II.

43

35

approach focuses on service demands and takesinto account the elements of manpower utilizationand the organization of the delivery system.

The first step in applying the method is to setthe target of the types and quantity of service re-quired by the population in the local area. Theconstraints of the particular program and of thehealth delivery system afti taken into account inselecting the service goals. For example, let ustake the Early and Periodic Screening, Diagnosisand Treatment Program mandated by Congress in1967 as an amendment to the Social Security Act.Persons under 21 eligibie for medical assistanceare to be screened periodically and treated forhealth problems. The program is administered bythe states, which, following national guidelines,individually determine the health evaluation pro-cedures to be used and the mechanisms neededto carry out the program. Given these guidelines,the service targets in each state may be deter-mined. If 10,000 children are eligible and screen-ing for health assessment must take place an-nually, the target is 10,000 visits. However, thedecision may be reached to phase the program,taking one-third of the eligibles each year in thelight of available resources. Or it may be decidedto provide more frequent screening to youngerchildren, less frequent to the older. In That case,the population distribution by age determines thenumber of visits for the screening target. To setthe service target for treatment, one would needinformation on the incidence of conditions re-quiring medical treatment that are covered by theprogram.

With the service targets quantified, manpowerrequirements are derived by applying factors re-lated to manpower staffing and productivity.Neither of these factors is easily quantified. Forpurposes of estimating manpower requirements,we would like to know what proportions of theservices the tasks done in a health assess-ment of a young child, for example are per-formed by the pediatrician, the pediatric nursepractitioner, or the practical nurse. Manpower

4 4

38

staffing practices are analyzed by means of taskanalysis, and this entails a detailed study of jobperformance in which job functions are identifiedand separated into tasks. Finally, the worker'sactivities are studied to pinpoint the tasks he un-dertakes and the time allotted to each one. Thesignificance of the staffing pattern is that thedelegation of tasks from dentist to dental aux-iliary, or from pharmacist to pharmaceutical tech-nician alters the manpower requirements foreach affected occupation.

Productivity (output per unit of mahpoWer) alsohas a direct and important bearing on the numberof workers needed. If one can increase produc-tivity on the average by ten percent, one has infact proportionately lowered the requirements forpersonnel. (The unit of output for measurementof health-manpower productivity is a problem stillto be resolved.) 5

To illustrate the methodology of the servicetargets approach with a simple example, let ussuppose that the question the planner needs toanswer is, "How many emergency medical techni-cians (EMT's) are required for adequate service inour community?"

First, we define the service target: say that thenumber of ambulances required is established asone ambulance per 10 10 people, based on theanalysis of the past yea's emergency calls andthe organization of ambulance services in thearea. The service target is validated by the expertJudgment of emergency medical physicians.Thus, this community, with a population of 50,000would require five ambulances.

Second, we determine the staffing pattern: letus assume it is agreed that each ambulanceshould carry two EMT's, and should be manned24-hours-a-day, seven-days-a-week.

Third, we decide upon tne level of productivity:in this case, the output of emergency personnel ismeasured by the hours they are available to re-

5. See the dIscuaelon In Chapter HI.

45

37

spond to emergency calls. Each EMT will man theambulance for an 8-hour day, 40-hour week.

Finally, we calculate the manpower require-ments: each ambulance will operate round-the-clock, a 168-hour week. Dividing 168 hours by a 40-hour week for each man equals four men. Multi-plying by two men per shift equals eight men,which, when multiplied by five ambulances,equals 40 men. Adding one additional EMT as arelief man gives the community's total manpowerrequirement for EMT's 41 men.

The primary focus of the service target ap-proach is on the delivery of care. In estimatingmanpower, one must consider what types andquantity of service are to be provided to the com-munity and, in turn, the type.% and amount of ser-vice each worker will offer. ) he planner is led toconsider alternative forms of organization andstaffing and to recognize that manpower require-ments vary with different delivery systems.

Following this approach, however, has itsdangers. The planner can easily become boggeddown, because a huge volume of data, consider-able Investment of time and money, and a highdegree of statistical expertise may be required touse this method successfully. Another hazard Isthat the end product, the estimates based onnorms or seMce targets, may not be defensible Interms of economic realities or the felt needs ofthe community.

Another methodological approach uses the The Health Needshealth status of the local population as the start- Approaching point and estimates manpower requirementson the basis of the care needed to attain and main-tain good health. A study is made of the healthstatus of various demographic and socioeconom-ic groups using statistics and professional opin-ions on morbidity and the proper care for ilinessand health maintenance. The community's needfor health services is translated into manpowerrequl rements.

"Good," bsing subject to definition, may meanin this context either minimal, adequate or

optimal well-being. The standards of goodmedical care and the pathways of care are set byprofessionals doctors, dentists, therapists, andother medical personnel. They are asked tospecify the standard of good care for each serviceand diagnosis under various health conditionsthe nature and content of the care, the number ofvisits, the time required for each visit, the kinds ofpersonnel delivering the care and so on. For aspecific program such as good medical care forchildren under 17, detailed standards of care areset for well babies, acute and chronic child ail-ments and other problems. With quantitativeinformation on the size of population, frequencyof illness and health conditions, services to beoffered and time required to deliver these ser-vices, the planner can define the health needsspecified by a particular program as a basis forestimating personnel requirements.°

A mass of information is needed. On the ser-vices side, detailed data are needed on the healthconditions of the population and the volume ofservices estimated to be needed to provide carefor specific conditions and health needs. On themanpower side, detailed data are needed on theamount of time It takes to perform the necessaryservices by each health occupation and on thep iuuctivity of each health worker. The healthneeds approach may be viewed as an extension ofthJ service targets approach in which the targetsaro ot by the biologic needs of the community, asprofessionally determined.

Let us illustrate the distinctive character of thehe. 'n needs approach by reverting to theexaloplo of an emergency medical technician. Theproblem Is to estimate the number of EMT'snor,dt d for ambulance service in a comprehensiveert .,,ncy medical service system.

:;chonfeld, ot al., "The development of standards forthe audit and planning of medical care: Good pediatric careprogram content and method of estimating needed per-

Araorican Journal of Public Health. November 1968,

4 7

31

The health needs are prescribed by the opinionsof experts as to the standard of medical care re-quired by the critically injured, heart attack andstroke victims, and other emergency cases. Thesestandards may stipulate that physician and tech-nician services would be available 24-hours-a-day,to provide professional emergency care within 25minutes of call, for an ambulance trip of no morethan one hour within a geographic radius of 50miles, employing a two-way radio-communicationsystem and hospital-based ambulances. The num-ber of ambulances necessary to meet these stan-dards of need can be calculated by taking intoaccount the community's population characteris-tics, geography, travel conditions, location ofhospital emergency rooms and emergency equip-ment, and the numbers and types of accidentsthat require emergency treatment within a certaintime period.

The manpower requirements the number ofEMT's needed are derived in much the sameway as In the service target approach: first, thestandard of "needed" emergency care must bedefined; then the volume of services needed mustbe quantified. In this particular case, the commu-nity is divided into emergency care zones; terrainand road conditions are taken into account in thecreation of these zones so that the standards of a50-mile radius and a maximum trip duration ofone hour are met. Within each zone, a study ofthe number and characteristics of the populationand the record of accidents and other emergen-cies is made to provide a basis for quantifyingthe emergency care load. The hospitals at whichthe specially equipped ambulances should beplaced are pinpointed. Let us assume that threeemergency care zones are designated, requiring atotal of five ambulances.

If the manpower components the staffingpattern and productivity are envisioned as inthe service targets example, we will come up withthe same estimate of 41 men. But let us assumethat our objective is to meet the community's

48

minimal needs. We might then alter the staffingpattern to man each ambulance with one driverand one EMT. Our conclusion then would be thatthe community needs only 21 drivers and 21EMT's; given the salary differential, this staffingpattern would represent a considerably lowerfinancial burden.

On the other hand, if needs are defined in termsof optimal care, the staffing pattern might be twoEMT's per ambulance for each shift, one with abasic 80 hours of training and the second with 2e0hours of advanced training a total requirementof 41 men, as we have already calculated. In addi-tion a backup emergency medical physicianwould be required in each zone to guide and directthe personnel in the ambulance. If each doctorworks a 50hour week, each zone would requirethree and one-third doctors. This would mean aminimum of ten emergency medical physiciansfor the community.

The main advantage of the health needsapproach for the health planner is its logical basisin "what ought to be" as the reference point.TNs is especially appealing for longrange plan-ning, or highpriority categorical health programs,and for environmental health programs. It couldbe the proper basis for a health system plan whichsets community goals in line with the nationalguidelines to be established under the NationalHealth Planning and Resources Development Actof 1974.

One disadvantage Stems from the difficulty in

defining health needs, since expert opinionsdiffer. Moreover, this approach requires detaileddisaggregated data and sophisticated comr..!ia-tional techniques to quantify needs, servIcer ndmanpower requirements and therefore !s Jik& o

be most costly and time consuming. In addiwthe health needs approach, if based upon stan-dards set too high, will over-estimate the rnal-power that the commimity is in a position finance. Policies enc., programs to supply this 143ve!of manpower are likely to result in fk:turfiunemployment,

4 9

41

This methodological approach focuses on the The Economic (Effective)"effective" demand for health care the willing- Demand Approachness and ability of consumers or the communityto pay for health services as the basic deter-minant of the demand for manpower. Manpowerrequirements are derived from an estimate of themonies available from all sources to pay for care,including wages and salaries, oi from an estimateof the services consumers are willing to buy,taking into account the tasks performed and theproductivity of health personnel. Effectivedemand for manpower may be elicited from em-ployers or analytically deduced from healthexpenditures or service utilization data.

To illustrate the distinctive concept that under-lies this approach: let us again use the example ofthe emergency medical technician. It is signifi-cant that the problem is now defined as thenumber of EMI's required to fill the jobs in acommunity-wide ambulance service.

The initial task is to quantify the effectivedemand for emergency ambulance services. First,the funds available to operate the service must bedetermined from a study of the gross revenuegenerated from emergency calls in the past orfrom an analysis of the factors that account forthe revenue the number of emergency runs, thedistance traveled, fees paid and uncollectedcharges, government funding support and so on.Information would also be needed on the averagecost of operating an ambulance. The number ofambulances the community will support can bedetermined by dividing the average cost of anambulance into the funds available.

Alternatively, the effective demand for servicesmight be identified from utilization records, sinceit can be reasoned that the use of the service indi-cates the level of consumer demand, representingthe need, willingness and ability to buy.

Another approach to measuring economicdemand is to go directly to the operators ofambulance services the hospitals, fire depart-ments, funeral directors, police, or whoever in the

5 0

42

community performs this service and to ask fortheir opinions of requirements, present andfuture. This task produces a direct estimate ofmanpower requirement, based on employmentopportunities, although the employer may himselfbe deriving the requirements estimate from hisjudgment of the services demanded or the rev-enues generated.

The conversion of the demand for ambulancesinto the demand for EMT's would proceed, as wehave described in the service targets approaches,based on data on the staffing patterns and pro-ductivity. The computation divides the totalamount of services, however defined, by theaverage output of one health worker.

As we have indicated, a variety of techniquesmay be used to calculate the economic effectivedemand and to convert this demand for senlicesinto manpower requirements. A survey of em-ployers or an area skill survey is one approach.Another is to analyze empirical evidence of theutilization of services by population groups inrelation to measures of disease or iliness, per-sonal income ana other health system determi-nants. The most sophisticated techniques employmathematical models, using regression analysis,mathematical programming and simulationmethodologies.

The economic (effective) demand approacheshave the advantage of estimating requirementsthat relate to job opportunities. When the plan-ner's frame of referenoe is the manpowerdemanded in the labor market, he properly turnsto these methodologies. However, situationsexist where service target or health need, noteffective demand, is the key determinant, as Inassessing the impact of a proposed program Inpublic or environmental health. Moreover, most ofthe economic demand methodologies requiresubstantial data, technical expertise and financialresources.

Future Requirements Manpower estimates for present or future re-quirements may be based on the same method-

5 1

43

ological approaches: health manpower/populationrates; service targets; health needs; economic(effective) demand. The time frame, not the tech-nique, produces the distinction between presentand future requirements. In calculating futurerequirements, one incorporates the changesexpected or estimated to occur over time. Thevalidity of the requirements projection will reflectthe realism of the assumption about the futureand the quality of the data that describe the pastand the present.

In general, the analytical approach to projec-tions starts with a study of the past or the present,identifying the forces affecting manpower thatproduce change, and then calculates future re-quirements based on the impact of these changesover time.

The methodological process underlying all pro-jections may be described briefly as follows: First,one must determine the factors that affect man-power requirements, and study the way in whichthey have operated. If possible, one should deter-mine any economic variables which measure, arerelated to, or can serve as a proxy tor each factorand which are capable of being projected inde-pendently, and study their past relationship to theoccupation. Second, one must estimate the futurelevels of the relevant variables and their futurerelationship to manpower requirements. Third, therequirements must be projected to the target yearon the basis of these relationships.'

A brief summary of each method presentedearlier in the discussion of present requirementsillustrates [tow each concept is adapted for use inprojecting manpower requirement.

1. Health manpower/population ratios (ratiomethod): This method involves the identifica-tion of a suitable health manpower/popu;a-

7. Goldstein, Harold M., "Methods of p,ojecting supply anddemand In high level occupations," paper delivered at AnnualConvention of the American Statistical Association, Phila.delphia, Pennsylvania, September 8, 1965. Washington, D.C.:American Statistical Association, 1965.

5 2

44

T

tion ratio for a future point in time and thenthe application of this ratio to the projectedpopulation to derive manpower require-ments ....

2. Service targets approach (normative ap-proach): This method emphasizes the de-velopment of detailed standards for the pro-vision of different kinds of services, stan-dards which are then used to derive targetsfor the production of these services. Man-power staffing and productivity standardsare then used to convert service targets intothe manpower required to attain such stan-dards, including those used in the healthneeds. [For projection purposes, the targets,manpower staffing and productivity stan-dards are future-oriented.]

3. Health needs (or biologic needs): Thismethod seeks to determine, based on expertopinion and taking into account [healthstatus, medical knowledge andj availabletechnology, what kinds, amounts and qualitylevels of services are required to attain andmaintain a healthy population. Service tar-gets are then converted into manpowerrequirements by means of staffing and pro-ductivity standards .... [Again, for projec-tion purposes, measures are quantified inaccordance with assumptions about thefuture.]

4. Economic demand (or effective economicdemand): This method concerns the mea-surement and projection of what health ser-vice people are willing and able to pay for,irrespective of the quality of the specific ser-vices obtained or of their need for them ....[One variant of]the method consists of corre-lating the receipt of services with selectedeconomic and other variables, and then ofprojecting the changes likeiy to occur regard-ing these variables in order to derive theimpact of these changes on the demand forservices and ultimately on health manpowerrequirements ....a

Other variants are employer surveys, in whichinformation is obtained on current and expected

8. Hall, Thomas L., op. cit.

5 3

employment opportunities; and the input-outputtechnique, in which an industry-occupation em-ployment matrix is used, and estimated propor-tions in specific occupations are applied to theprojected total labor force to determine futurerequirements.

Manpower requirements are influenced bymany economic, political and social factors thatinteract in a complex fashion we cannot fullymeasure. Econometric models, attempting to de-scribe the functional relationships among themany variables, use dozens of equations, it iseven more difficult to discern and to express inprecise terms the dynamic changes which arelikely to occur in the future and to measure theirimpact. The data we do have about the past andpresent may not be adequate for our needs. In theface of these underlying difficulties, it is not sur-prising that all of the alternative methodologiesare subject to criticism; their strengths, however,should not be overlooked.

The manpower/population ratio method isquick, cheap, easy to do and to understand; un-fortunately, it may also be dangerously simplistic,and may use inappropriate ratios that fail to takeinto account any influence other than populationsize. The service target approach recognizesand this is its strength that manpower require-ments are derived from the demand for services,however defined, and that the way manpower isused (the ,taffing pattern) and how much serviceeach health worker produces (productivity) deter-mine the number of health personnel required. Aproblem arises when the service targets are set byconsumer needs or wants, irrespective of the ef-fective demand for such services. The estimatewill most likely overstate the future ability of themarket to fund jobs and lead to erroneous policydecisions that could result in unemployment.Another major weakness of the service target ap-proach for projection purposes is its extensivedata requirement of variables difficult to quantify

5 4

Strengths andWeaknesses ofAlternativeMethodologies

45

46

and the real possibility that the planner willbecome enmeshed in a longer term, more com-plex, and more costly effort than he realized at thestart.

The same criticisms may be directed at the useof the health needs approach for projecting man-power requirements. If the service targets of thefuture are set according to the standard of"need," the two methodologies converge. Thecritical step in the health needs approach, that ofsetting standards for good health care and ofestablishing the appropriate mode of care for thefuture, makes the process of projecting require-ments even more difficult. Still, the future healthneeds of the community are the relevant measurein the context of public health and environmentalprograms.

The economic (effective) demand approachdoes relate to job opportunities, preventing theoverstatement of manpower requirements whichmay result from other approaches and producingmore accurate guidelines for educational plan-ning and vocational counseling. However, none ofthe specific effective demand approaches is. freefrom technical limitations or problems. For exam-ple, the employer survey produces an estimatefrom the perspective of existing institutions,ignoring new organizations that may come intobeing and the unstated but assuredly variedassumptions about the future that underlie theresponses.

Since each approacn has its unique strengthsand weaknesses, the planner with sufficient re-sources may find it to his advantage to make esti-mates based on several methodologies. He wouldhave at his disposal a range of estimates thatwould probably prove more useful as a basis forhealth resources planning than any single esti-mate would be.

5 5

47

V. MethodologicalApproaches toEstimating Supply

Measurement of current supply is frequently Current Supplyundertakei. In conjunction with the measurementof requirements in order to assess the presentsituation am.; to develop a plan that will correctthe imbalances. The gap between supply and re-quirements may be due to a shortage or surplus,but, alternatively, it may be the consequence ofmaldistribution and poor utilization. In the shortrun,.changes in relative wages, in labor produc-tivity and in worker mobility may bring about anequilibrium without any alteration in the numbersof health personnel. Supply is modified moreimmediately by geographic and occupational re-distribution and greater output per workerthrough the incentive of increased wages than itis by production of new graduates.

Measurement of the current supply is neces-sary also for projecting future supply, an essentialdatum for long-range planning to assure adequateresources to meet the anticipated demand forhealth care. Educational planning and vocationalcounseling am guided by supply projections toassure that future graduates will find job oppor-tunities and the economy will have neededworkers. An understandir: of the factors influ-encing the supply and the operation of the occu-pational labor market, especially the demand and

56

48

supply elasticities, is vital for formulating effec-tive manpower and educational policies.

Meaningful data on current supply go beyond ahead count of the number of persons. For pur-poses of supply analysis, we should understandthe process and structure of supply: how peopleenter the field; how they acquire the necesaryskills; why they leave. Furthermore, we shouldknow the institutional characteristics of thecurrent supply: who the employers are; how jobsare titled; how people are used; how much servicethey produce; how many hours they work; wherethey are located; and what they are paid.' In addi-tion, the personal characteristics of the currentsupply age, sex, education, race, mobilityare also important factors in assessing the stateof cur, supply.2

We , i to know about the supply.structure inorder to properly measure the sources of currentsupply. We need to know the institutional charac-teristics because the possibilities foradjustments in hiring standards, utilization, pro-ductivity, location and salary levels to correctgaps between supply and requirements can bediscerned only from a detailed knowledge of exist-ing practices. We need to know personal charac-teristics to properly assess likely labor forceparticipation. Since women are predominant inmany health occupations, the distribution by sex,age and race is a clue to the separation betweenactive and inactive workers in the years ahead.Entry and separation rates are necessary to esti-mate future supply.

The planner finds that he must function withconsiderably less information than he ideally

1. Wages relative to earnings in related occupations and themovement of relative wages are significant indications of

supply imbalance.

2. For a more detailed discussion, see U.S. Department ofLabor, Bureau of Labor Statistics, Occupational Supply:Concepts and Sources of Data for Manpower Analysis. BLSBulletin 1816. Washington, D.C.: U.S. Government PrintingOf fice, 1974.

5 7

49

would like to have. Time and cost constraintsplace a limit on the detail with which currentsupply is measured. The planner is not concernedwith occupations whose skill level requiresminimal training acquired in short courses or onthe job and whose workers easily transfer in andout in response to change in relative wages. Theoccupations that do concern him are those forwhich supply is relatively inelastic in the short runbecause they require substantial education andtraining and legal and professional qualifications.Forward planning is needed to effect a change inthe supply for these occupations. The plannermay be especially concerned with occupations forwhich educational programs exist in his area.

Current supply is defined as the number ofactive practitioners in the labor force, that is, acount of those employed or actively seeking em-ployment in a health occupation at a given time.Job seekers include the unemployed or newgraduates, but they may also be professionalswho have been inactive and are seeking to re-enterthe labor market. 3 The given time may be thepresent or a recent year that is used as a baselineor benchmark.

A distinction is made between "active" supply,composed of the number of practitioners present-ly employed or actively seeking work, and the"potential" supply, consisting of the activeworkers plus those inactive workers who might beattracted back into the labor force or into theoccupation if working conditions or personalcircumstances were different. The data concern-ing inactive workers which is of interest for plan-ning purposes are their personal and professionalcharacteristics and the threshold conditions thatwould return them to active status. These inactivepersons represent a reserve force which may besizable and which may provide an immediate andleast-cost solution to the inadequacy of the cur-rent supply.

3. This is only one of the several definitions of supply that arediscussed in Chapter II.

5 8

50

Current SupplyMethodologies

Supply methodologies estimate separately thecomponents of supply and add these componentestimates to measure total supply. With respectto current supply, these components are the em-ployed and the unemployed. Current supply isoften measured in terms of the number employed,ignoring the unemployed (both those betweenjobs and new graduates actively seeking work)because this group is so difficult to identify. Thisexpedient adjustment in current supply may be

justified at times when the rate of unemploymentis low. Nevertheless, whenever the unemployedare omitted from this measurement, currentsupply is underestimated. To measure potentialsupply, the number of inactive workers is added to

the current supply.Different methodologies may be used to esti-

mate the components of current supply, depend-ing upon the availability of data. Data can be ob-tained from secondary sources, which have muchto recommend them since they provide relativelyquick and easy answers. Examples of secondarysources which can provide data useful in esti-mating current supply are the Bureau of theCensus (especially their Census of Populationand Current Population Surveys), the NationalCenter for Health Statistics, the U.S. Bureau ofLabor Statistics, state employment securityagencies, state departments of higher education,professional associations and trade unions.

However, caution must be exercised in usingsecondary sources.

In using estimates of the number of indi-viduals licensed, registered, or certified for aspecific category of health personnel as ameasure of current supply, the reader is cau-tioned that adjustments should be made toaccount for the fact that some of the indi-viduals are employed part-time and some areinactive. In addition, further adjustmentsshould be made on the assumption that notall inactive personnel are equally availablefor employment. For example, some are re-

5

51

tired, disabled, or no longer qualified, whileothers are only temporarily inactive. 4

When one gathers supply data from differentsources, problems of noncomparability areinevitable and must be resolved Uy making themost logical linkages and adjustments. The prob-lems the planner will face arise from inconsistenttime frames, from different supply concepts, fromvarying occupational definitions, from differentgeographical areas, and from gaps in coverage.

Planners should ascertain that secondarysources will not supply usable answers beforethey embark on the collection of data fromprimary sources. When it has been determinedthat it is necessary to collect data at first hand,the planner has two options: (1) an inventory orsurvey of health care providers; (2) a survey ofhealth professionals.

The emplcyer survey will be useful for themeasurement of the employed and the determina-tion of employment settings characteristics; theprofessional survey will gather data on the activeand inactive members and their personal charac-teristics. Surveys designed to obtain informationabout the current supply produce data that arerelevant for both supply and requirement esti-mates and for future supply projections.

Future supply estimates must measure the Future Supplymovement of personnel into and out of the currentsupply (see Fig. 1). The additions or inflows tosupply consist largely of graduates from educa-tional and training programs, but they may also betransferees from other occupations or geographicareas or workers re-entering the labor force. Thelosses or outflows from current supply are causedby death, retirements, and transference to otheroccupations and other areas.

4. U.S. Department of Health, Education, and Welfare,National Center for Health Statistics, Health Resources Statis.tics, 1971, p. 4. Washington, D.C.: U.S. Government PrintingOffice, February 1972.

6 0

52

Fig. 1. Flows ofWorkers into and out

of an Occupation8. ENTRIES

Speci!ictraining

2. Other 1training

3. Otheroccupations

4. Outside thelabor force

5. Immigration

A. CURRENT SUPPLY

EMPLOYEDplus

UNEMPLOYED

C. SEPARATIONS

1. Otheroccupations

2. Outside thelabor force,incl udingretirements

3. Deaths

N 4. Emigration

Source: U.S. Department of Labor, Bureau o Labor Statistics,Occupational Supply: Concepts and Sources of Data

for Manpower Analysis, BLS Bulletin 1816. Washing-

ton, D.C.: U.S. Government Printing Office, 1974.

To make projections, one must estimate theannual additions and losses during the timeperiod of the projection. In practice th followinoprocedure is usual:

First, a current supply estimate is estab-lished as the base of the projection. Then theannual number of entrant :sm all sources isdeveloped for the period .at the projectionis to cover. Third, the base current supply isaggregated with estimates of the annualnumber of entrants and annual occupationallosses are deducted.5

However, the lack of data on specific inflowsand outflows, e.g., occupationai and geographicmobility, leads to modifications in the estimatingprocedures. Judgments and proxy measures aresubstituted for specific data.

5. Rosenthal, Neal, "Projections of manpower supply in aspecific occupation." Monthly Labor Review, November 1966.

61

The annual number of graduates, the majorsource of additions to the current supply of healthprofessions, is usually available. Data on educa-tion enrollment are available from such sources asthe U.S. Department of Health, Education, andWelfare and the professional associations. Forany particular area, retained graduates are the netaddition to the supply; those who move to otherlocations or other occupations should be ex-cluded from the count of entrants.

To project supply on the basis of future classsize, one must allow for student attrition in everyyear of the program. To make this important cor-rection one analysis proceeded to "adopt the re-cent attrition experience for schools within anoccupation and maintain this rate for the length ofthe projection period .... The impact upon thesupply projections of using alternative estimatesof student attrition" was also studied.6

One measure of inflows that has a built-in ad-justment for student attrition and retention of newgraduates is the number of new licentiates. Newlicentiate data provide a basis for estimating netadditions to the stock of supply. It is not a com-pletely accurate measure, for among the annualrecipients of a state license are some graduateswho, for only one of a variety of reasons, will notpractice their profession in the area.

Occupations with long and arduous preparation Occupational Transference'lave few dropouts and few transferees. Physi-:lens, for example, typically remain in medicineafter graduation, going on to residencies. How-ever, those who elect to go into research, teachingand administration may be considered to be"lost" from patient care. In addition, a few physi-cians do shift from one medical specialty toanother during their working lives.

Estimating InflowsThe Production ofNew Graduates

53

6. U.S. Department of Health, Education, and Welfare, Bureauof Health Planning and Resources Development, The Supply ofHealth Manpower, 1970 Profiles and Projections to 1990, p. 9.DHEW Pub. No. (HRA) 75.38. Washington, D.C.: U.S. Govern-ment Printing Of f ice, Dacember 1974.

6 2

14

Geographic Transference

In the health fiek, croenljalir.g requirementsare major barriers to occupational transference.Since licensure laws and certification standardstypically stipulate graduation from an approvededucational program in the particular occupationas one of the necessary conditions, training orexperience in other health occupations is not con-sidered sufficient proof of competence. Thi3obstacle to occupational transference may belowered or removed by the rise of proficiency andequivalency examinations.

For the measurement of occupational mobility,existing data are unfortunately inadequate. Infor-mation on occupational transference (to and froman occupation) may be acquired in a multiphaselongitudinal study, in which the same people areinterviewed at intervals about their occupation or,as in the 1970 census, in a sample survey in whichrespondents are asked to recall occupationalactivities five years earlier. The 1970 census dataare being analyzed by the Bureau of Labor Statis-tics to develop tables showing mobility rates byoccupation. VVhen these rates are available, it willbe possible to calculate average annual flows dueto occupational mobility. Another approach tomeasurement is the residual method, in which theamount of occupational transference is estimatedfrom historical data on the total supply and on thechange in all other inflows and outflows. To usethe residual method, the planner must know totalsupply at two dates and of the additions andlosses to supply within that period of time, otherthan that due to occupational mobility. The differ-ences between total supply not accounted for bythe other flows is attributed to occupationalmobihty.

Movement in and out of the area may also bemeasured by longitudinal data and may be ap-proximated by the residual roathod. Geographicand occupational mobility often occur at the sametime. VVomen may move because their husbands'work requires it and may have to take whatever

6 3

55

Job is available. Applications for licenses to begranted on the basis of reciprocity or endorse-ment may be used as a proxy measure of geo-graphic movement between states. In the case ofphysicians, foreign med:.,n1 graduates are a majorsource of additions to supply. Their number ismonitored by the American Medical Associationand medical licensirg boardG. Projections of thesupply of physicians are strongly influenced bythe assumptions as to the level of the flow offoreign medical graduates into the United States.

Women characteristically move in and out of Re-entrance to the Labor Force

the labor force in response to changes in theirfamily obligations. female-dominated health oc-cupations, such as nursing and occupationaltherapy, have substantial reserves of inactive pro-fessionals. The age of children and the husband'sincome are decisive factors in the women's laborforce participation. NaConwide labor force partici-pation rates for women are available, derived fromworking-life tables, but they are not available byoccupation. However, the American Nurses Asso-ciation does collect data on the labor force mobil-ity pattern of RN's, and the nurse's pattern may bea basis for estimates in other women-dominatedhealth occupations. In general, information on re-entry for specific occupations is very meager.

As we have already mentioned, data limitationsmake it difficult to estimate losses caused byshifts out of an occupation or area. The nationaloffice or local affiliates of professional associa-tions may be able to approximate the number ofmembers who have dropped their license or cer-tification. However, many professionals, especial-ly women, retain their credentials even when theyhave withdrawn from the labor force.

[Outflows] from separations can be cal-culated in several ways. A simple way is todetermine the average working life of mem-bers of a particular occupation; in an exam-

6 4

Estimating OutflowsOccupation and Geo-graphic Transference

Separations Due toDeaths andRetirements

56

ple working life of 40 years, then 1140 or 2 112percent are assumed on the average to retireor die each year. This percentac a would bevalid if the numbers at each Cy.: level wereequal, as in an occupation v.:doh has notgrown for 40 years and which lied steadyinflux of workers each year. The ratu could bemultiplied by the number of wo!kers in theoccupation to obtain the actual number whomay be expected to leave the occupationeach year ,

A more refined technique for estimatingdeaths and retirements is based on "tablesof working life." These tables are statisticalor actuarial devices for summarizing themortality experience of the popuiation atsome particular period of time, i.e., the deathrates, by age, over a one-year period. A lifetable starts with a hypothetical group ofpersons usually 100,000 born alive andfollows the death rates of the real populationat each age. Tables of working life also indi-cate labor force participation of the initialgroup of 100,000 from 16 years of age; itshows attrition caused by withdrawals fromthe labor force as well as by mortality. Tablesof working life, which have been set up on anactuarial basis for both men and women,account for deaths and retirements (sepa-rately) at each age level.'

The Bureau of Labor Statistics recently issued abulletin, entitled "Estimating Occupational Sepa-rations From the Labor Forces for States," de-signed to aid state manpower analysts in esti-mating occupational separations from the laborforce. Individual state labor force separation ratesare covered for 1970 and 1985 for over 420 occupa-tions.

Strengths and Supply methodologies vary with the concept ofWeaknesses of supply and with the component of supply being

Supply Methodologies

7. U.S. Department of Labo', Bureau of Labor Statistics,Tomorrow's Manpower Needs: Estimating Occupational Seca.rations From the Labor Force for Slates, Supplement Nc. 4.Report prepared by McElroy, Michael P. and Cangialosi,Joseph S. Washington, D.C.: U.S. (lovernment Printing Office,1974.

6 5

57

measured. The strengths and weaknesses ofsupply methodologies reflect the availability,comparability, consistency and validity of the dataused, as well as the number of components ofinflows and outflows included in the supply esti-mate.

Active supply turns to employment records;potential supply must also measure inactive work-ers and therefore seeks information from indi-v:dual practitioners and professional associa-tiorm Current supply methodologies are designedto collect necessary data or to adapt availablesecondary data from different sources.

Future supply methodologies estimate specificadditions and losses to supply during the projec-tion period. To quantify the various inflows andoutflows from the stock of supply, data are col-lected or secondary sources and proxy data areanalyzed by a variety of statistical methods.Census and sample surveys, trend analysis andsimple statistical measures are widely used; thus,supply methodologies incorporate the strong andweak points of these techniques.

In general, data for the independent health pro-fessions physicians (MD's and DO's), dentists,optometrists, pharmacists, podiatrists, veterinar-ians, registered nurses and credentialed alliedhealth occupations, such as dental hygienists anddietitians are much more readily available andof better quality than for most allied health andpublic health occupations. Similarly, estimates ofsome elements of future supply are better thanothers; for example, the inflow of graduates isbetter reported than the outflow due to geograph-ic or occupational mobility.

One of the strengths of supply methodologiesis that, in general, data are available for the majorinputs; those elements for vihich information isdeficient are by and large relatively small. Forexample, estimates of current supply typicallyignore the unemployed, and estimates of futuresupply frequently do not take into account occu-pational mobility. These "omissions" do not seri-

6 6

58

ously affect the final supply estimates since theyare of small account in most places, at most timesand for most health occupations.

The weaknesses of supply methodologies aresignificant. First, they produce a head count over-looking the importance of the mix of manpowerand the quantity of services supplied. It does notfollow automatically that a larger number ofpersons will provide more services, or the con-verse, that a smaller number of personnel willreduce the quantity of services. Labor productivityand the organization of the delivery system affectthe supply of services that health workers provide,but supply methodologies ignora these effects.Moreover, even when the head count is of full-time equivalent personnel, the estimate may be

misleading if it is based, not on the number ofhours worked, but on some arbitrary assumptionthat two part-time workers equal one full-timeworker. Furthermore, the head count does not suf-ficiently describe the workers by such character-istics as age, sex, race, mobility, educational at-tainment, to enable the planner to logicallyevaivate the estimates of supply.

Second, the methodological approaches esti-mate the supply of labor independent of thedemand for labor. As we have discussed earlier,

e amount of supply the number of healthworkers is respo- s v e to changes in the labormarket, such as relative wages or working condi-tions. An obvious example is the number of inac-tive nurses who would return to active participa-tion in the labor force if wages became sufficient-ly attractive, or if part-time work or flexible hourscould be arranged.

Third, aeveral methodological approaches haveinherent weaknesses, which are discussed inChapter IV, Vol. II. The estimate of additions tosupply, for example, may be based on the outputof educationa, programs, new graduates, or theannual addition to licensed practitioners. If thefigures of new graduates or new licentiates areused without adjustment, the attrition due to geo-

6 7

59

graphic mobility, occupational transference, orlabor force withdrawal is ignored.

All supply methodologies have inherentstrengths and weaknesses. The planner mustrecognize the qualities of the methodology he isusing, adjust the estimate to the best of his abilityto take into account the likely error, and use hisfinal estimate with a full awareness of the implica-tions of the estimating process.

6 8

60

VI. HealthManpower Issues:

Uses andLimitations of

Statistics

Introduction Manpower statistics are a basic element in area'malth planning. One writer, who sees more ade-

quate statistics as a key element in the solution ofmanpower problems, has characterized themanifold uses of these statistics as:

1. assessing present and prospective healthmanpower supply relative to demand or re-quirements;

2 f .)rmulating goals for recruitment, train-ing, and alternative uses of health man-power;

3. appraising the feasibility of implementingspecific pieces of health legislation;

4, health manpower planning on the region-al, state and local levels with respect to in-equality of access to health care services;and

5, health manpower research, such as thediagnos s of shortages, productivity studies,the impact of new technology, and the like.

1. For an insightful discussion of the uses and limitations ofcurrent manpower data, 5 e e Butter, Irene, "Improved statisticsare required." Hospitals, July 1, 1972.

61

From another perspective, the President's Com-mittee on Manpower pointed to several specificpurposes which manpower projections mightserve, such as:

To alert govern: )nt (and other interestedartiet;) to emerging manpower problems;

commonly, an imbalance between the de-mand for and supply of workers in the labor

.ce.To help choose between alternative pro-posed policies.To assist administering specific govern-ment programsTo provide an essential element for develop-ing other gen )ral types of projections bygovsrnment and private organizations. 2

Thus, manpower statistics are the raw materialsof planning, necessary inputs to the developmentof educational and training programs and to voca-tional counseling. Appropriately analyzed, theyhelp the planner to probe the constraints to in-creased supply and to identify the proper focus ofrecruitment and retention efforts. They affect hisevaluation of the feasibility of new health policiesand programs and of the likelihood of achievingcertain standards of health care. Manpower datainfluence his recommendations on proposed newand expanded health care facilities as well as onnew educational Intl lotions. In this way, theplanner becomes one of the forces working toalign health manpower resources with health caredemands and needs,

Area health planner, find they must have Problems In Dataanswers to a multitude of questions related to Gatheringmanpower supply and requirements: What typosof personnel are needed'? What educational andtraining programs are b- offered now whatkinds of occupations are [icing trained; whore aro

2, U.S. Department of Labor, E3uroau of Labor Statistics,Occupational Manpower and Training Neoda: Information torPlanning Training Programa tor the 1970s. BLS BuilMln 1701.Washington, D,C.: U.S, Goverrirnor,t Printing Of lice, 1971,

7 0

62

the programs located; how many students aregraduated; where do they find work? What othersources of supply exist in the area? How will man-power requirements change over time? What willhappen to demand in the future, if, for example,National Health Insurance becomes law? How willit affect personnel requirements in the area? Willthe present supply sources be adequate for thecommunity's needs in the future? What changesshould be made?

The health planner must cope with a number ofproblems in measuring local supply and require-ments, First, accurate estimates must be de-veloped from an analysis of the detailed character-istics of the population and its demand for healthcare; of the organization of the health deliverysystem and the pattern of health manpower staf-fing and utilization; of the inflows and outflows ofsources of supply and the functioning of the labormarket. However, with regard to each field ofanalysis, area statistics (and in many instances,national data) are limited; for example, area dataon healtn care needs, services used, medical ex-penditures, labor productivity, the personalcharacteristics of workers and their relative wagesare seldom available, Not only are current datalacking, but reliable information about the pastmay not exist and that which is available may notbe consistent or comparable over time because ofdifferent definitions, coverage, time arins anddata collection methods. Moreover, the , aces ofinformation may be misleading; for example, cer-tification registries and liconsure records over-state the supply by including professionals not inthe labor force but omit employed practitionerswho am not crodentialed, The data problem ismagnified by the proliferation of health occupa-tions and titles --- numbering in the hundredswith the r-tme job title at times used to describedi fforont sets of functions and responsibilities. Tocompound the problem, dlsaggregated data forrelevant labor market and health non/ Ice areas arewoefully inadequate: most manpower data have

7

63

been collected at the state and national levels.While some relevant statistics are available for thecounty and standard metropolitan statistical area(SMSA), small area estimates and projections ofessential information are notably missing.

Second, no simple guideline or consensusexists to aid in the identification of appropriatemethodologies. The lack of agreement reflectsthe general lack of knowledge about the forcesinfluencing the demand for health care, about theconversion of that demand into manpower re-quirements, and about the supply responses tochanging circumstances. It also reflects differingopinions about concepts and definitions. Evenwhen there is agreement on factors that must bemeasured, we are not certain how best to proceed.For example, we know that head counts by dis-crete occupations are misleading; that the healthprofessions function in a complex web of cornple-mentarity and substitutability. But we do not haveoperationally practical techniques to measure themanpower input in health teams or other staffingpatterns. We have no accepted method for mea-suring the output of health workers. In otherwords, we have no Input and output measures totranslate manpower demand and supply Into thedemand and supply of services, and vice versa.

Third, projecting local conditions into the futureis hazardous. Futuro requirements for manpowerin the health service industry will bo influenced bypolicies and programs (o,g., the introduction of aNational Health Insurance plan and the spread ofhealth maintenance organizations) the impacts ofwhich can only be dimly seen. Similarly, the sup-ply of manpower will be influenced by new ef-forts, such as the direction of Federal support forhealth manpowor training and education and theIncreased labor forco participation of women. Inaddition to forces of national signitn ance, thelocal planner must take into account 'specificlocal factors and their effect upon healthmanpower in the future. Moreover, the adequacyof tho future supply in relation to future demand

7 2

64

for health manpower will depend in large part ongeographic and spPcialty distribution, subjectsabout which there limited understanding andinformation 3

Distribution of Health The maldistribution of health manpower hasManpower been identified as one of the key issues the health

planner must address in the 1970's. This problemwas not apparent in the 1960s when the numberand variety of paramedical personnel proliferatedand shortages were prevalent. It is estimated thatat least 1,000 communities in the United Statesare without adequate health care. Areas of scar-city suffering from maldistribution of personnelrefer not only to geographic area:, but also tominority populations, to types of physicians, andto types of services. Today the inadequacy ofhealth care and the short supply of health workersare especially evident In outlying rural areas andin the inner-city slums. Of special concern is thedeclining proportion of physicians providing primary caro, indicative of the rnaldistribution amongmedical specialties.

Health planners rely heavily on the manpower/population ratio mothod for Identifying areas ofscarcity, It is a gross measurement, generally ofthe number of physicians, nurses, pharmacists,etc., per 100,000 population, that does not explic-itly take into account the determinant factorsother than changing age composition and urban,-iation. Be5Ides ignoring causative relationships,thi5 approach to studying rnakfistributlonpre5entn the problem of deciding what shouldappropriately go into the numerator and what Intothe denominator, and what in tho proper ratio to11110;111a 5tandard. However, when ama populationfigure5 and 5upply data ;lie available, a comp; ri

3 111; NoTartmont ol Mo31111, 1(localion, rind Iliiro, 111ihon;11 Non tor tor I l.iiIIh 111i101(;), actoth Ptlicfitm I oGidlott 0 Pttdo.iwonal Mt111110WW: A litpvlow

I dtmdli(ei ;)I II 1/! No. (1-111A) PO. liockvillo, ML. Nil.hrliarterittot t1 Ii,Ih thiciltit,n, itrul Witlfilre), 19/4,

7 3

65

son of the ratios in various areas does distinguishthe poorly served relative either to other areas orto some standard of a well-served cOmmunity,although it does not explain the differences orpoint to possible remedial actions. 4

Moreover, rnaldistribution is a problem that maynot be capable of resolution at the local level. It isnot within the local planner's ability, for example,to change the immigration laws to increase ordecrease the number of foreign medicalgraduates in his area, nor can he change the en-vironment of the area to make it more attractivefor doctors and their wives.' Doctors operate in anational labor market; they have freedom tochoose where in the United States they will prac-tice. The planner can, however, locate the under-served area, pinpoint the skills in short supply,identify the constraints on supply that must belowered or removed, and explain the policy op-tions available to his community in its attompt tocope with the problem. An accurate diagnosis ofthe nature of the maldistribution and the properprescription of effective remedies are the respon-sibilities of the planner.

Effective manpower utilization, anuther critical Effective Monpowerissue, may be the answer to maldistribution in Utilizationsome areas. In fact, some authorities believe themaldistribution proolem is in reality a malutili7a-tion problem. Concern with tho utilization of man-

4, Soo tho of tho Wilt of Population rot lori fw piirpo8e5 of anaiy9f in U.S, Dvparlmoot of Hoolth, ucal ion andWoltruo, BUreall of Hoalth firriollrc118 DOVOIwnent, OPari.bullan al Modica/ Spry:rally Manpowar, FH f iD/FIAS Howl No.74.192, Bottionda, Md,; U.S. Doportmont of Hoolth, Filw;ation,and Wol faro, April 26, 1974, p. 27

5. Soo U S. Dopartmont of Hoalth, Education, and Milton:,Nal tonal Coring for Hoalth actorss qlluvoringPoselleis tor:Masa ol Poster:shoal/ Munpownr: A litsvwwol Oar:aura, DHEW Pun, No, (H RA) 75.3, Md 11.

Dormrtmont ol Hwi It h, Educallon, and Milton], 1074.

7 4

88

power is associated also with efforts tc 1-,P,C)siethe quality of care and to control medlca:

Just as demand for health servicei manpowelis derived from the demand for health clarvices, it

is possible to think of supply of rr lo!,ower in

terms of the health services that Can 6, providedby a given pool of manpower utilized ,.t.h reason-able efficiency. This is another wa\i f 5ayinc thatthe approach to problems of suppii is throughproductivity (output of services pe 'Hilt of man-power). This places the emphasis nc,i on tables oforganization but on the most econ.rnical way toprovide the specified level and qua:: j of service.One way Is through substitution of e.'jipment foradditional manpower: Given the costs Of man-power and equipment, is it ecc,nort'cal to use anautoanalyzer Instead of additional .huoratory per-sonnel? Another approach is tt.ro,igi: task deler-tion, to economize on the usc c te si:arcer,costly manpower by delegating lcs exrrorltasks to less costly personnel as `lr Ps coo-sistent with good medical p. :Tic- , :hur in-

creasing the effective supply of vices tr.,.t Canbe provided by the more expensive typutilization requires a higher order nf r:lt,Itn re-sources planning and better manaprorA t than isgenerally practiced in the health sr.rvices sector.(In fact, it may require an exprAnsion the supplyof trained health servicer; idrnellt person-nel.)

Techniques of job are being devel-oped. The concept of remgnment of tasks to,r.eate more effective tar,i' clusters and more ra-tional occupational roles .n a ''medical team" isbeir g implemented. The complementary nature ofecdupatinns and the most efficient assignment of

Seo Departmen, Health, Education, and Welfare,National C, nter for Health Services, Reseercn and Develop.ment, The Ut larillon of Health RervIces; Indices and (Awe-lutes, A 13, search Blbllograph,,, ra,IEW Pub. No, (HMS) 73.3003(Rockville, Md. U.S. Departme. of Health, Education, andWelfare, 1972),

7 5

work among a group of health workers arestressed in effective utilization. The functioningof the operating room team is an example of com-plementarity; the use of physicians' assistants inan illustration of substitutability. The possibilitiesof substituting more readily Or lowercost health workers for scarce era; 3xpersive pro-fessionals have implications 'ci trc quantity,quality and cost of care, as well cis 1.07' manpowerplanning.

The transfer of functions from professional toparaprofessional task delegation is one ofthe major mechanisms for increasing the supplyof medical services. In the last few years a spateof new occupations, collectively called "physicianextenders," has emerged under such titles as phy-sician assistant, Medex, child health associate,clinical associate, nurse practitioner, Primex,aimed at coping with the doctor shortage in pri-mary care and in underserved geographic areas.

The complementary nature of occupationalroles and the possibilities for substitution add anadditional important dimension to estimatingsupply and requirements. Estimates of discreteoccupations, measured without regard to their re-lationship to other professions, overlook thepossibilities for Increasing productivity and thusreducing manpower requireme, and increasingeffective manpower supply.

Basic to the development of the physician as-sistant as a new health occupation, for example,is the assumption that the assistants will increasethe productivity of the physician and that togetherthe doctor and his assistant will deliver a signifi-cantly higher quantity of medical care, with nosacrifice of quality. No estimate of future require-ments or supply of physicians can ignore the roleof the physician assistant and his impact onphysician productivity. Unfortunately, the analysisof labor productivity in the health care sector hasproved to be one of the more difficult tasks under-taken by researchers in this field. The cause of

7 6

67

Productivity: The"Physician Extenders"

68

this difficulty is the uncertainty underlying thespecification and measure of the "output" ofmedical personnel. In recent years two distinctapproaches have developed toward what physi-cians and other health manpower actually "pro-duce." One approach defines service output asthe total health profile of the consumers of suchservices. To this end researchers have used mor-tality and morbidity rates. The other approachconsiders that "Inputs" to maintaining the healthof a population (e.g., visits to the doctor, patient-days in hospital, or specific detailed health ser-vice activities) are actually the output of healthservice personnel.

What galn in physician productivity occurswhen he uses assistants? One study concludedthat:

The average American physician could profit-ably employ roughly twice the number ofaides he currently employs and thus in-crease his hourly rate of output by about 25percent above Its current level. This figuretakes on added meaning when it is recalledthat a mere increase of four percent in aver.age physician productivity In the UnitedStates would add more to the aggregatesupply of medical services than would theentire current graduating class from Ameri-can medical schools.'

The estimation of the supply of services avail-able from a given manpower supply requiresexplicit specification as to the deHvery system forhealth care in the area. A given number of physi-cians, nurses, and technicians will supply onequantity and mix of services through a deliverysystem based on physicians in solo practice; anda quite different one through a system based onclinics, HMO's and hospital outpatient services.Even if the quality of care is assumed to be thesame, there are differences in the utilization andproductivity of manpower, as well as of capital.

7. See PL 92-803, sec. 241, entitled "Program for DeterminingQualifications for Certain Health Care Personnel."

7 "I

In the determination of manpower require-ments and in the evaluation of manpower supply,productivity is too important a consideration tobe ignored. Faced by a lack of data, the healthplanner must rely on his best judgment to developreasonable assumptions In his projections.

Existing credentialing practices have been QualificationIdentified as a major obstacle to Increasing pro- of Manpowerductivity in the health occupation's. One criticismof present credentialing practices is that theyprescribe qualifications beyond those required forsatisfactory and reliable performance, thus artifi-cially constraining the supply of manpower andthe efficient organization and operation of medi-cal delivery systems. Still another criticism is thatcredentialing today constrains the developmentof new occupations. For example, the functions ofthe family nurse practitioner and the pediatricnurse practitioner may involve functions re-stricted by law to the physician. Many doctors aredeterred from delegating duties to subordinatesbecause of the legal limitations on their scope ofpractice. The Federal Government has acted tolower the credentialing barriers by supporting thedevelopment of equivalency tests to permitpersons with knowledge acquired on the Job orthrough nonacademic routes to challenge courserequirements and receive advanced standing, andthus reduce the time needed to achieve academiccredentials. The Government is also funding thedevelopment of proficiency examinations, en-abling persons who have acquired their skillsthrough nontraditional routes to demonstratetheir ability to do the work and thus be creden-tlaled. a The area health planner may attack thecredentialing constraints on supply by working tochange licensing laws and by encouraging em-ployers to reassess their hiring standards and touse equivalency and proficiency examinations.

7 88. Ibid.

69

70

Technologicalinnovation and

New Modesof Delivery

Productivity increases may stem from techno-logical advances as well as from the realignmentof tasks. In fact, these two forces often go hand inhand. The revolution in the medical laboratory, forexample, has been wrought by automatic equip-ment and computers in the hands of such spe-cialized personnel as cytotechnologists andhematologists. The extraordinary rise in thenumber of laboratory tests performed has onlybeen possible through the downward transfer offunctions from the pathologist to the medicallaboratory technologist to the technician to theaide. Until recently, it was feared that an inade-quate supply of laboratory personnel would limitthe potential contribution of pathology to medicalcare. Similarly, the revolution in communicationstechnology has made possible the great advancesin emergency medical services by enabling alliedhealth personnel, such as emergency medicaltechnicians and physician assistants, to functionunder the supervision of a physician, althoughphysically remote. Designers of communicationsatellites envision their use for medical consulta-tions to remote areas via two-way, color, satellitetransmission.

Technological improvements impact on require-ments as well as on supply. In the recent past, wehave experienced, in part as a result of techno-logical and biomedical advances, an explosion ofexpectations which have increased the demandfor health services and compelled changes in thedelivery system. New patterns of care broughtabout by technology and research call for newconcepts in health personnel; new occupationsand changes in existing occupations through in-creased specialization are probably the mostdirect manpower effect of technological ad-vances. 9

9. U.S. Department of Health, Education, and Welfare, Tech .nology and Health Care Systems In he 1980's. DHEW Pub. No.(HSM) 73.3016. Washington, D.C.: U.S. Government PrintingOf lice, 1972,

71

Projections of future supply and requirementscannot ignore the inevitable changes that willoccur in health care demands and manpower pro-ductivity due to new technology. Assumptionswith regard to its impact should be formulated; aset of assumptions high, medium, low impacts

would provide more useful guidelines than apoint estimate.

In recent decades, new modes of delivery haveemerged in the fac;9 of rising demands and risingcosts. The most taked about is the health main-tenance organization, which, through economiesof organization and scale, both encourages andfacilitates combinations of medical and alliedhealth personnel not feasible in the typicaldoctor's office; for example, the use of nurse prac-titioners acting, under the supervision of physi-cians, as primary care providers in an HMO clin-lc.10 Similarly hospital delivery systems are be-ing modified in ways that alter staffing patternsand open new potentials for paramedical person-nel. Outpatient departments and emergencyrooms are being restructured and reorganized asprimary providers. These modes of delivery lendthemselves to team approaches with consequentopportunities for increased use of allied healthpersonnel. The health planner must be cognizantof the changing pattern of health delivery andmanpowar utilization evolving in his area andmust estimate manpower requirements for thatdeveloping system with its staffing patterns.

One study of the impact of HMO's .on futuremanpower requirements found that fewer physi-cians and nurses, but more eye-care workers,dentists, health administrators and some allied

10, U.S. Department of Health, Education, and Welfare,National Institutes of Health, HM0'.:; Their Potential impacton Health Manpower Requirements, prepared by Swift, JohnL., Montalvo, Ramiro A. and Ward, John R. GEOMET Report No,HF.239. Washington, D.C.: U.S. Government Printing Of fice,1973.

80

72

health workers, would be needed. " The method-ological approach was a constt utilization ratewith a changing population model. Briefly, thesteps followed were:

1. Manpower requirements were estimatedseparately for two types of HMO's group prac-tice HMO's and individual practice HMO's andfor all other organizations.

2. It was assumed that future manpower re-quirements in all segments were a function ofhow manpower is used (staff utilization ratio) andthe total health care services demanded in thefuture.

3. The staff utilization ratio in HMO's is theratio of the base-year manpower requirements andthe base-year number of patir -It visits (visits perunit of manpower), information obtained from asample of existing HMO's.

4. The manpower requirements for HMO's werederived directly, and all other manpower require-ments were derived by the residual method.

5. Three scenarios were assumed, based on thelevel of Federal involvement in the spread ofHMO's minimal, moderate and substantial. En-rollment growth is estimated at each level ofFederal support and is applied to the data on utili-zation rates obtained from a sample of existingHMO's to project the future level of patient visits.

6. The staff utilization ratio is applied to thefuture level of patient visits to estimate futuremar.power requirements of HMO's.

Al! ty:lalth planners must take cognizance of theimpact c,f the National Health Insurance programthat :ippears certain to be enacted in some form inthe next few years. The lowering of the financialbarriers to care will inevitably shift the demand formecIical services and the demand for manpowerupward, but not necessarily equally across theboard. The differential impact on services and onmanpower will reflect the coverage, beneficiary

11. Ibid.81

contributions, and provider reimbursement for-mulas that are adopted, on the one hand, and theprice elasticity of demand, on the other. In theface of the unceitainties of the kind of NationalHealth Insurance program that will finally emergeand of the changes in utilization as consumersreact to lowered medical prices, the conceptualproblems the planner faces in adjusting his man-power estimates for the impact of National HealthInsurance are very great. However, one study ofthe impact of National Health Insurance on futurehealth manpower requirements applied an effec-tive demand methodological approach, as-suming constant utilization rates and changingpopulation and income.12 Three different demandmodels were structured. Several types of demandcurves, alternative nationai health plans, andseveral levels of price elasticity were assumed.Changes in the utilization rate without NationalHealth Insurance and with the archetypal healthinsurance plans were determined. Demand shiftfactors were calculated and applied to target- yearmanpower requirements, in order to estimate theimpact of the alternative health insuranceschemes on manpower. The study found thatthere is no conceivable, practical way to producethe number of physicians required by the likelyexpanded demand for health services if presentmanpower utilization patterns continue to be fol-lowed.

The new manpower issues that need to be ad-dressed in the 1970's maldistribution, utiliza-tion, National Health Insurance, and so on addurgency to the health planner's task. The impor-tance of accurate and timely area statistics onhealth manpower supply and requirements is

12. U.S. Department of Health, Education, and Welfare,Bureau of Health Resources Development, Assessment andEvaluation of Archetypal National Health Insurance Plans onU.S. Hea Manpower Requirements, prepared by Huang,Llen-f u, and Shorno, Elwood W. (Robert R. Nathan Asso-iates,Inc.1. DREW Pub. NO. (HRA) 751. Rockville, Md.: U.S. 1;(..part-ment of Health, Educat ion, and Welfare, 1974.

73

74

und-ricor.2- by Public Law 93-641, the NaLionalHealth Plarining and Resources Development Actof 1974. It provides that national guidelines "re-flecting the appropriate supply distribution andorganization of health resources" be set by1976. Health planners will need to quantify themanpower supply and requirements of their areain order to establish local priorities in accordancewith the national guidAnes and to resolve thedemand-supply imbalances that exist or are de-veloping.

MJ.S. GOVERNMENT PRINTING O.FICE: 1976 624-438/ Z 64 1-3

8 3

13. P.1. 93-641, Sec. 1501 (b) (1).