benefits, burdens, and quality of postharvest technology and marketing economics research

7/29/2019 Benefits, Burdens, and Quality of Postharvest Technology and Marketing Economics Research

1/24

Chapter 4

Benefits, Burdens, and Quality

of Postharvest Technology and

Marketing Economics Research


2/24

Contents

Page

Benefits From PHTME Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42Reduced Food Processin g an d Marke tin g Cost s . . . . . . . . . . . . . . . . . . . . . . . . . . 42

Enhanced Quality, Nutritional Value, and Safety of Food . . . . . . . . . . . . . . . . . 46New or Improved Food Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

Reduction of Food Losses Due to Waste . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47In formation for Decision makin g . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

Beneficiaries of PHTME Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48Farm Producers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48Con su mers and General Econ om y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48Marketing Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50Labor an d Oth er Input Su ppliers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

Funding and Flow of Benefi ts From PHTME Research . . . . . . . . . . . . . . . . . . . . . . 50Private Sector Funding Rela ted to Flow of Benefits . . . . . . . . . . . . . . . . . . . . . . . 50State Govern men t Fun ding Related to Flow of Benefits . . . . . . . . . . . . . . . . . . . 51Federal Government Funding Related to Flow of Benefits . . . . . . . . . . . . . . . . . 52

Burdens of PHTME Research. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54Environmental Contamination of Food . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55Disp lacemen t of Labor/Sh ifts in Employm en t . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56Burdens on Consumers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

Quality of PHTME Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57Evaluation of PHTME Research Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57Peer Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

Principal Findings . . . . . . . . . . . . . . . . . . . . . q. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

Chapter p references . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

List of Tables

Table No.$. Relationship Between Food and Agricultural Research Benefits and

Family Income . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499. Distribution of Benefits From and Funding for Production-Oriented

Agr ic u l tu r a lResearch and Extension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5210. Distribution of Benefits From Production-Oriented Agricultural Research

and Extens ion Inves tmen t in Regions of th e Unit ed States, 1976-88 . . . . . . . 53

List of Figure

Figure No.24.Estimated Labor Productivity Growth Rate in the Food Processing Sector,

1954-72 and 1973-79 . . . . . . . q . , , . . , . , . . . . , . . q .+. . . . . . . . . . 4325. Estimated Labor Productivity Growth Rates in the Food Distribution

Sector, 1954-72 and 1973-79. . . . . . . . . . . . . . . . . . . . . . . . . . +... . . . . . . . . . 4426. Real Price of Food, 1947-80 ..,,.. q .....* .*..*.. . . . . . . . . . . . . . . . . 4527. Regional Distribution of Benefits Resulting From an Increase in

. . .

Agricultu ral Research and Extension Expenditures.......,......,........ 55


3/24

Chapter 4

Benefits, Burdens, and Quality

of Postharvest Technology andMarketing Economics Research*

The rationale for public funding of postharvesttechnology and marketing economics (PHTME)research, Federal fun ding in par ticular, is that suchresearch yields benefits to society that w ould n otbe realized in the absence of pu blic funding, andfurthermore, that these societal benefits are wellin excess of societal costs. This chapter discussesthe nature of the benefits from PHTME researchto farmers, marketing firms, consumers, andothers.

The benefits of PHTME research to society areyielded over time and , an app ropriate timeframeis required to evaluate su ch benefits. Time lagsoccur that can slow the benefits from reachingsociety (7). Such lag period s may includ e in part :

q

q

q

the length of the research itself,the time between when research results be-come available and the time when societystarts to use them, andthe tim e it takes for a m ajor sector of societyto use or adap t the research to their needs.

The benefits society receives from increasedprod uctivity resulting from PH TME research fol-lows a bell-shaped curv e overtime, one similarto production-oriented agricultural research whereincreased productivity reaches its maximum dur-ing the sixth year (7).

Moreover, many of the benefits of food andagricultura l research accrue to p arties (e. g., farmprodu cers or consumers) other than the p artiesfunding or conducting the research. Improve-men ts in storage, processing, retailing, and tran s-

q The material in this section draws heavily on Nature and Flow

of Benefits From Ag-Food Research, prepared by Fred C. White,B. R. Eddleman, and J. C. Purcell; and An Evaluation of Methodsfor Examining the Quality of Agricultural Research, prepared byRobert E. Evenson an d Bryan D. Wright in An As s es s m en t of th eUnited S tates Food and Agriculural Research Sy stem , Vol. II, PartC (Washington, D. C.: Office of Technology Assessment, U.S. Con-gress, April 1982).

portations systems, for example, can benefit farmproducers by allowing them to sell their productsin off seasons or in nonproducing areas of thecountry. Such improvements similarly can benefitconsumers by increasing food availability andlowering food costs if such benefits are passedalong. Benefits from activities of a private or pub-lic organ ization for which the orga nization is notcompensated are term ed positive externalities orspillovers. The association of p ositive externalities

with PHTME research has significant implicationsfor the fun ding of such research in both the privateand the public sectors.

Because they are generally motivated by profit,private firms have an incentive to invest in thosetypes of PHTME research for which they antici-pate reap ing a sufficiently large return on theirinvestment. Factors that help a firm capture bene-fits from its own research include patentability,patent enforceability, and obstacles to the imita-tion of such research (26). These factors ar e p res-ent, in the area of mechanization research, for ex-

ample, a research field traditionally conducted bythe p rivate sector.

In some cases, however, private firms have littleincentive to invest in research because they antici-pate being u nable to capture a large share of thebenefits from their own research. It may be thatmost of the benefits will accrue to p roducers, otherfirms, or consum ers. This maybe a problem evenwith patent enforceability, because of factors suchas imperfect consumer know ledge, prod uct emula-tion, and the p rohibitive costs of collecting ben e-fits from research. From a societal perspective the

research m ay be of considerable imp ortance, butif a firm is una ble to reap th e benefits it w ill hav elittle incentive to conduct the research. Thus, tohelp ensure the optimum level of societal invest-ment in suchbe needed.

research, pu blic sector sup port m ay

41

9 8 - 9 5 2 - - 4


4/24

42

However, the problem of spillovers, as dis-cussed further in this chapter, can also arise withinthe public sector. Benefits resulting from the re-search of State agricultural experiment stations(SAES), for example, frequently are reaped notonly by producers in the State in which the re-

search is conducted bu t also by produ cers in otherStates (25). These benefits similarly maybe reapednot only by consumers in the State wh ere the re-search is conducted but by consumers in otherStates, as well. From the standpoint of equity, itmay be desirable to provide Federal fund ing forsuch research.

Overall, a high degree of un certainty abou t thepossible outcomes of research increases the likeli-hood of underinvestment in research relative tothe optimum level for society (36). Some food and

agricultural research projects und ertaken d o notturn out to be economically successful. A sm allfirm that is able to pur sue on ly a few pr ojects facesa high probability that none of its projects willbe successful. Historically, however, when all suc-cessful and unsuccessful projects in food and agri-

cultural research are considered, the average orexpected rate of return on inv estment has beenfavorable (9). In a large firm or in a large aggre-gate such as the United States, the pr odu ctivityof food an d agricultural research can be viewedin terms of a framework of risk rather than of pureuncertainty.

The following sections of this chapter examinethe benefits, burdens, and quality of PHTME re-search. Chap ter 5 further d elineates the roles ofpu blic and private research participan ts.

BENEFITS FROM PHTME RESEARCH

PHTME research is designed to provide prod -ucts where and when needed, in the form desired,and with maximum economic efficiency. PHTMEresearch can make more food available by im-proving processing, upgrading products, prevent-ing waste, and p roviding for use of produ cts pre-viously not considered usable. Such research alsocan reduce marketing costs by improving efficien-cy in storage and transp ortation of food and by

imp roving efficiency of use of resour ces.The primary aims of PHTME research are to:

1) increase prod uctivity in the food pr ocessing an dmarketing sectors and reduce the real cost of food;2) maintain or enhance the nutritional value, qual-ity, and safety of food; 3) develop new or im-proved food prod ucts; and 4) prov ide informa-tion that policymakers and others can use indecisionmaking.

Reduced Food Processing and

Marketing Costs

An estimated 70 percent of food costs to con-sumers is attributable to assembling, processing,transporting, and distributing food; the remain-ing 30 percent goes to farmers an d their supp liers(see also ch. 2). Furthermore, rising marketing

costs are the m ain cause of rising food costs. Byleading to increases in labor productivity andother developments, PHTME research can reducethe costs of processing a nd marketing food, andthus it may lower food costs to consumers.

Reducing Costs Through IncreasedLabor Productivity

The largest cost component in the food process-ing and marketing sector by far is direct labor,accounting for 45 percent of total marketing costs,followed by food packaging costs (12 percent) andthe costs of shipping food by rail and truck (8 per-cent). Thus, research that leads to increases inlabor prod uctivit

yoffers the g reatest p otential for

constraining increases in consumer food prices (s).

For most industries that process and d istributefarm products, the rate of labor productivitygrowth (output per em ployee hour) since 1972 hasdeclined in comparison to the rate for previous

years (see ch. 2). Figur e 24 show s trend s in laborprod uctivity growth amon g processors of farmproducts. During the period 1973 to 1979, in-creases in rates of growth of labor prod uctivityin the food processing sector occurr ed in grainmilling and soft drinks; whereas in cereal and


5/24

43

Figure 24. Estimated Labor Productivity Growth Rates (output per employee hour) in the Food ProcessingSector, 1954-72 and 1973-79

SlC Industry Annual Growth Rate Percentcode

-1 -2 0 2 4 6 8

II I I I I

2046 Wet corn milling 1963-72

1973-78

2082 Malt beverages 1954-72

1973-79

2086 Bottled and canned 1958-72soft drinks

1973-79

2044 Rice milling 1963-72

1973-78

2041 Flour and other 1954-72grain mill products

1973-79

2026 Fluid milk

203 Preserved fruits

and vegetables

2061 Raw and refined2062 cane sugar

2063 Beet sugar

2043 Cereal andbreakfast food

2065 Candy - otherconfectionery products

2045 Blended and

prepared flour

1953-72

1973-79

1954-72

1971-78

1958-72

1973-78

1958-72

1973-78

1963-72

1972-78

1964-72

1973-78

1963-72

9.8

1.7

] 2.5

\ 3.8L

J 3 . 5

2.7

3.5

1.5

t3.4

[ 0.6

( 2.2

j 0.8

3.6

1 I 1973-78

5.9

SOURCE Bureau of Labor Statistics, Productivity Indexes for Selected Industries, 1979 Edition, Bul. No. 2054, December 1979 [Tiegen, 1981]


6/24

44

breakfast foods, beet sugar, and candy m anu fac-turing, labor produ ctivity remained almost con-stant and the blended flour industry declined atan annual rate of 4.0 percent. Although the Bu-reau of Labor Statistics d oes not rep ort produ c-tivity in meatpacking, output per employee hour

as computed from the Industrial Production In-dexes of the Federal Reserve Board showed nochange from 1971 to 1978 (28).

All labor p rodu ctivity growth rates in the fooddistribution sector have declined significantlysince 1972 (see fig. 25). In p articular , labor pro-du ctivity in food stores and eating and drinkingestablishments had a negative growth rate. In food

stores, longer shopping hours essentially ac-counted for the same volume of sales and is con-sidered a major cause of decline; automatedcheckout systems have not yet affected labor pro-du ctivity in food retailing. Although fast-foodchains have ma de organ izational and technologi-

cal imp rovements, increases in hou rs w orked an dthe number of small, marginal enterprises haveheld prod uctivity dow n (4).

Lagging labor pr odu ctivity growth r ates in thefood processing and distributing sectors have con-tributed significantly to increased rates of foodpr ice inflation (see ch. 2). In the period 1960 to1965, the annual rate of increase in food prices

Figure 25. Estimated Labor Productivity Growth Rates (output per employee hour) in the Food DistributionSector, 1954-72 and 1973-79

SICcode Industry Annual growth rate percent

2 o 2 4

4213PT Intercity truckinga

Intercity truckinga

4213PT (general freight)

401 Railroad (car miles)

20 5 Bakery products

1954-72

1973-78

1954-72

1973-78

1954-72

1973-79

1959-72

1973-79

54 Retail food stores 1958-72

I 2.6

1.1

2.1

1.4

3.8

2.7

1.0

1.0 1973-79

Eating and

58 drinking places 1958-72 1.2

aOutput per employee.

SOURCE: Bureau of Labor Statistics, Productivity Indexes for Selected Industries, 1979 Edition, Bulletin No. 2054, December 1979 (Teigen, 1981).


7/24

45

was 1.1 percent; in the period 1966 to 1972, theannual rate of increase in food prices rose to 3.9percent, and in 1973, it reached 7.7 percent. Al-though annual wage gains during the period 1950to 1965 (3.9 percent) and 1966 to 1972 (7.5 per-cent) outpa ced the an nua l food price gains, they

barely kept pace during the period 1973 to 1979(7.5 per cent). The r eal pr ice of food in the per iod1973 to 1980, in terms of hourly earnings, wasnot significantly different from 1967 level (see fig.26).

Reducing Costs Through Improvementsin Processing and Preservation

Many food products are wasted because of im-proper methods of processing and preservation.In some cases, a particular treatment can m akefood prod ucts available that would otherw ise be

wasted. Research to develop treatments thatwould be economically feasible to use is currentlybeing conducted (33). Reducing food losses dueto w astes could r esult in cost savings to consum-

ers. In the area of preservation, the products safe-ty and other factors mu st be taken into considera-tion. In addition, the food must be acceptable toconsumers.

Both better use of food materials through im-provem ents in processing and p reservation, andresearch that red uces energy and wa ter use forpr ocessing m ay lower food costs. Given the pres-ent level of resource use, such cost savings couldbe substan tial. For exam ple, drying pr ocesses areenergy-intensive and could be m ade m ore effi-cient. Fabricated foods can lower food costs bysubstituting cheaper grain ingredients for more ex-pensive livestock ingred ients. Improved know l-edge of physical and chemical properties of foodmay result in savings related to food storage,processing, and handling.

Reducing Costs Throu gh EfficientMarketing and Distribution

In the last 30 years, changes in th e m arketingand d istribution of food h ave been significant, as

Figure 26. Real Price of Fooda, 1947-80

160 -

150 -

140 -

-

80 -

o 1 I I 1 I I1950 1960 1970 1980

YearaConsumer price index for food relative to private nonagricultural hourly earnings, adjusted for overtime and interindustryshifts

SOURCE. Lloyd D Teigen, Productivity A Food and Agriculture Perspective, mimeo draft, U.S. Department of Agriculture,Economic Research Service, August 1981


8/24

46

evidenced by the expansion of supermarkets.Supermarkets have reduced the retail cost of foodto consumers by some 15 to 25 percent (15). Thesecost redu ctions w ere achieved through labor re-du ctions m ade possible throu gh self-service an dlarge-volume operations in transportation, stor-

age, and d istribution.Thus, research designed to improve efficiency

in food m arketing and d istribution can be usedto lower marketing costs. Improved coordinationof marketing activities, in the p rivate and pu blicsectors, is needed to move food more efficientlythrough marketing and distribution channels.

Enhanced Quality, Nutritional Value,and Safety of Food

The quality of food prod ucts prima rily refers

to such products esthetic characteristics, but mayalso refer to their n utrient value, wh ether natu ral-ly occurring or introduced by p rocessing an d pres-ervation technology. Operations relating to stor-age, handling, shipping, intermediate processing,packaging, delivery to merchants, shelf life, andfinal sale of food prod ucts can materially influencesuch products quality, nutritional value, andsafety.

Some n utrients, notably vitamins and fats butproteins and carbohydrates as w ell, are sensitiveto pH, oxygen, heat, and light, and are particular-ly susceptible to dam age in the p resence of cer-tain trace elements. PHTME research can helpminimize or negate the influence of these factorsand the effects of environm ental conditions ondevelopment of mycotoxins (naturally occurringtoxic contaminants in food produced by molds)and the effects of insect and rodent infestation onnu trient content and safety.

Fortification and enrichment of foods are meansto imp rove nu tritional value of food. Fortifica-tion and enrichm ent of foods generally increasethe foods retail cost, and this increase must beweighed against benefits derived from the pra c-

tice. Actions that ha ve help ed im pr ove U.S. dietsinclud e the ad dition of vitamins Bl, B2, and n iacinto cereal prod ucts, vitamin D to milk, iodine tosalt, and fluoride to drinking water, Iron fortifica-tion of foods is another examp le, although som e

controversy exists regard ing the benefits of thispractice.

Consu mer s almost exclusively benefit from r e-search that: 1) increases the supply of nutrients;2) determines the nutritional requirements andfood consump tion p ractices of various consum er

groups; 3) increases the nutritional value of foods;and 4) redu ces the potential hazard s of naturallyoccurring toxins, food preservatives, or other ad-ditives used in p rocessing, insect and rodent con-tamination, residual p esticides u sed in prod uction,and other inadvertent contaminants. Producersmay benefit from, be unaffected by, or be dis-adv antaged by such research (36).

In some cases, food p rodu cts may contributeto nu tritional problems. Examp les of such p rod-ucts are the high-calorie, low-nutrient snack foodsand food su bstitutes that are of lower n utritional

value than the foods they replace. There is a needto consider n utritional value as an integral partof food qu ality, since an increasing prop ortion ofthe food sup ply is modified by p ostharvest proc-essing.

New or Improved Food Products

PHTME research can contribute to develop-ment of new or improved food products that di-rectly benefit consumers. Such developments in-clude food products which substitute vegetablefor animal protein and use of other alternative

food sources.The importance of plant protein in the d iets of

people in industrialized countries such as theUnited States is expected to increase, relative tothe d emand for m eat, fish, and egg protein. Plantprotein prod ucts can be introdu ced into the dieteither as food ingred ients (e.g., textured soy flouras an extender added to ground beef) or as ersatzfoods (e.g., simu lated chicken or ham). Foodblends have been developed through researchfrom a variety of cereals and oilseeds, Lipid pro-tein concentrate d erived from soybeans can be

used as a beverage base. Raw soybeans can beused to pr odu ce soy flours, protein concentrates,soybean isolates, or textur ed p roducts.

Other possible alternative food sources include:1) fats and oils from controlled animal feeding and


9/24

47

breeding , controlled fish farming, single-cell culti-vation (yeast/ fungi), selective plant p ropagation,and synthesis from petroleum and enzymic proc-esses; 2) meat or simulated m eat prod ucts fromnew animal species and vegetable and microbialsources; and 3) carbohydrates such as glucose

syrup s from cellulose, chemically synthesizedsugar, or low energy substitutes replacing starch(13). Research on new or improved food prod uctsmay have the greatest p otential for p roviding foodnutrients to the world population at minimum orreasonable food costs (36).

Reduction of Food Losses Due to Waste

The National Academy of Sciences has esti-mated that research on reducing food losses du eto pests could save $1.5 billion ann ually in theUnited States (19). Research on reducing thestorage and transportation losses of fruits andvegetables, it estimated, could increase the supplyof these p rod ucts from 15 to 30 percent (19). Thebenefits from basic research in ar eas such as estab-lishing physiological and biological interactionsand analyzing metabolic processes of micro-orga-nisms, although more difficult to quantify, wouldvery likely include the prevention of food lossesand improved food quality.

As discussed earlier, many food products arewasted because of improp er method s of process-

ing and preservation. Particular treatments canmake food products available that would other-wise be wasted. Research to develop treatmentsthat would be economically feasible to use is cur-rently being cond ucted (33), Redu cing su ch foodlosses could result in cost savings to consumers.

Information for Decisionmaking

PHTME research can provide information onmarkets, prices, and government regulations thatfarmers, processors, distributors, consu mer s, and

policy m akers can u se in m aking d ecisions. Re-search on industry competitiveness can provideinformation to produ cer group s and th e pu blic onthe forces that are shaping this industry, informa-tion to assist p rodu cers in long-range p lanning,and information to p olicy makers on alternative

legislative proposals in relation to the projectedchange in the indu stry.

PHTME research that provides informationabout current and future market conditions, in-clud ing futu re supp ly and dem and , for example,can be u seful to individu al firm s. This is the case

in the area of grain marketing. Food an d indu strialuses of grain and grain pr odu cts are increasingrapidly, and the composition of products pro-du ced by U .S. grain processors has changed, re-flecting shifts in consumer preferences. In orderfor ind ividu al firms in the grain marketing indu s-try to m ake effective an d efficient ad justm ents,they may need information about how changesin economic and institutional factors affect the de-mand for grain and grain prod ucts (33).

PHTME research th at assesses the impact ofgovernmen t regulations can p rovide valuable in-

formation for policymakers. Regulations intendedto improve industry performance, for example,in the red meat indu stry, may sometimes have un-intentional and costly side effects, and the benefitsderived from the regulations may sometimes beeroded. This situation leads to n um erous ineffi-ciencies that a ffect the ab ility of firms to compete,adopt new technologies, and experience growthin outp ut and sales. Research th at assesses the im-pact of regulations provides policymakers withthe necessary information to assess the benefitsand costs of the regu lations.

PHTME research that measures the competitiverelationships among firms p rodu cing a similar setof produ cts considers both how the organizationof the industry affects its behavior and perform-ance and what determines how an industry isorganized and how it changes. Imp ortant aspectsof this research include the degree of market con-centration, barriers to entry, the types of compet-ing organizations, and regulations that affect com-petitive behavior. Such research can be useful toindustry, consumers, and policy makers.

Recent studies ind icate that increasing concen-

tration in the food processing and retailing subsec-tors may increase real food p rices to consumers.Parker and Conner (24) found that the following

monopolistic trends in the food manufacturing

subsector, which were identified by the National

Commission on Food Marketing in 1966, have


10/24

4 8

continued: 1) a decline in company numbers;2) an increase in concentration; 3) a substantialincrease in the conglomerate nature of leadingfood man ufacturing firms; 4) an increase in thenumber of large acquisitions by the larger com-panies; 5) substantial increases in product differen-tiation expenditu res by large food m anufacturers;

and 6) a growing differential between profitabilityof large versus medium and small food manufac-turers. The consumer loss attributable to thesemonopolistic characteristics was estimated to beapproximately $1 2 billion to $14 billion annually,or as mu ch as 25 percent of the value ad ded tofood produ cts by food m anu facturers (24). *

*This conclusion was rebutted by ORourke and Greig (23) andBullock (l).

Net profits and food prices of large food chainswere positively and significantly related to marketconcentration and a cha ins relative mar ket share(17).** Marions find ings tend to r efute the no-tion that higher profits for dominant firms in con-centrated markets are due to such firms efficiencyand lower costs. Increased profits by firms in non-

competitively structured markets account forabout one-third of the increases in food prices(17). Thus, policy-oriented PHTME research di-rected toward improving competition in the foodprocessing and retailing su bsectors p otentiallycould result in substantial benefits (i.e., lowercosts) to consumers and farm prod ucers.

BENEFICIARIES OF PHTME RESEARCH

** This study used cross-sectional data, not time series data. Thus,efficiency and costs could not be accounted for over time.

Analysis of the flow of benefits from food a ndagricultural research including PHTME researchfocuses primarily on the distribution of benefitsbetween domestic consumers and producers. Theanalytical fram ewor k is the concept of economicsurplus an d the partitioning of this surplus intothe portion accruing to buyers (i.e., consumer sur-plus) an d the p ortion accruing to sellers (i.e., pro-ducer su rp lus) (2,38).

Farm Producers

In the United States, the demand for food isonly slightly respon sive to changes in the p riceof food and rising personal income has almost noeffect on per capita food consumption. Thus, thedom estic deman d for food increases at about thesame rate as the popu lation growth. In this situa-tion, technology th at increases outp ut in excessof that needed to meet deman d grow th results indepressed prices.

The food pr ocessing and m arketing system en-hances the value of farm commodities by chang-ing their form an d distributing the prod ucts over

time and space. For examp le, few households bu ywheat, but some buy flour and most buy bread.The demand for wheat is largely derived from thedemand for flour, and the d emand for flour islargely derived from th e demand for bread.

PHTME research may lead to an improvementin the food pr ocessing sector which red uces coststo that sector; the redu ction in costs to the foodprocessing sector may or may n ot be passed onto consumers. An improvement, for example, intransportation or distribution technology in thefood marketing sector might red uce costs to farm-ers and could increase their income.

Consumers and General Economy

Benefits to consumers from PHTME research

may includ e increased availability of food andredu ced retail food costs. A technological changethat reduces the costs of marketing services mayredu ce the retail price to consum ers. The farmprice could be expected to increase, but the reta ilprice would be expected to d ecline with red ucedmarketing margins. *

Consumers (and farmers, too) also may benefitfrom research to improve market or price infor-

These relationships hold for a competitive market structure inwhich changes in margins are reflected throughout the marketingsystem. Both the derived su pp ly at retail, which is dep enden t on

farm supply and the marketing margin, and the derived demandat the farm level, which is dependent on the retail demand and themarketing margin, would shift in a competitive market as a resultof a reduction in marketing costs. The impacts on consumers of atechnological change in the marketing sector are dependent on theprice elasticities of supply and demand and the magnitude of costsavings resultin g from th e technological change.


11/24

49

mation. If producers incorrectly estimate the priceof a commodity above the market equilibriumprice, they will produce a qu antity larger than themarket equilibrium quantity. The resulting price,how ever, will be lower than p rodu cers expectedand result in a net loss to society. If marketing

economics research such as econometric modelinglead to a price forecast close to the market equi-librium, this n et social loss w ill be minimized (22).

Consumers also may benefit from improvedquality, nutrition, safety, and convenience offoods. Some of these benefits seem to be concen-trated among certain groups of families. It is esti-mated th at the diets of20 percent of the house-holds in the Un ited States provide less than tw o-thirds of the recomm ended daily allowance of oneor more nutrients (15), Malnutrition is more evi-den t amon g low-income families, and relation-

ships between diet and chronic illnesses have beenidentified (30). PHTME research that improvesthe quality and safety of food products is likelyto affect consumers in all income categories bene-ficially.

From the stand point of equity, the distribution-al impacts of agricultural researchincludingPHTME research-are impor tant. In the analysisthat follows, the distribution of benefits from agri-cultura l research is estimated on the basis of con-sum er food expend itures. Each d ollar of food ex-

pend iture is assumed to be related to the sameamou nt of research benefits, whether the expend i-tur e is made by high- or low-income families.

The relationship between family income andagricultural research ben efits for average size fam-ilies in six family income categories are g iven intable 8. Family income ra nges from un der $5,000in the lowest class to ov er $20,000 in th e highest;average family size ranges from 2.93 persons inthe lowest income class to 3.79 in the highest. Thepresent value of average benefits per family forthe various income classes show n in table 8m aybe interpr eted as th e benefits accru ing to eachfamily as a result of agricultural research expendi-tures in that year. * Average benefits per family,which increase with the level of family income,ran ge from $16.20 in th e lowest incom e categoryto $30.74 in the highest.

The ratio of benefits to family income is almostfour times higher for the lowest income class thanfor the highest, indicating that food and agricul-tural research has a greater beneficial impact onlow-income families than on high-income familiesin relation to family income. For some commod-ities, the distribution of consumer benefits fromresearch t ends to be biased in favor of families

q Even though benefits would be realized through time, this streamof benefits was discounted with a 10-percent discount rate to findthe present value of benefits (8).

Table 8. Relationship Between Food and AgriculturalResearch Benefits and Family Income

AverageFamily Distribution Average Average benefitsincome of size family per family

c

class population a familyb income (present value)

Under $5,000 . . . . . . . . . . . . . 18.190/o 2.93 $3,981

$5,000 -$8,000 . . . . . . . . . . . . .

$16.2014.14 3.15 7,922 19.06

$8,000-$ 12,000 . . . . . . . . . . . . 21.17 3.28 10,528 20.13$12,000-$ 15,000 . . . . . . . . . . . 14.47 3.48 13,458 22.63

$15,000-$20,000 . . . . . . . . . . . 16.07 3.68 17,371 25.91Over $20,000 . . . . . . . . . . . . . 15.96 3.79 28,953 30.74aAnthony E. Gallo and William T. Boehm, Food Expenditures by Income Group, National Food Review, NFT-3, USDA, ESCS,

Washington, D. C., June 1978.bU.S. Department of Commerce, Bureau of the Census, Current Population Reports, Series P-60, No. 101, Money Income in

1974 of Families and Persons in the U. S. (Washington, D.C.: U.S. Government Printing Office, 1976).c

Total consumer benefits are calculated according to th e equation:T BC

= X MVP R X RE X D

where TBCis total consumer benefits from ag-food research; MVP

Ris marginal value product of research (Davis); RE is pro-

duction oriented research expenditures in 1974 (Budget of the US Government; USDA, Inventory of Agricultural Research,U.S. Department of the Treasury); and D is the discount factor over 13 years at 10 percent (Lu, Cline, and Quance) Total con-sumer benefits are allocated to income classes according to the level of food expenditures.

SOURCE: Fred C. White, B. R., Eddleman, and J. C Purcell, Nature and Flow of Benefits From Agriculture-Food Research,in AnAs s es s m en t of th e Un ite d States Food and Agricultural Research System, Vol. 11, Part C (Washington, D.C.Office of Technology Assessment, U.S. Congress, April 1982).


12/24

50

w ith higher income. The larger the quan tity con-sum ed in the higher income categories relative tothe lower income categories (e.g., pork, poultry,beef, and dairy p rodu cts), the m ore favorable thedistribution of benefits to higher income families.

Marketing Firms

Competition am ong m arketing firms results inlower costs of marketing services being passed onto consumers and prod ucers. As indicated earlierin this chapter, however, there is evidence thatthe food marketing system does not conform toa comp etitive market, par ticularly at the regionaland local market level (17,24).

Marketing firms w ith some m onopolistic powermay r etain a par t of the cost savings from tech-

nological change in the form of increased profits(10). Monopolies, because of profits captured anddesire to retain monopolistic control, maybe will-ing to invest in p rivate research.

The more comp etitive the ind ustr y, the less in-centive there is for private research, because thebenefits accrue to consum ers and farmers. In aless competitive ind ustry, how ever, private re-search is more profitable for the individual firm,and private research may reduce the level of com-petition.

Public research may enhan ce comp etition orredu ce market power in a particular subsector.The benefits from pu blic food and agricultu ralresearch in a highly competitive marketing sec-tor accrue to consumers and to farmers.

Labor and Other Input Suppliers

Technology that changes the relative productiv-ity of resources shifts the distribution of incomeamong resources (12). These changes have re-duced the proportion of total food and agricul-

tural income attributed to labor and increased theproportion attributed to capital.

Changes in the p roportion of income attribut-able to labor dep ends on the type of technology.Labor-saving technology redu ces the dem and forlabor. Mechanical technology th at has been de-veloped almost entirely by the private sector gen-erally can be characterized as labor-saving (6).

Research that develops m ore prod uctive post-harvest technologies usually makes possible in-creased labor productivity and provides opportu-nity for increasing wages and salaries without nec-

essarily placing an upward pressure on retailprices. The process may displace part of the laborforce thus requiring them to find other employ-ment.

Research that prov ides for m ore efficient u seof energy and capital provides more residual cap-ital for increased wages and salaries in real terms.Conversely, rising prices of energy and investmentfunds (through rising interest rates) place d own -ward pressures on salaries and wages, upwardpressures on retail prices, and dow nward pres-sures on farm prices. These are the conditions that

dom inated du ring most of the past d ecade. Suchpressures as rising energy costs can be mitigatedby improved efficiencies from PHTME research.Increases in wages an d salaries, without compen-sating increases in labor efficiency, result d irectlyin inflation.

FUNDING AND FLOW OF BENEFITS FROM PHTME RESEARCH

From th e standp oint of equity, an important Private Sector Funding Related toconsideration is the extent to w hich the flow of Flow of Benefits

benefits from PH TME research is related to thesources of funding for such research. Ideally, the Who captu res the benefits from pu blic sectorbeneficiaries of research should pay the research and private sector food and agricultural researchcosts. As discussed further below, however, the is a prominent issue in PHTME research. Thebeneficiaries of PHTME research sometimes do question is whether a particular research problemnot pay the research costs. area should be addressed through public research


13/24

51

if the gains from the research are embodied in pri-vate firms prod ucts. In general, there are sp ill-overs or indirect benefits both from public re-search to the private sector and from the privatesector research to society.

If benefits from the results of private sectorPHTME research can be captured by the privatesector, then th ere is an incentive for private firmsto invest in resea rch activities. Research a ctivitiescan be d istinguished in terms of patentability andpaten t enforceability of the prod uct, techniqu e,or pr ocess that results from the r esearch; the eco-nom ic life of the techn ique or process; the tech-nological versus pecuniary effects of the techniqueor process; and the ability of rival companies toinitiate the research and development pr ocess (36).These characteristics determine whether net bene-fits of the research activities can be captured by

the private sector.To the extent that the benefits of research can

be captured by firms in the private sector, public

sector involvement in the types of research activ-

ities from which such benefits flow would be a

form of subsidy to private firms. In some cases,

however, private sector research activities yieldspillover effects and indirect benefits to society.Although no specific case studies have been donefor agricultural input or food-processing indus-tries, stud ies by Mansfield, et al. (16), Terleckyj(29), and Griliches (11) of the distribution of gains

from private research activities in manufacturingand nonm anufacturing ind ustries indicate that thespillover effects are at least as larg e as the d irectbenefits going to firms conducting the research.Thus, the social returns from ind ustry r esearchare roughly double the private investment returns.In this regard, substantial social benefits are de-rived from private industry investments in re-search activities.

State Government Funding Related

to Flow of Benefits

Food and agricultural research financed by oneState m ay benefit the residents of other States.For examp le, a more energy-efficient food p rocess-ing technology developed in one State may be

adopted in neighboring States to increase efficien-cy at the same or higher total outpu t.

State bound aries do not coincide with h omoge-nous agricultural regions. PHTME research proj-ects in on e State that are ad dr essed to sp ecific local

problems likely will produce results applicable toother States. Applied research focused on a spe-cific local problem may be adapted for more gen-eral purposes to help m eet the needs in otherregions. Furthermore, knowledge gained frombasic research is d isseminated without regard togeographic boundaries.

Spillover benefits generated by State A whichaccrue to the residents of State B generally are notaccounted for by State A policymakers. The argu-ment concerning neglect of these externalities hasbeen that State A w ill provide a smaller level of

research expenditures than would be the efficientfrom societys perspective. Given the possibilityof negotiation betw een States, State B may findit advan tageous to p ay A to increase its level ofresearch activities. Such a subsidy will reduce Asresearch costs and lead to a higher level of researchactivities. The negotiation process likely will becomplicated by th e fact that sp illovers flow inboth d irections betw een the two States. Further-more, the outcome will depend on the relative bar-gaining strength of the two States and will notnecessarily lead to an efficient solution to the ben-efit problem (18).

If only a few States have an interest in a p ar-ticular commodity or segment of the marketingsystem, one of the States might cond uct the re-search, with the research effort being supportedby the other States. However, attempting to coor-dinate these activities involves d ecisionm akingcosts that includ e the value of time, effort, anddirect outlays related to the bargaining p rocess.For those cases in which external benefits fromagricultural research affect a large number of deci-sionmaking units, total decisionmaking costs ofeffective coordinated action are likel y to be large.

When the impact on consumers is considered, alarge number of States wou ld be concerned withalmost all aspects of food and agricultural re-search.


14/24

52

Federal Government Funding Related

to Flow of Benefits

Pub lic goods w ith significant externalities maybe classified as either national or quasi-nationalpu blic goods. A national pu blic good that is con-

sumed equally by all residents (e. g., nationaldefense) may be provided more efficiently by theFederal Government than by State or local gov-ernments. Quasi-national public goods, on theother hand, are consumed on a less comprehen-sive basis throughout the Nation.

Agricultural r esearch, and especially PH TMEresearch, serves as an excellent examp le of a quasi-national public good. Financing th is research atthe State level produces benefits that are con-sum ed by the States residents, but also providesbenefits that p ass to other States and n ations in

the form of externalities. While financing can beprod uced at either the Federal or State level ofgovernment, the presence of externalities indicatesa need for the coordination of research amongvarious States.

Partial fund ing by the Federal Government a f-fords one solution to attaining the nationally d e-sired level of research expend itures. An often-usedtechnique to increase State expenditures for gov-ernment services is the matching grant, in whichthe recipient State government is required tomatch Federal funds with funds from its ownsources according to some specified formula.While some Federa l grants to States for food andagricultural research (including PHTME research)require matching fund s, most States invest morein food and agricultural research than just theamount required to match Federal grants, asshown in chapter 3.

The formu la for matching fund ing should bebased on the relative imp ortance of external andinternal benefits. Properly designed Federal grantprograms direct State expenditures toward levelsconsidered optimal from societys perspective(rather than from th e States p erspective) by fi-

nancing the cost of the external benefits.

An app ropriate matching grant program re-quires identifying and quantifying State benefitsand spillovers from agricultural research expendi-tures. An estimated 55 percent of the change in

productivity attributed to technology-oriented re-search conducted by a State is realized w ithin theState conducting the research, whereas the re-m aining 45 percent is realized in oth er States (9).Table 9 shows the estimated spillover benefitsfrom food and agricultural research conducted

from 1949 to 1972 in 10 d ifferent regions of thecountry (37). The ratio of external benefits (i. e.,spillovers to other regions) to internal benefits(i.e., benefits realized within the region in whichthe research w as conducted) range from a low of1.31 to 1 in th e N orthea st region to a h igh of 2.80to 1 in the Southern Plains region. The aggregateratio for all regions is 1.73 to 1.

Table 9 also show s the ra tio of Fed eral to Stateexpenditures for production-oriented food andagricultural research in each of the 10 regions andfor all regions combined. By comparing these

ratios to the spillover ratios, one can determinewhether the Federal Government actually fi-nanced the spillovers. The aggregate ratio of Fed-eral to State expenditures is only 1.38, comparedto aggregate spillover ratio of 1.73.

Table 9.Distribution of Benefits From andFunding for Production-Oriented Agricuiturai

Research and Extension

Ratio of Ratio ofspi llovers Federal to State

to regional expendituresb

Region benefits a 1949-72

Northeast . . . . . . . . . . . . . . . 1.31Lake States . . . . . . . . . . . . . 2.73Corn Belt . . . . . . . . . . . . . . . 2.04Northern Plains . . . . . . . . . . 1.40Appalachian . . . . . . . . . . . . . 1.19Southeast . . . . . . . . . . . . . . . 1.40Delta . . . . . . . . . . . . . . . . . . . 2.48Southern Plains. . . . . . . . . . 2.80Mountain . . . . . . . . . . . . . . . 1.60Pacific. . . . . . . . . . . . . . . . . . 1.89

All regions . . . . . . . . . . . . 1.73

0.971.101.251.631.601.371.802.102.350.901.38

aThe values measure the benefits accruing to farmers outside the region relativeto the benefits accruing to farmers within the region.

bFederal expenditures are not limited to these funds going to the State agricul-

tural experiment stations and cooperative extension services under formula andgrant programs; they also include Federal funding of under formula and grantprograms; they also include Federal funding of production-oriented agriculturalresearch and extension in each region through USDA/ARS, USDA/ERS, and

USDA/SCS, The values measure the level of Federal funding from all thesesources relative to State funding for agricultural research and extension in eachregion during 1949 to 1972. The 1949.72 period was used for estimating the rela-tionship between productivity growth and agricultural research/extension in-vestments, based on a 13-year distributive lag estimation technique.

SOURCE: Fred C. White and Joseph Havlicek, Jr., lnterregional Spillover of Agri-cultural Research Results and Intergovernmental Finance: Some Pre-liminary Results, in Evaluation of AgriculturalRe s ea rch , MinnesotaAgricultural Experiment Station Miscellaneous Publication No, 8, 1981


15/24

53

This analysis suggests that the the Federal Gov-

ernment did not finance the spillover benefits to

the Nation as a whole of production-oriented agri-

cultural research conducted in the various regions.

In order to align aggregate Federal expend itureswith national benefits from prod uction-oriented

agricultural research conducted in the variousregions, Federal expenditures would have to beincreased abou t 25 percent, Several regions wouldrequire a greater percentage increase to yield anequitable distribution among regions.

It could be inferred that a similar analysis forPHTME research might result in the same conclu-sion. The benefits of PHTME research accrue toregions and to the N ation, not to local areas asis characteristic of production-oriented research,so the ratio of spillovers to regional benefits fromPHTME research would likely be at least the

production-oriented research ratio of 1.73 orhigher. As discussed earlier in chapter 3, almost50 percent of the PHTME research is conductedby the States, and over 50 per cent of SAES fund sare from State app ropr iations for this research.Thus, the ratio of Federal expend itures to Stateexpenditures for PHTME research would likelybeat least as low as the 1.38 ratio discussed above,if not lower. Thus, assuming these estimates arerealistic, the Federal Governments contributionto PHTME research w ould h ave to be increased

at least 25 percent to align fund ing with na tionalbenefits.

The benefits from increased productivity offood and agricultural research are divided betweenproducers and consumers. When demand is in-elastic and growing slowly, as in the United Statesdu ring most of the p ast 50 years, a large shareof the gains from innovation are passed on to con-sumers in the form of lower real food prices.When consum er benefits are combined with p ro-ducer benefits, the magnitude of spillovers to re-gional benefits is affected d ram atically, as show nin table 10.

Table 10 shows th e discounted total net benefitsthat accrue to consum ers plus prod ucers per dollarof prod uction-oriented agricultural research and

extension (R&E) investm ent: the intern al benefitsare p resented in the first colum n of the table, andthe spillover benefits are presented in the secondcolumn. The third column of table 10 shows theratio of external benefits to benefits accruing with-in the region, i.e., the spillover ratio per dollarof R&E investm ent, and the final colum n of table10 shows the actual ratio of Federal expendituresto State expenditures for the production-orientedagricultu ral R&E within each region (repeatedfrom table 9).

Table 10.Distribution of Benefits From Production-Oriented Agricultural Researchand Extension Investment in Regions of the United States, 1976=88

Total producer and consumer benefitsper $1 R&E investment

Inside the Outside theRegion region region

Northeast . . . . . . . . . . . $28.39 $13.14Lake States . . . . . . . . . 7.93 36.82Corn Belt . . . . . . . . . . . 5.19 37.95Northern Plains. . . . . . 1.20 47.96Appalachian . . . . . . . . . 8.19 34.01Southeast. . . . . . . . . . . 7.98 34.45Delta States. . . . . . . . . 3.38 39.38Southern Plains . . . . . 8.05 37.99

Mountain . . . . . . . . . . . 2.72 40.35Pacific ... , . . . . . . . . . 7.88 34.76

All regions . . . . . . . . 8.62 34.84

Actual ratioRatio of of Federal

spillovers to State R&E

to regional expendituresb

benefits 1949-72

0.464.647,32

40,104.154.32

11.654,72

14.854,414.04

0.971.101.251.631.601.371.802.10

2.350.901.38

aDiscounted at 10 percent.blncludes Federal funding of production-oriented agricultural research and extension in each region through CSRS, ARS, ERS,

SCS, and Cooperative Extension relative to State expenditures within the region,

SOURCE Rod. F. Zimmer, F. C. White, and P. L. Cline, Regional Welfare and Agricultural Research and Extension in the U.S.,Agricultural Administration, vol. 9, 1982.


16/24

54

In all regions except the Northeast, the benefitsaccruing to residents outside the region cond uct-ing the prod uction-oriented agricultural R&E areat least four times larger than the benefits accru-ing to the residents within the region. The regionswith the highest ratios of spillovers to regional

benefits are the Northern Plains (40.10), Moun-tain (14.85), Delta States (11.65), and Corn Belt(7.32). Three of these regions (Northern Plains,Mountain, and Delta States) have relatively fewerconsumers tha n most others; thus, a major pro-portion of net benefits accrue to consumers inother regions through interstate flows of foodprod ucts and lower consumer p rices in the recipi-ent r egions. The Lake States (4.64), Appalachian(4.15), Southeast (4.32), Southern Plains (4.72),and Pacific (4.41) regions have spillover ratiosnear the national average for all regions (4.04).

From an equity perspective, the spillover ratiosindicate that the Federal Governm ents share ofinvestment should be m ore in the Northern Plains,Mountain, Delta States, and Corn Belt regions(e.g., farm prod ucing regions) than in other r e-gions of the Nation that have spillover ratios nearthe national average. A comp arison of the ratioof actual Federal expenditures to State expendi-tures in each region to the ratio of spillovers tointernal benefits in each region indicates that inevery region except the Northeast, the spilloverratio is more than dou ble the ratio of Federal ex-penditures to State expenditures.

Figure 27 show s the regional distribu tion of ben-efits resulting from an increase in ag ricultural R&Eexpenditures. Distribution of consumer benefitsis highly correlated in a positive manner with foodpu rchasing patterns and pop ulation d ensity. TheNortheast receives 38 percent o f all benefits result-ing from agricultural R&E investments madethroughout the Nation. Other major beneficiariesare the Corn Belt, Pacific, Appalachian, and

BURDENS OF PHTME RESEARCH

Achievements from PHTME research have con-tributed to the economic stature of the UnitedStates, but th ese achievements ha ve not been at-

South east regions. Each of these regions receivesmore th an 10 percent of the total benefits.

Even though specific data for PHTME researchis not available, it could again be inferred w ithconfidence that a similar an alysis w ould result in

the same conclusion given the above rationale.For examp le, the d evelopm ent of a food process-ing technology w hich lengthens th e shelf life ofa food p rodu ct or the development of a comp uter-ized food w arehousing system w hich is based onresearch cond ucted in Iowa is just as ap plicableif used in N ew York, California, or Texas. Thu s,because of the high tra nsferability of the PHTMEtechnology the benefits of PHTME research arehighly correlated with food purchasing patternsand popu lation distribution and could result inthe sam e, if not gr eater, distribution of benefitsas above. However, the benefits to consumers in

all these major recipient regions are partitionedinto such small amounts tha t the individual con-sumer cannot feel the connection with increasedfarm and PHTME prod uctivity realized as a con-sequ ence of R&E. Thu s, agricultura l R&E rem ainund ervalued by consumers as w ell as by farmers(3,9,39)

From an equity perspective, the spillover ratiosindicate that the Federal Governmen ts share ofresearch investment should be larger. When suchpublic benefits have an impact on residents of theNation, funding for the research can be moreequitably provided by the Federal Government.The bulk of research investments is financed byStates in th e major farm prod ucing regions. How -ever, the m ajority of food consum ersespeciallythose w ho benefit from PHTME researchliveoutside these major agricultural States. Thus, tax-payers in major agricultural States are subsidiz-ing PH TME research for consum ers in less inten-sive agricultural States.

tained w ithout certain costs and bu rdens to soci-ety. For th at reason, the benefits accruing fromsuch research m ust be w eighed not only against


17/24

55

Figure 27. Regional Distribution of Benefits Resulting From anExtension Expenditures

Increase in Agricultural Research and

Pacific11.59%

Lake States7.77%

m,

I

Appalachian10.44%

east

%

SOURCE: Rod F. Zimmer, F C White, and P. L. Cline, Regional Welfare and Agricultural Research and Extension in the U.S., Agricultural Administration, vol. 9, 1982

the magnitude of whatever dollar and scientificman pow er investments are required, but againstthese other costs and burd ens.

Environmental Contamination of Food*

H istorically, chemicals such as sa lt, suga r, andwood smoke have been used to preserve foods.Modern food technology relies extensively on the

use of chemicals not only for preservation but also*For a more in-depth discussion of this topic see Environmental

Contaminants in Food, Office of Technology Assessment, U.S. Con-gress, December 1979.

to produce appealing colors, flavors, aromas, andtextures.

In the United States food law s have been d e-signed to permit the use of such chemicals in foodunder conditions judged to be safe. These chemi-cals are not considered ad ulterants or contami-nants an d are classed as intentional additives.

Environmental contaminants includ e su bstances

from natural sources or from industry and agri-culture. Many n aturally occurring contaminantsin food are of microbiological origin and consistof harmful bacteria, bacterial toxins or fungal tox-


18/24

56

ins. The second category of environmental con-taminants includes organic chemicals and heavymetals.

Environmental contamination of food is the re-sult of our m oder n, high-technology society. TheUnited States prod uces and consumes large vol-

umes of a wide variety of substances, some ofwhich are toxic. It is estimated that 70,000chemicals may currently be in commercial produc-tion in the United States and that sO of these chem-icals are man ufactured in qu antities greater than1.3 billion pounds per year (32). During the pro-du ction, use, and disposal of these substances,there are opp ortunities for losses to the environ-ment.

Chemicals contaminate foods through differentroutes d epend ing on th e chemical, its use, andsource of contamination. A pesticide becomes an

environmental contaminant w hen it is present infoods for which the ap plication or use of the sub-stance has not been ap proved . Imp roperly fum i-gated railroad cars, trucks, ships, or storage build-ings used for the transport or storage of humanfood or animal feed are sources of contamination.The interiors are sprayed or fum igated with p esti-cid es, and if not sufficiently aired, contam inationof the food or feed occurs.

Polychlorinated biphenyls (PCBS) were widelyused in transformers and capacitors, as heat trans-fer fluid s, and as an add itive in some dyes, car-bon paper, paint, pesticides, and plastics. PCBSoccur in food as the result of environmental con-tamination leading to accumu lation in the foodchain, direct contact with food or animal feeds,or contact with food packaging materials madefrom recycled paper containing PCBS (35). Sev-eral comprehensive literature review s have beenpublished detailing the acute and chronic toxic ef-fects of PCBS in animals and hu mans (32).

Displacement of Labor/Shiftsin Employment

The adop tion of mechanization technology inthe m arketing sector has increased labor prod uc-tivity, but it has also displaced labor and causedshifts in employment. Moreover, new productdevelopmen t, although beneficial to consum ers,

may have adverse effects on others. For example,following d evelopment of m argarine from vege-table oils, butter w as to a great extent r eplacedby margarine. Consumers benefited from this de-velopment through lower cost of food and poten-tially better health, but it was d isadvanta geousto the d airy industry. How ever, jobs lost in but-ter m anufacturing plants became jobs gained inmargar ine plants, and dairy farmers, at least inthe Midwest, now can p rodu ce soybeans as a cashcrop to provide the oil for mar garine prod uction.Future developments and consumer acceptance ofsynthetic milk made from vegetable proteinsand fats also may h ave an adverse effect on dairyfarmers, as the development of meat analoguesbased on vegetable proteins would h ave on themeat industry. The magnitude of such effects willdepend on the demand-supply relationships forthese produ cts when competitive produ cts become

available (14).

Burdens on Consumers

Some burdens from PHTME research have beenplaced on consumers. As noted earlier in thischapter, postharvest technology research hasresulted in some prod ucts that actually contributeto nutritional p roblems. For examp le, high-cal-orie, low-nutr ient snack foods and food su bsti-tutes are of lower nutritional value than the foodsthey replace. Nu tritional value needs to be an in-tegral part of food quality, because an increas-ing prop ortion of the food sup ply is mod ified bypostharvest technology.

Among the rap id changes that PH TME researchhas brought to the processing and packaging offood, the use of chemicals in these processes islooked on w ith disfavor by certain segments ofsociety. Others eschew highly processed foodprod ucts in favor of more na tural foods. Foodattitudes are deeply rooted even in a technologicalculture such as that pr evailing in the Un ited States.

During the last three decades packaging of food

has u nd ergone m ajor changes. In bottling, for ex-ample, there has been a shift from use of refillableglass bottles to the use of nonreturnable glass andplastic bottles and metal cans. A result of thistrend has been that discarded food containers


19/24

57

have become significant components of both lit-ter and mu nicipal solid w aste (31). In add ition,economies of scale in bottling and transportation,especially using lightweight nonreturnable con-tainers, have favored a trend toward centraliza-tion of processing and bottling with fewer proces-

sors and fewer brands available.

QUALITY OF PHTME RESEARCH

Quality is an important aspect of all research.While the qu ality of research is difficult to meas-ure qu antitatively, most scientists wou ld agreethat essential aspects of quality include a logicalresearch plan, dealing with adequ ate num bers ofsamples, reproducing results, recording data sothat the material can be un derstood an d evaluatedby others, and organizing an d conducting researchso that it is amena ble to statistical analyses.

In the d ebate between the executive branch andCongress on funding PHTME research, critics fre-quen tly point to declining qu ality of the researchas reason for not supporting p ublic PHTME re-search. The p erception of low q ua lity is based onpast informal or qu asi-formal evaluation m ethods,including those contained in the so-called PoundReport (20). * A rev iew of th e literatur e failed tofind any for m al methods used in the past forevaluating the qua lity of agricultu ral research.Newer, more formal evaluation method s were ex-

amined for their usefulness.

Evaluation of PHTME Research Quality

OTA commissioned a study to examine newways of measuring the quality of PHTME researchand to evaluate these techniqu es (8). The stud yevaluated research qu ality by examining the nu m-ber of citations of: 1) postharvest technologypatents and 2) postharvest technology articles inrefereed (i.e., peer reviewed) journals. The resultsof the stud y are discussed furth er below.

q For a detailed discussion of the Pound Report and other reportssee An As s es s m en t of th e Un ite d S ta te s Food and Agricu ltu ra lRe s ea rch Sy s te m (Washington, D. C.: Office of Technology Assess-ment , U.S. Congress, 1981).

Transportation and storage requirements of ourmass d istribution system h ave mad e necessary thedevelopm ent of plant varieties resistant to bruis-ing and with long shelf life. Some sensory qu alitieswere relinquished in order for consumers to haveyear-round availability of fruits and vegetables.

Patent Citations From PostharvestTechnology Research

The objectives of postharvest technology re-search programs are generally stated in terms ofthe development of new and improved products,testing of food and food add itives. Much of thisresearch w ill prod uce new technology wh ich is

paten table, so in evaluating the qu ality of post-harvest technology research it is important to lookat patents in p ostharvest technology areas grantedto pu blic agencies and private firms.

Even though p atents provide a suggestive meas-ure of innovative activity and quality, at least twoimportant p roblems w ith using patents as such ameasu re should be noted. One problem is that apatent does not necessarily represent a discretestep forward in the innovation pr ocess, nor is theextent of its contribution directly indicated by thepatenting process. The quality of a patent

w hether measu red by the research behind it, thecreativity of the invention, its economic value,or the clarity of its expositionvaries widely. ThePatent and Trademar k Office does not makein fact, is forbidden to makeany evaluation ofan inven tions contribution to the curr ent stockof knowledge, beyond its jud gment that the inven-tion is original (i.e., has not previously beenpatented).

The second problem with using patents as ameasure of innovation reflects the process ofpa tenting itself. The cost of the patenting process,the app ropriability of the return s from an inven-tion, the pr ospects of litigation, and incentives topatent vary over time. Thus, the number of pat-ents granted will fluctuate regardless of innovativeactivity. Also, the procedural attributes of a pat-ent (e.g., the number and kinds of claims allowed,


20/24

58

the nu mber and kinds of patents an examiner citesin his or her examination) may vary som ewhatfrom examiner to examiner, from class to class,as well as from time to time as the Patent andTrademark Office changes its directives. Thus,time series analysis of pat ents cannot claim to hold

all variables in the patenting process constant. Forthese reasons, the evaluation of patenting activityover time mu st be interpreted as a broad stud yof trends r ather th an a precise analysis (8).

Given these caveats, the study of patenting canbe set in a gen eral context. The investigator s rec-ognized that p atenting was only a partial measureof the output of postharvest technology research,but a very important on e. Furtherm ore, since pu b-lic postharvest technology research is conductedin the presence of rather large private investmentsin postharvest technology research, the public and

private research should be related. High-qualitypu blic posthar vest technology research shouldproduce pilot inventions. Public patents shouldbe cited frequently by subsequent private patentsas the pr ivate sector condu cts furth er research.Public publications should also be cited by sub-sequent private publications and patents.

Detailed data was collected for the years 1966-68and 1977-79 on all USDA patents in four posthar-vest technology research fields (food, textiles,other chemicals, other), and on a samp le of 80pr ivate U.S. firms an d 61 patents to foreign firmsin two of these fields (food and textiles). Informa-

tion was obtained on the class of patent, on thepaten ts and other references that were cited bypatent examiners as next best art,* and on thenum ber of times a p atent w as subsequently citedby other p atents.

Most of the U.S. patents in postharvest tech-nology were found to be from private firms.USDA labs accounted for only 5 percent of repre-sentative postharvest technology patenting. SAESaccounted for even fewer patents. The number ofpatents per dollar expended on research is higherfor USDA than for SAES. No comparative data

exists for private resear ch, but because of the im-portance of a patent to the p rivate sector a crud e

q Next best art is th e citation for a p atent that is closest to theone under consideration. Such a citation indicates the patent underexamination represents an identifiable step forward.

estimate is that the n um ber of patents per d ollarof research is substantially higher for the privatesector than the nu mber for the public sector.

A high p roportion of public and private post-harvest technology patents were found to be sub-sequently cited in other pu blic and private patents.

Of food patents granted in 1966-68, only 22 per-cent of the USDA patents and 30 percent of pri-vate p atents were n ot subsequently cited. For tex-tile patents, only 34 percent of the U SDA p atentsand 18 percent of the private patents were notsubsequently cited in the same time period. Pri-vate patents in both areas w ere cited slightly m oreoften than USDA patents, but the data suggestthat most patented innovations do contribute totechnology improvement.

The data also showed that 28 percent of thepatents cited as next best art for USDA p atents

in 1966-68 were other USDA patents and that 35percent were other USDA patents in 1977-79.Domestic-origin private patents cited USDA pat-ents as next best art 19 percent of the time in1977-79. The latter finding indicates that USDApaten ts were a significant part of the inventionstructure of the private sector. Foreign-originpatents cited USDA patents less frequently. USDApatents also cited fewer total patents than eitherU.S. private or foreign paten ts.

It is difficult to draw general conclusions re-garding research quality based on patent data

alone. Nevertheless, the standards of patentabilitytechnique are applied outside the agriculturalresearch system as well. This wider use of thetechnique involving patent d ata lends supp ort tothe methods credibility.

Publication Citations of PostharvestTechnology Research

A large p art of pu blic postharvest technologyresearch does not lead directly to patented inven-tions, but instead provides research potential,which may subsequently lead to inventions andtechnology. One measure of output of this re-search is publications. With data on publicationsthat h ave been su bject to quality screening, it ispossible to und ertake analysis of differences inprod uctivity for d ifferent research environmentsand different time periods. However, good data


21/24

59

on quality-screened publications are difficult toobtain. Since publication data reported in datafiles generally are not stand ard ized for qu ality,it was concluded that citations in referred jour-nals were a superior measure of quality.

The nu mber of citations is an ind ex of the im-portance of a research contribution. Perhapsmore importantly, citation d ata can ind icate theusefulness of one research sp ecialization to otherspecializations. Thus, a case in which public post-harvest technology research publications are sub-sequently cited only by public postharvest tech-nology researchers w ould be cause for concern.Citation an alysis to d ate has been u sed little inagricultural research evaluation.

The citation-of-publications study is more com-prehensive than the patent study described abovein two respects. First, for this study, the investiga-

tors obtained an SAES sample in add ition to aUSDA sample; and second ly, pu blished ou tpu tis a more complete measure of researcher outp utthan p atents are.

Data sour ces dictated tha t da ta be collected byindividual researchers. A list of research projectprinciple investigators was obtained from CurrentResearch Information System data for selectedUSDA and SAES postharvest technology researchstud ies. With this list of nam es for different p eri-ods, the citations abstract was used to collect dataon the number of times the publications of these

scientists were cited in 1979. Data were compiledboth on citations of papers published fairly recent-ly as well as on citations of pap ers pu blished inearlier p eriods. Both pu blications and citationswere screened in the sense that m ajor journalsin four areas covered by postharvest technologyresearch (i.e., food science, agricult ure, m aterialsscience, and other) were chosen from which toselect citation abstracts.

Statistical mod els were d eveloped to test wh eth-er recent citations were related to institutionalaffiliationsi.e., USDA or SAES. When citations

of earlier papers and age of investigator were usedas control variables, there was no significant ef-fect of institutional setting on total citations perscientist-year or on citations per publication.When expenditures per scientist-year were in-cluded as a variable measuring the support per

scientist in terms of equipment, assistants, etc.,they d id not affect pu blication per scientist-year,but d id p ositively affect citations per p ublicationand total pu blications per scientist-year.

Citations of early w ork w ere considered to bean ind ex of personal produ ctivity or quality ofthe researcher in question, and these were signifi-cant determinants of citations of later work. Thediver sity of citations (i.e., citations in fields o therthan the field of publication) was also affected bypersonal characteristics. Younger scientists havemore influence outside their fields than do olderscientists, and scientists with high early prod uctiv-ity and narrow influence have narrower currentinfluence.

Conclusions

These two attempts to look at p ostharvest tech-

nology research quality by examining p atents andjourn al cita tions do not provide conclusive an -swers about the qu ality of pu blic postharvest tech-nology research. In fact, one USDA center havingthe lowest journal citation score had the strong-est performance in p atenting. These efforts do,however, add to the pool of other less formal andmore judgmental evidence on research quality.

The data are consistent with the view that man ypu blic postharvest technology researchers maynot be aw are of relevant research in closely relateddisciplines and that some public research pro-

grams may be poorly organized. How ever, thedata are not consistent with the highly criticalview that public postharvest technology researchis prod ucing little of value. Pub lic postharvesttechnology research is p roviding patents that aresubsequently cited by the private sector. It appearsto be provid ing a bod y of scientific literature th atis roughly comp arable to that produ ced in otherapplied sciences. Furthermore, the SAES systemdoes not ap pear to be p roviding research environ-ments that are superior to those provided byUSDA in this research area.

Peer Review

Assessments of published output of scientistshave been used to evaluate certain aspects of agri-cultural research, most notably productivity (e.g.,


22/24

60

27), but th e app lication of this technique or varia-tions of it for evaluating the qu ality ofagricul-tural research is fairly recent.

The quality of research is not necessarily a fun c-

tion of numbers of publications or patents that

result from the research. The National Academy

of Sciences considers peer review probably thebest method of estimating quality. Attempting to

use the same scientists to evaluate basic and ap-

plied r esearch, or research in different disciplines,how ever, seems h azard ous. Estimating quality ofany research requ ires great care.

PRINCIPAL FINDINGS

q Benefits from public PHTME research may in-clud e: 1) increased p rodu ctivity and redu ced

real cost of food; 2) imp roved qu ality, safety,and nutrient content of food; 3) new or im-proved food prod ucts; 4) information for deci-sionmaking; and 5) information on industrycompetitiveness.

q The beneficiaries of PHTME research can in-clude: 1) farm producers, through the enhance-ment of the value of farm comm odities and im-proved marketing services; 2) consumers,through expand ed qu antity of food p roductsand lower prices and improved quality, nutri-

tion, safety, and convenience; 3) marketingfirms, through research results that are availableto small and large firms alike; and 4) labor,through increased labor efficiency that providesopportunity for increased wages and salarieswithout placing an upward pressure on retailprices.

CHAPTER 4 REFERENCES

1. Bullock, J. Bruce, Estimates of Consumer LossDue to M onopoly in U.S. Food Manu facturing In-dustries: Comment,Am erican Journal of Agricul-tural Econom ics, vol. 63, No. 2, May 1981, pp.290-292.

2. Currie, John Martin, Murphy, John A., andSchmitz, Andrew, The Concept of Economic Sur-

Peer review should be review of a scientists re-search only by researchers within the same generalarea, discipline, or mission. For example, in thecontinuu m of basic to app lied research, peers ofscientists w orking in basic research can effectivelyreview quality of scientists working in similar ba-

sic areas of research. Scientists wor king in app liedareas can evaluate quality of other scientists work-ing on similar applied problems.

qThe majority of the benefits from PHTME re-search flow to those regions and States with

high concentrations of population. However,the bulk of PHTME research is financed byStates in the major farm producing regions,w ith the majority of the research sup ported byState app ropriations. This means that taxpayersin major agricultural States are subsidizingPHTME research for consumers in less inten-sive agricultural States. When such public bene-fits have an impact on residents of the N ation,funding for research can be m ore equitably p ro-vided by th e Federal Government.

q Publicly funded postharvest technology re-

search seem s to pr ovide a bod y of scientific lit-erature w hich is comparable with that of otherapplied sciences. Some past analyses of thequality of pu blic research suggested th e researchprod uced little of value (20). Today, however,evaluation based on analyses of patents andpublications (8) does not sup port this app raisal.

plus and Its Use in Economic Analysis, TheEcon om ic Journal, vol. 81, No. 5, pp. 741-799,1971.

3. Edd leman , B. R.,Im pa cts of Reduced Fed era l Ex-pen ditures for Agri cultura l Research an d Educa-tion, IR-6 Information Report No. 60, May, 1982.

4. Eddlem an, B. R., Importan ce of Resources to Pro-


23/24

61

5.

6.

7,

8.

9.

10.

11.

12.

13.

14

benefits, burdens, and quality of postharvest technology and marketing economics research

Documents