Chapter Economic Considerations

Chapter 11

Economic Considerations

It would not be reasonable to make decisions on alternatives to animal use without havingsome idea of the consequences to the health and welfare of the public.

Kennerly H. DiggesNational Highway Traffic Safety Administration

U.S. Department of TransportationMarch 20, 1985

CONTENTS

P a g e

H o w Much Does Animal Use Cost? . . . . . . . . . . . . . . . . . . . . . .................243

Costs and Benefits in Research. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .245Biomedical Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........245Supporting Patent Claims , . . . . . . . . . . . . . . . . . . . . . . . . . . ..................247

Costs and Benefits in Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , ..248Testing pesticides for Toxicity. . . . . . . . . . . . . . . . . . . . . . ....................248Testing and Product Liability. . . . . . . . . . . . . . . . . . . . . . . ...................249Testing Costs of Animals and the Alternatives . . . . . . , . . . . . . . . ., . . . . . . . . . .249

National Expenditures for Research and Testing . . . , . . . . . . . . . . . . . . . . . . . . . . .250Toxicological Testing Services . . . ., . . . . . . . .“. . . . . . . . . . . . . . . . . . . . . . . . . , , .251Government Toxicological Research and Testing . . . . . . . . . . . . . . . . . . . . .. ...251

Protecting Proprietary Interests , . . . . . . . . . . . . . . . . . . . . . . . .................252Cooperative Research and Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . .........252Toxicity Testing Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .............252

Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ...253

Chapter 11 References. . . . . , , . . . . . . . . . . . . . . . . , . . . . . . . . , . . . . . . . . . . . . . ..254

List of TablesTable No.. Page

11-1. Total Savings Attributable to Biomedical Research . . . . . . . . . ............24611-z. Estimated Biomedical Research Outlays, Selected Years, 1900-75” . . . . . . , . .24711-3. Selected Federal Expenditures Related to Toxicological Testing

and Research, 1984-86 . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . .251

List ofFiguresFigure No, Page

11-1. Relation Between Number of Animals Used and Cost of Animal Use . . . . . .24411-2. Development of a Typical Pesticide for Agriculture , . , . , . . . . . . . . . . . . . . .249

Chapter 11

Economic Considerations

Economic considerations play an important rolein decisions on the use of animals in research, test-ing, and, to a lesser extent, education. It is demon-strable that many valuable techniques, pharma-ceuticals, pesticides, and other products havebeen developed or tested using animals. Yet ani-mal use is often very expensive and time-consum-ing. A new pesticide, for example, may require$5 million worth of testing with animals beforeit can be registered. Even higher animal costs maybe incurred in developing a new drug. Large in-centives thus exist to find alternatives that reducethe cost and time involved in animal research andtesting while maintaining the ability to improvehuman health, assess and manage the risks oftoxic substances, and acquire fundamental newbiomedical knowledge. Considerable investmentsare required to develop and validate such alter-natives before they can be implemented with con-fidence.

This chapter examines costs and benefits sur-rounding animal use and the development, vali-

dation, and implementation of alternatives inresearch and testing. Data are provided for bio-medical research as it relates to human health anddisease, but a precise determination of costs andeconomic benefits of animal use within biomedi-cal research is elusive. Several aspects of usinganimals in toxicological testing are examined,including the development of pesticides, the eco-nomic incentives to develop and validate nonanimaltests, and the extent of liability a manufacturermight incur for insufficient product testing.

In education, it is hard to put a price tag on theuse of animals. At the college and graduate levels,benefits of animal use include the training of bi-ologists, psychologists, toxicologists, physicians,and veterinarians. Education involving animalscontributes indirectly to research and testing bytraining those who eventually carry out this work.The benefits of using animals in primary and sec-ondary education include increasing students fa-miliarity with animal behavior and care (see ch. 9).

HOW MUCH DOES ANIMAL USE COST?

In either research or testing, the principal costassociated with animal use is that of human la-bor. Animals must be fed, watered, and have theircages cleaned. They require attendant veterinarycare and are housed in facilities needing labor-intensive sanitation. Such labor costs are the ma-jor component of both the expense of producinganimals in breeding facilities and the cost of main-taining them in laboratory facilities prior to andduring research and testing.

The total cost of animal use is the sum of thecost of acquisition of the animals and that of main-taining the animals prior to and during their use.Acquisition expenses vary widely among species.Mice, for example, cost on average about $2apiece, hamsters about $5, and guinea pigs about$19. Dogs range in price from $5 for a pound ani-mal to several hundred dollars for a purpose-bredanimal. Primates can cost from $400 to more than

$2,000. The actual cost for a particular speciesvaries with the sex, strain, weight, age, quantityordered, method of shipping, and distance shipped.

Maintenance costs also vary. Maintaining amouse, for example, costs about 5 cents per day,a hamster about 11 cents per day, and a guineapig about 40 cents per day. The actual cost variesamong different laboratory facilities, depending,for example, on accounting practices and local la-bor costs. The total maintenance cost of an animalis directly related to its length of stay in the lab-oratory. It is important to note that maintenanceexpenses can quickly exceed and even dwarf ac-quisition costs. A 2-month-old hamster costing $5,for example, used in research until the age of 10months costs $26 to maintain.

Figure 11-1 illustrates the relation between thenumber of animals used in research and testing

243

244 ● Alternatives to Animal Use in Research, Testing, and Education

Figure 11-1 .—Relation Between Number of Animals Used and Cost of Animal Use

Acquisition Maintenance Totali h. .

II

+ ‘x

I

The total cost of animal acquisition and maintenance equals the sum of the acquisition cost and the maintenance cost.maintenance cost depends-on the animal’s length of stay in the animal facility.)

Acquisition Maintenance

I Number ofanimals

used

x

I

+

Number ofanimals

used

x

Total

I

—

(The

Using fewer animals will yield a decrease in the total cost of animal acquisition and maintenance, but the proportionate sav-ings ‘will be less than the decrease in the number of animals used. Both the price of each animal and the cost of maintenanceper animal can be expected to increase to support the operating costs of breeding facilities and animal facilities.

SOURCE: Office of Technology Assessment.

and the total cost of animal use. Although thenumbers and species of animals used (see ch. 3)and the price per animal can be estimated, it iscurrently impossible to estimate with any ac-curacy the laboratory lifetime-and hence the to-tal maintenance costs-of animals used in theUnited States. Therefore no actual dollar figurecan be affixed to the cost of animal acquisitionand maintenance in research and testing. Begin-ning in 1986, the Public Health Service (PHS) willrequire reports on the average daily census of allspecies housed in PHS-funded facilities (see app.C). These data may permit an estimate of the to-tal cost of animal use in a sizable portion of ani-mal research—namely, that conducted in PHS-funded facilities.

The relationship shown in figure 11-1 empha-sizes several aspects of the economics of animaluse. If the number of animals used is reduced,the total cost of animal acquisition and mainte-nance will decline. But the proportional decreasein total cost will not match the proportional de-crease in the number of animals used. Reducinganimal use by 15 percent, for example, will noteffect a cost savings of 15 percent; the savingswill be somewhat less, for two reasons, First, ifthe number of animals used decreases, the costof acquiring each animal can be expected to in-crease somewhat. (A temporary drop in price forsome species that are in immediate oversupplymay occur, but this would last only through thelaboratory-useful lifespan of animals already on

Ch.11 -Economic Considerations Ž 245

hand and ready for sale.) With reduced demand, would have to spread the cost of operation overvendors would have to raise prices to cover their fewer animals. In both breeding and laboratoryoverhead. Second, if the number of animals used maintenance of animals, there are economies ofdecreases, the expense of maintaining each re- scale such that breeding and maintenance of mar-maining animal in a laboratory facility can be ex- ginally fewer animals does not yield a correspond-pected to increase. Laboratory-animal facilities ing decrease in costs.

COSTS AND BENEFITS IN RESEARCH

The many important economic contributions ofresearch with animals are difficult to character-ize. First, research does not lend itself to suchanalysis. Normally, one experiment will drawfrom many others and contribute to future re-search, making allocation of costs and benefits toa particular activity virtually impossible. Second,the outcome of each experiment is uncertain, andthe experiences in one program would not nec-essarily apply to others. Third, the delay betweenresearch and commercialization is long, reachinga decade or more, with payoff taking even longer.Thus, it is not possible to evaluate with any rea-sonable confidence the costs and benefits of cur-rent or even recent animal and nonanimal re-search practices,

Biomedical Research

This section discusses biomedical research ingeneral, which unavoidably averages many di-verse research experiences. Biomedical researchis of interest because it is a major user of animals,because it affects human health, and because itaffects an important sector of the economy—thehealth care industry. As with most areas of re-search, many of the contributions are indirect andmany are not easily quantified in economic terms(see ch. 5). Most benefits are realized in the healthcare industry, which in 1983 accounted for $355.4billion (10.8 percent) of the gross national prod-uct (9). Drugs, which require both biomedical re-search and toxicological testing in their develop-ment, have annual sales of about $30 billion andcontribute about 20,000 jobs to the economy (29).

The first medical discovery that was largely aresult of research with animals was diphtheriaantitoxin at the end of the 19th century. Its usereduced the likelihood of death for those contract-

ing diphtheria from 40 to 10 percent (28). Ani-mals eventually came to be used in all phases ofbiomedical research and in the development ofmedical products such as drugs and devices andof services such as surgery and diagnostic tech-niques.

Research with animals that leads to practicalapplications can last from a few days to manyyears. It may involve inexpensive equipment orhundreds of thousands of dollars’ worth of in-strumentation, may be performed by a laboratorytechnician with little supervision or by a team ofhighly educated scientists, and may be done withfruit flies or with primates. The costs will varyaccordingly.

The benefits and rates of return on a given ex-periment vary widely. The rate of return for agiven research program can only be determinedreliably many years after commercialization. Inthe case of products with high research and test-ing costs and long lead times to commercializa-tion, which applies to many of the products of bio-medical research, the lag can be several decades.

A 1972 study on the rates of return for six largepharmaceutical companies for research they con-ducted in 1954 through 1961, when animals werewidely used, estimated the pretax private rate ofreturn to be 25 to 30 percent (2). The social rateof return—the benefits to the public, was esti-mated to be at least twice as high (20).

Another approach to gauging costs and bene-fits involves looking at expenditures from 1900to 1975 and comparing them with the benefitsof medical advances in preventing sickness anddeath in the work force over the same period (4).All data were adjusted to 1975 conditions. Anal-ogous comparisons were made for 1930 to 1975.

246 . Alternatives to Animal Use in Research, Testing, and Education

Of course, this approach ignores both the costsof research before 1900 (or 1930) that contrib-uted benefits after those years and any researchconducted before 1975 for which the benefits hadnot fully accrued. The latter is more confound-ing because much more research was done in thelast decade of the study than any other, so thebenefits could not be fully counted. Undervaluealso occurs because it is impossible to measurethe value of not being ill. Yet, other assumptionsin the study may overvalue the benefits by ne-glecting changes in nutrition, lifestyle, and work-ing conditions.

Benefits exceeded costs in this study by factorsranging from 4 to 16, depending on the assump-tions made in calculating benefits and whetherthe time period is from 1900 to 1975 or 1930 to1975. Expressing the results in another way, thesavings in health-related costs due to the increasein knowledge was estimated to be $115 billion to$407 billion (4,20).

Savings were also calculated for various diseasecategories. As can be seen in table 11-1, researchcosts in certain areas have exceeded benefits ac-cruing over the same period (as indicated by aminus sign). Table 11-2 provides related informa-

tion about levels of funding in selected years,showing how funding grew during the sameperiod. The research budget for 1975 correspondedto almost 9 percent of the costs associated withneoplasms (data not shown in tables), whereas allother budget-to-cost ratios were below 1.4 per-cent (20).

Although economic and financial data such asthese are useful in making policy judgments, mostdecisions about using animals are more compli-cated and take into account political and technicalconsiderations, as well as economic ones. Mone-tary costs and benefits must be balanced with fac-tors such as scientists’ desire to be certain andsociety’s desire to have animals treated in a hu-mane fashion. Economic analysis focuses on onlypart of the equation, and cannot be the sole basisfor decisions.

one example of the tension between financialand other criteria is exemplified by the questionof whether pound animals should be used in lab-oratory studies. A recent survey indicates thatabout three times as many dogs and cats are ob-tained from pounds and dealers (who often pur-chase from pounds) as are purpose-bred for lab-oratory use (18). Scientists have argued that

Table 11=1.—Total Savings Attributable to Biomedical Researcha (in billions)

Disease categorv 1900-75 simulation 1930-75 simulation- .Totalb . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $299.37 to $479.83 $145.65 to $167.76

Infective and parasitic diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . .Neoplasms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Endocrine, nutritional, and metabolic diseases . . . . . . . . . . . . . . .Diseases of the blood and blood-forming organs . . . . . . . . . . . . .Mental disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Diseases of the nervous system and sense organs. . . . . . . . . . . .Diseases of the circulatory system . . . . . . . . . . . . . . . . . . . . . . . . .Diseases of the respiratory system . . . . . . . . . . . . . . . . . . . . . . . . .Diseases of the digestive system, oral cavity, salivary glands,

and jaws. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Diseases of the genitourinary system . . . . . . . . . . . . . . . . . . . . . . .Complications of pregnancy, childbirth, and postpartum . . . . . . .Diseases of the skin and subcutaneous tissue . . . . . . . . . . . . . . .Diseases of the musculoskeletal system and

connective tissue. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Congenital anomalies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Certain causes of perinatal morbidity and mortality . . . . . . . . . . .Svmptoms and ill-defined conditions . . . . . . . . . . . . . . . . . . . . . . . .Accidents, poisonings, and violence . . . . . . . . . . . . . . . . . . . . . . . .aMinus signs indicate costs exceeded benefits.%otals may not add due to rounding.

118.35–2.66

0.280.769.20

14,7610.6871.75

12.9623.584.590.93

to 174,16to –3.17to 0.95to 1,27to 21.88to 24.31to 18.11to 116.38

to 21.62to 37.28to 7.88to 1.76

to 14.59to 10.08to 3.73to 19.80to 9.39

63.15– 1.17– 1.57

5.223.73

–3.32–6.4223.44

23.539.27

10.922.98

6.197.562.80

14.464.08

– 11.04– 1.41

4.38–2.1026.05

to 69,23to – 1.40to –0.99to 5.45to 3.72to –2.70to –4.91to 27.56

to 26.53to 10.96to 11.62to 3.04

to – 10.98to –1.11to 6.58to – 1.67to 26.86

SOURCES: 1900.75 slmulatlon: S.J. Mushkin, Biomedical Research: Costs and Benefits (Cambridge, MA: Ballinger Publishing Co., 1979); 1930.75 simulation: A. Berkand L.C. Paringer, Econornlc Costs of Illness, 1930-1975 (Washington, DC: Public Services Laboratory, Georgetown University, 1977).

Ch. 11-Economic Considerations . 247

Table 11-2.—Estimated Biomedical Research Outlays, Selected Years, 1900-75 (in millions)

Disease category 1900 1930 1963 1975Total a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

- - - - - - - - - - - - - - . - - - - - - - - - -

Infective and parasitic diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Neoplasms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Endocrine, nutritional, and metabolic diseases . . . . . . . . . . . . . . . . . . . . . . .Diseases of the blood and blood-forming organs . . . . . . . . . . . . . . . . . . . . .Mental disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Diseases of the nervous system and sense organs. . . . . . . . . . . . . . . . . . . .Diseases of the circulatory system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Diseases of the respiratory system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Diseases of the digestive system, oral cavity, salivary glands, and jaws .Diseases of the genitourinary system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Complications of pregnancy, childbirth, and postpartum . . . . . . . . . . . . . . .Diseases of the skin and subcutaneous tissue . . . . . . . . . . . . . . . . . . . . . . .Diseases of the musculoskeletal system and connective tissue . . . . . . . .Congenital anomalies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Certain causes of perinatal morbidity and mortality . . . . . . . . . . . . . . . . . . .Symptoms and ill-defined conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Accidents, poisonings, and violence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .aTOtalS may not add due to rounding.

$0.15700.04760.01950.00060.00030.00050.00720.00390.03050.00600.01290.00090.00020.0006

—0.00850.00960.0068

$l0.0180

1.03292.82910.27750.07010.04610.21640.91361.06090.72730.93380.11420.01600.02400,10920.44480.27250.9377

$1,561.0

17.2847.6

33.24.54.1

14.2252.4

96.562.325.8

1.21.93.3

18.450.019.4

109.0

$4,640.0

37.62,464.8

109.012.522.741.3

876.0261.7175.466.8

0.94.6

12.132.068.774.2

379.6

SOURCEDatafrom S.J. Mushkin, Biomedica/ F?esearch:CosM and Benefits (Cambridge, MA’ Ballinger Publishing Cov 1979)

pound animals are much cheaper than purpose-bred ones and that it would be wasteful to de-stroy them when they could be used. The differ-ence in price between a purpose-bred and apound dog ranges from $200 to $500 per animal.Estimates of the impact on research of a ban onusing pound animals range from a tenfold in-crease in costs to effectively`stopping researchin Los Angeles County (25). Others have arguedthat pound animals are poorly suited to most lab-oratory work because they are often in poorhealth and their genetic background is usually un-certain (25).

It may seem ethically desirable to make use ofanimals that would be killed anyway, but an ani-mal that had been a pet may find laboratory con-ditions more stressful than a purpose-bred ani-mal would. Other nonpecuniary considerationsare that people may hesitate to bring their ani-mals to a pound if they oppose laboratory use ofpound animals and that those using pound ani-mals will see them as cheap, disposable experi-mental tools that need not be conserved (22).

Supporting Patent Claims

Data derived from animal research have pro-prietary value and are often used to support pat-ent applications for drugs or devices for humans.

Patents give the inventor an exclusive right tomake and sell the patented invention, thus pro-viding an incentive to invent, which in turn fuelsa growing economy. Thus, animal use can haveimportant economic consequences in addition toimprovement in health.

To obtain a patent, an inventor must show thatthe invention is novel and useful and must dis-close how to make it and use it. Data from studieswith humans are normally obtained to supporta patent on an invention to be used by humans,but data on animals can provide evidence of util-ity as well (12,15). And because they are normallyobtained before research is done on humans, suchdata sometimes play a crucial role in determin-ing the date of an invention, which could deter-mine who gets the patent in the case of two com-peting inventors.

Utility can be demonstrated with animal studies,but only if the data would convince someone ofordinary skill in the art that the same effect wouldbe observed in humans (14). The character andquantity of evidence needed to show utility de-pend, in part, on whether the results agree withestablished beliefs (13). Courts recognize that ananimal may respond differently than a humanwould (16), and in demonstrating the utility of aninvention it is not necessary to demonstrate safety(11)17).


In vitro experiments are sometimes sufficientto demonstrate utility for patent purposes. In onerecent case (7), in vitro tests showed that thechemical to be patented, an imidazole derivative,inhibited thromboxane synthetase in blood plate-lets. The activity of thromboxane synthetase wasthought to be related to hypertension, pulmonaryvasoconstriction, and other cardiovascular dis-eases, and the demonstration of the chemical’sability to inhibit it was sufficient to show utility.Data showing therapeutic use were not requiredin showing that an invention had taken place. Inanother case, the fact that the inventor had given

a detailed description of how the substance to bepatented would behave was enough to supporta showing of utility, thus fixing the date of inven-tion (24).

Although the use of alternatives to support pat-ents is interesting, it does not have much practi-cal effect on the use of animals in developing med-ical products because safety and efficacy must bedemonstrated to satisfy regulatory requirements(see ch. 7). These patent cases might have someapplication, however, in demonstrating the suffi-ciency of alternatives in other areas.

COSTS AND BENEFITS IN TESTING

There are several major economic benefits tousing animals in toxicological testing. Drugs, foodadditives, pesticides, and many consumer prod-ucts are tested for toxicity or other kinds of haz-ards before they can be marketed and begin togenerate income for the manufacturer. This isoften done to meet regulatory requirements, butthe tests are also done to avoid marketing unsafeproducts. In addition, testing is done to confirmthat a product does in fact confer a benefit.

Testing Pesticides for Toxicity

Over a billion pounds of pesticides are used inthe United States annually, corresponding to over$4 billion in sales. About 130 firms produce theactive ingredients in pesticides. Thirty of theseproduce common products in high volume; theothers tend to produce specialty pesticides. Mostof the pesticides are used in the agricultural sec-tor. About 7 percent are purchased by consumersfor home and garden use, while industrial andinstitutional use account for about 20 percent (31).

Because pesticides are designed to be biologi-cal poisons, they are among the most toxic sub-stances commercially available. Most of the haz-ards result from chronic, low-level exposure.Exposure and the risk of it are widespread. About2 million commercial farms in the United Statesuse pesticides, some of which remain in or on thefood and are eventually consumed. About 40,000commercial applicators use pesticides to treat

structures and facilities. The Environmental Pro-tection Agency (EPA) estimates that 90 percentof all households regularly use or have used pes-ticides in the home, garden, or yard (31). The re-sults of tests on animals are used by EPA to iden-tify hazards and to develop acceptable exposurelevels and safe handling and disposal practices (seech. 7). Thus, animal testing plays an importantrole in the protection of virtually the entire U.S.population.

Acute poisonings have been estimated to costover $15 million annually (1980 dollars), exclud-ing the value of saving lives or avoiding suffer-ing. The estimated cost of each death due to pes-ticide poisoning is $112,000, whereas the averagecost of a nonfatal poisoning is $200 (23,31). Thecosts of cancer, the most important chronic ef-fect, is over $34 billion in 1980 dollars, with eachcancer costing $52,000 (31). One research goal isto find new pesticides that are less toxic and moreeffective than those now in use, a search that en-tails animal testing.

There are over 48,000 registered pesticide for-mulations, with an estimated 1,400 to 1,500 ac-tive ingredients (5). There are between 5 and 20new registrations for active ingredients issued an-nually, each requiring a complete toxicologicalevaluation based on animal testing and other data.Another 1,500 to 2,000 new formulations or usesare also registered annually (5,31). These requirelittle additional testing, as a rule, and often relyon data in EPA’s files.

Ch. 11—Economic Considerations Ž 249

Testing costs vary with the product and its uses.Pesticides intended for food crops require muchmore testing than those for other uses. Some ofthe required testing depends on results obtainedin screening tests. The cost of testing can rangefrom $2 million to $5 million for a new active in-gredient. This represents a small fraction of thetotal developmental expenses, which may approach$100 million (31). Testing costs are incurred pri-marily in the beginning of the developmentalcycle (see fig. 11-2). As the figure illustrates, test-ing with animals-during testing for toxicity—isan integral part of the development of a new pes-ticide.

Testing and Product Liability

Toxicological testing of consumer productshelps keep unsafe products off the market, It alsomay sometimes allow liability for injuries to beavoided. The cost of product liability litigation canbe enormous, and companies are tending to droprisky products, as the current situation with vac-cines illustrates.

Most States have “strict liability, ” in that amanufacturer is liable for whatever injuries itsproducts cause. In most jurisdictions, there areexceptions, such as when the technology for de-termining that a product is unsafe does not ex-ist. There are also exceptions when the productis known to be dangerous but also to confer agreat benefit. Such is the case for rabies vaccine.A few jurisdictions merely require that a manu-facturer not be negligent (see ch. 7.) Manufac-turers are unlikely to adopt alternatives to ani-mal tests until they believe such methods offera level of assurance of product safety equal to thatoffered by animal testing.

Testing Costs of Animalsand the Alternatives

An estimated 80 percent of the cost of testing,whether whole-animal or in vitro, is for labor (6).Testing costs vary widely with the assay used andsomewhat with the facility. The cheapest, suchas for eye or skin irritation, can be done for un-

\ Figure 11 ”2.—Development of a Typical Pesticide for Agriculture(Note the integral role of animal toxicity testing in pesticide development, shown in boldface.)

~ Year “:” . “ ~ o - ~ ~ $: “ i g .$: & . 8

* ,& p H t” ’ ; , . , : : ‘ . , -tI

: m ; ~ ~ f ~$

7 .

-@& , ~ .,,, ‘

;J 4 ;?*

. ~ ,,. . -‘“ T- “--” , ““ { .? Full

Commercialization..

SOURCE: Off Ice of Technology Assessment, adapted from Haskell Laboratories, El. du Pent de Nemours & Company, Wilmington, DE, 1984.

38-750 0 - 86 - 9


der $1,000. An LD5O test can be performed forless than $2,000. Subchronic toxicity tests can costunder $100,000, and those for long-term toxic-ity or carcinogenicity for two species can be donefor less than $1 million, and perhaps for under$500,000 (10,31). As a rule, the cheaper tests re-quire fewer animals, but more importantly theytake far less time at each of three stages—plan-ning, execution, and analysis. Another reason forlarge variations in testing costs is the species used,with maintenance costs approximating $0.05 perday for a mouse, $4 for a dog, and $11 for a chim-panzee. Most of the cost of maintaining animalsis attributable to labor expenses.

Various short-term in vitro tests for mutagenic-ity have been developed over the past 15 yearsin an effort to replace the more costly and time-consuming carcinogenicity test (see ch. 8). The

most popular mutagenicity test, and one of thefirst to be introduced, is the Salmonella typhimu-rium/microsome plate mutation assay (the Amestest), costing $1,000 to $2,000 (10). This assay hasthe most extensive database thus far (1). Usedalone, it does not appear to be as predictive ofhuman carcinogenicity as are animal tests.

If the Ames test, some yet-to-be developed test,or a battery of tests proves to be more predic-tive of carcinogenicity than testing with animals,the savings could be enormous. A battery of teststhat might indicate carcinogenicity has been sug-gested by the National Toxicology Program (30)and has shown some promise in preliminaryevaluations (see ch. 8). Most testing laboratoriescould conduct this particular battery of tests forunder $50,000, and costs would probably declineas the tests become more commonplace (10).

NATIONAL EXPENDITURES FOR RESEARCH AND TESTING

Research and testing in the United States arefinanced and conducted in a variety of ways. Thesources of research funding are Government andindustry. Some Government research funds sup-port Government laboratories, but a larger sharesupport research in academia. Industry researchis done primarily at in-house industry labora-tories, with some funds contracted to other lab-oratories and to academia.

Most testing is conducted by industry. Thechemical industry is the sector most directly af-fected by regulatory policies concerning toxico-logical testing. In 1982, this industry (StandardIndustrial Classification Code 28) had shipmentsworth over $170 billion and employed 866,000people, which represented 8.7 percent of all in-dustry shipments in the United States and 4.5 per-cent of the employees.

Drugs, soaps and toilet goods, and agriculturalchemicals account for the greatest use of animalsin testing, and constitute almost a third of theiruse by the chemical industry. The rest of the chem-ical industry, in order to satisfy transportation,disposal, and occupational health requirements,does simple tests such as the LD5O for substancesfor which the potential exposure is high (see ch. 7).

Corporate research and development (R&D) inthe chemical industry is large and concentratedin the industrial chemicals and drug sectors. Ex-penditures by the industry totaled $7.6 billion in1984 (8), a figure that includes in-house toxico-logical testing, research involving the use of ani-mals, and many other activities. It has been esti-mated that the toxicological testing industryaccounts for just under 10 percent of the R&Dexpenditures in the chemical industry (27), mak-ing testing an estimated $700 million expenditurein 1984. An unknown percentage is spent on re-search involving animals.

In the past 10 years, industry’s R&D expendi-tures have grown at about 13 percent per year,following a slight decline in the early 1970s. R&Dexpenditures for drugs, as a percentage of sales,are twice as high as the industry average, andhave grown at a slightly higher rate (8). Animaluse could be growing at a similar rate, althoughsurvey estimates (see ch. 3) and other factors (seech. 8) do not support this notion.

The Federal Government also plays a major rolein animal research and testing, with almost $6 bil-lion obligated for research in life sciences for1985. University research in the life sciences,which is funded largely by Government and some-

Ch. n-Economic Considerations ● 2 5 1

what by industry, will cost an estimated $2.9 bil-lion (8).

Projections of future expenditures depend ona number of factors, including the growth of thechemical industry and of R&D within it; the areasof R&D (e.g., new substances, new uses for oldsubstances, new processes for making old sub-stances); regulatory policies, both domestic andforeign; the growth of the overall economy; taxpolicy; and further developments in nonanimaltests.

International developments can have economicrepercussions. For example, Swiss voters defeatedin 1985a referendum virtually banning all animaltesting (see ch. 16). A number of companies havefacilities in Switzerland, and such a change couldhave shifted testing to another country. whetherU.S. labs could compete for that business dependson the strength of the dollar.

Toxicological Testing Services

In 1984, the toxicity testing industry in theUnited States was estimated to be worth about$650 million per year (27). Sixty-five percent ofthe testing is done by corporations in-house. Theremaining 35 percent (about $225 million annu-ally) is conducted by commercial laboratories,universities, and other organizations. Althoughthere are over 110 U.S. laboratories that sell test-ing services, most specialize in a small numberof assays and are not ‘(full service. ” Hazelton isthe largest of the full-service labs, with domesticsales of $36 million in 1983. Except for Hazeltonand several other large commercial labs, the in-dustry is a dispersed one, with the many smallcommercial firms accounting for approximatelytwo-thirds of the value of domestic sales (10).

The industry expanded its facilities in the 1970sin response to Federal regulatory changes and thepassage of the Toxic Substances Control Act. Test-ing did not increase as much as expected, how-ever, and in the early and mid-1980s the indus-try was operating at 60 to 70 percent capacity (27).This has led to fairly level prices over the pastfew years and, in some cases, price cutting tomaintain market position. Because of this com-petition, current prices reflect the actual costs oftesting. Testing laboratories often do not quoteset prices for some testing procedures or for par-

ticular batteries of tests, preferring to negotiateon a case-by-case basis.

Government ToxicologicalResearch and Testing

The U.S. Government programs with strong tiesto toxicological testing are EPA, the National Cen-ter for Toxicological Research in the Food andDrug Administration, the Centers for Disease Con-trol, and the National Institutes of Health (see table11-3). other programs are not identified with sep-arate budget line items and are dispersed amongvarious agencies and departments.

Table n-3.-Selected Federal Expenditures Reiatedto Toxicoiogicai Testing and Research, 1984-86

(in thousands)

1984 1985a 1986a

Environmental Protection Agency:Program expenses . . . . . . . $327,145 $380,341$376,074

Toxic substances . . . . . . 34,484 39,341 38,660Pesticides. . . . . . . . . . . . . 32,772 37,805 36,948

Research anddevelopment. . . . . . . . . 144,903 195,449 212,061

Toxic substances . . . . . . 12,327 14,450 26,358Pesticides. . . . . . . . . . . . . 1,738 5,121 6,938Interdisciplinary. . . . . . . . 18,522 22,423 14,876

Food and Drug Administration:National Center for

Toxicological Research . 21,132 21,575 22,284Drug program ., . . . . . . . . . . 138,248 153,112 152,430Food program . . . . . . . . . . . 115,541 109,538 113,907Devices and radiologic

products . . . . . . . . . . . . . . 62,568 67,081 68,368

Centers for Disease Control:Occupational safety and

health research . . . . . . . . 54,740 54,863 57,645Research on chronic and

environmental disease . . 25,953 28,568 23,726

National Institutes of Health:National Cancer Institute:

Cause and prevention. . . 276,075 301,655 285,844Detection and

diagnosis . . . . . . . . . . . 63,182 70,524 66,839Treatment . . . . . . . . . . . . . 340,041 367,940 351,683

National Institute of Environmental andHealth Sciences:

Characterization ofenvironmentalhazards . . . . . . . . . . . . . 19,152 21,136 21,601

Applied toxicologicalresearch and testing. . 57,781 57,303 56,737

Intramural research . . . . . . . 48.643 55.051 52.536aEstimates.

SOURCE: U.S. Executive Office of the President, Office of Management and Budg-et, Budget of the United States Government, Fiscal Year IQ% (Washing-ton, DC: U.S. Government Printing Office, 1985),


PROTECTING PROPRIETARY INTERESTS

In commercializing a particular product, animaluse may be limited to toxicity testing, but manyproducts rely on animals in the initial researchphase as well. These research and testing resultshave proprietary value that is sometimes protectedby secrecy, other times by obtaining a patent. Thevalue of data that lead to a particular product maydepend more on the size of the market and its prof-itability than on the cost of obtaining them, par-ticularly when it takes a long time to generate thedata.

Cooperative Research and Testing

Two major competing factors influence thesharing of research and testing costs— the desireto keep information that has proprietary valuesecret and the desire to share the very large ex-penses that may be involved in generating it.These business decisions are only slightly influ-enced by Government policies. Another factor isantitrust law, however, which is greatly affectedby such policies.

In considering the role of antitrust law, it is im-portant to recognize that the results of researchand testing enable society to use resources effi-ciently. Antitrust laws help ensure that theseefficiencies benefit consumers, by preventingmanufacturers, for example, from colluding tomaintain high prices. However, these statuteshave sometimes been applied in a way that im-pedes technological development (3) by makingit difficult for companies to pool resources forresearch so expensive that none would undertakeit alone.

In recent years, antitrust policies have beenchanged or clarified so that resources can bepooled more easily (19). One component of thisis the National Cooperative Research Act of 1984(public Law 98-462). The Sherman and Claytonantitrust acts still apply, but damages in privatesuits are reduced from three times the value ofthe unfair advantage to the actual value. This willcertainly lower the risks involved in collaborat-ing, and probably the likelihood of being sued aswell.

Testing costs can be most equitably shared ifpotential participants can interact before testingbegins rather than after it is completed, becausea party who has already tested may have an un-fair advantage (or disadvantage) in negotiatingcompensation. It is easiest to identify potentialsponsors for a particular chemical when testingis required by a regulatory agency, because it isknown that testing will take place and who is re-quired to test. When industry forms testing con-sortia to share costs, it is most easily done throughexisting trade associations, such as the ChemicalManufacturers Association. Cooperative testingis also conducted by industry through the Chem-ical Industry Institute of Toxicology.

Many testing consortia have been put togetherto negotiate agreements in anticipation of re-quired testing under Section 4 of the Toxic Sub-stances Control Act. Such negotiations were ruledinvalid in a recent case (2 I). Despite this ruling,testing consortia will continue to have appeal solong as testing is expensive and the results havelittle or no proprietary value other than in ful-fillment of regulatory requirements.

Toxicity Testing Data

Many companies begin making other financialcommitments to the commercialization of a prod-uct before testing is completed. Plant design andsmall-scale production may coincide with long-term toxicity testing. The practical costs of ful-filling lengthy testing requirements may greatlyexceed the costs of testing. Thus, it is advanta-geous to be able to use any existing data gener-ated by another laboratory in order to avoid thedelays and uncertainties of testing. Conversely,this provides an incentive to prevent data frombeing made available to competitors.

The protection of pesticide testing data has beenthe subject of much litigation and several amend-ments to the Federal Insecticide, Fungicide, andRodenticide Act (FIFRA). The most recent changesprovide that data submitted after September 30,1978, are protected from uncompensated use for15 years. There are two kinds of protection. One

Ch. 11-Economic Considerations ● 2 5 3

requires that data be shared as long as compen-sation is offered. The terms of the compensationare subject to arbitration if the parties cannotagree. The other protection only applies to newpesticides (new active ingredients), not to new for-mulations of old ingredients. It gives exclusive useof the data to the data owner for 10 years unlessthe data owner explicitly agrees to sell the rightto use the data.

The Supreme Court recently decided in Ruck-elshaus v. Monsanto (26) that these provisions ofFIFRA are constitutional. For data submitted be-fore 1972 or after 1978, there is no expectationof a proprietary interest, thus nothing is taken;for data submitted between those years, the com-pensation and arbitration provision, in combina-tion with the Tucker Act, provides adequate com-pensation. (See also the Environmental ProtectionAgency’s regulations at 40 CFR 1984 ed. 152; 40FR 30884.)

Congress has recognized the important businessinterest in keeping information from competitors,but it also supports the public’s ‘(right to know”and the Federal Government’s need to know. Animportant barrier to the sharing of confidentialbusiness information among agencies is the differ-ing standards and procedures for handling it. Thead hoc interagency Toxic Substances StrategyCommittee, coordinated by the Council on Envi-ronmental Quality, thought it would be necessaryto pass legislation permitting the sharing of con-fidential data between health and environmentalagencies (32). Such legislation would establish aneed-to-know standard, require uniform securityprocedures for the data to be shared, impose uni-form penalties for disclosure, and provide fornotification of the data submitted by the dataholder at least 10 days prior to transfer.

SUMMARY AND CONCLUSIONS

The total dollar cost of animal acquisition andmaintenance is directly related to the length oftime animals stay in the laboratory. With no ac-curate source of data on various species’ lengthof stay, it is impossible to calculate the total costof animal use. Analysis of the factors involved inthe costs of animal acquisition and maintenanceindicates that a reduction in animal use will beaccompanied by a reduction in cost—although theproportionate savings will be less than the propor-tionate decrease in the number of animals used.

Many of the issues involved with using animalsin research and testing have economic implica-tions, although they do not lend themselves wellto rigorous quantitative economic analysis be-cause many considerations are nonmonetary. Ahighly contested concern, for example, is thepropriety of using unclaimed pound animals inlaboratory studies.

An area of animal use that is of major economicimportance is biomedical research, which contrib-utes to health care through the development ofdrugs, medical devices, diagnostic techniques, andsurgical procedures. Health care accounts for

over 10 percent of the Nation’s gross nationalproduct, or $355 billion in 1983. The results ofresearch with animals might also reach the pub-lic through patented products. Although data onhumans may also be required, and althoughnonanimal or in vitro methods are sometimessufficient, many such patent applications use ani-mals to show that the invention is useful.

Another use of animals with economic impor-tance is toxicological testing, used to ensure thatnew products are sufficiently safe. One type ofproduct for which such testing is of major con-sequence to public health is in the developmentof pesticides, which affect virtually all Americansthrough the production and contamination offood. The Environmental Protection Agency hasestimated that 90 percent of all households usesome pesticide product.

Whole-animal tests can be far more costly thanin vitro and nonanimal alternatives, largely be-cause they are labor-intensive. The incentives tofind alternatives to the LD5O and Draize tests areprimarily nonmonetary, however, as these testscan be performed for $1,000 to $2,000. This is

254 . Alternatives to Animal Use in Research, Testing, and Education

in the price range of the cheaper, currently avail-able in vitro and nonanimal replacements.

Most research and testing in the United Statesis financed by Government or industry. Thechemical industry, including the production ofdrugs, has annual sales of over $170 billion andspends over $7 billion on research and develop-ment. An unknown fraction is spent on researchinvolving animals and about $700 million is spenton toxicity testing.

The Federal Government sponsors much bio-medical research and testing involving animals(see ch. 12). An unknown amount leads to the de-velopment or use of alternatives. The Governmentalso has many programs related to testing, includ-ing the evaluation of testing data generated in

other sectors. Agencies with significant budgetsfor such activities include the Environmental Pro-tection Agency, the Food and Drug Administra-tion, the Centers for Disease Control, and the Na-tional Institutes of Health.

The Federal Government also has a special rolein the sharing of data derived from animal use,as the data have proprietary value. First, antitrustlaws and policies affect industry’s ability to sharedata and the costs of generating it. Such sharingis facilitated by the passage of the National Co-operative Research Act of 1984. It is also facili-tated under the Federal Insecticide, Fungicide,and Rodenticide Act and the Toxic SubstancesControl Act.

CHAPTER 11 REFERENCES

1. Auletta, A., Genetic Toxicologist, U.S. Environ-mental Protection Agency, Washington, DC, per-sonal communication, 1984.

2. Baily, M. N., “Research and Development Costs andReturns, the U.S. Pharmaceutical Industry, ’’.l. Po]it.Econ. (Jan. -Feb. ):7O-85, 1972.

3. Baxter, W .F., “Antitrust Law and Technological In-novation,” 1ss. Sci. Tech. 1(2):80-91, 1985.

4. Berk, A., and Paringer, L.C., Economic Costs of W-ness, 1930-1975 (Washington, DC: Public ServicesLaboratory, Georgetown University, May 1977).

5. Bishop, F., Chief, Registration Support and Emer-gency Response Branch, Office of Pesticide Pro-grams, U.S. Environmental Protection Agency,Washington, DC, personal communication, 1985.

6. Brusick, D.J., Director, Molecular Sciences Direc-torate, Litton Bionetics, Kensington, MD, personalcommunication, 1984.

7. Cross v. Izuka, 224 USPQ 745, 1985.8. Chem. Engin. News, “Facts and Figures for Chemi-

cal R&D,” 53(29):28, 1985.9. Gibson, R. M., Levit, K. R., Lazenby, H., et al., “Na-

tional Health Expenditures, 1983,” Health Care Fi-nan. Rev. 6(2):1-29, 1984.

10. Hertzfeld, H. R., and Myers, T.D., “Economic andPolicy Considerations,” contract report preparedfor the Office of Technology Assessment, U.S. Con-gress, 1985.

11, In re Anthony, 56 CCPA 1443,414 F.2d 1383, 162USPQ 594 (1969).

12. In reBergeZ, 48CCPA 1102, 292F.2d 955, 130USPQ206 (1961).

13. ~n re Chihnvsky, 43 CCPA 775, 229 F.2d 457, 108USPQ 321 (1956).

14. In re Irons, 52 CCPA 938, 340 F.2d 974, 144 USPQ351 (1965).

15. In re Jo]]es, 628 F.2d 1322, 206 USPQ 885 (CCPA1980).

16. In re Krimmel, 48 CCPA 1116, 292 F.2d 948, 130USPQ 215 (1961).

17. In re Watson, 517 F.2d 465, 186 USPQ 11 ([Cust.and Pat. App.1 1975).

18, Institute of Laboratory Animal Resources, NationalSurvey of Laboratory Animal Facilities and Re-sources (Washington, DC: National Academy ofSciences, 1980).

19. McGrath, J. P., Remarks at the 18th Annual New Eng-land Antitrust Conference, Boston, MA, Nov. 2,1984.

20. Mushkin, S.J., Biomedical Research: Costs and Bene-fits (Cambridge, MA: Ballinger Publishing Co., 1979).

21. Natural Resources Defense Council v. EPA, 595F. Supp. 1255 (S.D.N.Y., 1984).

22. Payton, N., Massachusetts Society for the Preven-tion of Cruelty to Animals, Testimony in Supportof H1245, Boston, MA, February 1981.

Ch. n-Economic Considerations ● 2 5 5

23.

24.

25.

26.27.

28.

Pimentel, D., Andow, D., Dyson-Hudson, R., et al.,“Environmental and Social Costs of Pesticides: A Pre-liminary Assessment,” Oikos 34:127-140, 1980.Rey-Bel]et v. Englehardt, 493 F.2d 1380, 181 USPQ453 (1974).Rowan, A. N., Of Mice, Models, & Men: A CriticalEvaluation of Am”mal Research (Albany, NY: StateUniversity of New York Press, 1984).Ruckelshaus v. Monsanto, 104 S. Ct. 2862 (1984).Stock Market Surveuy for Hazelton Labs (New York,NY: Silverberg, Rosenthal, and Company, Apr. 9,1984).Turner, J., Reckoni~ With the Beast: Animals, Painand Humam”ty in the Victorian ~“nd (Baltimore, MD:Johns Hopkins University Press, 1980).

29. U.S. Department of Commerce, 1982 Census ofManufactures, Drugs (Washington, DC: U.S. Gov-ernment Printing Office, 1985).

30,

31.

32.

U.S. Department of Health and Human Sert~ices,Public Health Service, Food and Drug Administrat-ion, National Toxicology Program, Board of Scien-tific Counselors, Draft Report-Ad Hoc Panel onChemical Carcinogenesis Testing and E~raluation(Bethesda, MD: Feb. 15, 1984).U.S. Environmental Protection Agency, Regulatoqt~Impact Analysis: Data Requirements for Register-ing Pesticides Under the Federal Insecticide, Fungi-cide and Rodenticide Act, EPA 54019-82-013 (Wash-ington, DC: 1982).U.S. Executive Office of the President, Council onEnvironmental Quality, Toxic Substances StrategyCommittee, Toxic Chemicals and Public Protection(Washington, DC: U.S. Government Printing Office,1980).

Documents

Chapter Economic Considerations