PIGS, PEOPLE, AND PATHOGENS: A SOCIAL WELFARE FRAMEWORK FOR THE ANALYSIS OF ANIMAL ANTIBIOTIC USE POLICY

Transcription

1 PIGS, PEOPLE, AND PATHOGENS: A SOCIAL WELFARE FRAMEWORK FOR THE ANALYSIS OF ANIMAL ANTIBIOTIC USE POLICY PAUL E. MCNAMARA AND GAY Y. MILLER...the US and all other nations should follow Europe and ban all antibiotic use in animal feed. (Daily and Erlich, p. 334) The magnitude of the actual animal-tohuman transfer problem and associated development of disease is poorly characterized and varies greatly because of food-processing and consumer handling practices that are separate from animal production or antibiotic use in food production. (National Research Council, p. 143) How will new infections be controlled in food animals if not with increased availability of antibiotics? (National Research Council, p. 143) Use of antibiotics in animals faces increasing public scrutiny. Antibiotics are at the center of a swirling debate, whose participants include a wide range of actors, including public health officials, agricultural producers and farm groups, drug companies, doctors, veterinarians, elected representatives, government agencies, ecologists, environmentalists, and scientists. While the concerns expressed about animal antibiotic use (AAU) have a number of origins, the primary source of concern arises from the public health community, which faces increasing cases of antibiotic-resistant infectious diseases and increasing health care costs (not to mention the value of increased mortality and morbidity) associated with the more expensive treatments Paul E. McNamara is assistant professor in the Department of Agricultural and Consumer Economics, University of Illinois at Urbana Champaign; Gay Y. Miller is professor in the Departments of Veterinary Pathobiology, Agricultural and Consumer Economics, and Veterinary Clinical Medicine, University of Illinois at Urbana Champaign. Our research program on antibiotic use in pork production is supported by the Illinois Council on Food and Agricultural Research. This article was presented in a principal paper session at the AAEA annual meeting (Long Beach, CA, July 2002). The articles in these sessions are not subjected to the journal s standard refereeing process. necessary to deal with multiple-drug resistant infections. While agricultural economists have contributed to the debate so far, especially through studies which estimate the impact on producers of a ban on feedgrade use of antibiotics, the central argument of this paper is that the toolkit of agricultural and resource economists has much more to offer policy makers involved in the AAU question. Specifically, we propose that the social welfare framework can serve as an analytic framework for examining whether or not specific policy proposals to modify AAU would benefit society, and we outline the issues that need to be addressed in order to construct detailed welfare analyses of such policy proposals. Antibiotic resistance differs qualitatively from most other foodborne illness problems, in large part because the public health concern not only focuses on the foodborne illness but also on bacteria, their susceptibility to various antibiotics, as well as their ability to move through a variety of environments and pathways and to transfer their acquired resistance. As titles such as Emergence, Spread, and Environmental Effect of Antimicrobial Resistance: How Use of An Antimicrobial Anywhere can Increase Resistance to any Antimicrobial Anywhere Else indicate (O Brien), microbiologists emphasize that the resistance issue is essentially one of microbial ecology. In the ecological perspective, food animals and farms serve as only one of many reservoirs of bacteria, with billions and billions of genetic events occurring (replication and gene transfer) on farms. These farm reservoirs are a small fraction of the microbial ecological environment. In the face of antimicrobial use, each genetic event creates some possibility of the development of a genetic expression of resistance. The ability for antibiotic-resistant bacteria strains to spread around the world poses one portion of the risk, while the ability of the antibiotic-resistant bacteria strain to spread resistance genes to other Amer. J. Agr. Econ. 84 (Number 5, 2002): Copyright 2002 American Agricultural Economics Association

2 1294 Number 5, 2002 Amer. J. Agr. Econ. strains of bacteria represents another portion of the risk. Another wrinkle in this debate over AAU concerns terms and definitions. We define antibiotic use as subtherapeutic if the use is to improve growth performance and feed efficiency (growth promotion), and therapeutic if the drug is used to treat a specific infectious disease (National Research Council). However, others use slightly different terms or propose new terms. McEwen and Fedorka-Cray, for instance, recently proposed using the term nontherapeutic to take the place of subtherapeutic, where nontherapeutic includes use for growth promotion and feed efficiency, as well as disease prophylaxis. Their definition of nontherapeutic (or subtherapeutic) refers to uses of antibiotics in feeds at levels typically below 200 grams per ton for periods of time exceeding two weeks. Along with terms and definitions, information such as volume of antibiotics used in pork production in the United States has also been the subject of debate. As The Union of Concerned Scientists reported in Hogging It: Estimates of Antibiotic Use in Livestock (Mellon, Benbrook, and Benbrook) that the swine industry applied roughly 10.3 million pounds of nontherapeutic antimicrobials annually, with 70.3% of these antimicrobials being used at the finishing stage of pounds in the late 1990s. The Union of Concerned Scientists report estimates that overall, on an annual basis, AAU is 24.6 million pounds of antimicrobials (nontherapeutically or in the absence of evidence of disease), with pork production accounting for 41.9% of AAU. The report estimates that the poultry sector uses 10.5 million pounds, while the beef sector uses 3.7 million pounds nontherapeutically. Levy reports that over 50 million pounds of antibiotics are produced in the United States (as of the late 1990s) and that human treatments account for approximately half of all antibiotics consumed in the United States. While feeding antibiotics at subtherapeutic levels would appear to generate exactly the right conditions to select for resistance, producers find that even at relatively low levels of application, the bacteriocidal and bacteriostatic effects lead to economically significant gains in growth and feed use efficiency in pork production (Zimmerman). Dewey et al. (1999), using National Animal Health Monitoring System data, report 88% of swine producers used antimicrobials in their feeds in Producers were more likely to use antimicrobials with young pigs because of their heightened risk for disease and greater growth potential. They also found that the most commonly used antibiotics were tetracyclines, carbadox, and bacitracin and that most antimicrobials were fed on a continuous basis (Dewey et al., 1999). The clinical and economic context of veterinary practice raises other important dimensions of the AAU issue. As opposed to most human medicine, the veterinarian cannot talk with the patient in developing a diagnosis. Furthermore, given the sheer numbers of animals involved as well as the cost of veterinary services, treatment often occurs at the level of a herd or flock (Gustafson and Bowen). That is, when the farmer or veterinarian detects that a significant portion of the herd is affected with a disease where antibiotics are one of the appropriate therapies, the treatment is often delivered to the entire herd. The value at risk is the entire herd, especially if the disease is contagious. Also, there may be a lower per unit cost of administering drugs to the entire herd rather than by individually injecting animals. A primary contribution of applied economists is an analytic framework that incorporates both producer and consumer welfare, and that introduces an explicit role for public policy. We outline a basic social welfare maximization framework below and discuss the issues involved with extending this framework to additional levels of detail, including epidemiological detail and information on economic and biological dynamics. We then use the social welfare framework to organize the discussion of issues faced by producers, consumers, and policy makers as they make decisions concerning AAU. Since our research program focuses on antibiotic use in pork production, we use the pork system as a model throughout our analysis. The paper concludes with observations about policy options and future research needs in the area of economic epidemiology of the public health risk posed by AAU. Social Welfare Maximization Problem Viewing the susceptibility of bacteria to antimicrobial drugs as an open access resource shared by both producers and consumers leads to economic models with a social planner in the role of determining production and consumption plans so that utility is maximized in the economy. Such a social welfare maximization approach highlights the externality associated

3 McNamara and Miller Pigs, People, and Pathogens 1295 with a production technology like antibiotic use and it serves as an organizing framework for examining the entire antibiotic use issue. Brown and Layton develop a dynamic model of the optimal use of antibiotics in people and animals and discuss the social planner s problem. In our static model of the susceptibility resource, we assume there is one strain of bacteria and one type of antibiotic. The strain of bacteria is characterized by a susceptibility level S and the antibiotic (A a denotes AAU, and A h denotes human antibiotic use) is assumed to be more costly to administer as the bacteria develop resistance. Furthermore, we assume that there is one pool of susceptibility and it is shared across animal use and human use. Consumer s Problem The consumer is assumed to maximize a wellbehaved utility function that is a function of health status (H) and food (F): (1) max U = U(F, H(A h(s))) A h,f s.t. F + p A A h = Y. Here, p A is the price of antibiotics per dose, A h denotes the amount of antibiotics (doses) used in producing health status, and Y represents income. Prices have been normalized so that the price of a unit of food is equal to 1. Producer s Problem The producer is assumed to produce food animals according to a neoclassical production technology that is a function of grain (G) and antibiotics (A a ). The producer takes prices parametrically and is assumed to maximize profits. The producer s problem is (2) max = F p A A a p G G A a,g s.t. F = f (G, A a ). Here, p A is the price of antibiotics per dose and p G is the price of grain per unit. Social Planner s Solution The social planner aims to maximize net societal welfare in this Robinson Crusoe-type economy taking into account the susceptibility externality. We assume that the depletion of susceptibility occurs according to a linear function of A a and A h : (3) S = A h A a, with + > 0. The social planner takes the externality into account and finds a consumption plan for the consumer and a production plan for the producer that will maximize net benefits for society. To find this solution the social planner maximizes the following Lagrangian: (4) max L = U(F, H(A h(s))) A h,a a,g,f + 1 (Y F p A A h ) + 2 ( f (G, A a (S)) p A A a p G G) + 3 (S A h A a ). The first-order necessary conditions yield expressions which characterize the optimal solution and they have an economic interpretation. The Lagrangian multiplier 3 provides the implicit social value of relaxing the susceptibility constraint marginally. The 3 term figures centrally in the comparison of the social planner s optimal solution with the production and consumption plans determined myopically by the farmer or the consumer. Compared to the myopic consumption optimum, the social planner s consumption plan will take into account the marginal social cost of human antibiotic use, weighted by the parameter, which shows the incremental decline in susceptibility due to use of one dose of antibiotics in humans. Similarly, the social planner s production plan adjusts for the marginal social cost of AAU, weighted by the parameter, which shows the incremental decline in susceptibility due to use of one dose of antibiotics in animals. Overall, even this simple model of the antibiotic externality raises a number of important issues for policy discussion. First, compared to the myopic consumption and production plans, the social planner s use of antibiotics is likely to be lower because the marginal social costs of antibiotic use are taken into account. Self-interested consumers and producers will ignore the social costs of their choices, and a social planner can improve overall welfare by defining production and consumption plans, consider the antibiotic use externality, or by crafting a set of prices that have the same effect. Second, achieving the optimal resource allocations in the social planner s problem requires knowledge of the biology of the externality (for e.g., the size of alpha and beta) and that is precisely what is at debate. The model embodies a number of critical environmental and biological assumptions: the common

4 1296 Number 5, 2002 Amer. J. Agr. Econ. susceptibility pool; the single antibiotic; the functional form and parameter values for the decline in susceptibility function; and, the relationship between susceptibility and antibiotic productivity in producing human health and in animal production. The model allows for AAU to generate a burden for consumers producing human health as well as vice versa, where antibiotic use in the production of human health might generate a decrease in susceptibility that disproportionately affects agriculture. Extending this framework to consider dynamic issues allows a number of further insights. Brown and Layton develop a dynamic model of bacterial susceptibility as an exhaustible resource that considers the potential impact of AAU, as well as antibiotic use in human medicine. Their analysis leads to the conclusion that the social planner will allocate antibiotics at a slower rate than use when the social costs of antibiotic use are not considered. They point out that the social planner also needs to know how resistance transfers from animals to people (or vice versa), as well as the marginal benefits of AAU and antibiotic use in humans. Other dynamic models of antibiotic use also point out timing strategies and optimal antibiotic use when there are multiple antibiotics available to treat one disease. Laxminarayan and Brown show that in the human medicine case of treating a bacterial infectious disease for which two antibiotics are available, the optimal use plan is a function of the difference in rates at which the bacteria develop resistance to each drug and the relative difference in drug costs. Another analysis (Laxminarayan and Weitzman) argues that standard uniform treatment guidelines that do not take into account the development of drug resistance may be dominated by a mixed treatment policy using multiple drugs. In their population ecology model of growth of human commensal bacteria, Smith et al. model mathematically the impacts of using the same antibiotics in human medicine and animal agriculture. They give an illustrative example using their model of the spread of vancomycin-resistant enterococci in hospitals, and their analysis treats agricultural use of avoparcin (the same antibiotic as vancomycin used for animal growth promotion) as well as vancomycin use in the medical context for humans. They report the results of a simulation where, relative to the base case without avoparcin use (and with medical use of vancomycin), the agricultural use of avoparcin hastens the development of an equilibrium vancomycin-resistant enterococci population in humans by three years. They note that since the differences between equilibrium prevalence in the pristine state (no AAU) and the AAU scenario are small, ordinary surveillance data of prevalence may not have sufficient statistical power to identify an increase in antibiotic resistance that AAU generates. Furthermore, because of the nature of the microbial population dynamics, once an increase in antibiotic resistance prevalence was detected by a monitoring system, it may be too late to take meaningful measures (change AAU patterns or some other policy change) in response. Thus, these authors argue that public health benefits may be generated by restricting AAU until widespread antibiotic resistance is observed in humans, and they emphasize that this principle holds particularly for new classes of drugs. While they acknowledge that their study is not a quantitative risk assessment, their paper is important since it develops a starting point for economic-based risk assessments of AAU and draws attention to the importance of ecological dynamics. Despite their study s limitations, such as undefended parameter estimates and no explicit consideration of the stochastic nature of genetic events and their role in microbial ecology, it makes a contribution through the application of mathematical population biology to demonstrate how population dynamics might influence optimal drug policy. Toward a Social Welfare Analysis Our interest is in developing analyses of public policy proposals regarding AAU that quantify the net social benefits of a change in policy and identify economically efficient means of managing risks. How far are we from being able to conduct a cost-benefit analysis of a proposed policy change? What missing information, if obtained, would ensure a credible study? Producer s Side The economic model of the pork producer s side of the social planner s problem requires information on the derived demand for antibiotics at the farm level as well as information on the market for pork. Fortunately, a number of studies estimate the economic impact on producers of a ban on AAU in feeds (Gilliam and Martin, Mann and Paulsen, Wade and Barkley,

5 McNamara and Miller Pigs, People, and Pathogens 1297 Hayes et al.). Gilliam and Martin assume that producers would maintain pre-ban production levels and use production and market conditions in 1973 to estimate the impact on production costs. They further assume that for pigs between 15 and 40 pounds, growth rates would decline by roughly 23% and feed-use efficiency would drop by approximately 6.5%. In the grower-finisher range (hogs weighing more than 40 pounds), they assumed growth rates would decline by 5.5% with a ban and that feed-use efficiency would drop by about 2%. Based upon these assumptions, their analysis estimates that producer costs would increase by $533 million ($6.94 per hog), based upon cost and market conditions in Mann and Paulsen s analysis assumed that growth rates after a ban on antibiotics in pork production would drop by 10.7% and feed efficiency would decline by 3.8%. They assume that a drop in production after the ban would increase hog prices, and that this effect would more than offset the production cost increase, leading to an increase in profits of producers of 4.5% ($4.42 per hog) in the short run. After industry adjustments, they projected that producer profitability would return to pre-ban levels. Hayes et al. conducted a study of the economic impact of a ban on antimicrobial feed additives and based the productivity estimates on the experience following the ban on overthe-counter antibiotics in Sweden in Their most likely scenario assumes that average daily gain would decline following a ban by 1.3% for pigs weighing pounds, and 1.8% for pigs above 100 pounds. They further assumed that for pigs in the pound range feed efficiency would drop by 1.7%, while for pigs above 100 pounds feed efficiency would drop by 1.5%. They also assumed that a ban would lead to mortality increases of 1.5 percentage points for piglets and 0.04 percentage points for grower-finishers. Their mostlikely scenario predicts that a ban on feedgrade antibiotics would raise costs by $6.05 per hog in the short run and that over a period of ten years the impact would change to $5.24 per head. They estimate that producer profits would decline initially by $4.17 per hog, but that after adjustments the impact on profits would decline to $0.79 per hog after ten years. Another study found improved productivity and improved profits ($0.59 per hog marketed) from the use of antibiotics in feed when considering only the finisher stage of production (Miller et al.). On an industry level, a decline of $0.79 per hog represents a decline in net profits of roughly $73.5 million (assuming 93 million marketed pigs) annually. These studies assume that the value to producers of the feedgrade antibiotic input can be measured from ex post changes in productivity, but it is likely that antibiotics modify the risks to herd health that a producer faces. Gustafson and Bowen point out that even subtherapeutic antibiotic use may have a disease-preventing effect, through the bacteriocidal and bacteriostatic effects on harmful bacteria, and if this effect modifies the distribution of net returns that producers may realize, the estimates of the value to the industry of the input may be understated by the producer s willingness to pay to reduce risk. Consumer s Side Medical experts recognize that the most important source of the antibiotic resistance problem facing health care consumers is inappropriate antibiotic use in human medicine. For example, Robert Besser, Medical Director of the Centers for Disease Control and Prevention s National Campaign for Appropriate Antibiotic Use, states: The biggest problem is inappropriate prescribing of antibiotics. Up to 40% of antibiotics prescribed in doctor s offices are for viral infections, which are not treatable with antibiotics. There are many reasons for this, including demand from patients, time pressure on physicians, and diagnostic uncertainty (Besser) The magnitude of the resistance problem in human medicine is significant and appears to be growing. In public health, numerous examples exist of diseases which once were easily treatable by antibiotics but now pose significant challenges because of antibiotic resistance. Infectious diseases such as tuberculosis and gonorrhea serve as examples of this trend. For instance, the treatment response to gonorrhea has been complicated by the emergence of drug-resistant Neissera gonorrhoeae, which has raised treatment costs by a factor of 15 from a base of roughly a few dollars per infection (Office of Technology Assessment). Hospitals have also seen an increase in antibiotic-resistant nosocomial infections, including the well-known cases of methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant Enterococci (VRE). Phelps investigated the costs

6 1298 Number 5, 2002 Amer. J. Agr. Econ. of antibiotic-resistant bacteria and estimated that they cost the United States between $0.1 billion and $30.0 billion annually, depending on the value of human life used. The CDC has reported that the cost of nosocomial infections is $4.5 billion per year, including both antibiotic-resistant and susceptible cases (Office of Technology Assessment). With respect to consumer behavior and foods that present a potential food safety hazard, evidence exists suggesting that consumers have preferences for avoiding or reducing risks that relate to biological hazards from foods. Miller and Unnevehr found using survey data of over 600 households that consumers with children and older consumers were two groups with increased concerns about the safety of pork products. Additionally, they found that older consumers who had higher concern about pork safety also had a positive willingness-to-pay for an enhanced safety pork product. Findings of this type suggest that groups which in fact have been documented to be at higher risk for foodborne illness, that is, both very young and elder populations, also have increased concern because they recognize that their risk is not just perception, but a reflection of a real increase in vulnerability of household members. General food safety has important implications related to antimicrobial resistance. One mechanism for exposure to resistant organisms is through direct consumption of resistant microbes from contaminated products. Another, and probably more important, hazard is the possibility of selection for a resistant microbe that is present in a person s flora (from foodborne or some other exposure source) because of treatment with an antibiotic. While evidence concerning the link between subtherapeutic use of antibiotics in animal agriculture and human health is in great debate, consumer food safety practices play an important role in reducing any risk of exposure to antibioticresistant bacteria. Social Planner s Problem The debate and uncertainty over the biology of the resistance-transfer issue is the fundamental roadblock for constructing a benefitcost analysis or social welfare analysis of AAU. While scientists can detect and identify genetically antibiotic-resistant bacteria, proving the source and direction of movement of antibiotic-resistant bacteria within the bacterial biosphere is extremely difficult, if not impossible. Thus, scientists disagree about the magnitude of the risk to human health posed by AAU. In this environment of fundamental uncertainty and lack of scientific consensus, what direction should social welfare analyses of agricultural antibiotic use take? Several possible responses exist. First, applied economists interested in the antibiotic and food system issue should realize that many research questions exist that do not require addressing the precise nature of the resistance-transfer risk. Addressing questions such as the identification of optimal on-farm strategies for reducing antibiotic resistance transfer to humans will allow the construction of a knowledge base for future economic research. Second, scenarios might be constructed using different scientific views on the nature of the resistance-transfer risk. Third, social welfare analyses might be conducted in collaboration with microbiologists and epidemiologists to build biologically detailed models incorporating potential risks and a wide breadth of parameter values to reflect the scientific disagreement. However, at this stage, without strong agreement among scientists about the risk posed by AAU, any conclusions from a social welfare analysis will necessarily reflect the lack of scientific consensus on the issue. Given the historical value of antibiotics in human medicine that is potentially at risk, investing in studies concerning microbial ecology and the development of antibiotic resistance would appear to be economically justified. To that end, the Food and Drug Administration (FDA), US Department of Agriculture, and the Centers for Disease Control and Prevention (CDC) have implemented the National Antimicrobial Resistance Monitoring System (NARMS) that monitors changes in antibiotic resistance in zoonotic pathogens from specimens collected from humans and animals (including specimens obtained from slaughter and processing plants). Additionally, scientists from veterinarian medicine, microbiology, animal science, and other disciplines are conducting myriad research projects in the area of antibiotic resistance and animal agriculture. Susceptibility management has parallels to other open access resource management problems though with the added complication of significant uncertainty. In this context economists have observed the development of cooperative responses. There are cooperative

7 McNamara and Miller Pigs, People, and Pathogens 1299 responses which are being used for enhancing food safety and indirectly influencing the development of antimicrobial resistance. Some cooperative responses being used by the swine industry include USDA/AMS Quality Systems Certification Program to verify aspects of the production process, Pork Quality Assurance (PQA) Program to educate producers and guide production practices that will contribute to the reduction/control of chemical and physical hazards, and system-specific on-farm Hazard Analysis Critical Control Programs (HACCP) (Unnevehr, Miller, and Gomez). Other cooperative responses target the development of antimicrobial resistance directly such as the judicious use guidelines for antimicrobials, developed by FDA s Center for Veterinary Medicine in cooperation with the American Veterinary Medical Association (AVMA), for swine veterinarians and pork producers, in addition to a number of other similar publications of this style for other important groups. Conclusions Agricultural economists have the ability to make a unique contribution to the AAU debate by analysis of policy proposals that would restrict or modify antibiotic use. While the lack of a scientific consensus may lead to welfare change estimates over a very wide range (including no human health impact and benefit), this paper has argued that the social welfare framework highlights many important economic aspects of the debate. Further, we have argued that policy analyses that estimate the potential value of a policy change are necessary to help place the antibiotic resistance and AAU issue in a context where it can be compared with other societal risk management questions. Enhancing our economic and epidemiological models yields a number of observations that may move the debate further. First, a number of policy alternatives exist concerning AAU. Our simple model and its extensions point out a number of possible strategies, such as education for farmers and veterinarians, resistance monitoring, moving from over-the-counter use to veterinary supervised/ prescribed use, permits, taxes, substitution strategies so that other inputs are substituted for AAU, certification programs for producers including both use and antibiotic-free certification programs, targeted bans, protection of new antibiotics for human use only, and antibiotic pricing policies (user fees) which effectively limit or decrease antibiotic use, both for humans and for agriculture. Second, the existing models (economic or biological) of the potential externality associated with AAU deserve a rigorous critique. Model conclusions rest squarely upon biological assumptions that have not been proven and are the subject of intense debate among scientists. It appears possible to reconfigure these models with parameter values such that animal well-being is affected by overuse of human antibiotics or so that the potential affect on human use antibiotics is so small that the loss in value associated with a ban of AAU would outweigh the benefits. Last, the segmentation of the antibiotic resistance debate into more focused debates on protocols for preventing overuse of antibiotics in humans, reducing nosocomial infections, and AAU obscures solutions to extending the life of our best antibiotics, such as an understanding that microbes do not recognize country borders, and that antibiotics are also used in pets and for crop production. Antimicrobial use guidelines implemented in the United States do not prevent the development of resistance in other areas of the world. Transmission of these microbes to populations of humans and animals in the United States may represent a significant infectious disease risk. References Besser, R. Personal statement from the Medical Director for CDC s National Campaign for Appropriate Antibiotic Use. Available at accessed on 7 April Brown, G., and D.F. Layton. Resistance Economics: Social Cost and the Evolution of Antibiotic Resistance. Environ. Develop. Econ. 1(July 1996): Daily, G.C., and P.R. Erlich. Impacts of Development and Global Change on the Epidemiological Environment. Environ. Develop. Econ. 1(July 1996): Dewey, C.E., B.D. Cox, B.E. Straw, E.J. Bush, and H.S. Hurd. Use of Antimicrobials in Swine Feeds in the United States. Swine Health and Prod. 7(1999): Gilliam, H.C., and J.R. Martin. Economic Importance of Antibiotics in Feeds to Producers and Consumers of Pork, Beef and Veal. J. Anim. Sci. 40(1975): Gustafson, R.H., and R.E. Bowen. Antibiotic Use in Animal Agriculture. J. Appl. Microbiol. 83(1997):

8 1300 Number 5, 2002 Amer. J. Agr. Econ. Hayes, D.J, H.H. Jensen, L. Backstrom, and J. Fabiosa. Economic Impact of a Ban on the Use of Over-the-Counter Antibiotics in U.S. Swine Rations. Staff Report 99 SR90, Center for Agricultural and Rural Development, Iowa State University, December Laxminarayan, R., and G.M. Brown. Economics of Antibiotic Resistance: A Theory of Optimal Use. J. Environ. Econ. and Manage. 42(September 2001): Laxminarayan, R., and M.L. Weitzman. On the Implications of Endogenous Resistance to Medications. J. Health Econ. 21(July 2002): Levy, S.B. The Challenge of Antibiotic Resistance. Sci. Amer. 278(March1998): Mann, T., and A. Paulsen. Economic Impact of Restricting Feed Additives in Livestock and Poultry Production. Amer. J. Agr. Econ. 58(February 1976) McEwen, S.A., and P.J. Fedorka-Cray. Antimicrobial Use and Resistance in Animals. Clin. Infect. Dis. 34(2002):S93 S106. Mellon, M., C. Benbrook, and K.L. Benbrook. Hogging It: Estimates of Antimicrobial Abuse in Livestock. Report issued by the Union of Concerned Scientists, Available at accessed on 10 March Miller, G.Y., K.A. Algozin, P.E. McNamara, and E.J. Bush. Productivity and Economic Impacts of Feed-Grade Antibiotic Use in U.S. Pork Production. Working paper, University of Illinois, Department of Agricultural and Consumer Economics, Miller, G.Y., and L.J. Unnevehr. Characteristics of Consumers Demanding and Their Willingness to Pay for Certified Safer Pork. J. Agribus. 19(Fall 2001): National Research Council, Board on Agriculture, Panel on Animal Health, Food Safety, and Public Health, Committee on Drug Use in Food Animals. The Use of Drugs in Food Animals: Benefits and Risks. Washington DC: National Academy Press, O Brien, T.F. Emergence, Spread, and Environmental Effect of Antimicrobial Resistance: How Use of An Antimicrobial Anywhere Can Increase Resistance to Any Antimicrobial Anywhere Else. Clin. Infect. Dis. 34(2002):S78 S84. Office of Technology Assessment. Impacts of Antibiotic-Resistant Bacteria. Congress of the United States, OTA-H-629, September Phelps, C.E. Bug/Drug Resistance. Sometimes Less is More. Med. Care 27(1989): Smith, D.L., A.D. Harris, J.A. Johnson, E.K. Silbergeld, and J.G., Morris, Jr. Animal Antibiotic Use has an Early but Important Impact on the Emergence of Antibiotic Resistance in Human Commensal Bacteria. Proceed. Nat. Acad. Sci. 99(2002):1 6. Unnevehr, L.J., G.Y. Miller, and M.I. Gomez. Ensuring Food Safety and Quality in Farm-level Production: Emerging Lessons from the Pork Industry. Amer. J. Agr. Econ. 81(December 1999): Wade, M.A., and A.P. Barkley. The Economic Impacts of a Ban on Subtherapeutic Antibiotics in Swine Production. Agribusiness 8(1992): Zimmerman, D.R. Role of Subtherapeutic Levels of Antimicrobials in Pig Production. J. Anim. Sci. 62(1986):6 17.