Behavioral incentives, equilibrium endemic disease, and health management policy for farmed animals.

Infectious animal diseases that are endemic, or common, in a region generate a variety of significant adverse economic consequences for that region. Most directly, mortality, morbidity, barrenness, and miscarriage in production animals reduce technical efficiency. Costly treatments and altered management practices to ameliorate these losses also reduce profitability. Bennett and IJpelaar (2005) use an accounting methodology to estimate such direct and treatment costs for thirty-four animal diseases in Great Britain. At 180 [pounds sterling] (sterling) per year, they estimate that mastitis in cows creates the largest costs in these categories. Opportunities for trade within and between regions may also be curtailed. In addition, some infectious animal diseases, such as bovine tuberculosis, brucellosis, and possibly Johne's disease, have adverse consequences for human health (Myers and Steele 1969; National Academies of Sciences 2003, 2005).

For these and other reasons, most countries invest in veterinary public health infrastructure. At the transnational level, the United Nations, through its Food and Agricultural Organization and World Health Organization units, seeks to facilitate better management of infectious animal diseases. The OIE, also called the World Organization for Animal Health, is also funded by countries and has more emphatic objectives in this regard. Many public control policies, such as animal quarantine, human movement controls, border inspections, vaccinations, and mandatory testing schemes also involve economic losses. Economic losses can arise in disparate ways. For example, Horan and Wolf (2005) show how feeding wild deer can increase both deer use value and bovine TB prevalence. A recent National Academies report (2005) has identified several areas of concern in the United States animal health infrastructure, including the need to develop models of biosecurity systems and strategies for indigenous and exotic diseases. (1)

A large applied modeling literature has emerged at the interface of preventative veterinary medicine and economics. For instance, Mahul and Gohin (1999) seek to understand cost control and eradication strategies upon the event of an epidemic outbreak of Foot and Mouth disease in Brittany, France. Destroying infected herds, while costly, may be preferred to vaccination. This is because a region using a vaccination strategy will subsequently have difficulty selling produce into external markets. Chi et al. (2002), on the other hand, compare the effectiveness of prevention strategies for four endemic cattle diseases in the maritime provinces of Canada to find that privately optimal strategies vary by disease.

Bicknell, Wilen, and Howitt (1999) include private incentives in their analysis of public intervention to control bovine tuberculosis in New Zealand. They find that the policy mix then in place, together with private actions, reduces disease prevalence relative to private actions alone. Kuchler and Hamm (2000) study the effectiveness of an indemnity for reporting infection in a sheep scrapie eradication program that operated in the United States for forty years until 1992. They show that the supply of reported infections was elastic to indemnity price offered. Surprisingly, with the exception of these latter two articles and a pair of papers to be discussed below, scholarship appears to have been silent on characterizing the economic nature of the equilibrium extent of infectious endemic disease. (2)

That there is an economic dimension to endemic animal disease becomes apparent upon perusing any introductory animal/poultry production book, such as Ensminger (1992) or Gillespie (2002). Costly management strategies such as selective purchasing of feeder animals, implementing labor-consuming hygiene regimes, and timely equipment replacement are advocated. Beyond this, infection is an externality of a very public variety. The private and social benefits of protecting against a disease are likely to diverge so that the marginal private value of reducing the probability of contracting a disease is lower than the marginal social value of reducing this probability. Each grower is, when determining the risk of contracting a disease, a private provider of the public bad that is the stock of disease in a region. This perspective might suggest that game theory methods intended to model a large number of players in the presence of public externalities should be relevant to the problem at hand. Specifically, these methods should hold promise for better understanding the extent of endemic animal disease and how to manage it. The intent of this article is to build a model to this end.

Hennessy (2005) has considered private actions to guard against spatial spread of a disease already in a region to conclude that the way in which farm actions behave as local substitutes can lead to peculiar spatial patterns in taking protective actions. That article also considered the risk of disease entry into a region. Then efforts by producers are more likely to complement, so that policies to promote inter-farm communication should be beneficial.

The work most closely related to the content of the present article is Hennessy, Roosen, and Jensen (2005), in which two models are developed to address the strategy of internally supplying feeder animals for fattening. One model looks at the externalities created by trading to take private advantage of feeder animal production cost differentials. The other looks at the internal organization of production to protect against the risk of disease entry into a farm. Both models are nontemporal in structure, viewing static farm decisions where no distinction is made between farm disease statuses. This is an important limitation because in reality farms differ in the extent of disease. Farms transition between disease-free and diseased conditions over time. Much of public disease management policy is intended to alter the probabilities of transition, i.e., to reduce the probability of transition to diseased status and to increase the probability of transition to disease-free status. As a result of this nontemporal structure, the models are very limited in what they can say about the nature of incentives to protect against disease and the consequences of such control practices as testing and movement controls.

In this article, we will develop a continuous-time dynamic model of farm-level capital values in which disease status is influenced by farm actions, but is still stochastic. The approach is to use a stochastic model of transitions between two disease states in order to value farms in either state, and so to characterize incentives to change the state transition probabilities. Similar models have been used elsewhere in economics, where the best-known application is perhaps that of efficiency wage and involuntary unemployment by Shapiro and Stiglitz (1984). (3)

Our analysis points to the possibility of a multiplicity of equilibrium disease levels. It also suggests that public disease management programs could conceivably improve social welfare even if they have no direct or spillover impacts on the extent of a disease. This could occur by encouraging farmers to protect against becoming entangled in the bureaucracy of acquiring disease-free status. We show it is also possible that one class of disease management innovations could reduce social welfare. This class is comprised of innovations that increase the probability of transition from diseased status to disease-free status. The anomalous effect is due to a reduction in the loss expected from becoming diseased when externalities ensure that the level of protection against disease is socially inadequate.

We also apply our model to better understanding indemnity payments to report infection when consuming the produce poses a human health risk. Conditional on biosecurity effort levels, the indemnity can be set to optimally trade social gain from more marketed output with social loss from allowing suspect produce onto the market for consumption. But the indemnity reduces the private incentive to biosecure, and so the overall effect of the policy is unclear.

The article's outline is as follows. We will first explain what sorts of animal disease our model is intended to shed light on. The model is then presented, and policy implications are developed. Suggestions are provided concerning empirical implementation using duration analysis. A brief discussion concludes.

Disease Issues

Diseases causing economic harm to farmed animals can take many forms. Some diseases, such as those arising from nutritional imbalance or genetic abnormalities, are not at all infectious. In other cases, environment, genetic make-up, and infectious agents may all contribute as causal factors. For infectious diseases, the presence of the infectious agent is a pre-requisite and the local environment can be very important in determining the likelihood of presence. Some infectious diseases of economic importance are exotic to a region due to climate or geography. A prior successful public eradication strategy may determine the disease's exotic status, as may man's influence on the environment. The production characteristics in the region (e.g., low animal density), and prevention strategies employed in the region may affect the risk of entry. But while these diseases can be disastrous for producers in a region, if entry probability is low then growers generally will not be concerned about controlling for the disease on a daily basis. Growers may be especially reassured if they have confidence that the government will provide catastrophic insurance in the event of an outbreak.