Modeling the Costs of Food Safety Regulation

Abstract

Benefit-cost analysis has long been part of the economist's toolkit. Arising out of the need to provide guidance for decision makers about the optimal level of public investment in public goods and services, the principles of benefitcost analysis have been used (and, some would argue, abused) for decades. Since much of the use of benefit-cost analysis has been in realm of public goods, environmental services, and the like, the professional literature on benefitcost analysis has historically stressed two concerns: (a) the development of benefit estimates for goods, like reduction of safety risks or the preservation of natural habitats, that have no direct market valuations; and (b) the treatment of temporal effects through the discounting of benefit and cost estimates to present values. Relatively less emphasis has been placed on cost analysis. This allocation of effort should not be surprising, because benefit estimation and discounting each present a series of difficult analytical issues. In particular, the need to provide estimates of benefits for goods where no market exists (or where market failure is pervasive, such as food safety concerns) has stimulated a rich intellectual growth. By contrast, the principles embodied in cost analysis are well understood; it is in the application that things often go wrong.' Benefit-cost analyses are intended to be used for two related purposes. As Antle (1995) advocates, they can be used in a gatekeeper role by an oversight group charged with deciding whether a new regulation ought to be imposed, in which case the goal is to determine whether the regulation's likely societal benefits outweigh the costs. Benefit-cost analysis can also be used in a rule design role by a regulatory agency as a means of structuring the formation of a new rule. The agency would use the analysis to anticipate how firms will react to a new regulation; to assess the several components of a proposal and to ask what each component is expected to do, how the components relate to one another, and whether each component makes sense; and to assess whether the new proposal will actually cause the desired effects to occur.