Current And Future Implications

Because food safety is a combination of an experience good and a credence good, the private marketplace provides less than the socially optimal level of food safety. Since the turn of the twentieth century, the U.S. government has used its regulatory powers to remedy market failures associated with food safety. It remains important to reassess these regulations, however, to improve their efficiency in light of experience and of new information about the nature and extent of food safety problems. There are a number of areas where economic analysis can help. We highlight four here.

1. Set Priorities by Ranking Relative Risks and Costs for Microbial vs. Chemical Hazards

Ideally, the limited federal budget for food safety should be spent on the most cost-effective methods of controlling significant risks. In other words, spending on food safety regulation should be allocated to achieve the greatest feasible level of safety from any given level of overall expenditure across hazards and agencies. Comparison between microbial and chemical food risks becomes relevant in this context. The literature clearly indicates that the foodborne risks from microbial hazards are much greater than the risks from chemical hazards. The ERS estimates that foodborne illness associated with five pathogens cost the U.S. society $6.9 billion (Aug. 2000 dollars) in medical charges, lost productivity, and value of premature death each year. Because the risks are so low, the economic costs of foodborne chemical risks have not been estimated. These differences in current costs to society suggest that too much chemical safety and too little microbial safety may be provided currently. Ranking risks among pathogens has been started in Tables 7.2 and 7.3 but needs to be extended to the remaining pathogens identified by CAST.

In 1998, the National Research Council Committee to Ensure Safe Food from Production to Consumption recommended that a comprehensive food safety plan be developed and that funds for food safety programs (including research and education programs) reflect science-based assessments of risk and potential benefit (Institute of Medicine, 1998, p. 11). In response, the Risk Assessment Consortium (RAC) was established as part of the President's Food Safety Initiative (http.'Hwww.foodriskclearinghouse.umd.edujRisk_Assessment_ Consortium.htm). The RAC has proposed a project to rank relative risks for all sources of foodborne disease and to conduct benefit-cost analyses of potential control options for the most significant foodborne disease risks.

2. Improve Economic Valuation of Microbial Risks with WTP Methods

A second area of economic research targets consumer concerns about food safety risks. As noted above, the appropriate level of safety equates the marginal benefit of reducing foodborne risk with the marginal cost of risk reduc tion. One important component of assessing regulation is thus the benefit of risk reduction, that is, what consumers are willing to pay for reduced food safety risks. Two federal agencies, the USDA's ERS and the DHHS's CDC, have allocated Food Safety Initiative money to develop new valuation estimates for reducing risks from microbial pathogens.

In 1998, the CDC awarded funds for a cooperative agreement to study consumer demand for food safety. The study, which may continue for up to five years, is designed to estimate the value that consumers place on reducing the risk associated with specific microbial foodborne illnesses for which interventions already exist. The effect on consumers' value of alternative combinations of private and collective risk reduction strategies is also being assessed. The study provides opportunities to advance the techniques used in risk communication, conceptual and empirical economic modeling, and value estimation in the public health setting. Another important contribution of the project is the education of consumers with respect to the risk of foodborne illness and the available public and private efforts to reduce risk. The agreement opens the way to new collaborative efforts between the CDC and the economic research community.

The ERS also is investigating techniques to develop ex ante values for the willingness to pay to avoid risks associated with foodborne pathogens in cooperative agreements awarded in fiscal year 1999. Several valuation techniques can be used:

• The contingent valuation method is a stated-preference technique in which the consumer's WTP for non-market goods is revealed with surveys (Buzby et al., 1998).

• Experimental auction markets use an artificial choice situation with real choices. For example, in a staged experiment, participants bid real money to buy an irradiated chicken sandwich that poses lower food safety risks (Shogren et al., 1999).

For a comparison of methods, see Buzby et al. (1999) and Golan et al. (2001). The other three objectives of the ERS program are: 1) to evaluate the validity and effectiveness of different methods that model the process by which consumers assess changes in probability and risk, 2) to test whether the presentation of distinct pathogen-specific and symptom-specific scenarios result in different consumer valuations, and 3) to examine how alternative combinations of private and collective risk reduction strategies affect consumer valuation of safer food.

The results of these studies will be used to improve valuation methods in the regulatory agencies. Many valuation issues were discussed at the conference at the University of Maryland in September, 2000 which was cosponsored by the Risk Assessment Consortium, Federal agencies, NE-165 regional research group of economists {www.umass.edujnel65D, and others (for conference proceedings, see http:/jwww. jpublicationsjmpl570l).

3. Incorporate Uncertainty into Regulatory Impact Analyses for Food Safety

A third area involves the treatment of uncertainty in regulatory decision making. Quantitative assessments of risks of foodborne illness from pathogens and chemicals are subject to considerable uncertainty. Susceptibility to both pathogens and chemicals varies across individuals in the population. Some of that variability is linked to observable attributes (e.g., age, sex), making it possible to devise policies that address identifiable subpopulations. Some of that variability is not easily observable and can only be treated as random in analyzing regulatory impacts (for example, see Havelaar et al., 2000). In addition, gaps in scientific understanding of the mechanisms by which chemicals (and, in some cases, pathogens) cause adverse health effects mean that estimated cause-effect relationships are subject to considerable uncertainty. Gaps in scientific understanding of correspondences between animal and human responses to chemicals and pathogens mean that the use of animal models adds further uncertainty.

As noted above, regulators tend to be sensitive to these uncertainties, in particular, to the prospect of declaring a compound to be safe when it actually poses a significant risk. At present, they adjust quantitative characterizations of risk by using arbitrary "conservative" assumptions, with a number of negative effects discussed above. An alternative approach is to use probabilistic risk assessment methods that incorporate uncertainty formally and explicitly. Lichtenberg and Zilberman (1988a) present a method of estimating uncertainty-adjusted regulatory costs based on such probabilistic risk assessments. This Lichtenberg-Zilberman approach involves minimizing the cost of meeting a nominal risk standard while holding violations to a given (low) probability. Cost-minimizing strategies consist of combinations of measures, some of which are relatively more effective in reducing risk on average, whereas, others are relatively more effective in reducing uncertainty about risk. Monte Carlo methods can then be used to generate regulatory cost as a function of the nominal standard and probability of violation as well as to explore changes in efficient combinations of risk reduction measures. Empirical applications of this approach include cases involving the cost of reducing the risk of cancer from pesticide contamination of drinking water (Lichtenberg et al., 1989), the cost of mitigating the risk of gastroenteritis from consumption of shellfish contaminated by dairy wastes (Lichtenberg and Zilberman, 1988b), the cost of reducing farm workers' cancer risk from insecticide exposure (Harper and Zilberman, 1992); and the cost of meeting nitrate standards in well water (Lichtenberg and Penn, 2003).

There are two principal directions for extension of this approach in the context of food safety. First, the general approach can be used to take uncertainties into account in devising HACCP strategies. Factors contributing to risk differ in terms of uncertainty about (or unobserved variability in) their effects on risk. The Lichtenberg-Zilberman approach allows incorporation of that uncertainty into formulation of cost-minimizing HACCP strategies. Second, use of the approach implies a need to consider demand for uncertainty reduction as a component of the value of life saving. The degree of reliability with which safety is attained is likely important to consumers and regulators. Methods of incorporating willingness to pay for added reliability (reduced uncertainty) would permit comparison of uncertainty-adjusted benefit and cost.

4. Use Economic Incentives in Designing Public Control Strategies

Economic incentives are an important part of any regulatory strategy. Economic theory suggests that reliance on incentives allows achievement of regulatory goals at low cost in many situations. Moreover, regulations may create incentives that give rise to unintended consequences that undermine the effectiveness of regulation. For example, both nominal food safety standards and the ways in which they are enforced alter the economic incentives facing food processing firms in ways that sometimes increase foodborne risks (see van Ravenswaay and Bylenga, 1991 for a case study of antibiotic and sulfa drug residues in veal). Improvement in the design of regulations can, in principle, increase the economic incentives for firms to produce safer food both in the short run and, via technological change, in the long run (Crutchfield et al., 1997). Gill (1999) found significant variability among beef slaughter plants in their practices and the resulting levels of generic E. coli on carcasses and trim destined for hamburgers. Minor modifications in worker practices in the skinning operation caused a significant reduction in generic E. coli levels on the carcass, suggesting that small increases in economic incentives may significantly reduce contamination in slaughter plants.

One approach is to shift from process to performance standards. Process standards specify the exact safety-enhancing procedures for firms to use, whereas performance standards allow firms to choose the mix of safety enhancement procedures that generates the regulated level of safety at least cost. In addition, performance standards encourage innovation because firms can keep any cost savings generated by technical improvements. Much of the success of the EPA's sulfur dioxide trading program, for example, can be attributed to substituting performance standards for process standards. Widespread use of low-sulfur coal, along with a number of process innovations, accounted for much of the low cost of meeting stricter sulfur dioxide emissions standards (Schmalensee et al., 1998). These measures would have been impossible under the process standards formerly used by the EPA, which mandated the use of scrubbers to meet emissions targets. Such a shift would not be entirely new for food safety, because HACCP-required testing for Salmonella and generic E. coli are performance standards (Powell et al., 2001). However, the economic incentives implicit in alternative HACCP approaches deserve further study.

Another possible use of incentives would be to create differentiated markets for safer foods through the use of certification and labeling (see Golan et al.'s

2001 paper on the economics of labeling: (http:Hwww.ers.usda.govlpublicationsl aer793/)). Certification is currently used to remedy problems associated with experience and credence goods in a number of markets. Underwriters Laboratories, for instance, is a private organization that certifies the safety of home electrical equipment, and shipping point inspection by the USDA's Agricultural Marketing Service, a government program that certifies the quality of fruits and vegetables. In the case of food safety, the analogy would be to label foods certified as having exceptionally low risk from microbial pathogens. The reliability of testing methods (and thus certification itself) is one key to the feasibility of this approach. If testing distinguishes risk from foodborne pathogens perfectly, then a certification program gives firms incentives to provide enhanced levels of safety—provided that the testing is conducted by a third party that is not beholden to the industry using its services (Viscusi, 1978). But if testing is imperfect, so that certification is not a reliable indicator of enhanced safety, then product differentiation will likely be infeasible (De and Nabar, 1991). In this latter case, the benefits from certification are unlikely to justify the costs. It is also possible that enhanced safety from certification may be undercut by the "lulling effect," that is, consumers taking fewer precautions because of the increased perception of safety, as has been observed in the case of safety caps on poisonous products (Viscusi, 1985). Enforcement is required to ensure that certification standards are met by industry (see, e.g., the counterfeit Underwriter Laboratories certification on electrical extension cords produced in China and illegally sold in the United States in 1999).

Research is needed to determine whether testing is sufficiently reliable to make certification a reasonable policy instrument. Research is also needed to determine the potential for certification in creating incentives for enhanced food safety, that is, on demand for food products differentiated as "safer." One aspect likely to be of major importance is consumer response to perceived risk of illness of foodborne pathogens and to uncertainty about that risk. If consumer behavior is strongly responsive to perceived risk, certification is likely to have substantial effects on demand, creating strong incentives for enhanced safety.

0 0

Post a comment