Was it something I ate? Implementation of the FDA seafood HACCP program.

[Received January 2006; accepted March 2007.]

References

Antle, J.M. 2000. "No Such Thing as a Free Safe Lunch: The Cost of Food Safety Regulation in the Meat Industry." American Journal of Agricultural Economics 82:310-22.

Carey, V., S.L. Zeger, and P. Diggle. 1993. "Modelling Multivariate Binary Data with Alternating Logistic Regressions." Biometrika 80(3): 517-26.

Caswell, J.A., and N.H. Hooker. 1996. "HACCP as an International Trade Standard." American Journal of Agricultural Economics 78:775-79.

Cato, J.C. 1998. "Seafood Safety: Economics of Hazard Analysis and Critical Control Point (HACCP) Programs." FAO Fisheries Technical Paper 381, Food and Agricultural Organization, Rome.

Centers for Disease Control. 2005. "Preliminary FoodNet Data on the Incidence of Infections with Pathogens Transmitted Commonly Through Food--10 Sites, United States, 2004." Morbidity and Mortality Weekly Report 54:352-56.

Cohen, M.A. 1999. "Monitoring and Enforcement of Environmental Policy." In H. Folmer and T. Tietenberg, eds. International Yearbook of Environmental and Resource Economics 1999/2000. Northampton, MA: Edward Elgar, pp. 44-106.

Diggle, P., K.-L. Liang, and S.L. Zeger. 1994. Analysis of Longitudinal Data. Oxford, U.K.: Oxford University Press Science Publications.

Fahrmeir, L., and G. Tutz. 1994. Multivariate Statistical Modelling Based on Generalized Linear Models. New York: Springer-Verlag.

Food and Drug Administration. 2000. "FDA's Evaluation of the Seafood HACCP Program for 1998/1999." Available at http://www.cfsan.fda. gov/~comm/seaeval.html.

--.. 2001. "FDA's Seafood HACCP Program: Mid-Course Correction." Available at http:// www.cfsan.fda.gov/~comm/shaccp1.html.

--. 2002. "FDA's Evaluation of the Seafood HACCP Program For Fiscal Years 2000/ 2001." Report available at http://www.cfsan. fda.gov/~comm/seaeval2.html.

General Accounting Office. 2001. "Food Safety: Federal Oversight of Seafood Does Not Sufficiently Protect Consumers." Report to the Committee on Agriculture, Nutrition, and Forestry, U.S. Senate. Report No. GAO-01-204. Available at http://www. gao.gov/new.items/d01204.pdf.

Huleback, K.L., and W. Schlosser. 2002. "Hazard Analysis and Critical Control Point (HACCP) History and Overview." Risk Analysis 22:547-52.

Mayes, T., and S. Mortimore. 2001. Making the Most of HACCP: Learning from Others' Experience. Cambridge, U.K.: Woodhead Publishing Ltd.

McEachern, V., A. Bungay, S.B. Ippolito, and S. Lee-Spiegelberg. 2001. "Enforcing Safety and Quality." In T. Mayes and S. Mortimore, eds. Making the Most of HACCP: Learning from Others' Experience. Cambridge, U.K.: Woodhead Publishing Ltd, pp. 165-82.

Muth, M.K., S.A. Karns, M.K. Wolgenant, and D.W. Anderson. 2002. "Exit of Meat Slaughter Plants during Implementation of the PR/HACCP Regulations." Journal of Agricultural and Resource Economics 27:187-203.

National Marine Fisheries Service. 2005. Fisheries of the United States, 2004. Silver Spring, MD, November.

Polinsky, A.M., and S. Shavell, 2000. "The Economic Theory of Public Enforcement of Law." Journal of Economic Literature 38:45-76.

Putnam, J.E, and J.E. Allshouse. 1999. "Food Consumption, Prices and Expenditures 1970-1997." Statistical Bulletin No. 965, Economic Research Service, U.S. Department of Agriculture, Washington DC.

Raymond, M. 2004. "Regulatory Compliance with Costly and Uncertain Litigation." Journal of Regulatory Economics 26:165-76.

Roberts, T., J.C. Buzby, and M. Ollinger. 1996. "Using Benefit and Cost Information to Evaluate a Food Safety Regulation: HACCP for Meat and Poultry." American Journal of Agricultural Economics 78:1297-1301.

Unnevehr, L.J., and H.H. Jensen. 1996. "HACCP as a Regulatory Innovation to Improve Food Safety in the Meat Industry." American Journal of Agricultural Economics 78:764-69.

Wooldridge, J.M. 2002. Econometric Analysis of Cross Section and Panel Data. Cambridge and London: MIT Press.

(1) We assume that the firm is not liable for social damage from inadequate precautionary effort since food poisonings are difficult to trace back to their source. This assumption clearly involves no loss of generality.

(2) We tested the reasonableness of this assumption by estimating lognormal mixed continuous/censored data model with and without random effects. Although a Weibull-accelerated life model fits the data best, a lognormal model provides an acceptable approximation to our data. The correlation between any two observations contributed by the same plant was very small (about 0.0477). The regression coefficients were virtually unchanged when we restricted the correlation coefficient to be zero.

(3) This approach effectively augments the dataset with one additional observation--the censored time between the most recent inspection and December 31, 2001--for each plant.

(4) Although the inspection outcome sample is a mix of cross-sectional and panel data (for plants that were inspected only once, and more than once, respectively), we treat all observations contributed by plants with more than one inspection as independent from one another. This is reasonable for three reasons. First, single inspections account for the majority of all inspections (58%). Second, we are interested in marginal models that predict the expected value of a certain outcome as a function of regressors (Diggle, Liang, and Zeger 1994, p. 146 and following), and in this context ignoring the potential correlation between observations contributed by the same plant does not bias the estimated coefficients--if anything, it results in a loss of efficiency (Fahrmeir and Tutz 1994). Third, since the correlation among observations is not of interest to us, we follow Fahrmeir and Tutz (1994, p. 113) in assuming independence.

(5) The statistical properties reported in footnote 4 extend to vectors of binary outcomes (Carey, Zeger, and Diggle 1993), suggesting that treating the various compliance outcomes as independent will not bias the estimated coefficients, and at worst results in a loss of efficiency. All of our models explicitly correct for heteroskedasticity.

(6) Using a Weibull distribution, we estimate mean duration between visits to be 723 days (s.e. around the mean 5.65 days), and median duration to be 633 days (s.e. around the median 5.41 days).

(7) Such a result is entirely consistent with the predictions of the theoretical framework in the section "A Model of HACCP Enforcement and Compliance."

(8) It is likely, for example, that plants with larger sales volume adhere to stricter sanitation standards in their processing operations because of economies of scale in sanitation equipment and technical expertise (e.g., specialized staff to oversee food safety) or because their products are more likely to be traceable in the marketing chain in the event of a foodborne illness outbreak (see Antle 2000).

(9) The coefficients on the dummies for plant size are, however, broadly consistent with the raw violation rates, which exhibit a roughly quadratic relationship with class size. The violation rates are lowest for SALES2 and SALES8 plants, and peak at about 50% for SALES6 plants.

Anna Alberini is associate professor and Erik Lichtenberg professor, Department of Agricultural and Resource Economics, University of Maryland, College Park. Dominic Mancini is economist, Office of Information and Regulatory Affairs, Office of Management and Budget, Washington, D.C. Gregmar Galinato is assistant professor, School of Economic Sciences, Washington State University, Pullman.

This research was funded in part by a grant from the University of Maryland/FDA Joint Institute for Food Safety and Nutrition. All views expressed are those of the authors alone and do not necessarily represent those of the Office of Management and Budget, the Executive Office of the President, or the University of Maryland/FDA Joint Institute for Food Safety and Nutrition. Table 1. Descriptive Statistics of the Inspection Data Variable Mean Standard

Deviation Compliance status at current inspection (N = 7,086) HACCP violation 0.4583 0.4983 Any sanitation violation 0.6169 0.4861 Observed sanitation deficiency 0.4274 0.4947 Sanitation records not

reflecting actual conditions 0.1580 0.3648 Inadequate or no monitoring records 0.4745 0.4994 Compliance status at the previous inspection (n = 3,011) HACCP violation in previous inspection 0.4749 0.4994 No HACCP plan when needed

in previous inspection 0.1438 0.3509 Observed sanitation

deficiency in previous inspection 0.4616 0.4986 Sanitation records not reflecting

actual conditions in

previous inspection 0.1776 0.3823 Inadequate or no monitoring