Testing errors, supplier segregation, and food safety

Diagnostic tests provide valuable information to buyers about credence attributes such as food safety and GMO content. Errors in testing, however, can mislead buyers and lead to problems such as adverse selection. The ability to segregate suppliers who can deliver safe food from suppliers who cannot depends on the accuracy of the test procedure. In this article we examine the effect of test sensitivity, specificity, and sampling error on the ability to segregate safe and unsafe suppliers. We find that there is a maximum level of error below which unsafe suppliers are deterred from accepting a utility maximizing buyer's bid price. The maximum error depends on the probability that a supplier's production is unsafe and the cost of producing an unsafe lot, among other things. Understanding this relationship makes it possible to design contracts and government regulations that discourage unsafe suppliers from trading.