Mandatory vs. Voluntary Approaches to Food Safety
operative, the likelihood the source of the contamination is detected and the
firm is held liable, and the severity of the public response (e.g., reduced
future purchases due either to a perceived reduction in the safety of the
product or to consumer boycotts as manifestations of public outcry).19
Given these costs and benefits, the firm's payoff from voluntarily
undertaking the protective measures is Bv+S-Cv-L if contamination still
occurs and Bv+S-Cv if no contamination occurs. Thus, the expected payoff
is simply Bv+S-Cv-qL.
If the firm does not undertake voluntary measures, or equivalently
chooses not to participate in a voluntary government program, then there is
some probability r (0≤r≤1) that the regulator will impose mandatory
controls or standards. If the imposition of mandatory standards is
guaranteed if a voluntary approach is not adopted, then r=1. Alternatively,
if there is no threat of imposition of mandatory standards, then r=0. More
generally, the firm might expect that there is some likelihood of mandatory
standards, i.e., it might perceive r to lie in the open interval (0,1).
We assume that the outcome of the voluntary measures (if adopted) is
the same as the outcome of the mandatory standards that might be
imposed. For example, if the protective measure under consideration is a
HACCP system, we assume that a HACCP system adopted voluntarily
would provide the same level of food safety as a possible mandatory
system. Alternatively, if the voluntary measures are keyed to performance
standards, we assume that any mandatory measures that might be imposed
would be designed to ensure the same level of performance.20 This
essentially assumes that the target level of performance is the same
regardless of whether the protection stems from voluntary measures or
mandatory policies. Thus, we do not allow the possibility that the firm
could forestall mandatory controls by undertaking a level of protection less
than the target level set by the government.21
In the context of Figure 1, the above assumption implies that the
probability of contamination would be the same for mandatory and
voluntary standards. Thus, as shown in Figure 1, if the firm does not
undertake voluntary measures and the government responds with
mandatory standards, a contamination episode will still occur with
probability q. If contamination occurs, the firm's payoff is Bm-Cm-L, where
Bm is the benefit of meeting the mandatory standard and Cm is the
associated cost. The payoff when no contamination occurs is simply Bm-
Cm. If the mandatory standard is identical to the one that might have been
19 Caswell and Henson (1997) argue that the loss of reputation and market sales
are likely to be of more importance to firms that the direct damage costs imposed
through liability.
20 This is not necessarily the case when firms negotiate with regulators over
levels of protection. See Segerson and Miceli (1997).
21 This assumption could be easily relaxed by allowing the probabilities of
contamination to differ for the two approaches.