Kathleen Segerson
If the consumer were totally unaware of any potential contamination, then
=0. If producers are also unaware of potential Contaminahon, then β=0
as well.
The socially efficient values for у and q maximize the hue expected net
benefit from production and sale of the good given by
(4) B(y) - C(q,y) - qDy.
The last term in (4) is the expected damages from Consumpfion of у units
of the good. Note that total expected damages can be reduced in two ways,
either through an increase in product safety, i.e., an decrease in q, or
through a reduction in consumption. For example, the expected damages
from Consumpfion of contaminated seafood can be reduced either through
increased care in processing and handling or through decreased
consumption. Likewise, the expected damages from ingestion of pesticide
residues can be reduced either through a reduction in residue per unit of
production or by a reduction in the consumption of foods with residues.
Maximization of (4) yields the following efficiency conditions:
(5) B'(y)-Cy-qD=O
and
(6) -C4-Dy = O.
These are the usual marginal benefit equal marginal cost conditions for an
interior solution. We should note, however, that in some cases the efficient
outcome may be a comer solution, with respect to either у or q. For
example, if D is extremely large, it may be efficient to produce a product
that is completely free of that risk (i.e., a product with q=0). If this is
extremely costly, it may be efficient not to produce the product at all
(y*=0). For many cases, though, it is likely to be efficient to reduce the risk
of contamination but not necessarily to eliminate it completely (unless this
can be done at a reasonable cost).33
Given the efficient levels of output and safety, the question is whether
the market equilibrium will yields these levels. Under risk neutrality,34 the
33 While we assume interior solutions throughout, our qualitative conclusions
would continue to hold when the efficient output or safety level is a comer solution.
34 The assumption of risk neutral consumers is perhaps quite restrictive in the
context of food safety, where some of the possible outcomes from consumption of
contaminated food may not be easily insured against. (See Antle (1998) for a
discussion of risk aversion in the context of food safety.) Risk aversion can be
introduced into product safety models (see, e.g., Spence 1977). However, with risk
aversion there are two potential distortions in the model, one from the damages
associated with contamination and the other from the inefficient allocation of risk.