Reputations, Market Structure, and the Choice of Quality Assurance Systems in the Food Industry

convenience that there is a continuum of stringency levels from which to choose.

Consumers are assumed to be homogeneous in their tastes and willingness to pay for
the value-added product. We also assume that consumers can observe whether a QAS is in
place, but are not able to infer the actual quality of the product from the particular QASs used.
This implies that the QAS implemented cannot be used as a signal of quality by processors to
differentiate themselves from other certifying suppliers.

Monopolist Processor

We begin our analysis by examining the case where there is only one processor in the market
and the processor is trusted until proven wrong. Therefore, there are only two possible states of
the world, denoted by r = 1,2 . The first state denotes the periods in which the processor has a
good reputation and hence faces a positive demand. In state two, the demand for the high-
quality product is zero. Since there is only one processor, and profits are zero in the second
state of the world, the superscript of the per-period profit function will be dropped.

Let T denote the time when reputation is lost because consumers discover that they
purchased a product that does not meet the promised standards. A processor that moves from
state 1 to state 2 in period T has profits given by

TT1 _βT

(1)      ∏( 5^, y ) = ∑ ^{β π}( y^, s ; a ) = ^π( y^, s ; a )∑ ^β  = ^π( y, s ; a )-——

t=1t=11 ^- β

where π(i) represents the per-period profits of a processor that has a good reputation and β

is the relevant discount factor. However, quality is random so the processor cannot exert
perfect control over it. The processor’s expected profits are

<<   <   5   6   7   8   9   10   11   >   >>