Let Z contain indices of public information on the product safety of X , with the index zi linked
to the quality of good i. The index could be a measure of the extensiveness of product recalls or
intensity of media reports of food contaminations. A higher value of zi indicates increased adverse
publicity on the quality of good i implying ∂Xu∂z- < 0 (see, for example, Piggott and Marsh). If
preferences are time-separable, zi would have the immediate and full effect on xi by reducing the
level of its consumption. However, if consumption of xi is habitual, the effect of an increase in zi
on xi has several dimensions. Consumption responses to a change in product quality are different
between in the short run and in the long run. For a transitory increase in zi , holding the increment
in zi constant, the size of the immediate drop in consumption of xi is an increasing function of
the expected duration of the adverse publicity. In the long run, xi returns to its equilibrium level
before the negative quality shock. On the other hand, the quality deterioration that is expected
to be permanent would gradually reduce xi to a new long-run equilibrium at a lower level. The
sluggish adjustment in xi causes its long-run response to be larger than its short-run response, and
this difference increases with the degree of habit persistence. The cross-effects of food safety events
on other goods depend on the nature of these goods (substitutes or complements) and would be
interesting to investigate in the empirical analysis.
Using some forms of consumer cholesterol awareness index, several static demand system ap-
proaches have found statistically significant negative/positive impacts of this health hazard on
demand for beef /poultry and sometimes pork (e.g., McGuirk et al. 1995; Kinnucan et al. 1997).
Dahlgran and Fairchild and Piggott and Marsh noted that cholesterol related health effects require
long-term and repeated consumption, unlike food safety effects that would result in sudden and
acute illness. If this is true, it is plausible that the health effects would express themselves through
the habit stock term Xt-1 . Arguably, increased cholesterol information may have reduced/increased
the marginal effect of past consumption on the marginal utility of current consumption of red/white
meat. The strength of a habit for red/white meat could have consequently declined/increased over
time.