Hétérogénéité - 2. Etude des biais de représentativité (sous l'hypothèse d'exogénéité faible)

In microeconomic, the individual differences are taken into account by the introduction of explanatory variables but also by means of the parameters of interest. This parameter of heterogeneity is in general described by the p.d.f. of the parameter among the individuals (heterogeneity distribution) and it is often assumed that it is independent on the values of the explanatory variables (Assumption of weak exogeneity). Then we replace the initial disaggregate model by the model integrated with respect to the heterogeneity distribution (The so-called representative model). The aim of this paper is to discuss the biases, which are consequences of a direct use of the representative model, as if heterogeneity was absent. We establish in particular that this usual practice leads to an overestimation of the substitution phenomenas in consumption or production models, to an overestimation of the effect of technical progress. On the other hand the biases may be neglected if we are interested in determining cost functions.