ABSTRACT
If the general rule is a skewed size distribution, then both the level of approximation and the limit conditions under which deviations are expected to decrease remain unclear. We know that some versions of stochastic growth processes reproduce the limit size distribution in some industries better than others (see the case of pharmaceuticals described in Fu et al. 2005), but we cannot make predictions about whether and how they can be applied to other industries. Moreover, a common feature of those models is that they are compatible with a minimal role of differences among firms. Such a characteristic stems from the idea underlying the Law of Proportional Effects (Gibrat 1931) that, since its formulation, cast serious doubts about the theory of optimal size. However, even if we dismiss optimal size theory, we cannot dispose of differences among firms in driving the pattern of industry evolution (Nelson 1991). Indeed, a parallel set
of empirical regularities concerning the economic performances of business companies, outlines persistent differences in profitability even within narrow defined industrial sectors. The appearance of long-lasting profit differentials among firms has been interpreted as indicating that firm specific organizational capabilities do actually exist. Nonetheless, persistent heterogeneity among firms can barely be reconciled with a law postulating equal chances of growth behind the observed regularities in firm size distribution.1
