Preliminary and Incomplete

A growing body of empirical evidence suggests that, even in narrowly defined industries, there is significant heterogeneity in firm productivity and firm size. Moreover, consistent with core theoretical models of the size distribution of firms, the productivity and size distributions are closely related – that is, more productive firms are larger. However, the relationship between productivity and size varies across countries. A working conjecture is that this variation in the productivity/size relationship reflects differences in market distortions to allocative efficiency across countries. This is not a new hypothesis, but the burgeoning development of firm-level databases permits exploring it in richer and new ways. In this paper, we present empirical evidence of the within industry covariance between firm level productivity and market share using a widely-used decomposition proposed by Olley and Pakes (1996). We show that within the typical U.S. manufacturing industry, labor productivity is almost 50 percent higher than it would be if employment was allocated randomly. However, we also show that this empirical measure of the covariance between size and productivity is much lower on average in Western European countries and, in particular, in transition economies of Eastern Europe even if, in th