Plant Size Distribution and Cross‐Country Income Differences

Abstract

Cross‐country differences in income per worker are widely known to be enormous. Per capita income in the richest countries exceeds that in the poorest countries bymore than a factor of 50 (see KlenowandRodriguez‐ Clare 1997; Prescott 1998; Hall and Jones 1999; Caselli 2005). An important strand of the literature trying to understand cross‐country differences in per capita incomes has focused on the role of aggregate factor accumulation by abstracting from heterogeneity in the production units. But there is an emerging and growing body of research that takes a different approach, focusing instead on the misallocation of resources across plants (Restuccia and Rogerson 2007; Bartelsman, Haltiwanger, and Scarpetta 2008; Hsieh and Klenow 2009). Policies’ and institutions’ differential effects on the business climate broadly defined might significantly influence the allocation of resources across establishments. The working hypothesis in this literature is that not only the level of factor accumulation but also how these factors are allocated across heterogeneous production units matters in trying to understand income differences. Our paper contributes to this literature by performing a development accounting exercise using a new data set of more than 20 million establishments in 79 developing and industrialized countries. Specifically, we develop a simple model of heterogeneous production units that follows Melitz (2003). Plants’ dynamics and policy distortions are as in Restuccia and Rogerson (2007), but we assume that production units have constant returns to scale technologies and some degree of market power, as in Hsieh and Klenow (2009). We calibrate the model to match our data set. Our calibration exercise consists in finding the profile of output taxes and subsidies needed to match each country’s plant size distribution. These distortions should be interpreted as the different