Industry Dynamics and the Distribution of Firm Sizes: A Nonparametric Approach

Abstract

The aim of this paper is to analyze the evolution of the size distribution of young firms within some selected industries, trying to assess the empirical implications of different models of industry dynamics: the model of passive learning (Jovanovic 1982), the model of active learning (Ericson and Pakes, 1995), and the evolutionary model (Audretsch, 1995). We use a non-parametric technique, the Kernel density estimator, applied to a data set from the Italian National Institute for Social Security (INPS), consisting in 12 cohorts of new manufacturing firms followed for 6 years. Since the patterns of convergence to the limit distribution are different between industries, we conclude that the model of passive learning is consistent with some of them, the active exploration model with others, the evolutionary model with all of them.