Testing for Conditional Convergence in Variance and Skewness: The Firm Size Distribution Revisited

Abstract

This article derives generalized method of moment‐based Wald tests for conditional convergence in terms of the second and third central moments (conditional σ‐ and ζ‐convergence). Monte Carlo simulations indicate that the proposed tests are properly sized and equipped with enough power for sample sizes in excess of 1,000 observations. We apply the Wald tests to a comprehensive sample of Austrian firms. The estimation results indicate a substantial reduction in both the second and third central moments of log firm size over time for the younger age cohorts. For older ones, the log‐normal distribution is a valid description of their size distribution.