The Determinants of Distribution Dynamics: A Novel Methodology with an Application to the Cross-Country Distribution of Productivity Preliminary version. Please do not quote without permission)

In this paper we present a novel methodology for estimating the determinants of distribution dynamics and discuss an application to the distribution dynamics of (labour) productivity across a large sample of countries. We perform a Monte Carlo study of methodology taking as base Mankiw et al. (1992)’s model. We finally apply the methodology to a sample of 84 countries for the period 1960-2006. The level of output per worker in 1960 strongly contributed to reduce inequality. All the other variables (unexplained source of high-productivity, human capital, investment rate, employment growth) positively contributed (in decreasing order of importance) to inequality. Both the unexplained source of high-productivity and the level of output per worker in 1960 increased polarization. Also investment rate favoured polarization, but the such effect appears not statistically significant. Finally human capital and employment growth appear not to affect the polarization of distribution.