Refitting the Kuznets curve

Using Baltagi and Li (2002)’s semi-parametric fixed effects regression estimator, we investigate the existence of an aggregate ‘Kuznets curve’ in a sample of 113 countries over the 1960–2000 period. Our results show that misspecification of the functional form in a panel model with fixed effects, and the failure to control for the endogeneity of income per capita, can lead to dramatically different conclusions regarding the impact of economic development on inequality. While we find observational evidence of an inverted U relationship between inequality and economic development, this relationship does not seem to be causal. Once we account for endogeneity, inequality appears to decrease monotonically with income per capita.