Determinants of Income Inequality in the Early Twenty-First Century: A Cross-National Study

Abstract

This paper uses Solt’s Standardized World Income Inequality Database and attempts to explain variations in Gini coefficients for net household income across 142 developing, transitional and developed societies. The causal model contains three sets of explanatory variables: (1) economic dualism, (2) educational attainment and educational inequality, and (3) political and state influences on income inequality. The most important cause of inequality is still the Kuznets effect: societies at low and high levels of development have less inequality than those at intermediate levels. Population growth increases inequality. Rising labor productivity in the agricultural relative to that in the non-agricultural sector, and being a former Soviet society reduce inequality. Educational attainment has less effect than educational inequality on income inequality. Government income transfers sometimes reduce inequality, but have no effect when all variables are in the model. Finally, liberal democracy has no net effect on inequality