INCOME INEQUALITY, DEVELOPMENT, AND DUALISM - RESULTS FROM AN UNBALANCED CROSS-NATIONAL PANEL

We investigate the relationship between income inequality and economic development using an unbalanced cross-national data set that allows o bservations on inequality and development for several years for the sa me country. The 88 countries in this data set contribute 279 observati ons dated from 1952 to 1988. income inequality is measured in three wa ys-as income share of the top quintile of income-receiving units and w ith two estimates of the Gini coefficient (decile-based and quintile-b ased). The relationship between income inequality and development in t his data set exhibits the inverted-U shape characteristic of the Kuzne ts curve. Regression analyses using pooling techniques with the assump tions of a random effects model show that the curvilinearity is largel y accounted for by a model based on three major processes: labor force shifts from agriculture to industry; the demographic transition; and the spread of education. These processes are represented in regression models by four variables. The variables have significant effects on i ncome inequality in directions predicted by the model: sector dualism (positive effect); percent of labor force in agriculture (negative eff ect); natural rate of population increase (positive effect); and secon dary school enrollment (negative effect). The effects of political dem ocracy, Marxist-Leninist regime, horticultural or agrarian subsistence technology prior to industrialization, an indicator for Taiwan, and c alendar time are also estimated.