Impacts of Income Inequality on Economic Growth (in Japanese)

The objective of this paper is to empirically test the following: (a) the tendency of per capita income gaps between the countries worldwide, (b1) the relationship between the level of per capita income and the degree of income inequality to verify Kuznets's inverted-U curve hypothesis- inequality first rises and later falls as the economy develops, (b2) the relationship between the growth rate of per capita income and the degree of income inequality, and (b3) it further investigates whether income inequality promotes or curbs economic growth by running the cross-country growth regressions. The convergence theory is related to the international distribution of per capita income, while Kuznets's hypothesis focuses on the time profile of income distribution in the course of development in a country. One observes that the economies of the least development countries in Latin America and Africa have often been trapped in poverty for decades, while many countries in East Asia have taken off from poverty and on to the sustainable industrialization paths. The cross-country growth regressions by the generalized least square (GLS) method, based on the long-term data from 1820 to 1994 of 56 countries collected by Maddison (1995) revealed that the coefficient of the initial income was significantly negative for the period 1950-1994, suggesting that the convergence of income is pervasive. However, the trend of the ratios of individual-country income to regional-average income were not likely to converge, except for the developed countries. The cross-country analysis for the period 1960s-1990s-with a dependent variable being the income inequality presented by Gini coefficients, taken from the data collected by Deininger and Squire (1996), along with an explanatory variable being the logarithm of per capita GDP-proved that Kuznets's inverted-U curve hypothesis was plausible, since Gini coefficients fitted to a negative quadratic curve of the logarithmic income. Using data from the 1990s, the quadratic curve had a peak at $3080 of per capita GDP. When Gini coefficients were regressed to a positive cubic curve of the dollar-termed per capita GDP, the curve peaked at $3000, and bottomed at $17000, reflecting a recent recurrent increase in income inequality in the developed countries. A cross-country growth regression to some explanatory variables including Gini coefficients revealed that income inequality had a negative impact against growth in Asia, whereas it had a positive impact against growth in Latin America and Africa.