There has been a lot of debate about the relationship between inequality and economic growth. For example, Brad DeLong says:
The correlation between economic growth--defined as ten-year growth in GDP per capita--and income inequality that Greg Mankiw asserts exists? I certainly cannot see it in the data.
This paper, a study of inequality and growth in rural China, comes to a different conclusion - that higher inequality is associated with slower growth in income:
Inequality and Growth in Rural China: Does Higher Inequality Impede Growth?, by Dwayne Benjamin, Loren Brandt, and John Giles, SSRN No. 2344 [alternate]: Conclusions: Is the trajectory of household income growth adversely affected by the level of inequality in the village in which it lives? ... Using a rich household-level data set stratified by village, we estimate the empirical links between the level of inequality near the beginning of reforms (1986) and the growth of household incomes in the village through 1999.
Whether we use the sample of households that can be tracked all the way through the panel, or treat the villages as distinct cross-sections, we find robust evidence that initial inequality is significantly related to subsequent growth: Inequality “hurts” growth. Because we can estimate the relationship at the household level, and given a rich set of instruments that can be used to address measurement error, we are able to rule out a number of explanations. In particular, we can rule out the possibility that the relationship is an artifact of aggregation, measurement error, or other mechanical explanations based on non-linearities linking initial household income to growth. Furthermore, since we control for own resources in the household-level specification, we cast doubt on the imperfect credit-market class of explanations. In short, we find robust evidence that high village inequality exerts a negative “externality” on household economic growth. ... [V]illages with higher inequality grow significantly more slowly.