XUELI TANGy

We provide an empirically plausible endogenous growth model to prove analyti-cally that sometimes a progressive redistribution from rich to poor lowers the growth rate of consumption per capita in all subsequent periods. The model accommodates the growth retarding effect of income inequality by combining the assumptions of no credit market and a production technology with diminishing returns to the com-bined inputs of physical and human capital. Also, to make the model's assumptions consistent with the evidence reported by leading labor economists, we assume that the parental human capital sufciently improves the effectiveness of expenditure on a child's education, in order to induce increasing returns to scale in the education technology. A reduction in the progressivity of redistribution, under such educa-tion technology, enhances the average human capital of all future cohorts of parents, which in turn boosts the growth rate of average human capital. The immediate re-sulting gain in the growth rate of consumption per capita sufciently outweighs the subsequent growth loss due to the decline in TFP brought about by the associated increase in income inequality. Consequently, in our model, a policy of progressive redistribution is dynamically inefcient