Regional heterogeneity in the relationship between inequality and growth: Evidence from panel vector autoregressions

This note estimates the responses of income inequality to growth shocks, and growth to shocks to income inequality using a panel VAR model. Employing annual data from 1930 to 2012 for the 50 U.S. states and DC, we find that inequality and growth decrease following respective shocks to growth and inequality. However, we document significant heterogeneity in U.S. regional responses to inequality and growth shocks. Furthermore, variance decompositions show that regional income inequality explains a relatively small proportion of the variation in regional real per capita growth in the short run, but that the explanatory power of inequality for growth rises over time, explaining as much as 60% of the long run unpredictable fluctuations in growth in some regions. We also report substantial heterogeneity in the explanatory power of growth to movements in regional income inequality, with income growth explaining a little over 1% of the long run variation in income inequality in the Mid Atlantic region, but over 45% in the West North Central region.