Income Inequality and Economic Growth in United States Counties: 1990s, 2000s and 2010s

Using a common reduced-form regional growth model framework, an expanded geographic classification of counties, additional years of data, a trio of income inequality metrics, and multiple empirical specifications, this analysis confirms and builds upon the notion that the nature of the relationship between income inequality and economic growth varies across geography (Fallah and Partridge, 2007). A positive relationship between an income Gini coefficient and per capita income growth is observed only in central metro counties with population densities greater than 915 people per square mile or in about 5 percent of all counties, whereas previous research found a positive relationship in all metropolitan counties (27 percent of counties) and a negative relationship in nonmetropolitan counties. Where inequality is in the distribution is also shown to impact this relationship. Inequality in the top and bottom halves of the income distribution has a positive relationship with growth within this 5 percent of counties. However, in most locations (the other 95 percent of the counties), inequality in the bottom half of the income distribution has either no statistical relationship with growth or a positive relationship, while inequality in the top half of the income distribution tends to have a negative relationship. These patterns are relatively stable over time but tend to not be robust to the inclusion of county fixed effects. These results provide some evidence that the mechanisms explaining how this relationship varies across places are more likely associated with agglomeration and market incentives rather than social cohesion. This analysis also highlights the need for a robust research agenda focused on further refining the growth model along with incorporating new data sources and concepts of income inequality.