Income Inequality and Growth

Vaguely Right, who I belive is an Economics Ph.D. student (and should probably be working on his/her dissertation instead of blogging) has a good (long) post on income inequality in the US, and it’s based on Census data. The abbreviated version contains these nuggets:

  • Income Inequality in the US has increased substantially since 1967.
  • But viewed in levels, everyone–bottom, middle, top–is much better off now than them, even after adjusting for inflation.
  • If the rate of economic growth between 1967 and 2001 were unchanged (perhaps a big assumption), and the income distribution were unchanged, the bottom 20% would be making 14% more, the next 20% would be making 24% more, the middle 20% making 18% more, the second highest 20% making 5% more, and the top making 13% less.
  • But if keeping income shares constant has even a very modest impact on growth, then in levels the bottom 80% are better off having the inequality (and corresponding growth)
  • So is greater concentration of wealth at the top the price of higher economic growth for everyone (think trickle-down)? If this were true then you would expect that as income inequality increases, growth rates increase. Vaguely Right shows that this is not the case.

On the last point, I’d add that Alesina and Rodrik (Quarterly Journal of Economics, 1994) take a panel of countries and regress income growth 1960 to 1985 on control variables and two GINI coefficients in 1960, one measuring income inequality and one measuring inequality in the distribution of land. Countries that were more unequal in 1960 grew slower (and land inequality has more of an impact than income inequality)! Land reform in Asia that reduced inequality is one explanation for the “Asian Miracle” of economic growth.

Here’s a neat picture I stole from the University of Texas Inequality Project (green, light blue and dark blue are the more unequal; red is the most equal; click to enlarge):






AB

P.S. In a second post, Vaguely Right also comments on my rawls post.