Not Really a Post Looking at Income Inequality by President

This post is going to look like a post on income inequality by presidential administration. It even has a nice picture showing data going back to 1953. I’ll start by describing what I did, and then explain why its not a post on income inequality.

I was hoping data on the distribution of income per person. The IRS has data on idnvidual income, but it has several problems, not the least of which it doesn’t go back before 1987. (That’s OK, but not if other data is available.) The Census unfortunately only computes median and mean income for individuals. If you want to know how the income was distributed (e.g., bottom 20% get X, top 20% get Y, etc.) you have to make do with data on families or households.

So I decided OK, let’s go with families. See what it looks like. Well… the Census also goes ahead and computes the Gini Ratio for families in time series form. Basically… the closer to 0, the more equality, the closer to 1, the more inequality. (More information on Gini here.)

For the most part, the data fits with my intuition. But there’s a problem… a huge jump between the last year of GHW’s term and the start of Clinton’s. The same huge jump seems to exist with other series (e.g., percentage of income going to the top 5% jumps a lot, and we see the same thing with household data).

To me, that speaks for a break… a change in how the series was computed. Am I wrong about this? Did income distribution jump that much between GHW’s term and Clinton’s, or is something wrong with the data?