Chart of the Day: Are Changing Social Trends Causing Income Inequality?

When we hear of income inequality worsening, we think of rich people getting richer at the expense of the poor. Since data shows this having worsened over the past few decades, some are calling for reforms that would result in income being distributed more equally. For example, some say higher taxes on the rich would help. But what if the problem wasn't economic -- what if it boils down to the personal choices Americans make?

Let me explain what I mean. Take a look at the fascinating chart below, created by the Political Calculations blog. It shows the "Gini Coefficient" for Americans over time, grouped in three different ways. (A Gini Coefficient is a measure of income inequality. A value of zero indicates perfect equality where everyone has the same income, while a value of one indicates perfect inequality where one person has all the income and everybody else has none.)


As you can see, for individuals income equality has actually improved slightly over the period shown. Yet for families and households, it has worsened. According to Political Calculations, an economist has found that the Gini coefficient for individuals has remained nearly constant near 0.51% since 1960 -- so income inequality hasn't been worsening for individuals for 50 years.

Recommended Reading

What's causing these measures of income inequality to diverge? The blog post concludes:

In short, it's those social processes, which have driven demographic changes within U.S. households, that are almost exclusively behind the observed increase in family and household income inequality observed in the U.S. since the 1960s.

Put another way, the problem is social trends -- not economics.

Anecdotally, this makes pretty obvious sense. Single parent households have grown significantly over the past few decades, which almost certainly increased inequality among households.

But the gap also probably widens through the concentration of wealth in certain areas. Think about Manhattan, for example. It has quite a few very rich people. They go to many of the same restaurants and many of the same social events. If two of them meet and marry, then the household of each goes from very wealthy to very, very wealthy. If two middle class workers in the Midwest marry, their household income doesn't rise nearly as high on a relative scale.

Another possible cause could be that women are making more money these days than they were a few decades ago. Again, if similarly successful people tend to marry, then their combined household is simply much wealthier today than it was when women made less money or when more women became homemakers after marriage.

If income inequality among individuals isn't actually worsening, the growing gap doesn't seem as serious a concern. After all, how do you cure a problem caused by social trends and not economics? Short of efforts to encourage people to marry outside their income segment or remain married instead of divorcing, no "solution" would very well fit the heart of the "problem."

(Via: James Pethokoukis at AEI)


A few quick updates/concerns about this chart:

Justin Wolfers objects to putting too much weight in individual income inequality. He asserts that the variance here is explained by the "zeros" -- who work in the market versus the home. That is, since we have more people earning income today than a few decades ago, individual income appears to have kept up better, even though it doesn't if you assume those individuals before were working in the home and not earning income. 

- An economics professor e-mails me to raise a suspicion about the data used here, from the Current Population Survey. It may be top coded at $250,000, as implied by this economics paper. If that's the case, then it fails to fully account for full gains of some very wealthy individuals. But if that's a problem, I still wonder if that would change the variance we're seeing here between households and individuals. In fact, wouldn't households be even more skewed towards equality due to this problem since their incomes should be higher than those of individuals? If this is the case, however, then it would fail to capture the dramatic rise of some wealthy individuals.