What to Measure When Measuring Inequality?
Income inequality is a popular topic for many liberals. If you do a google search on Kevin Drum’s blog on income inequality you get quite a few hits. Crooked Timber also has a large number of posts on this as well. However, is income the best measure to use when looking at inequality?
Diana Furchtgott-Roth notes one reason why income may not be the best measure of inequality.
And spending is vital because it determines our current standard of living and our confidence in the future. It shows how much money Americans have. The usual pretax measures of income on which most inequality studies are based don’t show how much money Americans have and don’t provide an accurate measure of inequality.
The top 50% of earners pay 97% of income taxes, so all their income is not available for spending. Lower-income Americans receive transfers such as Food Stamps, housing vouchers, Medicaid, and Medicare, so they consume more than their stated income. Middle-income Americans have assets in pension and individual retirement accounts that are not included in income. Therefore, spending is a far better guide to well-being than pretax income.
The idea that one should use consumption spending vs. income is not new.
Consumption is often claimed to be a better measure of permanent income and thus well-being, but most studies of inequality and mobility using U.S. data focus on income. To examine inequality at a point in time, cross sectional data is sufficient. The evaluation of mobility, however, requires longitudinal data in order to follow particular individuals over time. One reason for the focus on income instead of inequality is that no longitudinal data has a complete measure of consumption.
What determines one’s standard of living is not income, although that is an important factor. Spending is much more what determines one’s standard of living. To the extent that one’s income is not available for spending (taxes) or that one’s income is supplemented (subsidies and vouchers) means that income is an imperfect measure for inequality.
John Quiggin thinks that the stability of consumption spending is a sign that people are using credit cards.
(Annual) Consumption inequality has not changed much since 1970. In my judgement, this reflects increased use of credit markets to smooth out short term fluctuations in income, which offsets increased long-run inequality
Frankly I don’t buy this. Borrowing is basically taking future income and using today. If one keeps doing that then eventually we’d expect to see a decline in future consumption. About the only way out of this problem is that people are declaring bankruptcy which is what Quiggin thinks is happening. However, I’m not convinced that bankruptcy is occuring frequently enough and has small enough costs/penalties to allow people to make up the difference in income. We are talking about a 30 year period here where incomes have become more unequal, but consumption has not. For example, Quiggin cites Todd Zywicki’s research on bankruptcy, but that data indicates that bankruptcies occur at a frequency of 14/1000 families in 2003. Further, Quiggin seems to mixing apples and oranges.
The typical story is that incomes are becoming less equal and, according to Quiggin, this difference in terms of consumption is being covered by credit. But, Quiggin also tells the story of people whose income has become more volatile and that they use credit to smooth out consumption. The second story, for which there might be reasonable evidence, does not mean the first story has to be true. So while it is true that bankruptcies have increased, I’m not sure it explains why consumption spending indicates much less inequality than does income.
Update: In comments John Quiggin explains his view of consumption and income inequality in a slightly different manner, one I find a bit more easily understood. His view is that there are two sources of “shocks” to a family’s income (and hence their consumption). One type of shock is transitory which is what is causing the increase in income volatility. The other is rising income inequality. Quiggin argues that using credit markets to try and cover the growing income gap is one possible reason that we have seen an increase in bankruptcies. It isn’t a bad story as economic stories go, I just have some problems with it.1
My problem is that growing income volatility can also lead to increased bankruptcies as well. For example, one might in a good year make $60,000 and in a bad year make $40,000. Thus the expected income is $40,000. But then things become more volatile and not only are the swings wider (say $70,000 in a good year and $30,000 in a bad) but the bad “year” is also variable going from say 6 months to 18 months. Miscalculating how long the “bad year” is going to be could lead to bankruptcy also. So disentangling the effects of increased volatility as well as rising income inequality is going to be problematic.
Another problem is that this would mean that people have some sort of persistant bias or irrationality that even when pointed out to them they don’t correct it. That is, you can’t use credit to cover the income inequality shock, only the income volatility shock. Given that people seemed to catch on to the Phillip’s curve employment/inflation trade off, it is going to take some pretty good data to convince me of Quiggin’s story.
1By story, I don’t mean anything negative. All economic theories are put into terms of a “story” or “narrative”. For example, rational expectations were first used in a “story” about agricultural markets and futures. That story didn’t seem to interest too many people. Robert E. Lucas and Leonard Rapping (and later others) used rational expectations, but changed the story to inflation and monetary policy. That story was much more interesting (probably because of recent macro-economic events) and it stuck.