Several years ago Bill Cosby chided poor blacks for spending their limited incomes on high-priced shoes and other items of conspicuous consumption instead of investing in education. Cosby was widely criticized but I went to the numbers, specifically Table 2100 of the Consumer Expenditure Survey and found the following for 2003:
Average income of whites and other races: $53,292.
Average income of blacks: $34,485.
Expenditures on footwear by whites and other races: $274
Expenditures on footwear by blacks: $440.
As I noted then "to do a proper comparison we
would have to correct for income and other demographic variables." The correction has now been done by three researchers in an NBER working paper (non-gated version). The results didn’t surprise me. How about you?
Using nationally representative data on consumption, we show that
Blacks and Hispanics devote larger shares of their expenditure bundles
to visible goods (clothing, jewelry, and cars) than do comparable
Whites. We demonstrate that these differences exist among virtually all
sub-populations, that they are relatively constant over time, and that
they are economically large.
To give the authors credit where credit is due they also show that the differences in conspicuous consumption are large and important. The differences in spending on clothing, jewelry, and cars, for example, can explain half of the differences in wealth between the races (conditional on permanent income) and a significant share of the differences in education and health spending.
Why do these differences exist? Aside from simple differences in preferences, signaling is one possible explanation. Suppose that high income confers status. Other people judge your income based on your conspicuous consumption and your group’s income. Under plausible conditions, the authors show that if your group’s income is already high conspicuous consumption has a low marginal product. Put differently a black man who wears a very expensive suit gets a bigger increase in status than a white man who wears the same expensive suit because the baseline income prediction is lower for the former.
The theory is plausible but I wonder if other groups haven’t converged on more efficient methods of signaling. Some groups, for example, use education as a signal. Other groups like to show how clever they are by writing pithy summaries of new economics research.