Wednesday, June 25, 2008

Over at Economists View Winslow R. says...

Robert Waldman wrote: "You know what else? Race exhibits a substantial, income-independent influence on consumption -- Black households consume less than White households with the same income."

Anne posted a article yesterday showing that blacks with the same credit scores, income etc. pay higher interest rates than whites. I don't believe subprime interest payments would be included in your consumption number.


It sure wouldn't be included. When I wrote consumption I should have written consumption of non durables and services so not even services from cars let alone houses.

The phenomenon noted by Anne can perhaps, in part, be explained if banks are (illegally) using Friedman's logic. That is called statistical discrimination. The idea is that bankers rationally decide that if there is a Black and a White family with the same current income, the expected value of future income of the Black family is lower, so they are a more risky borrower.

This is just as illegal and unfair to the individual as discrimination due to hostility or irrational prejudice. There is a fairly large literature on the possible magnitude of statistical discrimination in labor markets. It explains a tiny fraction of racial wage differentials.


I would guess it is that way for interest rates too.

Statistical discrimination can be distinguished from discrimination due to hostility (taste) or irrational prejudice using, you guessed it, statistics.

With a lot of data, one can estimate foreclosure rates as a function of current income, credit score, and race. Then estimate interest rates as a function of the estimated foreclosure rate and race. The second coefficient on race is an estimate of discrimination other than statistical (hostility, irrational prejudice, whatever). The effect of the first coefficient on race on average interest paid by race is an estimate of statistical discrimination.

Statistical discrimination is just as illegal as discrimination due to hostility (note the phrase appears in the law) and it is just as immoral and unfair. However, it is economically rational in the sense that the bank benefits from its tort while a banker motivated by racial hostility will have to pay for the pleasure of shafting African Americans. Thus market selection will not eliminate statistical discrimination.

No comments: