The L.A. Times reports:
A racial disparity in mortgage lending rates appears to be sharper in Los Angeles and other California metropolitan areas than the rest of the country, according to an analysis of federal data to be released today.
The study by the Assn. of Community Organizations for Reform Now, an advocacy group for the poor, looked at the percentage of higher-cost loans issued in minority communities compared with nonminority neighborhoods in the same metropolitan area.
My immediate reaction upon reading that was: of course. Assuming that there is a racial disparity in lending rates to begin with — a situation that is likely to result from non-racial factors — that disparity will naturally be magnified in an area where housing prices are higher, because higher-cost loans will be used much more commonly. So I was hardly shocked to see this:
California’s high housing costs are probably one reason behind the findings, said Raphael W. Bostic, a former economist for the Federal Reserve now at USC, who was not involved in the study.
Gee, ya think?
One paragraph temporarily seems to suggest that discrimination could be a factor:
The Federal Reserve, in its own study of the disclosure act data released this month, determined that African Americans and Hispanics were far more likely to get sub-prime loans than whites and said the gap could not be fully explained by factors such as income.
Until you learn this:
The federal loan data do not include information about borrowers’ creditworthiness and other basic factors that would provide a more complete picture of the lending process, Hobbs [Dustin Hobbs, a spokesman for the California Mortgage Bankers Assn] said.
In other words, there are no controls. This so-called study does a comparison with no controls. How in the world the Federal Reserve was able to determine that “the gap could not be fully explained by factors such as income” when they didn’t have access to basic control data like creditworthiness, I have no idea. And the intrepid reporter who wrote the story apparently didn’t ask, probably not wanting to gum up a story that he knew would be on the front page of its section.
I just have one question: does anybody ever do a study of alleged racial disparities, with respect to any topic, who knows how to do a proper study with proper controls?
Because it sure doesn’t seem that way. Race disparity researchers seem to be uniquely ignorant of the need for controls, if the widely reported racial gaps in areas like criminal justice and lending practices are any indication. It’s as if these researchers all go to schools where the statistics classes completely omit any discussion of the concept of controlling for relevant factors.
We see this all the time, of course.
I am used to seeing studies splashed all over the front pages of the newspaper discussing alleged racial disparities in areas like credit and criminal justice.
In the credit area, we are constantly treated to stories that say, for example, that banks turn down loans for blacks that they would approve for whites. In other words, banks are accused of being more selective with blacks than they are with whites. If this were true, blacks would have a lower default rate, since only those who clearly and indisputably qualify for the loans are getting them. In fact, the default rate among blacks is consistently slightly higher than that of whites. This indicates that the banks are, if anything, taking a chance with blacks slightly more often than they should according to the cold numbers. But you never see that little fact reported.
In areas like treatment by the criminal justice system, these “studies” almost universally fail to control for the one of the most important factors in a person’s treatment by the system: his criminal record. Prosecutors routinely give better deals to those who have no criminal record than they do to those with long rap sheets. I am confident that society wants it that way. Yet you see studies all the time — generally by liberal interest groups, though their affiliation is not always disclosed — detailing alleged gaps in the way that the system treats blacks and whites. These “studies” almost never control for the defendants’ criminal records.
But they’re generally splashed on the front page of the section, just as today’s story is on the front page of the Business section. You’ll never find a story highlighting alleged prejudice against blacks buried inside the paper. That would be foregoing an opportunity to inflame the black community. And we can’t have that.