Racism at work? - racism in mortgage loans

National Review, April 12, 1993 by Peter Brimelow

Everyone has heard of neoconservatism and neoliberalism, but the real issue in America today is neosocialism. Classical socialism called for direct state ownership of the means of production, distribution, and exchange. Neosocialism just aims at political control. Socialism claimed to be more efficient. Neosocialism claims to be more equitable. Above all, neosocialism professes to combat "racism," since this magic word cows all opposition. Apparent neosocialist objective of the season: commandeering the banking system and forcing it to subsidize key client constituencies.

And it's not going to be easy to stop. Late last year, the Wall Street Journal's news pages carried five campaigning stories in eight weeks alleging that residential mortgage lenders were discriminating against minorities. The Journal's evidence, raw rejection rates, was essentially worthless because it took no account of standard credit considerations like net worth and income. But then the Journal reported a Federal Reserve Bank of Boston study of a sample of mortgage applications that did correct for these criteria. And it found minorities were still rejected at a (slightly) higher rate. This difference, the Boston Fed concluded, could only be due to racism.

The ecstasy that greeted this conclusion ("Definitive--changes the landscape," claimed the Office of the Comptroller of the Currency) was always absurd. The Boston Fed study itself noted that rejected minority applications on average had "poorer objective qualifications" so that "a systematic bias in mortgage lending is very difficult to document." For that matter, Asian-Americans were rejected less often than whites.

At Forbes, Leslie Spencer and I got interested in a conspicuous omission from all these mortgage-discrimination stories: default rates. Had the Boston Fed study corrected for those?

Ah yes, said Alicia Munnell, then the Boston Fed's Research Director. Not of course in the study's sample, which was too recent, but by looking at default rates in black and white Census tracts. And they were equal.

Miss Munnell apparently assumed this meant blacks and whites were equal credit risks. But it didn't. Those black mortgage holders were, by definition, the ones who had already passed the mortgage approval process--which had presumably rejected the usual higher proportion of blacks on the way. Instead, the fact that black and white default rates were equal meant that the market was working. Mortgage lenders were somehow able to weed out the extra black credit risks, reducing defaults down to the same, apparently acceptable, rate as for whites.

"[That] is a sophisticated point," Miss Munnell told us. Our impression was she just hadn't thought of it. "I do believe discrimination occurs," she said. But she now readily conceded, "I do not have evidence ... no one has evidence."

They don't have evidence, but they sure have convictions. Our report of the Boston Fed study's fatal flaw appeared before Christmas. Perhaps the Associated Press's Rob Wells was too seasonally merry to adjust his version of Boston Fed ecstasy that appeared December 27. But this hardly applies to the Wall Street Journal's Albert R. Karr and John R. Wilke, or the New York Times's John H. Cushman, whose uncritical stories appeared on February 26 and March 7 respectively. Only the fact that Miss Munnell was nominated for a Treasury job, and syndicated columns by Llewellyn Rockwell and Paul Craig Roberts, saved our story from the memory hole.

Forbes did receive a letter from one Stephen M. Cross, deputy controller for compliance management, comptroller of the currency. Significantly, Cross made no effort to defend the claim that the Boston Fed study proved discrimination. Instead, he fell back to the position that we hadn't proved it didn't exist. And he offered this fantastic rationalization: Racism may be preventing qualified blacks from getting mortgages, tending to lower their default rate. But simultaneously, it may be causing premature foreclosure on black mortgage holders, tending to raise their default rate. The two forces may miraculously balance out. Which would result in a black default rate equal to that of whites.

This is indeed logically possible. It is also logically possible that the equal default rate is caused by green goblins stealing blacks' application forms. With a little imaginative effort of this sort, it will always be possible to allege racism.

Actually, of course, Miss Spencer and I never claimed to show that mortgage discrimination did not exist. We merely said the Boston Fed had failed to prove that it did. But Cross and his ilk do have this problem: The hypothesis that racism does not affect mortgage lending yields a prediction--equal default rates. This turns out to be true. Their alternative hypothesis, that racism is at work, implies unequal default rates. This tums out to be false. The racism hypothesis can only be sustained by ridiculous rationalizations. Accordingly, the presumption of normal scientific inquiry would be against it.