Racial Discrimination in Mortgage Lending?

Over the past several years, we have commented on numerous reports and studies that have alleged racial discrimination in lending, especially mortgage lending. Typically, the "evidence" of racial discrimination in these studies has been the fact that applications from minorities were denied more frequently than were applications from non-minorities. In all fairness, the authors of these studies did not have the credit scores of the applicants for home mortgages, but seldom did they mention this serious limitation. Unfortunately, the absence of this crucial information did not deter the authors of those studies, or the journalists that reported them, from claiming racial discrimination in mortgage lending.

Fortunately, we now have that critical information in a study published in Housing Policy Debate, a respected academic/policy journal published by the Fannie Mae Foundation. In the last issue of Spotlight we reviewed those findings of the study that evaluated the accuracy of human (or manual) evaluation of applications for mortgage loans versus evaluation by a computer program. You may recall that the computer won.

Among other data provided by the authors was a chart showing Fair Isaac & Company (FICO) credit scores of applicants for mortgage loans by minority and non-minority applicants in 2000.

Note: Details may not add to totals because of rounding. Source: Susan Wharton Gates, Vamessa Gail Perry; and Peter M. Zorn, "Automated Underwriting in Mortgage Lending: Good News for the Underserved?" Housing Policy Debate, vol. 13, no. 2, 2002, p. 386.

Assume that a mortgage lender establishes a policy of not approving loans from applicants that score less than 630 (about halfway into the second row). Roughly, this policy suggests that about 32 percent of the minority applicants would be rejected, whereas only 23 percent of applications from non-minorities would be rejected. Racial discrimination? Hardly. A critical factor in granting or rejecting applications for loans by consumers or business firms is the appraisal of risk. Researchers who do not have this measure are not in a position to evaluate whether past decisions in granting credit were discriminatory.