The Role of FHA Data in the Lending Discrimination Discussion

If minority loan applicants perceive discrimination by conventional lenders, they will crowd into the nondiscriminating Federal Housing Administration (FHA) sector. This movement to FHA will be particularly pronounced among minority borrowers who have difficulty providing easily observable evidence of their creditworthiness. The result will be lower default or loss on minority loans in the FHA portfolio. Thus testing for discrimination based on differential default or loss is a sound procedure, but it can only be done subject to the availability of data on FHA borrowers. The range of data, particularly credit history variables, maintained in FHA data files should be expanded. Ultimately, tests for adverse impact discrimination will require validation of FHA underwriting procedures based on these data. The interchange between Berkovec, Canner, Gabriel, and Hannan (BCGH) and Ross, Galster, and Yinger (RGY) recalls the old joke about laying all the economists in the world end to end and failing to reach a conclusion. While I do not accept this character-ization, it is certainly true that economists generally rely on nonexperimental data an