Mortgage lending in Boston - a response to the critics

New England Economic Review,  Sept-Oct, 1995  by Lynn Elaine Browne,  Geoffrey M.B. Tootell

• E-mail
• Print
• Link

Three years ago, the Federal Reserve Bank of Boston released an examination of racial patterns in mortgage denial rates in the Boston area (Munnell, Browne, McEneaney, and Tootell 1992 (MBMT)). The study was motivated by newly available data on mortgage applicants, showing that black and Hispanic applicants were two to three times as likely to be turned down for mortgages as white applicants. The study gathered additional data on applicants' debt burdens, credit histories, and other financial characteristics to see whether economic factors explained the racial difference in denial rates. Although the additional information did explain much of the difference, after taking account of economic factors the applicant's race still affected the probability of getting a mortgage.

The study was promptly given "landmark" status by some in the press, and in some respects the designation is warranted. The data collection was a major undertaking; thus, despite many calls for studies of racial lending patterns in other cities or sets of institutions, only one somewhat similar work had appeared as of mid-1995 (Stengel and Glennon 1995). It has also been influential. The study alerted both the mortgage industry and its regulators to the possibility of discrimination in mortgage lending. It has stimulated many financial institutions to re-examine their lending practices and has caused the federal supervisory authorities to change their examination procedures pertaining to fair lending. It has spurred efforts by the major secondary market agencies both to ensure that lenders do not interpret their credit guidelines excessively strictly and to reassess the appropriateness of some of these guidelines. The study may have provided some of the impetus to revise the Community Reinvestment Act regulations and it probably reinforced the Department of Justice's efforts to pursue fair lending more vigorously.

Given the attention the study has received, criticism is to be expected. Some of the criticism has been scholarly. Some has been strident, with one critic even hinting the study was "consciously fraudulent" (Roberts 1993). Much of the criticism seems to reflect a view that discrimination simply cannot occur in lending; much, especially some of the most vociferous, appears driven by concerns over policy directions that the study might inspire. These concerns have taken on new life in the past year in response to the Justice Department's more aggressive stance towards redlining and fair lending violations.

Thus, it seems appropriate to respond to the major criticisms of the study, showing why the study is sound and why its finding that discrimination and economic factors both contributed to the racial disparities in mortgage denials in Boston is solid. At the same time, it should be noted that the study itself did not advocate any specific remedial policies, simply concluding that "a serious problem exists in the market for mortgage loans" such that "lenders, community groups, and regulators must work together to ensure that minorities are treated fairly."

Subsequently, in testimony before the Senate Banking Committee, Richard Syron, then President of the Federal Reserve Bank of Boston, made clear that primary responsibility for addressing the problem of discrimination in mortgage lending lies with the industry, stating, "the most critical step is for mortgage lenders to acknowledge at least the possibility that the results of their lending process are discriminatory. As long as lenders sincerely believe their procedures are beyond reproach, efforts to get them to change will have limited success. . . . Lenders' reactions to the study suggest that they are now questioning what they always took for granted. They are starting to recognize that simply having a policy that prohibits discrimination does not prevent discrimination" (Syron 1993). Fostering this self-questioning was a major accomplishment of the Boston Fed study and it would be most unfortunate if it were reversed.

Although criticisms of the Boston Fed's findings in the media have been numerous, most of these repeat the arguments of three sources (Brimelow and Spencer 1993; Liebowitz 1993; and Horne 1994a and 1994b). In a separate category stands the more technical criticism by Yezer and various co-authors (for example, Yezer, Phillips, and Trost 1994) that negotiations between borrowers and lenders preclude finding discrimination. The appendix summarizes the issues raised by each of these and other major critics and provides point-by-point rebuttals.

The criticisms can be grouped into five categories:

1. Default rates - If discrimination exists, the average default rate of minority borrowers will be below that of white borrowers, whereas data in the Boston Fed study suggest that minority and white default rates are similar.

2. Omitted or missing variables - Variables have been omitted from the analysis that might explain the influence of race on the mortgage decision.

3. Misspecification - A different specification of the mortgage decision process might lead to a conclusion that discrimination is not occurring. The argument that the mortgage decision process is a negotiation is a specification issue.