Must Lenders Prove They Don't Discriminate?

Picture two freight trains. They are miles apart and traveling slowly. But they are on the same track on a collision course, and gathering speed. One train is carrying developments that could bring financial institutions under unprecedented criticism on the discrimination front. Chief among these is the requirement for public disclosure next year of expanded information reported under the Home Mortgage Disclosure Act (HMDA). The other train represents bank regulatory agencies' enforcement of nondiscrimination laws. HMDA effects. The new HMDA data will enable the public, for the first time, to compare the approval and denial rates of loan applications from minorities versus whites and from women versus men, holding income constant. It's possible that these disclosures will show that minorities and women are turned down far more often than whites and men of similar income. Such disparities have been found in four major studies in different parts of the country. These past analyses, ranging from the 1988 Pulitzer-Prize-winning Color of Money series by the Atlanta Journal-Constitution to last year's Boston Federal Reserve Bank study, found patterns suggesting that financial institutions lend far more in white census tracts versus black ones, even when income and property values are held constant. Do such patterns prove discrimination? Far from it. Statistics suggesting system-wide discrimination greatly overstate the case. As every lender knows, income is not a proxy for credit-worthiness. Studies that have gone a step further, looking at such factors as debt load, have found far less disparity between lending to minorities and whites. Nevertheless, the studies have raised serious questions about discrimination in the minds of many congressional leaders and civil rights advocates-and probably in the minds of many journalists. Agencies scrutinized. Now let's look at the agencies' enforcement of nondiscrimination laws. In May, the Senate Banking Committee Consumer Affairs Subcommittee held its fourth hearing in the past year covering discrimination, the Community Reinvestment Act, or both. The agencies testified that they rarely cite institutions for substantive violations of the Fair Housing Act or Equal Credit Opportunity Act (ECOA). They do cite technical and procedural violations. Because ECOA is very technical, lenders make inadvertent errors-errors that have nothing to do with discrimination. Even if they had not reported this fact at congressional hearings, the agencies' findings would be trickling into the public arena through the new CRA performance evaluations that the regulators began to make public beginning July 1. In these reports, the law requires the agencies to draw a conclusion regarding the bank's performance on every assessment factor in the CRA regulation. Two factors deal with discrimination. If official public data suggest that there may be widespread discrimination throughout the system, it could become difficult for agencies to report that they find virtually no discrimination in their examinations. Of course, the agencies might successfully defend the fact that they rarely cite violations. This could be done if they demonstrate to Congress that they thoroughly examine lending practices for discrimination and that the apparent discrimination patterns can be explained by valid credit factors. At this time, however, it appears that the regulators lack a really clear methodology for deciding what constitutes discrimination, beyond citing technical violations, such as writing husband instead of spouse on an application or sending out an unclear adverse action notice. It seems inevitable that, in the next year or so, the agencies will put a great deal of effort into strengthening their ability to determine whether institutions they examine are discriminating, especially in housing credit. …