The Next Fair-Lending Risk: Disparate Impact

Lenders trying to keep their sights on the moving target of fair-lending risks to take in Disparate-impact legal actions are in the works, particularly aiming at the expanding use of credit scoring. The regulators' 1994 interagency policy statement on fair-lending identifies three types of illegal credit discrimination. (1) Overt discrimination--where a lender directly and intentionally turns down a qualified applicant or sets adverse loan terms based on a prohibited factor like race or gender. (2) Disparate treatment--where similarly qualified customers are treated differently based on a prohibited factor. (3) Disparate impact--use of a policy or practice that is neutral on its face but has a disproportionate adverse effect on a group, and that cannot be justified by necessity. Illegal disparate treatment and disparate impact can both occur, the agencies say, even in cases where there is no intent to discriminate. Since the fair-lending issue exploded on the banking scene in the early 1990s, the spotlight has been on disparate treatment. All the Department of Justice actions have been primarily brought on those grounds, and the bank regulatory agencies' fair-lending procedures are geared towards disparate treatment. The Justice Department has said there are three types of disparate treatment--underwriting discrimination, pricing discrimination, and marketing discrimination (which covers access issues, such as lack of branches). Throughout this period, there has been little concern about overt discrimination, because it is so rare. And there has been very little focus on disparate impact. That may be about to change. DISPARATE IMPACT ARRIVES For all the difficulty lenders face in staying on the right side of the non-existent standards on disparate treatment, that issue is child's play compared with sorting out what is and is not illegal on disparate impact. The disparate-impact concept has been imported into the lending arena from employment law. In general, the concept involves a three-step test based on these questions: (1) Can the policy or practice be shown to produce a statistically significant adverse impact on a protected group? (2) If so, can the company show a business necessity or compelling need for the practice? (3) If so, is there another practice that would satisfy that and have a less- discriminatory effect? In the employment area there is some government guidance and a body of case law that helps clarify what these three tests mean in practice. In credit, there is almost no illuminating case law (a situation that some civil rights groups want to remedy, as noted below). This leaves lenders highly exposed to disparate-impact challenges, because every act/on a lender takes has a disparate impact on some group. Many decisions are disproportionately adverse to minorities, women, the old, the young, and the unmarried, not because the lender means to exclude them but as a by-product of strategies and operations adopted for other reasons. Thus, the opportunity for litigators to bring a disparate-impact case by triggering the first part of the test is wide open. Fair-lending groups acknowledge that this is easy. I recently asked a leading fair-lending litigation attorney whether he believed that, in the absence of credit discrimination, credit would be distributed equally among racial groups. No, he quickly said, explaining that because economic factors like wealth and income are not equally distributed, credit, logically, would not be, either. That means that the first hurdle--the one that gets the lender into the disparate-impact-litigation box--is very low. Almost anyone could show that a lender's practices have produced a disproportionately adverse effect in some way on groups. …