After a long career in mortgage finance and fair lending I’ve attended many panels, workshops, and conferences on fair lending. All too frequently at these events, I hear someone from the audience or even a panelist declare a “fair lending myth.” A fair lending myth is a statement about HMDA, ECOA, and FH Act regulatory enforcement that sounds reasonable on the surface, but does not withstand serious scrutiny. Here are some of the most common myths I’d like to dispel.
- Our hands are tied so we’ll do nothing. This myth states that with fair lending there is always a “Catch-22” situation. It says if we try to fix a problem, then the regulator will nail us for trying; so, we might as well not try. Usually this is crystallized in the phrase “we’re darned if we do and darned if we don’t.” This often comes up in the context of expanding outreach to potential minority borrowers. The flawed thought process is if we obtain more minority applicants, then we will deny more of them and therefore have a fair lending underwriting problem. While it is true that if you obtain higher levels of applications from any race you will have higher levels of denials, other things being equal. It does not follow, however, that fair lending issues unless the denial disparity increases without regard to qualifications. If a lender’s marketing targets a consistent borrower profile across all races, then this is very unlikely to occur. In fact, expanding lending to broader segments mitigates overall fair lending risk by lessening exposure to redlining risk.
- It’s better to have a low overall HMDA denial rate. While it appears to be a good thing on the surface, a low overall denial rate suggests two things: A.) The lender’s definition of an “application” for HMDA reporting may be too stringent, and/or B.) excessive pre-screening is potentially occurring. Either of these increases compliance risk.
- Fair lending is only about race and ethnicity. Lenders get preoccupied with this and forget about the other prohibited basis groups under ECOA and FH Act. Two others are gender and age for which hypothesis testing can be done with large HMDA and LOS datasets. Other protected groups include the following: persons with a disability, familial status, color, religion, national origin, marital status, income derived from public assistance, and applicants exercising a right under the Consumer Credit Protection Act. Any of the foregoing can appear in a lending context so a comprehensive fair lending monitoring program should seek to address them all.
- There is only one right way to rebut evidence of unfair lending. Nothing could be further from the truth. The best approach is a multi-pronged approach that includes redundancy of analyses. To obtain robust results a lender should corroborate fair lending outcomes through a variety of analyses before taking corrective action (e.g., assorted peer groups, different geographical cuts, and various statistical tests and models).
- Racial diversity of sales-side personnel and marketing does not matter. In fact, this is likely the single most cause of redlining patterns. A little improvement in this area goes a long way. Fair lending begins when you are planning to attract prospective applicants to apply for a mortgage.
My advice to lenders is to critically examine the statements you hear in casual conversations, conferences, and publications. The truth is that fair lending takes work and there are really no shortcuts. Please visit our website for additional information about our Fair Lending Consulting Services.