If the fair-lending vigilantes catch you, it’s your own fault

Practicing fair lending, and fair banking more broadly, must be a 24/7 job regardless of the political climate. This must also be the case with fair-lending enforcement.

This, unfortunately, was not the case under the Trump administration. The best proof is the Biden administration’s much celebrated TrustMark redlining and racial discrimination case, which should have been filed by the Department of Justice and other regulators four years earlier.

Banks and their regulators were comfortable with the Trump deregulatory environment, but this was not the case with community groups. They responded with what can best be called fair-lending vigilantism.

Traditional fair-lending analysis involves reviewing Home Mortgage Disclosure Act data to identify significant disparities in white vs. nonwhite lending applications and approvals. Since such statistical analysis is time-consuming and can always be challenged, the do-it-yourself approach to fair-lending enforcement is testing, also known as mystery shopping.

Government-financed testing through the Department of Housing and Urban Development’s Fair Housing Initiatives Program is monitored to make sure it is done right. One such audit found a conflict of interest with a national coalition improperly accepting $2.4 million in donations from 10 of the 38 banks that were being tested.

There are no such HUD audits if a community group’s testing is financed by the nearly $500 billion of “Community Benefit Agreements” made with merging megabanks.

While “matched pair testing” requires a scientific approach, no one can stop two people, one a minority and one not, from posing as loan applicants in a branch. They usually record differences in their greeting (e.g., with or without a smile, handshake, business card or small talk), questions about themselves or the subject house, and even suggestions about how to get better loan rates and terms.

Experienced loan officers often realize they were “tested” after the fact, but less experienced ones may have no clue. Bankers “passing” their fair-lending test receive no accolades because they are just doing their job. There is, of course, a totally different outcome when they do not do their fair-lending job.

The best example is the Liberty Bank case in 2018. The Connecticut Fair Housing Center used white “control” versus Black and Latino “protected” testers, with the latter having better incomes and credit scores. Six different tests concluded the protected testers were given significantly less information, inferior terms and were discouraged from applying, with some being “steered” into minority neighborhoods and subjected to unfortunate comments.

The center filed a Fair Housing Act lawsuit alleging racial discrimination and redlining, which forced a settlement, even though the bank had an outstanding Community Reinvestment Act rating and passing marks on its fair-lending exams. Besides derailing a pending merger, the 2019 settlement cost nearly $20 million counting legal and other fees, not to mention the reputational impact.

Community groups, realizing they could do the fair-lending enforcement job the government was not doing, planned similar vigilante efforts for 2020. However, as with nearly everything else, COVID-19 put an end to conventional fair-lending testing by closing most bank lobbies.

The pandemic did not stop committed, clever and cash-rich community coalitions from their fair-lending guerilla warfare. A national community coalition came up with two alternative testing techniques via telephone and email.

Their first wide-scale telephone testing focused on differential Paycheck Protection Program lending treatment of white versus Black callers. They used “racially identifiable names” perceived to be “highly correlated with gender and race. Citing linguistic profiling research that race can be determined “over the phone after a few sentences,” they hired experts to test the testers’ voices to ensure they were racially identifiable.

Testing began in April and May 2020 with 17 banks in D.C. and then expanded in July and August 2020 to 47 banks in Los Angeles, using white vs. Latino callers. Each bank was tested twice to determine a possible discrimination pattern. They found minority testers were treated less favorably in terms of PPP application encouragement, products offered and information provided.

Realizing the potential of telephone testing, they used it on over 20 Veterans Affairs home lenders in the Tacoma/Seattle area in June and October 2021. This resulted in a HUD racial discrimination complaint, since the Black tester (with stronger qualifications) was presented less favorable information, assistance and loan terms than the white tester. The coalition then used telephone testing on one lender in North Carolina in February and September 2021, resulting in another HUD complaint. HMDA data bolstered both complaints.

Instead of targeting a specific institution via in-person or telephone testing, virtual email testing targets a specific loan officer. To the best of my knowledge, these new testing techniques are being currently employed by community groups but not by federal bank regulatory agencies.

Using a loan officer’s email published on a bank website, similar loan information requests are sent from identifiably white vs. minority applicant emails over two different periods.

In the unfortunate event the officer responds more favorably to the white applicant on both testing dates or, worse yet, responds only to the white applicant and ignores the minority applicant, a HUD racial discrimination complaint and investigation likely follows. Banks may have to bear legal and other fees related to responding to massive HUD data requests. The worst-case scenario ends with a conciliation agreement with a cash settlement, and significant reputational damage.

One community banker who “failed” virtual testing and whose bank was served with a HUD racial discrimination complaint used the term “entrapment.” Another was very upset with the merging megabanks helping to finance virtual testing.

Instead of blaming the community groups doing virtual testing or the merging megabanks financing them, however, the blame belongs with the bankers who failed to do their fair-lending job.