Some White Borrowers May Get Refunds in a Racial Discrimination Settlement. Here's Why.

Even under normal circumstances, racial discrimination can be hard to spot. Identifying discrimination in car purchases and loans is particularly challenging, however – auto dealers don’t have to log the ethnicity of borrowers, and car loan terms are often bargained as part of a complex transaction.

Still, the Consumer Financial Protection Bureau promised to be aggressive about making sure that minorities had equal access to credit when buying a car, and back in 2013, it threw the book at Ally Financial, formerly General Motors’ lending arm. The agency said racial minorities were routinely charged higher interest rates than whites, costing 235,000 consumers an extra $300 or more during the life of their loans. Ally settled the allegations without admitting wrongdoing, and agreed to pay $80 million to consumers for damages.

Disbursing those “refunds” is a challenge, however, because neither Ally nor the CFPB knows which consumers belong to a minority group. Recall that, unlike mortgage applications, auto loan applications do not collect ethnicity.

Back in 2013, to calculate the potential number of victims, the CFPB used a fairly new and controversial method: it made educated guesses based on an algorithm known as “Bayesian Improved Surname Geocoding.” By plugging criteria such as surnames and home ZIP codes into a formula, the CFPB was able to calculate the likely ethnicity of the borrower.

This technique was criticized when it was used to estimate total victims to calculate the settlement. Now that it’s time to send out money (checks were supposed to start mailing out in January), criticism has grown much louder.

On Jan. 20, the Republican-led House Financial Services Committee released a report ringing the alarm bell that some non-minorities will receive payouts from the settlement fund.

Get matched with a personal loan that’s right for you today.

Learn more

“Sending remuneration checks to white borrowers as a means of remedying alleged discrimination against African-American, Hispanic and Asian borrowers is an unorthodox approach to fair lending enforcement, to say the least,” the report says.

Case ‘Invites Fraud on a Massive Scale’

Committee Chairman Jeb Hensarling (R-Texas) went further, saying the CFPB-designed method for sending out checks “invites fraud on a massive scale.”

In a statement, the CFPB said it is aware of the criticism. It declined to answer follow-up questions.

“We are reviewing the report from the majority staff of the House Financial Services Committee,” agency spokesman Sam Gilford said. “The CFPB’s goal has been, and continues to be, the elimination of illegal discrimination. Discrimination in auto lending has resulted in minority borrowers being unfairly charged higher interest rates on their loans. We will continue to fairly and consistently enforce the Equal Credit Opportunity Act to ensure borrowers harmed by discrimination receive the relief they deserve.”

Non-Minorities Asked to Opt Out

The disbursement controversy revolves around what to do with consumers that the algorithm is unsure about – a group of nearly 200,000 borrowers. For those consumers, Ally has been directed to send a simple “opt-out” notice – the consumer is told to respond only if she or he is not a minority. Consumers who do nothing will later receive a check. [Update: The CFPB declined to use an “opt-in” procedure for this group, which would have required consumers to affirmatively reply to the notice, confirming their minority status. The CFPB did use an “opt-in” procedure on consumers with a lower probability of minority status, but did not require them to sign an oath or supply any kind of affirmation.]

Not surprisingly, using that approach, only a tiny fraction – 0.46% – opted out of receiving disbursements. Hensarling believes that means non-minorities will receive checks, and has asked U.S. Attorney General Lorretta Lynch to suspend distribution of the money.

“It defies logic for federal agencies to distribute settlement funds without first verifying the eligibility of prospective recipients, particularly when the Bureau’s case is premised upon a flawed statistical analysis,” Hensarling wrote in a letter to Lynch.

The committee says internal CFPB documents claim that the opt-in method would have reduced the number of victims receiving checks to somewhere between 36,000 and 143,000, rather than the 235,000 expected by the CFPB.

“The Bureau employs a remuneration process designed to achieve political ends notwithstanding its acknowledgement that funds will be distributed to ineligible recipients. In this respect, the Bureau does a disservice to legitimate fair lending investigations undertaken by federal agencies,” the report concludes.

The House Financial Services Subcommittee on Oversight and Investigation promised to hold a hearing on the Ally issue in early February.