Editor's note: The last paragraph of this story has been amended to clarify NADA's position on CFPB guidance for identifying and avoiding discrimination in auto lending. NADA is not calling for formal rule-making as part of the process.
The Consumer Financial Protection Bureau has promised to provide members of Congress with a white paper later this summer that will explain in more detail how it detects discrimination in auto loans.
The bureau says it will issue the white paper in response to criticism from Congress, lenders and dealers that auto lenders don’t have enough specifics on the CFPB’s standards to police themselves.
“It’s been a source of, I think, some frustration to the Committee and to me, and to the Bureau, that we’ve been back on forth on different kinds of information about this, and we think we’re providing a lot of information, but people identify other information that they want,” CFPB Director Richard Cordray said last week at a House Financial Services Committee hearing.
“Partly as a result of that, we are going to put out a white paper on the proxy methodology to try to address that very directly later this summer. We’ll continue to try to be responsive on this.”
Meanwhile, the National Automobile Dealers Association said in a written statement that even if the CFPB white paper answers all questions about proxy methods, it almost certainly won’t be enough to solve the problem of what NADA calls a lack of transparency by the CFPB.
The proxy methodology Cordray referred to is the way lenders and the CFPB decide whether consumers belong to legally protected groups, such as minorities or women. Auto lenders aren’t allowed to collect race or ethnic data on borrowers. Instead, they use substitutes or proxies, such as the census tract where a borrower lives or first or last name.
The CFPB uses those results to determine whether protected groups on average pay higher rates for financing than other borrowers. Specifically, the CFPB says it believes lenders allow dealerships to discriminate against legally protected borrowers by allowing dealerships to set their own rate for dealer reserve, or dealer markup. The CFPB calls higher rates for protected groups a disparate impact.
However, there’s a wide margin of statistical error for identifying protected classes by proxy. For instance, some common first names can belong to a man or a woman. Or a woman could have a married name that’s identified with a protected minority when she’s not a member of a protected group, or the opposite could be true.
The CFPB acknowledges it uses proxies, but it hasn’t said much beyond that despite increasingly pointed questions from Congress.
Lenders also don’t like the fact that the CFPB has avoided saying whether there’s even a tiny acceptable level for disparate impact or whether zero is the only acceptable statistical result. With a margin of error, a zero result could mean protected groups actually paid less than everyone else.
NADA said there are many additional questions, such as how the CFPB determines it is comparing rates for “similarly situated” borrowers. That is, borrowers who are similarly situated in terms of credit history or other factors that could affect the rate, and different only in terms of whether they belong to a protected class.
NADA has encouraged its members to phone their representatives in Congress and ask them to co-sponsor a bill that would rescind the bureau's 2013 auto guidance and require more transparency before issuing future guidance. The bill also requires a public comment period and public access to documents supporting the guidance.