« The Weight To Be Given To Comments From Bankruptcy Judges On Proposed Bankruptcy Rules and Forms | Main | How Backpage Is Different from Choke Point »

The Politics of Indirect Auto Lending and the CFPB

posted by Adam Levitin

Steve Davidoff Solomon has a Dealbook column on the CFPB's attempts to regulate auto lending that unfortunately gives the wrong impression about what the agency is up to, but which does tee up a really interesting question about the agency's politics.

The auto lending story is really simple. There are two ways you can get an auto loan. You can get it straight from the dealer (direct lending) or you can get it from a third party (indirect lending), with the dealer brokering the loan.  When dealers make loans they often sell them to third parties (including securitization conduits), but they can also keep them on their books.  The CFPB has statutory authority for rulemaking, examination, and enforcement over indirect auto lenders (excluding community banks and credit unions).  The CFPB also has statutory authorityfor rulemaking, examination, and enforcement over dealers to the extent they do not routinely sell their loans. But for auto dealers that are making loans and routinely selling them, the CFPB has no authority (they are still subject to enforcement actions by state attorneys general and the FTC).

The CFPB's undertaken a bunch of enforcement actions against indirect auto lenders—not dealers—for discriminatory lending. The problem, it seems, is that when auto dealers act as loan brokers they are compensated based on the terms of the loan through something called a "dealer reserve." The higher the interest rate of the loan, the greater the dealer reserve. If this looks familiar it is because it is the same kickback arrangement as "yield spread premiums" in mortgage lending. Now, there's nothing inherently illegal about a dealer reserve--a dealer is permitted to mark up a loan it brokers--but if customers of color are routinely steered by the dealer into higher cost loans, then the dealer reserve has become an instrument of discriminatory lending, just like with yield spread premiums. While the CFPB cannot go after the dealers for accepting kickbacks, even discriminatory ones, it has full and clear authority--and should--go after indirect lenders that pay the kickbacks when the effect is discriminatory lending. (Indeed, no indirect lender has challenged the CFPB's authority in this regard.) 

Unfortunately, you wouldn't get this impression from the Davidoff piece. What the CFPB has done is not as Davidoff claims, "sidestepping the restriction on overseeing auto dealers". The CFPB hasn't sued any auto dealers that are acting as dealers. The deal cut in Dodd-Frank was that the CFPB would have clear authority to go after the indirect lenders, which everyone knew would have some indirect impact on dealers. This just a standard hydraulic regulatory effect that exists all the time in financial regulation. Regulating Fannie Mae and Freddie Mac, for example, will have downstream effects, but that's not legally the same as regulating the downstream entities in the housing finance market.  

Nor has the CFPB "acted aggressively to impose extensive regulation on these banks that lend indirectly to dealers." The CFPB hasn't imposed any new regulations on indirect auto lenders. It has merely issued regulatory guidance regarding how it interprets the Equal Credit Opportunity Act, a statute that has been around since the 1970s. That's not new regulation. The basic rule is the same it's always been: don't discriminate on the basis of race, sex, religion, etc. That's really not a hard rule to follow.  

But this just raises the question of how the CFPB knows there is discriminatory lending. The answer is statistics. Unfortunately there is no data reporting on the race of auto loan borrowers; there is no equivalent of the Home Mortgage Disclosure Act (an anti-redlining statute) for auto loans. Instead, the CFPB has to use statistical methods to determine borrower race. The method the CFPB uses is called Bayesian Improved Surname Geocoding. Consumer surnames correlate with race to some degree. This correlation can be improved through geographic adjustments. To illustrate, the name Gomez in Los Angeles is likely Hispanic, while the name Gomez in Fall River, Massachusetts is likely Caucasian (Portuguese).  BISG is a good methodology, but it's not perfect. It doesn't claim to have 100% accuracy. And this has CFPB critics upset. Not only do they object to the whole disparate impact theory of discrimination (which they claim is "controversial" despite it being upheld by the Supreme Court this year), but they are particularly upset about its application without 100% accurate data.  Never mind that the CFPB can control for the statistical probability of error in BISG and only proceed in cases where there is clear disparate impact controlling for the probability of error.

The House passed a bill, HR 1737, that would have the CFPB redo its indirect auto lending guidance (including a cost-benefit analysis). 88 Democrats voted for it. On its face, this bill seems to be about ensuring that the CFPB has carefully considered how to approach the problem of disparate impact in indirect auto lending, but what the bill really does is send the CFPB back to the drawing board at a few years at least, by which point there might be President Trump or other changes that clip the CFPB's wings. Put another way, HR 1737 will prevent the CFPB from policing discriminatory auto lending for several years under the guise of concern about a technical statistical methodology. (I wonder how many folks who voted for the bill can actually explain the methodology?) That's a major step backwards in civil rights. That's totally missing in the Davidoff presentation of the auto lending issue. 

Davidoff also argues that the CFPB's impending regulation of pre-dispute binding mandatory arbitration will end it in trouble with the Supreme Court. That's just wrong. The Supreme Court has taken a very broad view of the Federal Arbitration Act, but the CFPB has express authority to limit arbitration. The only question that is likely to be litigated is whether the CFPB had an adequate basis for its rulemaking, and that's a Chevron question. I don't see this one likely going up to the Supreme Court, and if it does it's not going to be decided on the basis of past SCOTUS arbitration jurisprudence. 

Davidoff is right that the CFPB is sailing into trickier political waters with arbitration, auto lending, and payday loans (and he hasn't seen the political storm that will happen when overdraft gets on the agenda). But the politics of auto lending aren't the same as those of payday or arbitration. Auto dealers have much, much more political sway in Congress than payday lenders (perhaps more than any other group, including community banks and credit unions), and the arbitration politics are more diffuse.

But here's the implicit choice suggested by Davidoff:  (1) the CFPB can pursue what it believes are the right policies, at the risk of political backlash (remember that a Democratic Congress defunded the FTC in 1980 over its attempt to regulate advertising targeting children until the FTC came back with new interpretations of "unfair" and "deceptive" in the FTC Act) or (2) the CFPB can compromise on policy to smooth out the politics. 

Most agencies take the latter position, perhaps consoling themselves that they'll live to fight another day. They'll compromise for political convenience. But I don't think the CFPB can. The CFPB is a mission-driven agency. If it compromises on that mission, its integrity is compromised. At that point it starts to resemble the captured agencies it was intended to replace, and it becomes just another federal financial regulator. There might be some wiggle room on the edges for the CFPB--look at how it has tried to make nice with community banks in the mortgage lending and servicing space, but I don't think the CFPB can substantially change its course. I think it will be damn the torpedoes for the CFPB, but the next couple of years will be interesting.  


As an auto lender I gotta ask: does the wildly higher default rate of African Americans mean anything to you? I would bet you have no idea that is true. It wouldn't surprise me if you even thought it wasn't true.
A non-monetary factor is that 95% of abusive behavior to my staff is from African Americans despite our loan book aligning almost perfectly with their proportion of the general population.
We don't treat African American customers differently but it is very easy to see why people with small operations might. With very few exceptions they are awful customers.
Despite the myth that dealers love to repo (only makes sense with high deposits which don't exist anymore) bringing a car back in results in a loss. The social justice juggernaut only cares about one side of the equation.

Two technical points, one for Adam and one for Mike.
1. For Adam: Chevron is not about rational basis. Rather, it is about agency interpretation of statutes. I think you're thinking of the Motor Vehicles case. Like Chevron, it is often honored in the breach.
2. For Mike: If you want to make a coherent argument, a high default rate is not enough. You would have to argue that African-American default rates are higher than those predicted by the ordinary credit metrics. (And for the record, my wife is African-American and the most terrible customer you could imagine--she does things like check out the dealer's lot to look for overstocks.)

1: For Mike: typically disparate impact analysis is not about default rates. Instead, it is about the terms of the loan. In the case of dealer reserves, the issue would probably be whether the dealer reserves were larger for people of color or women, etc. Given that the dealer doesn't hold the credit risk, the ultimate default rates shouldn't matter--in fact, if they were higher for people of color or women, that might suggest that lenders would prefer to purchase loans with _smaller_ dealer reserves.

2: For Ebenezer Scrooge: you're right, I had Motor Vehicles Ass'n v. State Farm in mind. Chevron is about statutory interpretation, while State Farm is about the basis for policy decisions. That said, it's pretty hard to distinguish between Chevron step two and a State Farm analysis, even if they are formally distinct.

The comments to this entry are closed.


Current Guests

Follow Us On Twitter

Like Us on Facebook

  • Like Us on Facebook

    By "Liking" us on Facebook, you will receive excerpts of our posts in your Facebook news feed. (If you change your mind, you can undo it later.) Note that this is different than "Liking" our Facebook page, although a "Like" in either place will get you Credit Slips post on your Facebook news feed.



  • As a public service, the University of Illinois College of Law operates Bankr-L, an e-mail list on which bankruptcy professionals can exchange information. Bankr-L is administered by one of the Credit Slips bloggers, Professor Robert M. Lawless of the University of Illinois. Although Bankr-L is a free service, membership is limited only to persons with a professional connection to the bankruptcy field (e.g., lawyer, accountant, academic, judge). To request a subscription on Bankr-L, click here to visit the page for the list and then click on the link for "Subscribe." After completing the information there, please also send an e-mail to Professor Lawless ([email protected]) with a short description of your professional connection to bankruptcy. A link to a URL with a professional bio or other identifying information would be great.