The White House’s Council of Economic Advisers has put out a crazy report about the supposed costs of the CFPB. It’s frankly embarrassing to see such shoddy legal and economic analysis come out of the CEA. 

Basically everything in it is wrong, starting with the first sentence, which is simply false:

The Consumer Financial Protection Bureau (CFPB) has steadily expanded its jurisdiction since inception, extending oversight across all consumer credit markets, including mortgages, auto lending, and credit cards.

The CFPB has not “expanded” its jurisdiction an iota since its inception in 2011—and it cannot. Its jurisdiction is fixed by statute and has always extended to all consumer credit markets. (I’m happy to share three chapters of my textbook on this very topic with the good folks at CEA…) The only possible argument here is that the CFPB has designated some firms as “larger participants” in their markets, making them subject to examination, but that’s just utilizing existing jurisdiction, not “expanding” it. If you’ve got a beef with CFPB jurisdiction, take it up with Congress. 

And then there’s this gem: 

from inception, the CFPB has avoided transparency and accountability, opting to regulate markets through its supervisory and enforcement authorities, which are not subject to the formal rulemaking process (or congressional review). In addition to the 400 final rules and formal advisory opinions, CFPB has avoided transparency.

What on earth is the CEA talking about? The CFPB has not issued anything close to 400 final rules and formal advisory opinions. It has issued 136 final rules, 11 interim final rules, and 17 advisory opinions (a total of 164). If the CEA can’t even get this basic number right, can it be trusted on anything? ChatGPT would spit out a more convincing argument. 

Regardless, what a self own:  after claiming that the CFPB has “avoided transparency and accountability” and acting outside of the “formal rule making process,” the next sentence admits that the CFPB has acted through the formal rule making process hundreds of times. (And for what it’s worth, all the advisory opinions were  issued under a policy adopted by the Trump-Kraninger CFPB, including one issued by the Trump-Vought CFPB!)

Mortgage Costs

But let’s get to the meat of the CEA’s claim, namely that the CFPB has resulted in hundreds of billions of additional costs to consumers because of regulatory burden. The econometrics here are bizarre and contorted. The starting point is with mortgages. The CFPB has heavily regulated the mortgage market–and it did so at express Congressional direction. The ability-to-repay requirement is not the CFPB’s. It is Congress’s. The CFPB actually eased the burden with a regulatory safe harbor. That important point was lost on the CEA. Instead, the CFPB noticed that mortgages with a DTI>43%, which were not eligible for the ability-to-repay safe harbor, had a higher interest rate than those with the safe harbor. Duh. It’s a surprisingly small difference, though, only 16 basis points, which is 4.3% of the total interest rate, on average. (I’ll note with some satisfaction that back in 2013, I predicted the impact would be 18 basis points!) The CEA extrapolates this difference to a total cost pass through to consumers by multiplying the difference by lending volumes.

What’s wrong with this? First, it appears that the CEA used the relative sizing of 4.3%, rather than the absolute size difference of 16bps for its extrapolation. That’s going to have the effect of goosing the impact when interest rates rise (as they did). I cannot be sure that CEA used the relative, rather than the absolute size because they didn’t show their work, but they come back to that 4.3% figure later, which makes me think that it is the number they used. 

Second, and this is the big issue, the CEA assumes that all of the difference in pricing is due to the CFPB’s regulation. But one would always expect higher rates for higher DTI mortgages–they are riskier all else being equal, and there’s no reason to think that rates would increase in linear fashion. Additionally, CEA ignores that there is no secondary market for non-QM mortgages. In 2014, the FHFA directed Fannie and Freddie to purchase only QM loans or loans exempt from the ability-to-repay requirements. That’s what explains the much smaller volume of loans with DTIs>43%–lenders do not want to be stuck with an inventory of risky loans. But the CEA is either unaware or purposefully ignores the effect of the FHFA directive, which would, of course, make it impossible to throw all of the blame at the CFPB. 

Auto Loans and Credit Cards

The problems with the mortgage figure are the least of the issues with the CEA report, however. Having concluded that the CFPB is somehow responsible for $116-$183 billion in additional borrower mortgage costs, the CEA then extrapolates from that to increased auto and credit card lending costs. How? By comparing the number of complaints per dollar of loan volume across credit products. In other words, figure out the complaint rate for mortgages, for auto loans, and for credit cards, and assume that if there were 5x as many complaints about credit cards as for mortgages, that the regulatory cost would be 5x as large. This is one of the most asinine analytical moves I’ve ever seen. 

The CEA is assuming that consumer complaint volumes are a proxy for intensity of regulation. They aren’t. Complaints are a sign of consumer unhappiness, which if anything is a sign of insufficient regulation. Moreover, a consumer with a mortgage problem might litigate. One with a credit card problem has little recourse beyond a complaint–the dollars involved won’t support litigation.

The problem with using complaints as a proxy should have been facially obvious:  anyone working in consumer finance knows that mortgages are massively more regulated than auto loans, but the complaint rate for the products is basically the same. In other words, complaints are a terrible proxy for regulatory intensity. A one minute conversation with anyone in the field would have flagged this for the CEA. 

What About the Savings?

The CEA is fast to claim all sorts of dubious costs associated with the CFPB. But it never considers the costs saved by the CFPB. The 2008 financial crisis is estimated to have cost trillions. We haven’t had a repeat, in part because of the CFPB. Shouldn’t that be included in the calculus? Or is that ignored because it produces the wrong result? 

What we have here is a political hatchet job tricked out with a bunch of enough shoddy econometrics to make it look respectable. But let’s not pretend that this is in any way a piece of legitimate scholarly inquiry.