In 2021 Illinois passed its Predatory Loan Prevention Act (PLPA), which imposes a 36% military APR (MAPR) cap on all loans made by non-bank or credit union or insurance company lenders. Not surprisingly, the law has not been popular with higher cost lenders who either have to change their offerings, cease doing business in Illinois, or figure out some way to team up with a bank that won't run afoul of the law's anti-evasion provision.
Recently, opponents of the PLPA have been making some noise, pointing to a study by a trio of economists—J. Brandon Bollen, Gregory Elliehausen, and Thomas Miller—about the impact of the PLPA. (The latter two are familiar scholars whose work consistently takes a dour view of consumer finance regulations: readers might recall my debunking of another recent study by Professor Miller, co-authored with Todd Zywicki, that was fundamentally flawed because of the miscalculation of loan caps in various states.)
Using credit bureau data, the Bollen et al. paper finds that the PLPA resulted in a 30% decrease in the number of unsecured installment loans to Illinois subprime borrowers and a 37% increase in the average installment loan size to Illinois subprime borrowers, which they attribute to the difficulty in making smaller loans profitable at 36% MAPR. Additionally, based on a lender-administered survey of 699 online borrowers (not necessarily of installment loans), the Bolen paper also reports a decline in borrower financial well-being following passage of the PLPA.
Unfortunately, the Bollen paper suffers from serious data and methodological problems such that it does not tell us anything meaningful about the wisdom of the PLPA. Here's why.
(1) The paper's findings are only about credit availability are about subprime installment loans, not small-dollar credit generally.
The credit bureau data relied upon by the paper provides an incomplete view of subprime credit. Many small dollar lenders—including most payday lenders—do not credit report. Accordingly, the Bolen paper’s findings cannot be generalized beyond (at most) unsecured installment loans. The Bolen paper tells us nothing about availability of subprime credit in Illinois more broadly. So when the paper claims (emphasis added) that its contribution is to find that the PLPA "significantly decreased the availability of small-dollar credit," that's simply not true.
An accurate summary of the paper's finding is that the PLPA "significantly decreased the availability of subprime installment loans." The paper simply cannot support a broader finding, and in fairness to its authors, it does elsewhere note the limitations of its data ("Our focus is on how the imposition of the 36 percent interest-rate cap in Illinois at the end of the first quarter of 2021 affected the use of unsecured installment loans. Consumer usage of other small-dollar nonbank loan products could be affected by Illinois’ interest-rate cap.") The paper's findings cannot be generalized to tell us about what happened to the availability of subprime credit in Illinois following the PLPA.
(2) The paper does not indicate anything about the number of borrowers getting subprime installment credit, so it is impossible to tell if fewer borrowers were able to get loans or if there was merely a shift in the number and size of loans to the same base of borrowers.
It's not actually so clear that the PLPA had much effect on the availability of subprime installment credit in Illinois. If you interact the decrease in the number of loans (-30%) and an increase in the loan size for the loans that are made (+37%), you'll realize that implication (not mentioned in the paper) is that there was only 4% reduction in the total dollar volume of subprime installment credit in Illinois following the PLPA.
Now, the Bolen paper only reports the number of loans, not the number of borrowers. That means it's impossible to tell from the paper what actually happened in the installment loan market. One possibility is that 30% of subprime borrowers were cut out of the installment loan market. But another is that there was basically no change in the number of borrowers, only a change in the number and size of loans they got: in some cases instead of getting two loans for $1000, a borrower got one loan for $2000 or the like (much as happened in Colorado after its payday loan reform). Or the answer could be somewhere in the middle. The point is that the paper doesn't really tell us what happened even in the subprime installment loan market.
(3) Credit availability is an incomplete measure of borrower welfare.
The availability of credit, standing alone, is an incomplete measure of borrower welfare (which is what really matters). Notably missing from the Bolen paper is anything about the terms of the loan other than their size. One thing we know for sure, though, is that the Illinois law eliminated non-bank loans at above 36% MAPR. A reduction in unaffordable, high-cost credit—the credit impacted by the PLPA—is likely to increase borrower welfare by the cost of the credit that is available and ultimately by reducing debt. The Bolen paper makes no attempt to factor in the welfare benefits of lower debt.
(4) A self-reported borrower survey, administered by lenders is not reliable evidence of borrower welfare.
Finally, the paper relies on a self-reported survey of online borrowers, administered by self-interested lenders (!), to suggest that the PLPA decreased consumer welfare. This sort of survey is a patently unreliable method for determining the PLPA’s effect on consumer welfare. There is no reason to believe the survey responses are representative, and I'd be really wary of relying on a survey that was administered by self-interested parties. The survey was only of consumers who borrow online, a population with particularly poor credit, and the response rate was less than 3% raising serious concerns about selection bias. Moreover, the borrowers covered by the survey may not be the same as in the credit bureau data: online loans might not be installment loans. Nor does the analysis of the survey attempt to control for macroeconomic conditions, such as the pandemic or inflation. The survey isn't a reliable measure of consumer welfare, nor can it provide casual evidence about the welfare impact of the PLPA.
So What Does the Bollen Paper Tell Us?
Put all of this together and what do we have? A paper that shows that after the PLPA the number of subprime installment loans in Illinois fell, while the size of those loans increased. And that's about it. But we cannot really tell what that means in terms of credit availability without knowing the number of unique borrowers: were there borrowers who couldn't get loans because of the PLPA and, if so, how many?
The real question about the PLPA is the impact on borrower welfare, but the survey evidence introduced is so unreliable for so many reasons that it should not be taken seriously as a basis for evaluating the PLPA. (I'd be surprised if a serious, peer-reviewed, economics journal would publish the survey evidence part of the paper...) We need reliable studies about the impact of the PLPA—it's a critical policy question—but the Bollen paper isn't one. The Bollen paper simply does not deliver the punch its anti-regulatory promoters think it does.
Nothing contributes so much to the prosperity and happiness of a country as high profits