Payday holiday: do we need a vacation from payday lending?

In a new staff study from the Federal Reserve Bank of New York, Donald Morgan and Michael Strain take another look at whether payday loans are predatory or, in Morgan’s terms, welfare enhancing. This time around, Morgan takes a look at the “afermath” when Georgia and North Carolina banned payday lending.

(Dr. Morgan has commented on my previous posts. See the related links, below.)

If the availability of credit is inherently welfare enhancing (as Morgan seems to take as a given), the only real issue is at what point credit becomes so expensive (due to interest rates, fees, and other costs) that consumers can no longer afford credit. At that point, if they get the credit anyway, it is welfare reducing. According to Morgan, in other words, credit is welfare enhancing as long as a household can afford the credit.

To determine whether this is the case, Morgan looks at the purported effects of banning payday lending in Georgia and North Carolina. The Center for Responsible Lending, which is mentioned several times in Morgan’s paper, posted a critique of his research, and I have added my own non-scientific thoughts after the jump.

Federal Reserve staff paper

Based on Morgan’s paper, the answer to the question whether payday lending reduces household welfare must be that we can’t be sure, although there may be some evidence both ways. Morgan states “Georgians and North Carolinans do not seem better off since their states outlawed payday credit . . . ,” but his conclusion is not warranted by his research.

The Center for Responsible Lending pointed out some fairly large holes in Morgan’s data, but as I am no statistician, I will not go down that path. I will point out that even if, as Morgan draws from his data, Georgians and North Carolinans “have bounced more checks, complained more about lenders and debt collectors, and have filed for Chapter 7 (”no asset”) bankruptcy at a higher rate” (while Chapter 13 bankruptcies fell at an even higher rate), this does not mean that the absence of payday loans reduced household welfare.

Instead, it means that banning payday loans caused shifts in the credit markets in Georgia and North Carolina.

Morgan’s findings seem consistent with multiple hypotheses from both “sides” of the debate, including that payday loans are worse than the alternatives. For example, if payday loans are worse, we should expect an overall decrease in bankruptcy filings (which Morgan found). If consumers cannot take out high-cost payday loans, they should have more money to service other debt (which was not discharged in bankruptcy), which would account for the corresponding increase in complaints to the FTC about debt collectors. We also would not be surprised if other credit provisions, like bounced check protection, see greater use.

The paper does not do a cost-benefit analysis of the costs of payday loans before the ban versus the cost of alternative credit after the ban, which might be more illuminating of the central question.

It does not look to me like Morgan’s data is strong or clear enough to support his conclusion. It could support more than one conclusion, and so the report may not add any particular value to the policy discussion surrounding the payday lending industry.

Center for Responsible Lending critique

In several places, Morgan mentions CRL and its policy positions. The two seems to be at odds. And so CRL issued a short, 3-page critique of Morgan’s staff paper.

According to CRL, the increases in FTC complaints, bounced checks, and bankruptcies may have other causes. CRL argues those numbers were actually fairly constant or part of a trend that extended to before the payday bans.

For example, CRL points out that the check processing centers in Atlanta and Charlotte were regional and therefore handled checks from states with payday lending. The Charlotte CPC also took on Louisiana check processing after Hurrican Katrina. If the other states that passed through Atlanta and Charlotte had constant check bounce rates during the study time periods, Morgan’s data may still be useful. If not, it is not.

CRL also points out that of the complaints received by the FTC, also on a steady rise, over 40% are complaints about identity theft, which is completely unrelated to payday lending. Further, complaints about payday lenders would have been reported to state agencies, not the FTC, prior to the payday lending bans, and so the increase in FTC complaints after the bans could be partially accounted for by the shift from payday loans to other sources of credit that the FTC does oversee.

Conclusion

In sum, the Federal Reserve staff paper by Morgan and Strain seems to have problems. And while their conclusion may be correct, the data underlying their conclusion does not seem strong enough to support it.

So is payday lending welfare enhancing, or does payday lending reduce household welfare? The answer may lie in statistics, but not these statistics.

Is the availability of credit, so long as a household can afford the cost of that credit, welfare enhancing? Morgan seems to take it as a given. I believe that credit, as a tool, is a powerful tool for enhancing personal welfare. Used poorly, or used when a household cannot afford the cost, credit is welfare reducing for more than just a household. Witness the subprime mortgage fiasco.

Related posts:

• Another rebuttal to “Defining and Detecting Predatory Lending”
• Required reading for the Fed
• Is the payday lending industry welfare enhancing, or is it predatory?
• NY Federal Reserve Bank staff decide payday loans are NOT predatory