August 25, 2010
Larry Garvin on Hawkins' Regulating on the Fringe
Posted by Christine Hurt

In Regulating on the Fringe: Reexamining the Link Between Fringe Banking and Financial Distress, Jim Hawkins takes on the conventional wisdom about payday loans, pawn loans, and rent-to-own leases. These commonly are thought to help push financially shaky clients into genuine financial distress. Professor Hawkins first looks at the fairly strong evidence that the promiscuous granting and use of credit cards can lead to distress. He identifies three aspects of credit card use that can bring this about. First, credit cards are painless to use, at least when compared with cash. The relative obscurity of mounting debt increases the user’s propensity to spend. Second, credit cards allow consumer to amass large amounts of debt, essentially mortgaging future income for present desires. Because credit cards are issued and limits sent largely on the basis of credit scores, drops in income will not generally affect credit limits. As a result, consumers lack most external constraints. In addition, because credit card issuers will grant credit lines for up to a fifth of annual income on a single card, it is easy to be awash in debt very quickly. Third, high credit limits provide a faulty heuristic about the amount of debt one can prudently incur. Consumers look at the credit limit as a gauge of the potential to repay in due course, even though the two are only loosely related.

Professor Hawkins then argues that the standard vehicles of fringe banking do not display these characteristics, certainly not to the degree shown by credit cards. Rent-to-own transactions (the subject of an excellent article by Professor Hawkins published in 2007) don’t pile up debt, because the lessee can walk away from the transaction at any time without penalty, and in any event they don’t provide the customer with a freely usable line of credit. Pawnbroking and auto title lending don’t pile up debt freely, because they are limited to a percentage of the value of the collateral. Secured credit cards (those where the customer deposits with the issuing bank a sum equal to the size of the credit line) similarly are limited to the amount of collateral provided, and in some respects aren’t really credit at all. (It occurs to me that secured credit cards resemble greatly evergreen retainers, perhaps with some of the same questions about trust and reliability.) Payday loans are a closer question, because there is no escape hatch if the borrower wishes to walk away from the transaction and because they are genuine credit transactions. But in many jurisdictions these are capped, and the average payday loan is only about $300. Moreover, payday lending is imperfectly linked to bankruptcy, given the fairly modest amount of added interest a typical borrower would pay.

This is not to say that these fringe banking areas need no further regulation, but rather that basing regulation on the assumption that fringe banking causes financial distress may yield bad regulation. There may be paternalistic reasons to regulate these financial products, and mandatory disclosure rules or cooling-off rules may thus be supported. Bans or impracticably low price caps may not. Finally, the tentative evidence for secondary effects of fringe banking, such as the financial consequences of pawning the tools of one’s trade, might justify appropriately tailored regulation.

Professor Hawkins writes clearly and makes a fairly persuasive case that the financial consequences of many fringe banking products are overstated. His analysis of secured lending is, with one caveat to come, particularly strong. But I remain uneasy, particularly about his analysis of payday lending, for several reasons. First, I think he too readily dismisses the effects of apparently modest added interest payments. For example: "It is highly unlikely that a $45 payment every other week [interest payments on a rolled-over payday loan for $300] will substantially exacerbate significant financial distress, although, admittedly, it will cost the borrower money." (p. 39) But what is the baseline? One recent study of payday loans found that payday customers who also had credit cards – presumably a relatively creditworthy subset of the payday customer population – had mean annual incomes of about $20,000 per year, or approximately $400 per week. $22.50 is not a trivial part of $400, particularly given the relatively low purchasing power commonly held by the poor. Second, and related, is the problem of studying the fringe banking products in isolation. The empirical studies I have seen suggest that payday borrowers usually have credit cards and frequently use other fringe products. It also seems empirically true, as work by Tobacman and colleagues has shown, that those using payday loans often are not using their credit cards because they are at or near the credit limit and are not current on payments, and that taking out a payday loan signals a marked increase in the likelihood of default on a credit card. To the extent that regulating just one product will simply mean that another product is used more heavily, the relation of these products might suggest a need to regulate across the board.

Third, Professor Hawkins downplays the significance of default in auto title lending. Doing so based on the frequency of default is plausible, but the data are incomplete. He notes that the default rate is eight to ten percent. But how frequently are these loans turned over? Professor Hawkins tells us that the usual length of a title loan is thirty days. In theory, then, the loan could be turned over as many as twelve times a year if allowed to go to maturity. If a typical title loan customer borrows even, say, three times in a year, that suggests an annual default rate per customer, rather than per loan, materially higher than eight to ten percent. Without data on the per-customer default rate, I don’t think one can minimize the consequences of title lending. Nor can it be done simply by looking at the typical loss when those loans go into default – according to the studies cited to by Professor Hawkins, around $700 in lost equity for the typical repossession. The absolute amount is small, but what about the amount as a percentage of annual income or of net assets? The article doesn’t state the mean income or asset portfolio of title loan borrowers. If they resemble payday customers, though, $700 is just under two weeks of income – one paycheck in the normal biweekly cycle. If you live paycheck to paycheck, losing the equivalent of almost one full check would seem disastrous, and very likely would lead to increased use of other fringe loans.

Fourth, Professor Hawkins might wish to incorporate some findings from the debiasing literature. A leading defense of heightened disclosure, a regulation Professor Hawkins mentions briefly in his conclusion, is that it produces its salutary effects with minimal cost and no real harm. But is this true? Professor Lauren Willis, for instance, has written about disclosure in real property transactions, pointing to the substantial literature that finds disclosure at best ineffective. Indeed, too much disclosure runs the risk of swamping important facts in a mass of irrelevant or minimally relevant information – the dark side of the availability heuristic, perhaps, or an instance of cognitive overload. But if there are ways to present the most salient facts vividly, debiasing may indeed work. There have been some very recent studies posted on SSRN that look at exactly this in the context of payday loans. (While I’m discussing the behavioral literature, I wonder whether rolling over loans of the types mentioned here is easy enough that it feeds into the biases Professor Hawkins invoked in his discussion of credit cards.)

Fifth, and blessedly last, I think Professor Hawkins might want to incorporate some of the recent empirical work on the effects of stringent payday regulation on such things as bankruptcy rates, use of other expensive forms of credit, defaults on credit, calls upon public assistance, and so forth. The modest portion I’ve read suggests at least on the surface that axing payday loans decreases the solvency of their erstwhile customers, though the studies do not all go in one direction and do not always examine representative populations or control for confounding effects. It does not necessarily follow that if payday lending costs a lot, then its abolition will yield a net benefit; the alternatives may be worse, and more to the point there may be none. The secondary effects of no access to needed credit have to be taken into account if one is looking at the secondary effects of access to needed, but costly credit. On a somewhat related point, do payday lenders have unusually high profit margins, or do their high interest rates merely reflect high default rates and high operating costs?

I suppose in the end I am persuaded that fringe banking products contribute materially less than credit cards to the financial distress of their users. This is an important point, and Professor Hawkins should be applauded for taking on the conventional wisdom so lucidly and fairly. I remain unpersuaded, however, that a materially smaller contribution is not worth serious attention. The marginal effects of high interest costs, as with all costs, mean more to those with low incomes and modest assets. Moreover, the close linkage of these fringe banking products makes it difficult for me to agree that high costs for any one of them will not have a material adverse effect on financial stability, whether directly or indirectly. But even here Professor Hawkins has raised important questions about the too-casual assumptions underlying much recent policymaking in this field. How sound empirical work will answer these questions, however, remains very much to be seen. In the meantime, Professor Hawkins’s call for greater attention to the secondary effects of fringe banking products seems appropriate and indeed urgent.

Junior Scholars | Bookmark

TrackBacks (0)

TrackBack URL for this entry:

Links to weblogs that reference Larry Garvin on Hawkins' Regulating on the Fringe:

Recent Comments
Popular Threads
Search The Glom
The Glom on Twitter
Archives by Topic
Archives by Date
January 2019
Sun Mon Tue Wed Thu Fri Sat
    1 2 3 4 5
6 7 8 9 10 11 12
13 14 15 16 17 18 19
20 21 22 23 24 25 26
27 28 29 30 31    
Miscellaneous Links