Why Now? The Faulty Economics of Credit Card Reform

Richard Epstein —  8 December 2009

Richard A. Epstein is the James Parker Hall Distinguished Service Professor of Law at the University of Chicago, the Peter and Kirstin Bedford Senior Fellow at the Hoover Institution, and a visiting professor at New York University Law School.

About four years ago, I worked for Visa in opposing the opposed limitations on interchange fees that the Australian government was about to impose on the credit card industry. The situation there, like the situation in the United States, seemed hardly propitious for reform.  The use of credit cards was rapidly expanding, and the rate of interest was being brought down by competition, the number of cards in circulation had increased.  What is there not to like?

The fear of monopoly apparently.  Everyone knows that the credit card industry is highly concentrated.  That point in and of itself is not necessarily a bad thing.  The credit cards only work in what is called two-sided markets.  Consumers are prepared to sign up for credit cards only if they are aware that merchants will be prepared to accept them.  Merchants will be prepared to accept these cards only if consumers are prepared to use them.  The huge number of players on each side of the market cry out for the use of an intermediary to forge the connections.  The fewer the intermediate parties, the easier it is to organize the grid.  That task does not come without cost, and that cost in turn is incurred by the credit card companies that invest in the large infrastructure that keeps the entire system humming.

Breaking up these companies has real efficiency losses, so naturally the thought of the aspiring regulator is to turn to price controls as the next best solution. Someone of course has to pay companies to cover that expense, and to make a profit as well, but who should bear that burden? The great siren of all service users is to insist that they be charged at marginal cost.  After all, once costs are raised above marginal cost, some consumer who could have used the service will be excluded, which counts, after a fashion, as a type of inefficiency.   Needless to say, marginal cost pricing is a game that all users of inputs would like to play.  But it is not a game at which all can succeed.  Let the merchants on the one side, and the cardholders on the other, pay marginal cost, and no one is left to pick up the fixed costs of running the entire system.

Those costs amount to a tidy sum that someone has to pick up if the network is to remain viable. But which side and in which proportions?  The usual response is to see which side is more inelastic in the demand, on the simple (but nasty) ground that it has fewer options to escape the costs in question.  One nice consequence of this strategy is that the network will tend to shrink less on one side, which offers an added inducement to remain on the other side.

But, comes the response, just how much over marginal cost should the credit card companies be able to charge in a two-sided market.  Coming up with a precise answer to that question offers enormous descriptive and normative challenges.  And so the question is whether this game is worth the candle.  The Australian experience is not all that reassuring, as the shifts away from merchants imposed higher costs of card users which tended to thin their ranks, without any clear evidence that the merchants did their part by passing on the reduced interchange fees to their customers.

It is therefore no surprise that the title of the GAO’s report is not exactly a clarion call to action:  “Rising Interchange Fees Have Increased Costs for Merchants, But Options For Reducing Fees Pose Problems.”  My believe is that the title was chosen by the editors of the Onion. The first part of this title is of course a necessary truth, but it tells us nothing as to whether the increased fees help or hurt the  overall operation of the credit card system.  The words after the “but” remind us that there no evidence that two simple risks of price controls won’t go away.  The first of these is that the cuts are too deep so that the system will experience a financial shock from which it cannot recover. Lawyers might think that this could even raise the question of whether the rate reductions are confiscatory.  The second risk is a bit less ominous.  The increases of the costs to customers reduce the number of potential customers while making the service less attractive to those who remain.

Needless to say the GAO report gives no assurance that the interchange fee cure is not worse than the disease.  So why wade into such muddy waters?   With the real estate mortgage market under siege, this is hardly the time to open a second front in the credit card wars.  The administrative costs and political risks of initiative are not worth the candle.

5 responses to Why Now? The Faulty Economics of Credit Card Reform

    Richard Epstein 8 December 2009 at 6:23 pm

    Omri’s recent comment about my initial remarks shows how difficult it is to deal with you own brand name. My basic position of course has long been opposed to overall regulation of the terms and conditions on which goods and services are traded. But, as I have become older and more fuzzy minded, I have become more willing to entertain various forms of regulation that counter a real risk of monopoly misbehavior, of which horizontal cartels are the main culprit. Yet at the same time, the various arrangements with network industries also hold out some risk of monopoly profit, but not enough to make me, as Omri suggests, “a secret supporter” of government regulation.

    Why? The explanation is clear enough. The gains from regulation are harder to identify, as there is a real risk that the effort to stop these price advantages could backfire. Second, the costs of regulation are far higher, as it is more difficult to isolate the abuses to be controlled. That all translates into placing a heavy burden, but not an absolute bar, against the government intervention.

    These gerneral remarks shape my answers to Omri’s particular question.

    First, I do believe that some of these price controls could cut “too deep”. If so, we could expect serious crimps in service. But if not too deep, not that much is gained. So I would weight the first error larger than the second, which again counsels caution.

    Second, I do believe that entry can help, but it is not clear how effective it can be in a two-sided market. But again even a small help tends to tip the balance away from intervention.

    Third, Omri is right that correcting small ills produces small gains, if we know what to do. But the capital available for these interventions is limited, and if theory suggests that overall gains are weak, why invest wealth here when it could be used to great effect in other areas, like labor law, where there is much that could be done to break down labor monopolies.

    So here, I think that the presumption against intervention sticks, at least if the GAO report is the best one can offer for further government involvement. I would not describe my opposition as “faint.” It is principled but not absolute. I suspect that the more specific the proposal the more I will return to my libertarian roots.


    Omri raises some good points. In Australia (which we will all be hearing about repeatedly, I am sure), there were cuts to interchange rates that, according to Joshua Gans and Bob Stillman, did not seem to affect usage very much, but reduced bank profits – perhaps suggesting that the rents were not being competed away. Before the Australian reforms, some of the parties involved expressed concerns about possible, potentially catastrophic, shocks to the system. None such appear to have occurred.

    In fact, the Reserve Bank of Australia has not only continued to endorse its decision over the years, but has extended its actions to include other payment methods that use interchange fees. One can certainly debate whether the Reserve Bank of Australia’s actions benefited consumers overall, but they also demonstrate that intervention is not necessarily a recipe for catastrophe.

    With respect to the tax analogy, the issues about moving sales back and forth across city lines can also be seen with payment methods. Changing the relative costs of payment methods can influence the usage of those payment methods. But I believe that’s a topic for tomorrow morning.

    Omri Ben-Shahar 8 December 2009 at 1:21 pm

    Richard, Coming from you, your post almost sounds like resounding support for some regulation. I don’t have to read between the lines to notice that this is NOT the strongest opposition to regulation that you can author.

    One risk you refer to is from price controls that are “too deep.” Well, surely there are some cuts that will not be too deep and should not lead to a unrecoverable shock to the system.

    Second, there was an element not featured in your analysis, which — when present — usually lends great force to the argument ‘leave the market alone.’ This is the claim that if super-competitive profits exist, they will be competed away, without the need for price controls. The two-sided network feature makes alternative systems more difficult to introduce. with less entry, the rents may not be competed away.

    Third, the point that Allan Shampine commented on — the plea that there are more burning matters to resolve — is not persuasive on the grounds that Allan mentions. Besides, the presence of an even bigger problem does not mean that small ones ought to go untreated.
    So am I correct to conclude, from you fairly faint opposition, that you are actually a secret supporter of some regulation here?


    I do love the Onion, but the stakes are certainly not a laughing matter. The GAO is clear on why one may wish to wade into these waters. Simply, there is a lot of money at stake. As the GAO points out, literally trillions of dollars of transactions are processed each year on credit and debit cards in the United States. 1.7% of each credit or debit card payment (to use the GAO’s example) may not seem like a huge amount on an individual transaction, but 1.7% of $2.5 trillion is real money, even by Washington budget standards. When one thinks about the situation globally, the amount of money at issue is even more substantial. The Reserve Bank of Australia, whose actions are being discussed elsewhere in this forum, was also concerned about the size of the credit and debit card industry and the potential effects of the industry’s policies and fees on Australia’s economy.

    At a local level for Prof. Epstein and I, there has been an immense amount of press on the Cook County Board’s efforts to roll back a sales tax increase by half a percentage point. One of the issues debated there was the degree to which shoppers were going to the suburbs to avoid the sales tax. The point of this anecdote is that transaction costs of this level matter, and people care a lot about them.
    Of course, there is a lot of debate about whether there is a problem with those fees, or, if there is a problem, whether the cure is worse than the disease, but I believe that the size of the fees is more than sufficient to justify having such a debate.

Trackbacks and Pingbacks:

  1. Richard Epstein - December 8, 2009

    See my response to Allan here.