Moving the Ball Forward: Macroeconomic Considerations

Ronald Mann —  8 December 2009

Ronald J. Mann is Professor of Law at Columbia University Law School.

What is most surprising about the GAO report is how little the analytical discussion of this subject has advanced in the last decade.  We all know that interchange rates might contribute to higher retail prices: customers that use cheaper payment products can be said to “subsidize” customers that use credit cards.  Starting with the Australian initiative, regulators for a decade have sought to understand the costs of credit-card processing and to push interchange fees down to the “competitive” level of the “costs” of providing the service.  Conversely, economists writing about two-sided networks have shown that the focus on costs reflects a fundamental misunderstanding of the two-sided network, in which the interchange fee is an allocation of burden between merchants and cardholders, not a charge for services rendered to one side or the other.

Yet despite the persistent attention to the issue and the steadily increasing lobbying and public relations campaigns on both sides of the question, the underlying issues interchange regulation raises receive little attention.  Given the important effects of credit cards on the national economy, regulatory intervention should rest on, or at least consider, the macroeconomic effects.  Consider the following two points.

First, the most important effect of interchange fees on credit card products is to fund rewards.  Networks increase their value if they take funds from merchants and give them to cardholders to encourage cardholders to use their cards, up to the point at which the marginal loss from declining merchant acceptance exceeds the marginal gain from increased rewards.  Since there is no obvious market trend toward declining merchant acceptance, we presumably have not yet reached the point at which the networks have an incentive to stop raising interchange fees to fund rewards.  Thus, a forced decline in interchange fees should shift product lines away from rewards cards.  If the rewards card is associated with increased spending (as seems likely), then the shift should bring with it a decline in consumer spending, and probably also a decline in consumer borrowing.  We can debate whether that decline would be good or bad, but if we are unsure we want a decline in consumer economic activity, we should think carefully before cutting interchange fees.

Second, consider the relation between interchange fees and market structure.  The market for credit card lending is concentrating rapidly; the top ten credit lenders now hold 88% of outstanding debt.  The remaining credit card issuers (hundreds of smaller banks and credit unions) typically market credit cards used as transaction vehicles.  Because those cards generate little in interest revenue, interchange fee revenues are much more important to their profitability than they are to the lending-driven products of the large issuers.  Hence, a decline in interchange fees shifts the balance of profitability from the smaller issuers to the larger issuers.  This, in turn, should foster increased concentration in the market.  Since the market for credit card issuance (as opposed to the market for credit card networks) traditionally has been quite competitive, market interventions that accelerate the concentration of that industry are noteworthy.

I offer these ideas not because they are the only, or even the most important, possible effects.  I proffer them to encourage a discussion of interchange fees that considers their role as part of a complex and important economic system rather than a simple contract term between merchants and their banks.

4 responses to Moving the Ball Forward: Macroeconomic Considerations


    As my later post suggests, I agree with Mark that the honor-all-cards rule is central to the current problem. The place where I part ways with Mark is the suggestion that credit cards are a commoditized product that is functionally identical from one issuer to the next. To be sure, the nature of the networks makes the service the merchant receives more or less a commodity, but even with the market concentrating the product offerings of issuers still differ substantially. Capitol One dominates subprime markets. Other issuers like Chase have tried without success to compete in that market. The successful affinity products come from a very small group of people with the know-how to make that product succeed. For me, the core competency here is information technology, which requires a massive capital investment that only a few of the largest issuers can execute with reliable success.


    Thanks for taking time to reply. I think the big difference between in-house credit and bank cards is that the merchant’s name is on the in-house arrangement. A merchant cross-subsidizing his credit customers from his cash customers is nevertheless funding his own marketing; the “rewards” he hands out induce credit customers to shop in his store. Heck, when the merchant buys radio ads his deaf customers cross-subsidize his hearing ones.

    Interchange fees are not a substitute for merchants’ internal cross-subsidy of credit customers because they fund bank marketing, not merchant marketing (I pass over “card partnership” arrangements between big merchants and banks; I never proposed to interfere with those anyway).

    Obviously for many merchants the cost of dealing with bank cards is lower than the cost of operating an in-house credit scheme. It does not automatically follow that banks should be excused from competing rather than colluding.

    I didn’t suggest regulating interchange fees, but merely preventing banks from colluding to keep them high. If card issuers competed for merchants’ business (that is, if merchants were permitted, whether or not they always chose, to discipline excessive interchange fees on certain cards by refusing or surcharging those cards while still accepting others on the same networks) then the market would set the level of interchange (and other) fees more effectively.

    I can’t say exactly what arrangements would fall out, but I don’t share the views of Prof. Zywicki and other writers that banks revenues (and profits) are fixed by cosmic forces so that the mix of fees they charge can change but not the total amount. (Cosmic forces no, political forces, very likely!)


    Mark: There is a grain of truth in what you say, but I’d argue with your characterization. It is the case, of course, that there are cross-subsidies in these markets. But these are not in themselves a problem, and merchants have not acted like they were a problem in the past. As a bit of powerful evidence, when retailers run their own credit programs, at significant cost to themselves, they don’t charge higher prices to customers purchasing with credit, even though it would be perfectly permissible for them to do so. Rather, they recognize the value of the cross-subsidy and the benefit to credit customers of access to credit and electronic payments, and they finance their credit systems (including, by the way, the discounts they offer to induce customers to sign up–just like offering rewards, and exactly the opposite of what you suggest) with revenue earned from both cash and credit customers. As Ron suggests, we can have a discussion over the macro question whether it is good to lower the price of consumer credit in this fashion, but that’s a policy question that has nothing to do with how the price of credit is being subsidized. On net, retailers’ willingness to incur the costs of maintaining in-house credit systems with no direct financing with surcharges to credit customers suggests these types of cross-subsidies are beneficial, even for merchants. (By the way–I’m sure for in-house credit as for bank credit, interest charges fund the lion’s share of the system, but then you are just subsidizing convenience cardholders with charges levied on revolvers–which doesn’t strike me as any more “honest.”)


    Prof. Mann writes: “First, the most important effect of interchange fees on credit card products is to fund rewards.”

    Thank you! That’s right, and that is the key to understanding the whole mess. “Rewards” are the means by which issuing banks differentiate themselves to compete for cardholder customers. Without “rewards,” they would have to compete on price and service[1] in what has become a commodity market (now that the credit-card infrastructure is fairly mature, though still capable of improvement). All sellers, including banks, hate participating in commodity markets, because competition tends to drive prices down toward costs, until a commodity seller cannot make either a comparatively or absolutely large return. Sellers always strive to differentiate their products, even if only by “brand identification” or other emotion-tugging qualities, to escape commodity pricing competition.

    However, interchange fees and the merchant fees which reflect them enable card issuers to differentiate their essentially-identical products at the expense of the merchants who sell possibly-non-commodity goods to cardholders. In concrete terms, the banks collude (through the networks) to make merchants pay for banks’ artificial differentiation. No wonder this is galling to merchants– the networks’ “honor all cards” rules force merchants to fund bank “rewards” instead of funding the merchants’ own marketing schemes– even though bank marketing does nothing for merchants or (in aggregate) cardholders! (“Rewards” schemes do move surplus from one cardholder to another.)

    I suggest the simplest reform likely to let the market find a natural balance would be to extirpate “honor all cards” rules and similar schemes which banks use to keep cardholders from realizing that they’re just buying their own “rewards.” All of those schemes are collusive (the networks coordinate the banks’ collusion in return for a rakeoff; a complex structure which has helped them evade antitrust enforcement). Let merchants accept cards but optionally refuse or surcharge cards with high interchange (merchant) fees. That will make those fees salient to cardholders. Merchants and cardholders (in aggregate) will be better off. Banks will merely have to compete honestly– which may trim their profits but who cares? There’s no reason we should give banks (another) special privilege (right to collude) in the marketplace.

    [1] Banks also compete on which can best trick cardholders into paying excessive fees and interest under incomprehensible and more-or-less crooked contract terms.