Viewpoint: Bank Interchange Fees Balance Dual Demand

Ever since Diners Club introduced the modern plastic credit card in 1950, consumers and merchants have shared the cost of this efficient and convenient payment method.

Consumers like cards because they don't have to carry around wads of cash or thick checkbooks and can defer payments for a few weeks. Retailers benefit because they sell more stuff and are able to meet a customer desire without having to offer their own finance programs.

Retailers are always angling for ways to shift most of the retailer costs to cardholders. An important 1984 antitrust decision - in NaBanco v. Visa - prevents merchants from getting a free ride. Unfortunately, the retailers have gotten new ammunition from articles by David Balto, a lawyer at the Federal Trade Commission, and Alan Frankel, a consulting economist who has worked against Visa. Mr. Balto and Mr. Frankel argue that the courts should overturn NaBanco. If that happened, cardholders would end up paying billions of dollars in higher fees.

The most popular card brands these days are from members of the bank associations. MasterCard and Visa use the "interchange fee" to make sure that their brands have the right balance between cardholder and merchant fees. Lowering the interchange fee shifts more of the costs to cardholders. Through intense competition, acquirers pass the costs on to merchants in the form of higher merchant fees, while issuers pass the revenues on to consumers in the form of lower cardholder fees.

Plastic cards are an example of a "two-sided market." For the market to even exist, you have to have two customers. An example is software like Adobe Acrobat, which needs people who want to read documents in Adobe as well people who want to write them in Adobe. No one would buy expensive Adobe software to create PDF files if no one could read them.

In the case of cards, there's no payment service unless both the customer and the merchant agree to buy it - the customers by presenting their card and the merchant by agreeing to take it.

Any entity selling to a two-sided market has to make sure that it has the prices right so that it attracts the right number of buyers on both sides. But it has another difficult pricing problem: Many two-sided markets have "network effects."

A product has a network effect if it becomes more valuable as more people use it. Merchants find plastic cards more valuable if more customers carry them; customers find plastic cards more valuable if more merchants take them.

When the Diners Club card was first introduced in New York to pay for restaurant meals, the restaurants picked up most of the tab. They agreed to pay 7% of the bill to Diners Club, which initially gave the cards away to a lot of well-off people it thought would use the cards. A few years later American Express entered this business and charged merchants a 5% fee. These proprietary systems recognized that they were providing a service to both merchants and cardholders, and they set their prices to ensure that they got enough merchants to make cardholders happy and enough cardholders to make merchants happy.

Few banks could start a national, much less a global, card network on their own. So when banks in the United States and Europe got into the plastic-card business in the late 1960s and early 1970s, they formed associations with other banks. The association was responsible for nurturing the brand and operating the big computer networks for processing card transactions. The bank was responsible for signing up cardholders and merchants. Some banks specialized in cardholders (issuers) while others specialized in merchants (acquirers). The associations wanted their members to compete with each other to get cardholders and merchants. The members decided what fees they would charge and which features they would offer.

But these bank brands needed a way to deal with the same two-sided network market that their unitary rivals dealt with naturally. They invented the interchange fee - it's the percentage of the transaction amount that the bank serving the merchant has to pay the bank serving the cardholder.

Merchants pay lower fees for transactions involving the associations' cards than for transactions involving the cards of other systems. In the early 1980s, a typical merchant paid 3.5% on the transaction when customers pulled out an American Express card but only 2.4% when they used a Visa card. Competition from the bank associations has resulted in that spread falling to 2.72% for American Express and 2.03% for Visa today. Cardholders have benefited from this competition as well.

So how could a feature of card brands that has helped balance the two sides of the market and harnessed network effects from the beginning come under attack? To get the interest of courts and competition agencies, adversaries of the bank associations claim that the banks are colluding.

In the United States, consumer-friendly courts have repeatedly rejected the cartel claim, because they've recognized that the associations need to make collective decisions to have a product. In 1979, National Bancard Corp. (NaBanco) filed an antitrust suit against Visa that claimed Visa was a cartel of card issuers that had fixed the prices it was charging acquirers. The trial court recognized, however, that Visa needed to have an interchange fee to balance cardholder and merchant demand and that it had to be set centrally. And that's not price-fixing. The 11th Circuit Court of Appeals agreed, and the Supreme Court refused to hear a further appeal from NaBanco. In affirming Visa's interchange fee, the courts recognized that it was necessary for Visa to be able to offer its product and, overall, that it was beneficial for consumers.

Yet Mr. Balto and Mr. Frankel use a deceptively simple argument to claim that it is time to overturn the NaBanco ruling.

When customers use one of their cards they impose a cost on the merchant - that is the merchant discount. It's hard for merchants to charge these costs back to the customer. Association rules discourage surcharges on card transactions. And there is some evidence that retailers wouldn't surcharge even if they could, though an increasing number attempt to steer customers away from a particular payment device - often with considerable success.

The result, Mr. Balto and Mr. Frankel argue, is that customers who use cash are subsidizing customers who use cards and that this results in plastic being used too much. The solution: Mandate a zero interchange fee so that merchants won't have to pass the cost along.

This criticism of the interchange fee sacrifices economic realities at the altar of classroom theory.

Mr. Balto and Mr. Frankel can't show that their solution makes consumers better off. A zero interchange fee would shift $14 billion of costs a year from merchants to cardholders in the United States alone.

Cardholders would pay higher prices but would not be able to use their cards at many more merchants, which would pay lower prices but have much fewer customers wanting to use their cards. Recent theoretical models of interchange fees that take explicit account of how these fees balance cardholder and merchant demand demonstrate, in fact, that there is no economic reason to believe that cardholders and merchants overall would gain from a zero interchange fee.

Mr. Balto and Mr. Frankel focus on a trivial market imperfection while ignoring others.

Retailers pass along all sorts of costs that don't benefit all customers to the same extent. All customers end up paying higher prices as a result of retailers offering parking, tailoring, escalators, convenient store hours, gift-wrapping, and many other amenities that are used by only some customers. Retailers don't try to charge for each of these services. So, just as with cards, some consumers pay higher retail prices without getting the benefit of these freebies.

Another real-world complexity these advocates of zero interchange fees have missed is that the government subsidizes cash and checks. So it is by no means clear that the "subsidy" to card users identified by Mr. Balto and Mr. Frankel even results in the overuse of cards: The card subsidy might just offset the effect of the cash and check subsidies, resulting in the more optimal use of all payment devices.

A mandated zero interchange fee would also put the associations at a serious competitive disadvantage with American Express and Discover, which would still be able to charge merchants as much as they wanted.

The private systems would also be able to cut deals with individual banks - "join us, and we'll give you a piece of the merchant fees that you've lost from the bank associations." In fact, American Express is already making just these sorts of deals in Europe. (The associations won't allow American Express to sign up members in this country; the Justice Department antitrust suit against MasterCard and Visa challenges that prohibition.) The loss of interchange fee revenues might also persuade banks to pull out of the associations and start their own private system. In the end, consumers could end up with far less competition among issuers and acquirers.

The ruling in the NaBanco case was one of several major decisions that have helped put U.S. antitrust laws firmly on the side of consumers in the past 25 years. The courts and competition authorities should heed its lessons and not cater to retailers and other special interests looking for things their customers never get: something for nothing.

Mr. Evans is a senior vice president at National Economic Research Associates and co-author of "Paying with Plastic." He has consulted for Visa U.S.A.