Subcommittee on Financial Institutions

Hearing on Competition and Innovation in the Credit Card Industry
at the Consumer and Network Level

10:00 a.m., Thursday, May 25, 2000 - Dirksen 538

Prepared Testimony of Mr. David Evans
Vice President
National Economic Research Association

Mr. Chairman, Members of the Committee and Subcommittee, my name is David S. Evans, and I am a Senior Vice President of National Economic Research Associates, Inc. I am the author (with Richard Schmalensee) of Paying with Plastic: The Digital Revolution in Paying and Borrowing (MIT Press, 1999), an economic treatise on the payment card industry. I have been asked by Visa U.S.A. Inc. ("Visa") and MasterCard International to address the subject of this hearing: "competition and innovation in the credit card industry." Thank you for the opportunity to appear before you today.


The American credit-card industry works extremely well from the standpoint of the consumer.(1) It is intensely competitive. Thousands of credit-card issuers compete for consumers by offering attractive prices, features, and service. And the credit-card industry has performed well. Prices have fallen and output has expanded. Cards have more features; more merchants take them; and technological advances have reduced the time it takes to complete a credit-card transaction. Credit cards have made it much easier for Americans to pay for goods and services and to borrow money. Most people and many small businesses have benefited from the ability to borrow conveniently on their credit cards. Of course, some people do not use their cards responsibly and some credit-card issuers have taken advantage of consumers. But those exceptions arise in every competitive industry.

The Visa and MasterCard associations have played important roles in promoting and ensuring competition in the credit-card industry. Almost every financial institution can join Visa and MasterCard and issue credit cards. Many new institutions join every year. Members compete with each other in every dimension that is important to consumers. But they also work together through the associations to achieve things that no member could on its own: nurturing the globally recognized brands; maintaining and improving the sophisticated computer networks for processing transactions; developing new payment methods; and expanding the areas in which consumers can use their cards. Members pay volume-related fees to cover the costs of doing these things. Visa and MasterCard do not earn profits or pay out dividends to members. Visa and MasterCard are open, not-for profit joint ventures. As brands, they compete with American Express and Discover/Novus, both of which are operated as parts of closed, for-profit, investor-owned corporations.

Although economists working at their blackboards could certainly argue about whether the credit-card industry is organized in the best of all possible manners, the credit-card industry works extraordinarily well for consumers. It is hardly an obvious candidate for radical industry restructuring and regulation--but that is exactly what the Justice Department has proposed in its lawsuit against Visa and MasterCard. The Justice Department wants to fundamentally alter ownership and control rights in both associations to create two classes of members: "governing" members who would be able to vote for board members and "issuing" members who would be essentially disenfranchised. Governing members would have to issue mainly the card of the association they govern, while issuing members could issue any card they want including American Express and Discover Cards. It appears that the Justice Department would leave many of the details of its restructuring plan to court-administered regulation.

There are many reasons to question why the Antitrust Division's regulators believe they can custom-tailor a better version of the payment-card industry. But the real problem is that the Department of Justice lawsuit ignores the economic purpose of antitrust laws in this country. Antitrust is not an agent of industry policy available to the Justice Department to fine-tune our country's competitive engine. Antitrust is designed to root out anticompetitive behavior that results in significantly higher prices and lower output for consumers. The Justice Department, notably, has not argued that Visa and MasterCard have engaged in any actions that have resulted in higher prices for consumers or that its redesigned industry would result in lower prices for consumers. Instead, the Justice Department claims that more innovation will ensue from a restructured and regulated credit-card industry. So much of our economic experience during the 20th century teaches us that this is a dubious proposition indeed.

American Express argues that consumers will benefit from bank-issued American Express cards if the Justice Department persuades the Court to adopt its plan. By the same logic, consumers might also be better off if we forced McDonald's to let its franchisees sell Whoppers on the side or if we forced American Express to make its card system available to all financial institutions. But we wouldn't think of invoking the antitrust laws to do this--even if property laws allowed us to--because we recognize that in the long run consumers are better off when businesses are able to decide these things themselves. Short of being an essential facility, to my knowledge the antitrust laws have never been invoked to force a competitor to do business with another competitor.

In seeking to enter into contracts with selected members of Visa and MasterCard, American Express apparently believes that it has found the Achilles' heel of the bank associations: the fact that they are organized as joint ventures. If American Express can enter into deals with selected members it can alter the incentives of members in ways that benefit American Express, harm other members, and weaken the ability of the joint venture to operate. American Express also, evidently, believes that the antitrust laws on joint ventures ought to create a potential opening for it that would not exist if Visa and MasterCard were organized as for-profit investor owned firms like American Express. The antitrust laws rightly worry that members of joint ventures will engage in conspiracies in restraint of trade. Unfortunately, as Visa has found throughout its history, this provides an opening for its competitors to persuade the courts to consider many issues that would not even get a quick look if Visa were a proprietary firm. Fortunately, Visa has been consistently successful in defending itself against these challenges. But, as the present suit demonstrates, it continues to be attacked.

The American consumer would lose if the Justice Department succeeds in restructuring and regulating the credit-card industry. Even well-intentioned and carefully crafted attempts to tinker with the economy seldom work out as planned. Redesigning an industry that works extremely well violates two cardinal rules of economic policy: if it ain't broke, don't fix it; if it ain't broke, don't break it.

II. The Credit-Card Industry Is Highly Competitive

Economists evaluate the extent to which an industry works well for consumers by looking at its structure--how close does the industry come to the textbook model of perfect competition?--and its performance--how much actual competition is there and does that competition tend to dissipate profits?

A. Structure

The textbook case of competition is an industry in which (a) a large number of competing firms vie for the consumer's dollar; (b) no seller is large enough to affect price significantly by itself; (c) firms can enter and exit the industry easily; (d) consumers have good information about choices available to them; and (e) consumers can switch vendors when better offers are available. Although all real industries depart from this ideal to some extent, the credit-card industry approaches a textbook example of competition. Richard Schmalensee and I have described this in detail in Paying with Plastic so I'll just touch on some of the important structural factors here.

There are a large number of competing issuers. A recent Federal Reserve System survey of 128 of the largest credit-card issuers in the United States found that 58 distributed their cards nationally.(2) The same survey reported that 40 additional issuers distribute cards regionally in areas encompassing more than one state. All told, more than 6,000 issuers offer credit cards to consumers (and even more offer debit cards). Many of these issuers could expand regionally or nationally if there was a market opportunity.

Compared with other industries, the largest credit-card issuers are small relative to the industry overall. The largest issuer is American Express which had a 20 percent share of all general purpose credit and charge card purchase volume in 1999.

Economists and antitrust experts use the Herfindhal-Hirschman Index (HHI) to measure the degree of structural competitiveness. This index combines information on the number of competitors and their relative sizes. According to the Justice Department merger guidelines, an index of less than 1000 indicates a competitive structure while an index of more than 1800 is a cause for concern (a gray area lies in between). The HHI for general purpose credit and charge-card issuers was less than 850 in 1999.

Lastly, we all know from personal experience that it is relatively easy to obtain information on credit card rates and features and that it is easy to switch cards.

B. Performance

The credit-card industry performs much as we would expect given its highly unconcentrated structure. Credit-card issuers compete aggressively on price and features. There are no data I could present on this that would be more persuasive than our common experiences. While we might complain about getting too much junk mail and seeing too many Internet ads, we can't really complain that credit-card issuers aren't competing for our business.

This competition has enabled more consumers to get cards and has made it more attractive for consumers to use cards. The percent of households with credit cards more than quadrupled from only 16 percent in 1970 to more than 65 percent in 1995, the most recent year for which data from the Federal Reserve's Survey of Consumer Finances are available. The share of consumer spending that is paid for with general-purpose credit and charge cards has increased from less than 3 percent in 1975 to 18.5 percent in 1999. Quality-adjusted prices have dropped precipitously. Richard Schmalensee and I have developed an index for measuring the overall cost of using a credit card. That index fell 20 percent (in real terms) between the first quarter of 1984 and the second quarter of 1999. However, that index does not account for the fact that consumers are getting more from their credit cards than they used to--things that can't be measured accurately. They get many more features today--airline miles and rebates for example. Their cards can be used at many more merchants. A decade ago consumers couldn't buy groceries with their credit cards. Now they can. And in our time-conscious society it is important to recognize that it takes much less time to consummate a card transaction at a merchant than it used to. Technological innovation has reduced the time it takes to authorize a card transaction from about two and a half minutes prior to the advent of electronic terminals to a few seconds today.

Many credit-card issuers are highly profitable because they have developed extremely efficient operations. Many others have done much less well. And some have left the industry because it hasn't been sufficiently profitable. Overall, however, the credit-card industry appears to have a roughly competitive rate of return after adjusting for the fact that credit-card lending is a risky business.

The credit-card industry performs quite well based on the usual criteria used by economists. Of course, I am not suggesting that the credit-card industry works like the auction market for red # 2 winter wheat. Few industries do. There have been some suggestions by my fellow economists that interest rates are too sticky in the credit-card industry. There have also been some suggestions that banks can exploit an apparent tendency by people to underestimate how much debt they have. And consumer advocates sometimes have argued that credit-card issuers have induced consumers to take on too much debt. We shouldn't let the small amount of truth in these assertions overwhelm the larger truth that this industry performs exceedingly well for consumers generally.

C. The Credit-Card Associations Are Largely Responsible for this Competitive Structure and Performance

We also shouldn't take this highly competitive structure and performance for granted. The American credit-card industry didn't have to evolve this way. Prior to the establishment of the credit-card associations, the primary sources of general-purpose payment cards were American Express and Diners Club. Both companies had proprietary card networks that they used for issuing charge cards. The industry was highly concentrated, as one would expect given the scale economies and network economies in operating a card network. (Network economies arise from the fact that consumers value cards if they can use them at more places; and merchants value more cards if more consumers have them.(3)) Consumers and merchants paid much more for these cards than they do today. If credit cards had been successfully offered only by proprietary companies like American Express, Diners Club, and Discover, it is likely that the credit-card industry would be highly concentrated today and behave more like an oligopolistic industry than a competitive one.

The banks that started Visa and MasterCard in the late 1960s and early 1970s came upon the idea of an open joint venture in which members would work together in limited ways but compete intensively in others. This wasn't altruism. What the banks lost as a result of intense competition at the issuing level, they gained from getting consumers and merchants to rally around the Visa and MasterCard flags. They also gained from sharing investments in the infrastructure for operating a card system.

But those open joint ventures continue to provide important vehicles for promoting competition in the card industry. Almost any financial institution can join these associations and issue cards and rely on the strength of the brand and network. And consumers have benefited substantially from entry by new issuers and expansion by existing issuers over the years.

III. American Express

There are two financial institutions that can't join the associations. Visa and MasterCard have each adopted rules that prevent American Express and Discover--the two major for-profit systems today--from joining their associations. That rule was upheld in the so-called MountainWest case. Also, Visa and MasterCard have adopted rules that prevent American Express and Discover from joining through the back door. They have rules that prevent member banks from entering into contracts to issue the cards of the American Express and Discover systems. It is those rules that the Justice Department has challenged in its current lawsuit. (The Justice Department has also challenged another feature of the joint ventures. Ever since litigation in the mid-1970s, banks had been able to issue both MasterCard and Visa cards under a system known as duality. During the 1990s, duality has declined as both associations tried to gain greater loyalty of their most significant members. Since duality has become essentially moot, it would appear that the Justice Department lawsuit is mainly about letting American Express and Discover enter into contracts with particular banks.)

There are sound economic reasons why Visa and MasterCard should be very concerned about letting American Express enter into selective contracts with their members--as should consumers of credit cards. Professors Patrick Rey and Jean Tirole have analyzed competition between open joint ventures like Visa and for-profit companies like American Express.(4) They show that it is common for cooperatives to have rules that give members an incentive not to abandon the cooperative, leaving the remaining members holding the bag for all the investments and obligations of the cooperative. The Visa and MasterCard rules being challenged by the Justice Department are examples of these kinds of "loyalty" rules. They don't prohibit members from leaving for American Express but they do make it harder because members would have to dispose of their Visa and MasterCard portfolios.

More importantly, Professors Rey and Tirole show that not-for-profit cooperatives like Visa and MasterCard are especially vulnerable to raids by for-profit competitors. The cooperatives have to pass the average costs of operation onto their members to break even and cannot treat members too differently without jeopardizing the cohesiveness of the cooperative. The for-profit investor-owned corporation, on the other hand, can make special deals. So long as it makes a little bit of money from the joint-venture member, it can defray some of the sunk costs of its investments. Unfortunately, the cooperatives will cut back on investments if they face the risk that the for-profit will pick off members that will no longer be available to contribute to funding those investments. The result is that cooperatives will invest too little in improving their businesses. In the extreme case, the defection of key members could destabilize the cooperative altogether.

The Rey-Tirole work explains rigorously in economic terms why letting a for-profit firm pick off members of a joint venture can destabilize the joint venture. Common experience also teaches us that most successful enterprises demand loyalty from their members. Law firms typically do not allow their partners to moonlight at another law firm. Many companies have non-compete agreements that prevent their employees from dividing their loyalties. Most franchisers prevent their franchisees from distributing competing products. Many manufacturers have exclusive distribution agreements.

IV. The Justice Department's Plan to Restructure and Regulate

The fact that the Justice Department isn't proposing to break-up Visa or MasterCard shouldn't obscure the fact that it is proposing a remedy that is extraordinarily radical and, I believe, unprecedented in the annals of antitrust. Visa and MasterCard members currently have ownership rights in the systems. These ownership rights allow them to vote for members of the Boards of Directors who then appoint management. As I understand the Justice Department's proposal, they want to create two classes of members. Governing members would still be able to vote for the board and serve on the board as well as various operational committees. Governors would have to issue the preponderance of their volume on the card system they help govern. Issuing (i.e., non-governing) members would not be able to vote for the board or serve on either the board or any operational committees. They would be disenfranchised. However, they could issue cards for any system they want, including American Express and Discover cards as well.

Neither the Justice Department nor its economist, Professor Michael Katz, has laid out the details of how all this would work. It is my understanding, however, that many of the details would simply be left for a court to consider as part of its oversight of the card industry if the Justice Department prevailed. It is also hard to predict at this point who would pick disenfranchisement over being a governor. My guess, though, is that most community banks, credit unions, and smaller banks would decide that having voting rights isn't worth giving up their flexibility to shift between systems. So the Justice Department's proposal would likely disenfranchise most of the smaller banks that belong to the associations. (At the same time, it is unlikely that American Express would enter into agreements with smaller banks.)

The Justice Department remedy is radical in at least two respects. First, although I have not completed a systematic search, I don't believe I have ever encountered a case in which the courts have required an industry with such a competitive structure to undergo major restructuring and ongoing regulation. Second, other than in essential facilities cases, court-ordered changes in ownership and control rights have generally occurred only in antitrust cases in which the company was the result of mergers and acquisitions.

Of course, there is nothing wrong with suggesting a radical course of action to deal with a serious competitive problem. However, there are no sound economic reasons for supposing that this adventure in reengineering an industry will end up benefiting consumers. American Express and the Justice Department claim that consumers will be able to get bank-issued American Express cards. But consumers today can choose from numerous card varieties issued by more than 6,000 members of Visa and MasterCard as well as many card varieties available from American Express and Discover. Any of these issuers can provide the consumer with virtually any feature for which there is demand. There is no sound basis in economics for suggesting that adding another variety of card to this mix will make consumers much better off--and the antitrust laws are only invoked when ending a particular practice gives rise to significant consumer benefits. It is, of course, always true that letting one business use the property of another business to sell products will make certain consumers better off. However, we don't let businesses get that kind of free ride because consumers overall will ultimately lose in the end.

Although I don't claim that the credit-card industry would collapse if the associations were obliged to repeal their loyalty rules, I do believe that consumers would lose far more than they could conceivably gain. That's because American Express could make Visa and MasterCard less competitive organizations. If that happened, even slightly, consumers would lose a tremendous amount of money.

Here's an example of how it could happen. American Express has had an extremely successful and profitable corporate card program. The popularity of its corporate cards is one of the cornerstones of American Express's success story. It charges high prices to merchants who need to accept that card because they cater to business travelers. Visa and MasterCard have challenged American Express in the last few years in the corporate card arena. This has increased choice for corporations. If American Express had been able to enter into contracts with Visa and MasterCard members it could have targeted the handful of banks that were behind this initiative and allowed them to issue American Express's corporate card instead.

V. Antitrust and Innovation

Why has the Justice Department sought to restructure and regulate an industry that, by Justice's own admission, hasn't artificially raised prices or restricted output? It argues that Visa and MasterCard have suppressed innovation and that its redesigned industry will spur innovation. I don't want to get into the specifics of the case, but I do want to raise two issues that I find troubling as an economist who works on antitrust matters.

First, the Justice Department and its economist, Professor Katz, are quite candid in saying that they can't identify the innovations that Visa and MasterCard suppressed and they can't say how much more innovation will occur if their proposed remedy was adopted. Their case is based primarily on the proposition that their remedy will promote more competition and more competition will promote more innovation. This proposition is not testable and therefore unverifiable. When confronted with Visa's superb history of innovation, the Justice Department and its economist have essentially responded: "but there would be more." When Visa has asked the Justice Department and its economist for proof that redesigning the industry would create more innovation, the response is that competition, by assumption, leads to more innovation. I do not believe that it is advisable for either the Justice Department or the courts to be relying so heavily on theories that can't be confronted with data.

Second, modern economics recognizes that there is no predicable relationship between market structure and the rate or direction of innovation. That is true for the commonsense reason that innovation is so unpredictable--it comes from the strangest places for the strangest reasons. That is also true because the very limited models of innovation that economists have been able to develop demonstrate that the effect of market structure on the rate or direction of innovation is completely ambiguous. Competition makes people race faster for the next great thing; but their competitors may be able to imitate their innovation and therefore free-ride on their efforts. Monopolies have the advantage that they can capture scale economies in research and development and they may be able to prevent others from capturing any of the returns; they have the disadvantage that they tend to restrict output. Antitrust cases that are based on restrictions on price and output are anchored in a well-developed body of theoretical and empirical work. While economists may disagree about the application of this work to the facts of a specific case they mostly agree on the general principles. Antitrust cases based on speculation about innovation have no such anchor.

I think the Justice Department's efforts to restructure and regulate the credit-card industry is based on sheer speculation that the government can do a better job of designing the credit card industry than its thousands of participants can. That is bad economics and bad policy.


1 In the case brought by the Justice Department against Visa and MasterCard it has alleged a credit and charge-card market. While I do not agree with that market definition, it is still useful to discuss competition in the industry and, for convenience, I will generally just refer to "credit cards" in talking about the business.

2 "Semiannual Credit Card Survey," Board of Governors of the Federal Reserve System, January 31, 2000, available at, visited May 23, 2000.

3 See, Evans and Schmalensee, Paying With Plastic: The Digital Revolution in Buying and Borrowing, Chapter 7.

4 See, Rey, Patrick and Jean Tirole, "Loyalty and Investment in Cooperative Credit Card Systems," February 2000. Professor Tirole is the author of Modern Industrial Organization, the leading graduate text in this subject, and a past president of the World Econometric Society. His work on this topic was funded by Visa U.S.A.

Home | Menu | Links | Info | Chairman's Page