My name is Bennett R. Katz. I am the Vice Chairman of the Management Executive Committee, Group Executive Vice President and General Counsel of VISA International. I am pleased to be here this morning to share some thoughts with you about joint ventures and the effect of federal antitrust laws on innovation involving collaboration.
VISA may well be the most successful joint venture in the history of American, if not world, commerce. During 1994, in the United States, more than $290 billion dollars worth of goods and services were purchased with more than 205 million VISA cards at the more than 3 million locations that accept the VISA card. Worldwide, the total volume of commerce is approximately $630 billion, the number of outstanding VISA cards is 390 million and the number of merchant locations that accept VISA cards as payment for goods and services is 12 million.
Those are, by any measure, impressive numbers. In fact, the general purpose charge card is such an integral part of the American business landscape that it is difficult to recall a time when the question "will that be cash, check or charge?" was not a part of our daily commercial life. In fact, however, the predecessor of today's VISA organization began less than 40 years ago as a small program of the Bank of America in Fresno, California. At the time VISA -- or BankAmericard as it then was known -- was launched, there was no such thing as a general purpose payment card. Local transactions were handled by cash or check, except for those relatively few merchants that offered an in-store charge-account program for their regular customers. Business travelers and vacationers faced an even more restrictive payment environment since most merchants would not accept out-of-area checks. For them, the solution was either to carry large amounts of cash or to purchase travelers checks. In the early 1950's, Diners Club, Carte Blanche and, shortly thereafter, American Express attempted to meet the needs of the business traveler by creating the so-called Travel and Entertainment, or T&E, card. Unlike the modern credit card, however, the T&E cards carried no credit feature. Moreover, their acceptance was limited almost exclusively to hotels, restaurants and a few upscale locations frequented by business travelers.
After some early, limited success with its own credit card program in California, Bank of America formed the predecessor of the current VISA joint venture, known as NBI, in 1970. From the beginning, the participants integrated only those functions that needed to be performed centrally while leaving members free to compete with one another in offering the VISA service to consumers and merchants. For example, VISA does not issue VISA cards or sign merchants to accept them. Rather, under the venture agreement, individual members are authorized to issue cards bearing the VISA name and to sign up merchants to accept those cards as payment for goods and services.
To launch VISA, the members had to take on some very substantial economic costs and risks. For example, the participating banks (principally the card-issuing banks) assumed the risk of fraud or non-payment by the cardholder. In addition, cardholders were given a so-called "free" period during which they could pay off their balance in full without incurring any finance charge. Thus, in effect, every cardholder got the benefit of a free short-term loan from his or her card-issuing bank. The "float" cost during that period also was born by the members.
There were substantial obstacles to VISA's success. One such obstacle -- which is a familiar problem in "network" industries -- is the "chicken and egg" problem: VISA cards were only of interest to consumers if they could be widely used. However, merchants were only willing to accept them if they were carried and used by many consumers. Since VISA was creating a previously unknown product, overcoming this "chicken and egg" problem was a difficult, time-consuming and expensive undertaking. Much red ink was spilled by the American banking industry in solving it.
VISA's members also had to bear the cost and effort of creating the various systems necessary for a complex national (and later international) network to operate. VISA was a pioneer in the use and creation of sophisticated telecommunications and computer systems. In the early days, those systems were necessary to permit the VISA product to exist at all. Today, VISA's authorization, clearing and settlement systems provide nearly instantaneous approval (or, where appropriate, disapproval) of transactions on a global basis, plus prompt settlement of accounts among members.
I have begun with a somewhat lengthy description of the VISA system, and its early history, because it is important for an understanding of the further comments that I wish to make about the impact of antitrust upon our business. VISA and the antitrust laws have been constant companions throughout our history. We have regularly communicated with both federal and state antitrust enforcers about many important issues. We, and our antitrust advisors, are constantly concerned with the potential antitrust implications of what we do. Moreover, since much of what we do is at the very forefront of technology, commerce and the law, we often find ourselves without clear guidance either in the very general language of the antitrust statutes or in the decided cases.
Even with careful planning, we have been the subject of a number of antitrust actions, not the least of which is the recently-concluded lawsuit brought by Dean Witter Discover in Utah seeking to compel VISA to admit a direct horizontal system competitor into the VISA joint venture. Fortunately, the Tenth Circuit rejected that attempt, just as the Eleventh Circuit rejected a challenge to our so-called "interchange" fee in the NaBanco litigation a decade earlier.
VISA's commercial and legal history teach us several things. First, they underscore the immense value that innovative joint ventures can produce for American consumers. The VISA card is a product that had to be created by a joint venture if it was going to exist at all. At the time VISA was formed, no bank could -- financially, technologically or legally -- have created a national, let alone international, payment card on its own. The problems were too great and the risks were far too high.
Second, the VISA joint venture had to take the risk that the antitrust laws would be applied in a sensible, rather than a mindless, fashion to practices that were necessary for the venture to exist. For example, VISA adopted its interchange fee (which is an internal transfer payment between issuing and merchant-signing banks) several years before the Supreme Court's decision in the BMI case. On its face, the interchange fee is an agreement among thousands of horizontal competitors fixing the price at which transactions are "interchanged" between card issuers and merchant-signers. That was precisely how it was described by plaintiffs in the NaBanco case when they challenged the interchange fee as per se illegal horizontal price-fixing. Having anticipated the Court's opinion in BMI, we were able to argue successfully that imposition of an interchange fee was necessary if the general purpose charge card was going to exist at all.
A decade later, in the Dean Witter litigation, we had a more difficult time sustaining a refusal to share our members' property with a direct competitor, even though such an argument if directed against a single firm would have been rejected summarily by virtually any court in the United States. Although we ultimately prevailed in that case in the Tenth Circuit, we did so only after suffering an adverse jury verdict and the prospect of potential damages of as much as $1 billion after mandatory trebling.
A principal reason for our lengthy period of anxiety in the Dean Witter case is that the antitrust laws have not adequately developed standards for evaluating joint venture practices under the rule of reason, including the importance of encouraging entrepreneurial risk-taking and investments in innovation. With the kind of uncertainty that such a rule of reason trial presents and with massive treble damages as the penalty for guessing incorrectly, there is little doubt in my mind that the antitrust laws have deterred or adversely affected the formation and operation of potentially important, cutting-edge joint ventures. Reforming the rule of reason is a topic that is broader than the subject of joint venture innovation. However, it is a matter of some urgency.
The cost of uncertainty is extremely high, and not just for the venture and its members. In the early days of the NBI joint venture, we decided to maintain at least partial membership exclusivity vis-a-vis banks that were members of the other major credit card system, MasterCard (then called Master Charge). As a result, we were sued by the Worthen bank in Arkansas that wanted to participate in both systems. Worthen claimed that our exclusivity rule constituted a per se illegal group boycott and the district court agreed.
Faced with the prospect of a treble damage exposure we could ill-afford at that time, we sought a business review letter from the Antitrust Division. The Division sat on the matter for 14 months before issuing an equivocal statement that gave us insufficient comfort to continue defending our policy or the litigation. Having warned a skeptical Justice Department that the consequence of its indecisiveness would be a virtually complete overlap of membership, NBI repealed its anti-duality rule. The result was exactly as predicated, with less vigorous competition between VISA and MasterCard as a result.
A number of government officials have since acknowledged that the Division's unwillingness to support us on this issue was an error. They have defended the government's indecisiveness on the ground that they simply did not know enough about the likely future of the industry or the impact of duality on competition to take a firm stand. Assume for the moment that that is true. The fact is that indecision is often decision. We had no choice but to change our rule and the consequence of that action quickly became irreversible, given the enormous investments made in reliance upon the new "duality" rule. Put otherwise, if an issue is too complex, or premature, for the experts in the DOJ or FTC to make up their mind about it, it certainly is too complex for us to be exposed to the vagaries of potential treble damage liability under the rule of reason.
Speaking more generally, I believe that careful attention needs to be given to the special problems presented by joint ventures in applying antitrust principles. This is not to deny that joint ventures can present opportunities for collusion or other anticompetitive practices. It is, rather, to say that we need more discriminating tools than now exist to distinguish clearly and in advance between practices that are, in fact, anti-competitive and practices that further the efficiencies of true integrations such as the VISA joint venture.
To take one example, traditional ways of looking at market power may need to be reconsidered, or at least more critically examined, in the joint venture context. In the Dean Witter case, for instance, the plaintiff argued that because VISA's members account for in excess of 70% of general purpose charge card volume, VISA's refusal to allow Dean Witter to issue a VISA card represented an exercise of market power notwithstanding the fact that VISA's 6,000 issuers compete actively with one another in credit card issuance. Similarly, when VISA entered the new POS market several years ago, would it have made sense to say that it had power over a product that, for all intents and purposes, did not then exist?
It also seems to me that insufficient attention has been paid to the distinction between intra- and inter-system competition in evaluating joint venture practices. As I mentioned previously, VISA, itself, neither issues cards nor signs merchants. Those functions are, instead, carried out by VISA's thousands of members who are licensed to use the VISA marks and systems, in competition with one another and with the so-called "proprietary" brands, principally Discover and American Express. In addition to that "intrabrand" competition, which is virtually atomistic, VISA competes at the "brand" or "system" level with a small handful of other credit card brands: namely, American Express and Discover and, to a far lesser extent, Diners Club and JCB. VISA also, of course, competes at the brand level with MasterCard, although that system-level competition is very much affected by the substantial overlap in membership between the two systems that was a fallout of the Worthen litigation discussed a few moments ago.
System competition is no less important than the intrabrand competition that takes place among the numerous VISA members. It involves such matters as technological innovation and brand promotion. Those of you who are familiar with VISA's "and they don't take American Express" advertising campaign will understand just how vigorous that competition is.
Unlike the many thousands of competitors at the issuing and merchant-signing level, however, there are only a handful of credit card systems. VISA attempted to make that point in defending against the Dean Witter lawsuit. Indeed, we actually asserted a counterclaim under Section 7 of the Clayton Act alleging that the partial merger that would result from allowing the owner of Discover to join VISA would have a materially adverse effect upon system-level competition in the general purpose charge card market. The district court (which acted as trier of fact of this equitable claim) found that VISA had the better of the argument on that point but that the magnitude of the adverse effect was insufficiently great to justify relief. Without arguing about the correctness of that determination, the point I wish to make is that in evaluating restrictions imposed by joint ventures, it is important to consider not only the effect of any limitation on competition within the venture but to take into account the broader context of competition between the venture and its other competitors.
Joint ventures remain a vital part of the payment services business. These ventures are as good an example as can be found of the benefits which flow from this type of industrial organization. Moreover, the financial services industry is continuing to innovate in ways that consumers value. ATM and point-of-sale (or "POS") networks are further instances in which joint ventures have produced products far different -- and more beneficial -- than what any individual financial institution could offer on its own.
At VISA, we are continuing to innovate in ways that will transform how financial transactions are conducted. We are pioneering numerous new products that involve improved chip technology, such as our new stored value card that is currently being tested. We are working to make the purchase of goods on the Internet secure against fraud. We have plans for sophisticated home banking services.
One of the more significant recent developments in our business is that VISA is not only a joint venture, itself, but it is engaging in more and more joint ventures with other firms. Such ventures -- with telecommunication or software firms, for example -- are going to be as indispensable to our future as the joint venture format was to the creation of the VISA system. We and our partners expect to bring complementary resources and expertise to the table. Combining those resources will allow us to do things that none of us alone could accomplish.
Each of these advances brings with it a new set of legal, as well as commercial and technological, issues. With whom may we share the results of our efforts? With whom must we share them? When? On what terms? May we set technical standards in collaboration with others? Must all of our competitors be included? What can we do to incent member (or third party) behavior in order to solve new versions of the "chicken and egg" problem? May we have interchange fees? Can we prohibit surcharges? What about "duality" for these new products and services?
In trying to answer these questions in a way that maximizes the possibility of commercial success, we remain cognizant of, and recognize the legitimacy of, antitrust considerations. But we need flexibility and sophistication in the application of the antitrust laws.
We also need predictability. Properly interpreted and properly applied, there is no reason why effective antitrust enforcement, vigorous competition and joint venture innovation cannot happily coexist. However, our inability to obtain prompt, reliable advance assurance about the antitrust analysis of our innovations remains a source of great concern. We are particularly concerned that in the hands of the treble damage plaintiff, the antitrust laws are a Damoclean sword over the heads of joint ventures such as VISA.
Without clearer legislative guidance and greater support from antitrust policy makers, I believe that innovation will be compromised as the threat, if not the actuality, of treble damage litigation conducted under rules that turn such litigation into little more than roulette. The consequence -- both to VISA and to many other existing and prospective joint ventures -- will be excessive caution and overdeterrence.
October 26, 1995