I’d like to thank Concurring Opinions for inviting me to blog about In re: Payment Card Interchange Fee and Merchant Discount Antitrust Litigation. This eight-year-old multi-district litigation has produced the largest proposed cash settlement in litigation history ($7.25 billion) along with what is perhaps the most extraordinary release from liability ever concocted. It may also be the most contentious. Over half the name plaintiffs and over 25% of the class, including most large merchants (think Walmart, Target) and most merchant organizations, have objected. On September 12, Eastern District of New York Judge John Gleesaon held a fairness hearing to consider the settlement, and the parties are awaiting his decision. An appeal is a virtual certainty.
This post will provide background on the credit card industry pricing mechanisms that led to this litigation, the legal issues in the case, and the structure of the settlement. (You can read more about the history of the credit card industry’s relationship to the antitrust laws here.) In subsequent posts, I’ll separately analyze the damages and relief provisions in the settlement. (If you can’t wait 
my working paper analyzing the settlement is here.) If there are particular issues that you’d like to read more about, let me know in the comments and I will respond in subsequent posts.
The credit card industry is atypical, but not unique, in that it competes in a two-sided market, i.e., one that serves two distinct customer bases. A card system like Visa provides both a purchasing device (credit cards) to consumers and a payment acceptance service to merchants. (By way of comparison, the legal blogging market is also two-sided. Concurring Opinions provides both an information forum to its readers and a platform to its advertisers.)
In a typical one-sided market, goods and services will optimally be priced at the marginal cost of providing them plus (some argue) a profit increment sufficient to attract investment given the riskiness of the industry. So, one might think, a two-sided market would optimally provide credit cards to consumers and acceptance services to merchants at the respective marginal costs (plus profit for risk) associated with each side of the business. It turns out, however, that efficient pricing in two-sided markets rarely works out that way. If the elasticity of demand for customers on each side differs, output is maximized by charging less than marginal cost to the customers with the more elastic demand. So, Concurring Opinions charges (a) its more elastic readers less than the cost of bringing them information; and (b) its somewhat less elastic advertisers more than the marginal cost of placing ads on the screen. Except for the price levels, the New York Times and any media outlet relying on advertising rather than subscription fees does the same thing.
In the credit card industry, consumers are the more elastic customer. They have many easily substitutable options, including the at least four credit card systems, debit cards, cash, and checks. While all those substitutes may not be perfect, credit card companies have rationally reasoned that their systems would grow more quickly if they charged cardholders less than the marginal cost of serving them. Given the incredible success of credit cards, and that every successful credit card system, including American Express, has followed this model, it has hard to argue with the card systems’ conclusions. Charging cardholders less than the cost of serving them, of course, means that merchants have had to pay more than the marginal cost of the card acceptance services they receive.
In the Visa and MasterCard systems, banks provide all of the customer functions on both sides of the market, and for any given transaction, the merchant and the cardholder are likely to work with different banks. To exploit the efficiencies inherent in charging less to the more elastic customers, the card systems needed a way to shift revenue from the merchant side to the cardholder side. To do this, Visa and MasterCard created a fee that the merchants’ banks must collect and pass on to the bank that issued a card used to make the purchase. This fee came to be called the interchange fee, and it sets a floor on the price that merchants must pay to accept a Visa or MasterCard card. The merchant’s bank tacks its own fee for providing card acceptance services on top of the interchange fee. The total fee paid by the merchant is called the merchant discount, i.e., the percentage less than the full purchase price that the merchant receives from the credit card system. Although interchange fees vary based on a variety of factors, they virtually always constitute the bulk of the total merchant fee with 75% constituting a reasonable rule of thumb.
To illustrate, if a merchant pays a 2% merchant discount to accept credit cards, the flow of funds would look roughly as follows:
(1) a customer purchases a good from the merchant for $100 presenting the credit card for payment;
(2) the merchant submits the receipt to its merchant bank which pays the merchant $98 (the purchase price less the merchant discount);
(3) the merchant bank then pays the interchange fee to the bank that issued the card used to make the purchase, which would likely be about $1.50 in this example;
(4) the card issuing bank bills the cardholder the full purchase price plus interest if the cardholder runs a balance.
The small per transaction amount belies the fact that these fees add up to over $30 billion annually. There is no shortage of complex economic analysis of interchange. You can find a short accessible economic and legal assessment here.
For several decades, a detente existed among participants in the industry. American Express served a niche market of wealthy customers, who didn’t need credit, and businesses. Merchants were generally willing to pay more to access the AmEx customer base, given that few run-of-the-mill consumers had American Express cards. Visa and MasterCard made hay by underpricing AmEx and providing a more ubiquitous purchasing option for the masses that included revolving credit and came with a much larger merchant base.
By the 1990s, however, this comfortable equilibrium began to unravel. The Discover Card demonstrated that a profitable mass market card system with revolving credit could operate with prices to both cardholders and merchants that were lower than those provided by the banks issuing Visa and MasterCard cards. And American Express grew restless with its position as a high-end niche player. It began a multi-front campaign to compete with Visa and MasterCard for typical consumers by adding a revolving credit option and price discriminating on the merchant side to attract more merchants to accept AmEx cards. Most controversially, American Express sought bank partners to issue its cards. Visa and MasterCard responded with exclusionary rules that prohibited any bank that issued American Express from also issuing either Visa or MasterCard. Banks could, however, continue to issue both Visa and MasterCard. Not surprisingly, no bank issued American Express cards (though several said they wanted to), because they could not afford to drop the two largest brands.
In 1999, the Department of Justice sued, attacking the Visa/MasterCard exclusionary rules and won. With the exclusionary rules out of the way, banks began issuing American Express cards at least in part because AmEx had higher merchant fees than Visa and MasterCard and thus could pay higher interchange fees to card issuing banks. This phenomenon exacerbated a reverse price war that had already begun between Visa and MasterCard. Each credit card company increased its interchange fee to merchants in order to make its credit cards more attractive to banks issuing them. From 1995 to 2005, interchange fees rose more than 25% despite concurrent technilogical advances that reduced the banks’ costs of running the credit card system and dramatic increases in volume and merchant base that should have supported economies of scale and scope. It was a nice deal for card issuing banks, and perhaps consumers too who received more generous rebates than ever before. But merchants saw it differently.
By 2005, merchants had filed more than a dozen cases against Visa, MasterCard, and the largest banks that issue their credit cards. The complaints advanced two primary modes of attack:
(1) interchange fees constituted illegal price fixing because Visa and MasterCard, albeit separately, set default interchange fees that all banks issuing cards for that system receive from the merchants; and
(2) the systems’ so-called anti-discrimination rules prohibiting merchants from surcharging credit card transactions and taking other steps to steer their customers toward less expensive means of payment constituted unreasonable restraints of trade.
Both the interchange fee and the non-discrimination rules, the merchants argued, were the product of horizontal agreements among the banks that should be competing to convince merchants to accept their cards just as they convince cardholders to use them. The Visa and MasterCard systems, the complaints alleged, were massive hub and spoke conspiracies with the card company at the hub and each bank at a spoke. By joining the card system and agreeing to the centrally set interchange fee and the anti-discrimination rules, the card issuing banks agreed not to compete on the price they charged merchants and to restrain merchants from adopting policies that would force the card systems to lower the merchants’ fees. The complaining merchants asked the court to enjoin the card systems from imposing interchange fees and non-discrimination rules, arguing that banks — like Citibank, Chase, CapitalOne, & Bank of America — should complete on the fees that they charge merchants just as they compete for cardholders and that merchants should be able to discriminate against the use of credit cards in order to drive prices down.
In particular, the merchants argued that Visa and MasterCard should be forced to abandoned their “honor all cards” rule. This rule required merchants to accept all cards within a brand or none. Merchants sought the ability to single out a particular bank, e.g. Wells Fargo, and refuse to accept that bank’s Visa cards (for example) unless the bank agreed to reduce its merchant’s fee. To be sure, any merchant could refuse to accept all Visa cards. But merchants argued, and the court found in the 1999 DOJ case, that Visa and MasterCard credit cards are so important to a merchant’s business that they could not realistically drop either brand entirely. But if they could single out an individual bank and refuse to accept its cards, a merchant could make a credible threat and leverage a lower fee.
The cases were soon consolidated in the Eastern District of New York before Judge Gleeson who, in the late 1990s, had overseen the successful settlement of a merchant class action involving the same parties dealing with the fees merchants paid to accept debit cards. In the fall 2012, after seven years of heated litigation, counsel for the named plaintiffs and the defendants proposed a settlement. Visa and MasterCard would pay the merchants over $7 billion dollars. The amount included a payment of just over $6 billion for past damages and a rebate of 10 bais points on interchange fees paid over the next eight months after the settlement was initially approved. Although the amount is large in an absolute sense, it constitutes a mere three months of interchange fee revenue for the card issuing banks, and nothing in the settlement would restrain Visa or MasterCard from raising their fees immediately to recoup the payment.
The settlement would also enable merchants to engage in some efforts to steer their customers toward less expensive payment mechanisms, though it stopped far short of the complete freedom that the merchants sought.
Equally important, and most controversially, are the things that the settlement does not do. It would permit the default interchange system and the honor all cards rules to remain in place in exactly the form that they exist today. And the release provisions would prevent any merchant (even those that don’t exist yet) from challenging forever the interchange fee, the honor all cards rule, or, breathtakingly, any other rule now existing – or any “substantially similar” rule or “the future effect” of any “conduct” of the defendants relating to those rules – even if the rule in question played no role whatsoever in the litigation.
Would this settlement be a good deal for the merchants, the card systems, the banks, and consumers? A lot of merchants didn’t think so, opposing it fiercely. The settlement included a threshold for opt outs, permitting the defendants to withdraw if it were exceeded. That threshold was met, but the defendants have decided to stand by the settlement. Next time, we will dig deeper into the settlement’s damages provisions to determine whether they meet the Federal Rule of Civil Procedure 23 standard of fairness, adequacy, and reasonableness.