The first judgments to apply the economic theory of multisided markets to the payment card industry have been recently adopted in the UK and in the US. This paper, available here, uses these cases to try to show that antitrust authorities should take into consideration the two-sided nature of the credit-card industry, and to explain how this can be done. Taking the multisided nature of payment systems into account is necessary to arrive at a realistic description of these markets, and to develop solid theories of harm and procompetitive justifications that can explain some business practices involving credit cards.
It is structured as follows:
Section II briefly describes the main economic features of multisided business models.
Different authors have defined multisided markets differently, as they looked at different markets and business models. Nonetheless, the following generic traits tend to characterise all such markets: the presence of indirect network externalities that cannot be internalised through a bilateral exchange (usage and membership externalities); the necessity for an intermediary (matchmaker) to intervene to resolve a transaction-cost issue, thereby generating value for at least one of the sides of the market; the need to bring a sufficient number of economic agents onto each and every side of the market so as to reach a critical mass to foster indirect network effects; and the imposition of asymmetrical prices on the different groups operating on the platform (skewed pricing) to foster participation in every side of that platform.
These peculiarities have obvious consequences for the analysis of multisided markets. If agents on each side of a market are interdependent, their welfare depends on the combination of outcomes on different sides of the platform. This means that businesses compete to attract two or more demands, making traditional single market-sided approaches unfit to represent the competitive dynamics at play and to evaluate the competitive impact of the platform’s conducts. In other words, multisided markets create difficulties for antitrust analysis, which tools were designed to address one-sided markets. These difficulties include, among other things, how to define markets, how to determine whether market power exists, or how to use price as a relevant parameter of competition.
Section III deals with the functioning of payment card systems.
The market for payment systems is one of the most notorious and studied examples of multisided markets, preceding current studies into many digital markets. The functionalities offered by payment systems and the structure of the demand on both sides of the market are strongly interdependent. On the one hand, cardholders obtain a benefit only if a vast number of merchants accepts payment with their card. Similarly, merchants derive an advantage from accepting a card only if this card is actually used by a sufficient number of cardholders.
To promote participation on both sides of the market, two models have been deployed by payment card systems. A first system is commonly called a four-party scheme and is exemplified by Visa and MasterCard. On one side of the market, the card issuer (usually a bank) operates as an intermediary between the merchant and the payment card service provider (i.e. Visa or MasterCard); on the other side, the acquirer (also a bank) plays the same role between the payment card service provider and the merchant – in effect, the role of both the card issuer and the acquirer can be played by the same financial institutions. When a cardholder makes a purchase using the card, the card issuer bank pays the purchasing price to the acquirer bank, who then transfers the sum to the merchant, minus a fee for itself (so-called merchant fee) and another to the issuer bank for managing the transaction (so-called interchange fee). A second model is normally called a three-party scheme, and is exemplified by American Express. This model does not involve banks. Instead, American Express deals directly with the cardholder and the merchant – and every time the card is used, it pays the merchant the value of the cardholder’s purchase, minus a fee for its management of the transaction.
Both systems require the cardholder to pay a membership fee, and the merchant to accept a usage-based discount fee for purchases made with cards. This fee has raised a number of antitrust questions. In four party-systems, the interchange fee the acquiring bank pays to the issuer bank can be seen as traditional horizontal price-fixing that sets a floor to price competition in fees charged to merchants via merchant fees. However, it is often argued that imposition of an interchange fee by the payment card provider (i.e. Visa or MasterCard) is necessary to maintain equilibrium between the two sides of a platform. This is because the interests of issuing and acquirer banks are not aligned – in particular, issuing banks have incentives to hold up the cardholder’s payment in order to extract higher fees from acquiring banks. Another practice that has raised antitrust concerns are antisteering clauses, which prevent merchants from applying different prices according to the payment method chosen by the consumer or from nudging consumers towards a given payment method. While such a clause eliminates competition between card brands and between card-issuing banks, it has been argued that such clauses may be justified by the fact that they ultimately ensure the survival of the entire payment system based on cards and alternatives to cash.
Section IV describes EU and US regulatory and antitrust interventions regarding credit-card payment systems.
The antitrust profiles of interchange fees and non-steering clauses triggered a number of antitrust and regulatory interventions in the EU and the US.
In the US, in the mid-1980s courts established that interchange fees were legitimate under antitrust law and should be subject to a rule of reason applied to a two-sided market. To date, no US court has declared that interchange fees infringe antitrust law. In Europe, on the other hand, interchange fees were originally said to be prima facie restrictions of competition justified on efficiency grounds. Following the 2007 sector inquiry, however, interchange fees began to be considered restrictions of competition by object. More recently, the CJEU in its Cartes Bancaires and MasterCard judgments clarified that is necessary to take into account both sides of a platform market – such as payment card markets – even if this does not necessarily preclude finding that interchange fees infringed competition.
Subsequently, EU Regulation 2015/751 introduced caps on interbank commissions based not on the cost incurred by the issuing banks but, rather, on a merchant indifference test. This test builds on the level of the fee that the merchant would be inclined to pay considering the costs it faces. The US has also adopted regulation in this regard. However, the Dodd-Frank Act mandates the Federal Reserve Board to regulate the interbank fee on debit cards so as to make it proportional to the cost faced by the issuer in a transaction, instead of focusing on the merchant’s perspective. Despite this difference, both EU and US regulatory interventions share the same underlying assumption, namely that interchange fees operate a value transfer from merchants and consumers to banks. However, some economic scholarship challenges this pass-through thesis. This scholarship argues that a two-sided market analysis demonstrates that interchange fees merely influence price structures, not their level. In other words, an increase in the value of an interchange fee does not correspond to a price increase for the end consumer, but, rather, to a cost reallocation between two categories of end consumers (merchants and cardholders), such that a reduction in the value of interchange fees does not necessarily mean higher benefits for cardholders.
The same regulatory instruments also regulate antisteering provisions in Europe and the US. In Europe, antisteering provisions are prohibited. Merchants can decide which payment system they want to favour; in addition, merchants are able to direct consumers towards one or another payment system, to inform consumers about the interbank fees applied by different payment methods, and to charge surcharges or grant discounts to consumers in amounts not higher than the costs incurred with the relevant means of payment. In the US, merchants can grant discounts or otherwise direct consumers towards a payment system different from a credit card (that is, towards checks, debit cards, cash).
Section V discusses EU and US court decisions on payment card systems.
I must emphasise that the judgments reviewed in this article are out-of-date, having both been subject to appeal – and I have reviewed the cases before here and here, so I will not do so here. What is important for the authors, in any event, is that the peculiarities of multisided platforms have obvious implications for the analysis of the relevant markets and business conduct – and that courts have started to incorporate such insights into their decision-making. All phases of the competition inquiry are influenced by the peculiar economic features of multisided markets, from the definition of the relevant market through the identification of market power to the evaluation of the competitive effects of the relevant conduct. The authors outline how the EU and US courts did this.
A particular concern in both cases was whether the restriction (i.e. interchange fees in Europe, antisteering provisions regarding credit card systems in the US) was necessary for the survival of the payment scheme, and whether pro-competitive or anticompetitive effects prevailed. In MasterCard, the English High Court adopted two-sided markets’ theory as a reliable description of reality. The court relied on hypotheses that can inferred from this theory, and tested them by reference to evidence collected to that effect – particularly in trying to determine what the relevant counterfactual was. The court recognised that interchange fees could amount both a necessary mechanism to manage the hold-up risk linked to issuers’ monopsony and a tool for rent extraction – and that the challenge lay in balancing between the two. In the US, American Express’ antisteering provisions were subject to a rule of reason. In addition, the Circuit Court adopted an approach to market definition that took the various sides of the market into account, and expressly balanced the system-sustaining possibilities of antisteering clauses – i.e. neutralising merchants’ free riding – against their potential anticompetitive restraints in multisided markets – i.e. preventing competition between different payment systems. Given the peculiarities of multisided markets, the coexistence of different business models, and the ambiguous effect of the business conduct under assessment, courts emphasised the need to identify the relevant counterfactual hypotheses in order to measure interchange fees and antisteering provision’s actual impact on competition.
Section VI concludes.
In short, the same reasons that lead the authors to praise the flexibility of the English and US courts’ approach also leads them to be critical of the ex ante regulation of payment systems adopted in the EU and the US. Empirical studies have shown that the cap on the interchange fee for domestic debit card transactions in the US have led card-issuing banks to mitigate lost interchange revenue by adjusting cardholder terms and fees accordingly. There are reasons to also doubt the effectiveness of EU regulation as adopted.
Any review of this paper must begin by acknowledging that the cases it reviews have been overturned on appeal, so their discussion is out of date. While it is a bit hard to read, the paper provides a good description of the challenges posed by multisided market for competition analysis and decision-making – including an interesting discussion of how legal doctrines, such as object versus effect analysis, and rule-of-reason versus ancillary restraints doctrines, impact economic analysis. On the other hand, and as in many papers by industrial economists, the paper ignores a number of institutional constraints, such as those posed by legal procedure or the relative costs of ‘flexible’ judicial decision-making and ‘inflexible’ regulation – questions which, I am happy to note, the next paper takes seriously.