Jean Tirole’s legacy
On October 13 the Royal Swedish Academy of Sciences awarded the Sveriges Risbank Prize in Economic Sciences in Memory of Alfred Nobel for 2014 to the French economist Jean Tirole for his analysis of market power and regulation.
On October 13 the Royal Swedish Academy of Sciences awarded the Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel for 2014 to the French economist Jean Tirole for his analysis of market power and regulation.
1 The numerous references to Laffont in the scientific background for the prize suggest that, were he still alive, he might have shared the prize with Tirole.
Jean Tirole, along with his mentor and coauthor Jean-Jacques Laffont,1 was one of the main scholars who contributed to the systematic use of game theory and mechanism design in Industrial Organization (IO), which fundamentally changed the way economists analyze oligopolistic competition, firms’ strategic behavior and market regulation. His 1988 textbook in IO is still considered “the Bible” in the field, after more than 25 years, and his 1991 textbook in game theory (coauthored with Drew Fudenberg) is still one of the main advanced references.
His contribution, however, goes well beyond elegant modelling and the robust analytical toolkit he helped to create. Several of his seminal papers had a massive impact on the way we approach relevant and delicate policy issues in the field of competition policy and market regulation. The following is a selection of some of his fundamental contributions to IO research. In a short blog post it would be impossible to describe all of Tirole’s contributions to IO and other fields (behavioral economics, corporate finance, organizational economics, etc.); however, the interested reader may visit Tirole’s personal webpage at the Institute d’Economie Industrielle, his page on IDEAS or read the slightly technical document with the scientific background for the prize.
Anticompetitive vertical mergers and restraints
Tirole developed a model in which vertical mergers (and vertical restraints) can result in market foreclosure
In his massive 82-page 1990 article with Hart, Tirole develops a model in which vertical mergers (and vertical restraints) can result in market foreclosure. Contrary to the then-common Chicago School view, that considered foreclosure an irrational strategy and efficiency gains and synergies the only motivations behind vertical practices, Tirole points out that, if contracting takes place in secret or contracts can be secretly renegotiated, the inability of an upstream supplier to make binding commitments with a downstream retailer reduces the supplier’s market power and its ability to keep prices high.
To understand why this is the case, consider an upstream supplier facing several downstream retailers and assume that there is no demand uncertainty so that both the supplier and retailers agree on the expected profits that can be made from selling the products. The supplier could promise exclusivity to one of the retailers in exchange of the profits (or parts of them, depending on the distribution of bargaining power between the two parties) that can be made from the sale of the products. However, once the deal is made, since there is no binding commitment behind the promise, the supplier has an incentive to renege on its exclusivity promise and opportunistically make a deal with another retailer (promising this time that there will be no further deals after that) for, let’s say, half of what was paid by the first retailer.
Of course, the inability to make binding commitments means that the supplier has an incentive to renege also on this promise, and iteratively repeat this scheme with all the retailers. However, anticipating the supplier’s opportunistic behavior, retailers would not accept the offer and thus the supplier loses the market power it might potentially have. A vertical merger, a legally binding exclusive contract or a retail price maintenance agreement allow the upstream supplier to “tie its hands” and retain its market power, to the detriment of consumers that have to face higher prices.
Tirole’s many contributions (see Rey and Tirole 2007, for a relatively recent one) are crucial to understand the framework in which competition authorities like the European Commission and economists more generally analyze vertical restraints and mergers which have vertical dimensions. This can clearly be seen in the guidelines of the European Commission on the assessment of vertical restraints and non-horizontal mergers.
2 The fact that in two-sided platform markets anticompetitive practices may be harder to identify calls for a precise definition of what a two-sided platform is, in order to avoid the case in which firms may simply try to redress the case as “two-sided platform-related” and hope in this way to avoid fines and continue their anticompetitive practices. This, however, might be a topic for another blog post.
The contributions of Tirole to the research in competition policy do not stop to vertical practices and foreclosure. He (with his coauthor Jean-Charles Rochet) kickstarted the literature on two-sided platforms (Rochet and Tirole 2003a, 2006). Two-sided platforms (multi-sided if dealing with more than two customer groups) are intermediaries which allow interactions between different parties who value each other participation (or where at least one party values the other participation, think of the case of advertisers and newspaper readership). Examples of two-sided platforms are credit/debit cards, operating systems, videogame consoles, dating clubs.
Contrary to other types of intermediaries, such as retailers, platforms leave the parties with control over their interactions/transactions. This results in a positive feedback loop between the demands of the two customer groups, in that a member of one group (usually) enjoys the presence on the platform of more members of the other group, representing for him potential possibilities of interactions. Think of payment cards, for example. Payment cards allow consumers to purchase from affiliated stores without the need to physically carry cash with themselves and are characterized by the positive feedback loop typical of two-sided platforms: the more are the merchants accepting card payments, the more are the consumers who value holding a payment card, which in turn implies that the more are the merchants willing to install card readers, and so forth.
The interdependence between the demands of the two customer groups usually tends to result in heavily skewed pricing structures
The interdependence between the demands of the two customer groups usually tends to result in heavily skewed pricing structures, where one side is used as a loss leader to attract the other side, on which the platform makes profits. This means that the two sides cannot be analyzed in isolation but must be considered jointly, otherwise a competition authority may end up wrongly punishing the platform for excessive pricing on one side and predatory pricing on the other (while in fact it is adopting normal pricing practices for its industry, which would be adopted in a competitive environment).2
Tirole did not just develop the theoretical concept of two-sided platforms, but he also studied the idiosyncrasies characterizing some of them, such as interchange fees in credit/debit card markets (e.g. Rochet and Tirole 2003b), which have been the focus of recent antitrust scrutiny by the European Commission (e.g. Visa and Mastercard cases).
3 In a recent article (Lerner and Tirole 2014), Tirole and his coauthor extend the framework developed in this paper to analyze standard essential patents and prove that price commitments prior to the selection of the standard deliver the same outcome as in the (fictitious) ex-ante competitive benchmark in which users need not match their technological choice with each other (meaning that in this fictitious environment a standard setting organization would not confer undue market power to holders of standard essential patents and would de facto not need to exist). However, such price commitments are unlikely to emerge in absence of regulation. The analysis in this paper provides useful insights to policy makers on how to address currently hot policy issues like standard setting organizations and standard essential patents. Recent decisions of the European Commission on standard essential patents involve Google-Motorola and Samsung (Rambus and IPCom are slightly older but still relevant).
Patent pools and cooperation arrangements
In a 2004 paper with Lerner, Tirole studies patent pools – agreements among patent owners to jointly license their patents – and provides simple but robust conditions that would help competition authorities screen anticompetitive patent pools. He first notices that, except for the extreme cases of perfect complements and substitutes, patent substitutability and complementarity depend on the current license fees. When license prices are high, patents tend to be substitutes since licensees may just want to use some of the patents, which thus compete with one another. An increase in the price of a license would then trigger the exclusion of the patent from the selected patent bundle. On the contrary, patents tend to be complements when the prices of licenses are low, with the potential licensee preferring to use all patents conditionally on the adoption of the technology. An increase in the price of a license would therefore not result in the exclusion of the patent from the basket of patents chosen by the licensee, but rather may result in a reduction in the demand for the whole basket. In such a case, he finds that patent pools increase welfare.
Determining whether patents are complements or substitutes may be tricky in practical terms, since usually competition authorities may have limited information on the demand for the patents in the pool. However, he notices that, if licensors are allowed to offer individual licenses in addition to jointly offer their patents in the pool, this fact could be used by competition authorities to screen patent pools that do not pose anticompetitive risks, since licensors would only be tempted to make such stand-alone offers in cases when bundling entails higher prices.
Although this paper focuses on patent pools and is therefore relevant for current discussions of this topic, the analysis is rather general and applies, mutatis mutandis, also to other cooperation arrangements like code-sharing agreements between airlines (examples of these arrangements are alliances such as Oneworld and Star Alliance, that have been subjects of recent antitrust investigations by the European Commission).3 Cooperation agreements are subject to specific European Commission’s guidelines.
Regulation of natural monopolies
One of the main contributions of Jean Tirole is to the field of regulation. Together with Jean-Jacques Laffont, he wrote a very influential paper in regulation (Laffont and Tirole 1986). In this paper he analyzes the regulation of a monopolist in a situation in which the regulator observes the firm’s production cost but the firm has private information on the determinants of its costs: the regulator cannot distinguish between a firm that is more efficient because, for example, it benefits from economies of scale and a firm that has actively invested in the past to make its production more efficient.
In such a situation, they find that a regulator cannot attain the first-best outcome because of the informational asymmetry between the firm and the regulator, but that the regulator can reach a satisfactory second-best outcome by carefully designing a menu of contracts – from which the firm then self-selects its contract by announcing its expected cost – minimizing the rents to the monopolist while still leaving incentives to the firm for cost reductions. The general form of the transfer in this menu of contracts consists of a fixed-price payment plus a linear (partial) cost overrun reimbursement. The contract designed for the most efficient type of firm involves just the fixed-price component (no cost-overrun reimbursement) in order to give the firm (might it happen to be that efficient) high-powered incentives to reduce costs. This is also the only contract of the menu that is able to replicate the first-best outcome.
4 Information rents are rents that the firm earns by exploiting the asymmetry of information existing between it and the regulator.
The contracts designed for less efficient types of firm offer a partial cost overrun reimbursement (and hence less powerful incentives for cost reductions) in order to reduce the information rents4 for the most efficient types. In fact, consider the case in which the regulator designs also the contracts intended for the less efficient types of firm as a fixed-price contract, in order to give them the most powerful incentives for cost reductions. In order for these potential types of firm to pick the contract, the fixed-payment should be larger, since they are less efficient. But then more efficient types of firm would have an incentive to announce a higher expected cost in order to receive a larger payment and earn additional information rents. The menu of contract is therefore optimally designed in order to trade-off the incentives to the firm to reduce its cost with the information rents for more efficient types.
Laffont and Tirole also extend this model to take into account that regulation is not a one-time occurrence but rather a process that changes over time (for instance, Laffont and Tirole 1988, which analyzes the ratchet effect arising when the regulator can only offer a short-term contract to the firm, and Laffont and Tirole 1990, which analyzes how a long-term contract between the firm and the regulator gets renegotiated over time) and the possibility of regulatory capture (Laffont and Tirole 1991). Many of these contributions are summarized in their book on regulation and procurement (Laffont and Tirole 1993).
The contributions of Tirole to research in regulation are not limited to general modelling of regulatory games, but deal also directly with regulation in specific (and very important) industries like communications (Laffont and Tirole 2001) and the financial sector (Dewatripont, Rochet and Tirole 2010).
Republishing and referencing
Bruegel considers itself a public good and takes no institutional standpoint. Anyone is free to republish and/or quote this post without prior consent. Please provide a full reference, clearly stating Bruegel and the relevant author as the source, and include a prominent hyperlink to the original post.