1
Introduction
1.1 Defining Collusion
Collusion is when firms in a market coordinate their behavior for the purpose of producing a supracompetitive outcome. A supracompetitive outcome is one in which price exceeds the price that would have occurred without the coordination among firms. In the economic theory of collusion, coordination is with respect to the strategies that firms use. A firm’s strategy prescribes its behavior (e.g., what price to set, how much to produce), and that behavior can be contingent on what has transpired in the market (e.g., the prices that firms recently charged) as well as on current market conditions (e.g., a firm’s cost and the strength of market demand). In laymen’s terms, the definition of collusion put forth by economic theory is:
Collusion is when firms use history-dependent strategies to sustain supracompetitive outcomes through a reward-punishment scheme that rewards a firm for abiding by the supracompetitive outcome and punishes it for departing from it.
If a firm abides by the collusive outcome—which could involve high prices, exclusive territories, customer allocation, and so forth—then it is rewarded in the future by rival firms continuing to abide by the collusive outcome (e.g., persisting with high prices); while if it departs from the collusive outcome (e.g., setting a low price, selling above its quota, serving another firm’s customers) then it is punished in the future by rival firms acting aggressively to reduce the deviating firm’s profits (e.g., lowering prices, selling to the deviating firm’s customers). Collusion involves an implicit or explicit understanding among firms that ties future rewards and punishments to current behavior and, by doing so, is able to induce compliance with regard to the supracompetitive outcome. This understanding can be viewed as contractual, though the penalties for acting contrary to the terms of the contract take the form of rival firms’ future (disciplining) behavior. For this arrangement to be effective, it must be self-enforcing, which means that each firm finds it in its best interest to abide by the arrangement as long as all other firms are expected to do so.
A collusive strategy comprises three fundamental elements. First, it describes the collusive outcome, such as what common price is to be charged or which firm is to serve which geographic area (when the scheme involves exclusive territories). Second, it describes the monitoring protocol; that is, how firms will monitor one another for compliance with the collusive outcome. If prices are observable and the collusive outcome is to set a common high price, then monitoring could be in terms of past prices. However, if the firms supply an intermediate good to industrial buyers (e.g., cement suppliers selling to construction companies) where price can be privately negotiated between a seller and a buyer, then price monitoring will not work. In that case, the supracompetitive outcome could involve an allocation of sales quotas (along with a common price to be set), with monitoring taking the form of comparing realized sales to those sales quotas. Third, the collusive strategy describes the punishment that occurs when there is evidence of noncompliance. Among other possibilities, the punishment could be a temporary or permanent reversion to pricing competitively, or it could be a focused price war that has firms charge low prices for the customers of the firm that apparently deviated.
Let us now state more formally, in the jargon of game theory, the meaning of “firms in a market coordinate their behavior for the purpose of producing a supracompetitive outcome.” As the competitive benchmark is typically defined to be a Nash equilibrium outcome for some static oligopoly game, a supracompetitive outcome involves higher prices than for a static Nash equilibrium. Coordination on a collusive strategy in a game-theoretic framework is taken to be a subgame perfect (or sequential) equilibrium in a repeated game (where the stage game is the original static oligopoly game), which produces supracompetitive outcomes through the use of history-dependent strategies. More specifically, an outcome with higher prices and profits is sustained by the threat that noncompliance with that outcome (or evidence consistent with noncompliance) is punished with a low continuation payoff, such as a temporary or permanent reversion to a stage game Nash equilibrium, or a finite number of periods with low prices and subsequent return to prices exceeding static Nash equilibrium prices.1
Thus far we have defined collusion. However, the focus of this survey is not broadly on collusion but rather on collusion that runs afoul of competition law, which leads us to ask: What is unlawful collusion? The answer depends on the particular law, the manner in which the law is interpreted (typically, by the courts), and the evidence that is required for determining when the law has been violated. As liability and evidentiary standards vary across time and space, I will attempt to provide a broadly applicable (but not universal) definition of unlawful collusion.
Competition law as it pertains to prohibiting collusion is generally dated from the Sherman Act in the United States in 1890 (though, in fact, Canada preempted the United States by instituting its competition law in 1889). Section 1 prohibits contracts, combinations, and conspiracies that unreasonably restrain trade.2 Subsequent judicial rulings have effectively replaced the reference to “contracts, combinations, and conspiracies” with the concept of “agreement.” It is now understood that firms are in violation of Section 1 when there is an agreement among competitors to limit competition. Though the term “agreement,” which is now so integral to defining liability, does not appear in the Sherman Act, many jurisdictions that arrived later to the enforcement game have put the term into their competition law. For example, in the European Union, Article 101 (1) of the TFEU (1999) states: “The following shall be prohibited: all agreements between undertakings, decisions by associations of undertakings and concerted practices which … have as their object or effect the prevention, restriction or distortion of competition.”
Our question “What is unlawful collusion?” has then become: “What is an agreement to limit competition?” Key judicial decisions by the U.S. Supreme Court have resulted in the interpretation that an agreement resides in a mutual understanding among firms to constrain competition. There is an agreement when firms have a “unity of purpose or a common design and understanding, or a meeting of minds”3 or “a conscious commitment to a common scheme designed to achieve an unlawful objective.”4 This perspective has been echoed by the European Union’s General Court, which has defined an agreement as or as requiring “joint intention”5 or a “concurrence of wills.”6
Reference to “meeting of minds,” “conscious commitment to a common scheme,” and “concurrence of wills” all focus on the same condition: There is a common understanding among firms that they will restrict competition in some fashion. The parallel between conditions on firms’ mutual beliefs and the notion of equilibrium has been noted in Yao and DeSanti (1993), Werden (2004), and Kaplow (2013):
Analysis of one-shot games provides the clear definition of self-interest necessary to allow evidence of action against self-interest to play a useful role in inferring the existence of an agreement. If there is a unique Nash, non-cooperative equilibrium to a particular game, as there is in conventional one-shot game oligopoly models, it follows that there is a unique action each player will take if [it] does not coordinate its actions with its rivals. These equilibrium actions are consistent with self-interest, and any other actions are not.… The existence of an agreement can be inferred from actions inconsistent with Nash, non-cooperative equilibrium in a one-shot game oligopoly model, even though they are consistent with Nash, non-cooperative equilibrium in an infinitely-repeated oligopoly game.7
Though these statements seem to make clear that mutual understanding to constrain competition is unlawful, the U.S. Supreme Court has been equally clear that more is required to draw such a conclusion. In particular, mutual understanding obtained through “shared economic interests” without some overt effort on the part of firms to create that mutual understanding is not in violation of Section 1 of the Sherman Act.
Courts have noted that the Sherman Act prohibits agreements, and they have almost uniformly held, at least in the pricing area, that such individual pricing decisions (even when each firm rests its own decisions upon its belief that competitors do the same) do not constitute an unlawful agreement under section 1 of the Sherman Act … [T]hat is not because such pricing is desirable (it is not), but because it is close to impossible to devise a judicially enforceable remedy for “interdependent” pricing. How does one order a firm to set its prices without regard to the likely reactions of its competitors?8
The court has gone on to require that firms have engaged in some expression of intent that results in reliance among themselves to coordinate to reduce competition. An agreement is not just a mutual understanding among firms to engage in coordinated suppression of competition but also an expression of that mutual understanding through the process of its creation:
By operationalizing the idea of an agreement, antitrust law clarified that the idea of an agreement describes a process that firms engage in, not merely the outcome that they reach. Not every parallel pricing outcome constitutes an agreement because not every such outcome was reached through the process to which the law objects: a negotiation that concludes when the firms convey mutual assurances that the understanding they reached will be carried out.9
In practice, firms must not only have mutual beliefs to restrain competition but those beliefs must have been achieved through some form of communication.
This legal approach has led to the recognition of three categories of collusion: (1) explicit collusion, (2) tacit collusion, and (3) conscious parallelism. If, for example, firms (or, more to the point, their representatives) reach a mutual understanding by speaking to each other regarding a plan to raise prices, then they have engaged in explicit collusion. Explicit collusion involves “easily observable proof that the defendants have exchanged assurances that they will pursue a common course of action.”10 In contrast, if, for example, a firm announces its plan to raise prices at an industry gathering, which, without any further communication, leads to mutual beliefs that all will raise their prices (as evidenced by subsequent behavior), then they have engaged in tacit collusion. Tacit collusion does not entail an express exchange of assurances but does require firms communicate “their intent to raise prices and their reliance on one another to do the same.”11 And if, for example, a firm raises its price and other firms match that price and there is no communication, that is an example of conscious parallelism, which is a process “not in itself unlawful, by which firms in a concentrated market might in effect share monopoly power, setting their prices at a profit-maximizing, supracompetitive level by recognizing their shared economic interests.”12
It should be clear from the preceding discussion that economic (or, equivalently, game-theoretic) collusion and unlawful collusion are distinct objects. Economic collusion focuses on firms’ outcomes and how those outcomes are supported by mutual beliefs with regard to firms’ strategies. In short, economic collusion focuses on the equilibrium. In contrast, unlawful collusion is also concerned with how the collusive equilibrium came about and thus with the equilibrating process. Economic collusion can be lawful. Putting aside excessive pricing laws,13 it is lawful to charge supracompetitive prices and to enforce such prices through the implicit threat of a punishment, so long as the threat did not involve any communication. At the same time, unlawful collusion need not be economic collusion. Communicating to coordinate a move from a static Nash equilibrium to a less competitive static Nash equilibrium is unlawful even though it is not a repeated game equilibrium.14
Terminological Tangent
Before leaving this topic, let me discuss an abuse of terminology routinely perpetrated by industrial organization economists. The equilibrium approach says: If firms have mutual beliefs regarding some collusive strategy profile then, under these conditions, that strategy profile is stable (in the sense of equilibrium) and thus collusion can persist. That approach is predicated on the presumption of mutual understanding and thus has nothing to say about how mutual understanding was achieved, which is exactly the focus of the law. Because the process by which an equilibrium is reached is not part of (almost) all economic theories of collusion, they cannot then encompass the legal and practical distinction between explicit and tacit collusion that focuses on the process by which firms move from a stage game Nash equilibrium to a collusive equilibrium for a repeated game. Equilibrium theories have nothing to say about the disequilibrium process and, on those grounds, it is not meaningful to refer to an equilibrium theory of collusion as either tacit or explicit, for it is mute on the matter.15
Therefore, regardless of the long history of oligopoly theorists referring to their theories as “tacit collusion” (for which I admit mea culpa), such a reference is misleading and unnecessary. It is misleading, because it does not deal with how firms achieved an equilibrium, which is the defining distinction between explicit and tacit collusion. Classifying a theory of collusion as “tacit” will either confuse legal scholars—as they wonder how the model is one of tacit collusion rather than explicit collusion, given that there is no discussion of communication—or worse yet, they will be misled into thinking that the theory is actually pertinent to understanding collusive behavior in the absence of communication.16 In addition, dropping the term “tacit” will not lose any substance for economists, because it has no formal meaning in a game-theoretic context. There will then will be no confusion among economists if the term “collusion” is used instead of “tacit collusion.” For these reasons, I propose that we dispense with the term “tacit” when referring to equilibrium theories of collusion.17
1.2 Overview of Book
In this book, I review theoretical research that addresses two broad questions: (1) What is the impact of competition law and enforcement on whether firms collude, how long they collude, and how much they collude? and (2) What is the optimal design of competition law and enforcement? Chapter 3 examines the first question. The second question is addressed in chapter 4 though, along the way, I will cover some research pertinent to the first question. As a starting point, chapter 2 discusses some general issues with regard to taking into account competition law and enforcement in models of collusion. In the context of these two questions, the goal is to describe some canonical models, review some key results and insights, and suggest future lines of inquiry.
Let me conclude this chapter by discussing what is not covered in the book. The focus is on theory, and thus I do not cover experimental and empirical research.18 In the domain of the theory of collusion, most of the models will have perfect monitoring, because there is very little research that encompasses competition law and enforcement in a model of collusion with imperfect monitoring. While many collusive practices can only be understood in the context of imperfect monitoring, much of the insight delivered thus far regarding the impact of competition law and enforcement is not tied to that feature.
Some of the theoretical literature on collusion and competition policy is also not covered in this survey. It is nearly universal to model the decision-making entity as a profit maximizer and to assume it is the firm that colludes and is penalized. In practice, it is the employees of the firms who collude, and they can be motivated by factors other than profit and be subject to penalties distinct from those levied on the firm. There is some research that encompasses agency issues that I do not cover.19 This is a critical area for future research. There is also a small literature examining the implications and design of competition policy in the context of auctions. To keep the discussion focused, I limit my attention to product markets but make reference to some relevant research on collusion and competition policy with regard to auctions. Finally, there are some competition policy issues dealing with vertical restraints and collusion which are not reviewed here.20
Notes
1. For coverage of the general theory of collusion in the repeated game framework, some useful references are Tirole (1988), Vives (1999), and Motta (2004).
2. “Every contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce among the several States, or with foreign nations, is declared to be illegal.”
3. American Tobacco Co. v. United States, 328 U.S. 781 (1946).
4. Monsanto Co. v. Spray-Rite Serv., 465 U.S. 752 (1984).
5. Judgment of the Court of July 15, 1970. ACF Chemiefarma NV v. Commission of the European Communities, Case 41–69.
6. Judgment of the Court of First Instance of October 26, 2000. Bayer AG v. Commission of the European Communities.
7. Werden (2004, 770, 779)
8. Clamp-All Corp. v. Cast Iron Soil Pipe, 851 F.2d at 484 (1st Cir. 1988).
9. Baker (1993, 179)
10. Kovacic (1993, 19)
11. Page (2009, 451)
12. Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209 (1993).
13. For a discussion of excessive pricing laws, see Evans and Padilla (2005).
14. For some discussion of this last point, see Harrington (2013b).
15. Some collusive theories have communication as part of an equilibrium (which are briefly discussed in section 3.3.3) and, depending on what is conveyed, that could be sufficient for illegality. Here, I am referring to communication to reach an equilibrium, which is the primary (though not exclusive) focus of the law.
16. Equilibrium theories are not theories of how firms can collude without communication. It is a non sequitur to say that assuming mutual beliefs in a game without communication shows how firms can achieve mutual beliefs without communication (and thus how firms can collude without communication).
17. If an economist thinks there is substance to referring to collusion as “tacit,” then what is an example of “explicit collusion” in economic theory? Some might say it is when collusion can be enforced through binding contracts, but that is an irrelevant category that runs contrary to law and practice. Collusive arrangements have never been enforceable in most countries and, to my knowledge, currently are enforceable in none.
18. For a review of some experimental work, see Choi and Gerlach (2015); for empirical work, see Levenstein and Suslow (2006, 2015).
19. Papers of which I am aware are Aubert, Rey, and Kovacic (2006); Angelucci and Han (2011); and Thêpot and Thêpot (2016).
20. For example, Jullien and Rey (2007).