Corresponding author: Svetlana V. Golovanova ( svgolovanova@hse.ru ) © 2020 Non-profit partnership “Voprosy Ekonomiki”.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits to copy and distribute the article for non-commercial purposes, provided that the article is not altered or modified and the original author and source are credited.
Citation:
Avdasheva SB, Golovanova SV (2020) Concerted practice enforcement in Russia: How judicial review shapes the standards of evidence and number of enforcement targets. Russian Journal of Economics 6(3): 239-257. https://doi.org/10.32609/j.ruje.63.51277
|
A judicial review of the infringement decisions of the competition authority substantially affects the standard of evidence in competition enforcement as well as the structure of cases that the competition authority takes. Enforcement against concerted practice in Russia represents a case-study of interaction between commercial courts of first instance, the Highest Court, the competition authority as enforcer, market participants and the legislator to influence the standards of liability under investigation of concerted practice. We examine the judicial review of infringement decisions on concerted practice and track the evolution of legal definition and sufficiency of evidence in such cases. We show, first, that in Russian enforcement, the ability of the Highest Court to influence the criteria of first instance courts is limited (in contrast to the ability of the first instance court to influence the strategy of enforcement by the competition authority). Second, the increase in the burden of proof motivates the competition authority to refrain from an investigation of concerted practice, in accordance with the prediction of the model of the selection of enforcement target by reputation-maximizing authority.
competition enforcement, legal standards, judicial review, tacit collusion, concerted practice, Russia.
Judicial review of decisions is an important part of administrative law enforcement. Judicial decisions influence actual and future enforcement in several ways. First, being appellate instance for administrative decisions and serving as an instrument of error-correction, judges directly reduce the probability of wrongful convictions (
The administrative authority also adopts the course of actions under the influence of the outcomes of judicial review. Adaptation of the authority towards judicial review takes different forms. The authority might make efforts to develop standards of evidence in order to meet the burden of proof established by the judges. Alternatively, the authority may try to persuade judges that standards of evidence applied by the authorities allow to distinguish between legal and illegal conducts. However, the authority may also adapt to the standards established by the courts by simply refraining from investigating the conduct, which requires great efforts to discover and prove the infringement. The effects of judicial review may substantially differ. They are, however, important not only for short-term adaptation of the authority but also for the evolution of evidentiary standards.
Among others, development of evidentiary standards is of special importance for competition law. Concepts of competition law — monopoly power, collusion and dominance — reflect widely accepted economic theory. The description of illegal conduct in competition law over the globe is also more or less harmonized. At the same time, targets and effects of competition or antitrust law enforcement differ a lot. Exploitative conducts are important enforcement targets in EU, but not in US. In Russia, the number of investigations on abuse of dominance exceeds other countries in Global Competition Review Enforcement Rating by more than one hundred times. BRICS countries apply very different approaches to distinguish between per se illegality and rule of reason for potentially illegal conducts (
For a long time, explanations of the difference of the approach to antitrust/ competition enforcement used the “out-of-economics” arguments, including specific legal traditions, constitutional goals (
This article, by contrast, is based on the model that explains the legal standard as an optimal choice of the competition authority facing conditional probability of annulment of an infringement decision under judicial review (
In this article we show another specific way to approach the enforcement target. The competition authority, after revealing the high cost of increasing legal standard, together with the high probability of annulment when using the highest legal standard available, refrains from the enforcement of specific provision. The competition authority’s inability to keep the balance of wrongful convictions and wrongful acquittals under a constraining budget — and citing this as a reason to abandon the enforcement of a particular rule — is not specific to Russia.
The structure of the article is the following. Section 2 notes again the problem of competition enforcement against concerted practice worldwide. Section 3 highlights specific features of competition enforcement (as part of administrative enforcement) in Russia, with implications for testing the prediction of the model of optimal choice of reputation maximizing authority. Section 4 describes the evolution of definition and evidentiary standards of concerted practice as anticompetitive conduct in Russian competition law. Section 5 provides statistical data regarding the judicial review of the decisions on concerted practice vis-à-vis decisions on price-fixing and market-sharing. Section 6 concludes.
Deciding on the illegality of concerted practice as a target of competition enforcement, the legislator and competition authority face a trade-off between (negative) welfare effects of market power enhancing conduct and balance of Type I and Type II legal errors (wrongful convictions and wrongful acquittals). Trying to prevent welfare losses by enforcing antitrust prohibitions on concerted practice, the authority should take the burden of complicated process to prove concerted practice. If the authority refuses to take this burden, it becomes unable to prevent welfare losses, which result from the anticompetitive conduct. This is true for both definitions of concerted practice, relevant for Russian competition enforcement, including “tacit collusion” and “agreements without direct evidence.”
Modern theory suggests that under certain circumstances (few sellers, homogenous product, publicly available information on prices, frequent price changes etc. — market structure known as tight oligopoly) tacit collusion might be sustainable
On the other hand, legal enforcement against tacit collusion is a problem. There is no satisfactory set of evidence sufficient to make a judgment if the practice in question represents tacit collusion or not. In contrast to explicit collusion, tacit collusion does not provide hard evidence. If a legislator decides to consider tacit collusion as illegal, it should make a choice of what kind of proof is sufficient to make judgment on law violation.
In the legal practice, there is no definitive answer to the question of sufficient evidence to prove tacit collusion. There were several attempts to design legal rules on illegal tacit collusion. One of them was to expand the meaning of agreement (illegal under Sherman law) on close communication between parties that have the intent to increase prices by refusal to compete. This option was close to that historically important US Supreme Court decision on the American Tobacco (1946) case. However, only a few years later, a decision on Theatre Enterprises (1951) took the completely opposite path, considering conscious parallelism as insufficient evidence of agreement. In legal literature, discussion between Turner and Posner reflects two opposite views on the very possibility of capturing tacit collusion within the framework of a particular legal concept (Wagner-
EU competition law applies an alternative approach by defining concerted practice as conduct that reflects most features of tacit collusion. Article 101 of the Treaty on the Functioning of the European Union (TFEU) prohibits concerted practice together with agreements that restrict competition. For the first time in Europe, a decision on concerted practices was made in 1969 against dyestuff suppliers — the Dyestuff Case (
The most notorious concerted actions case in the practice of the European Commission’s Competition Directorate was the Wood Pulp Case (1984) when antitrust charges were brought against bleached pulp producers. The exchange of information on prices was the key component of the sellers’ actions: they announced price increases on a quarterly basis at the same time following, in price growth rates, the market leader, i.e. the largest supplier. Price announcements on the bleached pulp market established the upper limit, and then thereafter individual negotiations started to reduce the price
A recent case whereby the competition authority applied a concept of concerted practice concerned an investigation into international container shipping companies where practice of “general rate increase (GRI)” is a reason for concern about refusal from competition in favor of tacit collusion. However, since investigations in 2016 resulted in settlement instead of an infringement decision, it is impossible to assess whether there is an evidence of anticompetitive effect of GRIs, or the evidence on GRIs themselves is sufficient to render a conclusion on concerted practice, or in the particular case competition concern is enough to elaborate remedies as a part of the settlement.
To summarize, European competition enforcement does not provide clear guidance on the appropriate legal definition of concerted practice. EU experience of enforcement towards concerted practice is contradictory. That is why this path of competition law development requires further analysis and assessment.
However, the issue of concerted practice is even more complicated because the enforcer can interpret this illegal conduct in a slightly different way: not as tacit collusion, but as agreement that should be prohibited without direct evidence. In 2006, OECD Roundtable “Prosecuting cartels without direct evidence”
Model of reputation-maximizing authority by
One of the predictions of the model is that if the court expects a very high legal standard in a particular group of decisions, the reputation-maximizing authority refrains from taking these decisions for investigation. The competition authority might make investigations under a lower standard than demanded by courts and refrain to comply when the court moves to a higher and more befitting legal standard. In this paper, we illustrate this prediction using the example of enforcement against concerted practice in Russia. To do that, we firstly track the history of enforcement informally and, secondly, analyze the changes in legal standards applied in the infringement decisions statistically using specially developed metrics. The metrics (
Commercial courts in Russia play an important role in the administrative adjudication. The rules of decision-making in commercial courts responsible for deciding on claims to annul administrative decisions provide the courts with significantly decisive power, combined with easy access of the parties to the courts. In the four-stage court system (first instance court, appellate court, cassation instance, and the Highest Court) only the Highest Court has the discretion of whether to take the case or not.
The most important characteristic of the system of claims to annul administrative decisions in Russia is that the burden of proof is imposed completely on the authority. In order to annul the infringement decision, the convicted party only has to persuade the judge(s) that the infringement decision is not sufficiently substantiated and supported by legal facts
Another reason that makes the Russian context suitable for illustrating the predictions of the model of reputation-maximizing authority is that the FAS is extremely strongly motivated on the number of decisions not annulled by the courts. Enforcement success assumes a more important place in the performance assessment in Russia than in other countries (
Russia is among the countries that borrowed the European concerted practice concept as a specific type of illegal conduct. The Russian experience contributes to the assessment of viability of this approach for several reasons. First, in Russia, competition enforcement is large-scale. Competition authority is organized as a network of regional offices (regional subdivisions, about 80 in total). Every regional subdivision makes decisions on the results of investigations in regional markets. The responsibilities of regional subdivisions are similar to those of the competition authorities in the Member States in the EU. The substantial difference is that competition authorities in the Member States act under the national competition legislation, while regional subdivisions of the FAS decide under the one law — Federal Law “On the protection of competition.” Central Office of the FAS obtains the responsibility of antitrust enforcement over the markets in geographical boundaries of the Russian Federation. At the same time, Central Office has the right of legislative initiative. Legislative initiative of the FAS necessarily depends on the outcomes of enforcement.
In this context, the development of legal definition and legal standards applied by the courts, as well as the strategy of the FAS towards concerted practice as enforcement target, is divided into two periods: from 2006 to 2011 and from 2012 onward. The borderline between the two periods was marked by amendments to competition law entered into force in 2012. These were adopted following pressure from the business and expert community.
Optimal legal standard to decide on the infringement like concerted practice is close to full effect-based (rule of reason). As
The first Russian competition law “On Competition and Limitation of Monopolistic Activity in Commodity Markets” was introduced in 1991. This law, attempting to harmonize competition rules with the EU approach, mentioned the “concerted practices” as a particular type of law violation, but did not include any explicit definition of concerted practice. In 2006, a new Russian Competition law (Law “On protection of competition,” Law-2006 hereafter) was introduced. The definition of the concerted practice was given in the Article 8:
“Concerted practices are actions of several undertakings in the market under absence of the agreement, which meet the following requirements:
The first part of the definition contains a short explanation of a grim trigger strategy except for the words “only under the condition” and “known in advance to each of them,” which explicitly refer to the communication between the parties under collusion. The second part contains alternative explanations of the parallel behavior that the competition authority should exclude in order to render a conclusion on concerted practice.
Highest Court (at that time — The Supreme Commercial Court of the Russian Federation) stresses the difference of concerted practice from the agreements, which are to be documented with hard evidence. On 30 June 2008, the Plenum of the Supreme Commercial Court of the Russian Federation in a resolution “On some issues arising in connection with application by commercial courts of the antimonopoly legislation” states: “Commercial courts should take into account that actions might be concerted in the absence of documentary confirmation of presence of agreement to commit them. The conclusion about the presence of one of the conditions to be established for the recognition of action to be concerted, namely: that each of the undertakings knew about committing such actions in advance — can be done based on the actual circumstances of their commit. For example, among other circumstances, the fact that actions are made by different market participants relatively uniformly and synchronously in the absence of objective reasons may indicate the presence of concerted practices.”
Clarification enables enforcers to look merely for parallelism and communication to prove concerted practices. The Supreme Court did not answer whether it is necessary to look at concerted practice as tacit collusion (first approach mentioned above) or infringement agreement decided without direct evidence, only using indirect proof.
The following examples illustrate the legal standards taken by Russian competition authorities.
The infringement decision against Rosneft–Stavropol and Lukoil–Yugnefteproduct (regional subsidiaries of two of four largest domestic oil companies)
The competition authority of St. Petersburg accused the largest grocery retailers in the city of concerted practices in the retail market of buckwheat during August-September 2010
The convicted companies tried to explore the following arguments. The period under investigation is too short, and price movements during the period are irregular and cannot be predicted by the collusion hypothesis.
In general, legal standards applied under court litigation on concerted practice infringement decisions may be assessed as low. Parallel pricing (and more precisely, price increase) was the most important evidence. Courts were satisfied with the evidence of parallel pricing during a very short period of time. There was no evidence on “communications” among market participants involved. Analysis of market structure (including facilitating practices etc.) was very basic. Though alternative explanations of coincidence in pricing were presented, their explanatory power was never tested. The discriminatory power of concerted practice hypothesis was also not tested.
However, in spite of non-annulments (or rare annulments) of infringement decisions on concerted practice during 2008–2010, litigations influence the perception of proper legal standards by judges. The very fact that opposite party (defending party in the first instance, competition authority) cannot reject the importance of contextual features of the affected market was an important reason for the courts to ask for theories of harm based on the analysis of market structure. Over time, judges expect more and more contextual analysis in spite of the fact that Clarifications of the Supreme Commercial Court do not require it.
In December 2011, Law-2006 and related regulations on penalties were amended by several rules. The most important changes for our analysis was the division of law violations into two groups: those that limit competition and those that only harm consumers or counterparty.
The definition of concerted practice was changed as well. First, the definition was amended by the clarification that “actions are known in advance to each of the participating economic entities because of the public announcement by one of them to commit such actions.” Importance of communication to sustain collusion, explicit or tacit, is one of the crucial take-outs of theoretical research models (
Second, by the same amendment, concerted practices are completely separated from agreements and contained in Article 11.1. Third, the legislator introduces a kind of De Minimis notice: a hypothesis of concerted practice is excluded as the joint share of market participants does not exceed twenty percent, and the share of each of the participants does not exceed eight percent. Changes and amendments in the Law-2006 reflect non-satisfaction of competition authority in the discriminatory power as an ability to distinguish legal and illegal actions that rules introduce.
After amendments, most investigations and decisions on concerted practice target relatively large market participants, the interaction of which goes beyond purely tacit collusion.
MTS and Vympelkom, which are the largest telecom operators in Russia, were accused in the concerted practice in the retail markets of IPhones (IPhone 4.16 GB and IPhone 4.32 GB). Evidence in favor of the conclusion on concerted practice consists of two parts: (1) identical retail prices on selected days during the period under investigation (from September 2010 to April 2011) and (2) the interview given by a manager of Vimpelkom to the newspaper in which he confirms that “The company will set the prices on IPhones… exactly the same as MTS.” Companies did not oppose the conclusion on concerted practice and reached a commitment decision with competition authority.
In 2014, the Russian competition authority, following the similar action of European Commission, opened an investigation against international liner container shipping companies, including Müller-Maersk, Evergreen Marine, Hyundai Merchant Marine, CMA CGM, Orient Overseas Container Line Limited, and nine more. This investigation centered on pre-announcements of pricing on shipping services in the form of a General Rate Increase (GRI). GRI was considered to be a tool of communication between competitors that supported tacit collusion. After the FAS adopted the infringement decision, the convicted companies submitted claims to annul it.
To conclude, investigations and decisions after 2011 have demonstrated that the competition authority has avoided serious errors in the motivation of infringement decisions. Contextual analysis of the market is normally present, as well as an explanation of strategic pricing in a particular case. However, the discriminatory power of the criteria applied is still weak. Applying more and more economic analysis, the FAS still seldom persuades judges about the sufficiency of arguments in favor of a conclusion on concerted practice.
In the next section we show how all these developments in commercial courts influence the choice of competition authorities on concerted practice as enforcement target.
In this section we want to explain the decision of the FAS to take concerted practice cases vis-à-vis other horizontal agreements cases, and legal standards applied under investigation of law violation, in the framework of the choice of legal standards by reputation-maximizing authority (
Judicial review of concerted practice cases vis-à-vis horizontal agreements illegal per se, 2008–2015.
2008 | 2009 | 2010 | 2011 | 2012 | 2013 | 2014 | 2015 | ||
Claims to annul | Horizontal per se | No | 28 | 75 | 23 | 28 | 37 | 87 | 56 |
Concerted practice | 21 | 46 | 31 | 31 | 12 | 0 | 1 | 1 | |
Number of instances that consider a case, average (st. dev. in parentheses) | Horizontal per se | No | 3.6 (1.9) | 2.5 (1.2) | 3.0 (1.3) | 3.6 (1.6) | 3.3 (1.8) | 2.6 (1.1) | 2.4 (1.0) |
Concerted practice | 3.2 (2.1) | 3.1 (1.5) | 3.5 (1.4) | 3.0 (1.4) | 2.8 (1.3) | – | 4 | 4 | |
Number of days to consider a case, average (st. dev. in parentheses) | Horizontal per se | No | 384 (291) | 302 (145) | 428 (290) | 564 (304) | 472 (308) | 311 (143) | 270 (139) |
Concerted practice | 530 (353) | 365 (206) | 417 (184) | 460 (261) | 442 (276) | – | 691 | 454 | |
Share of the decisions submitted to the Highest Court for supervisory review* | Horizontal per se | No | 0.29 | 0.19 | 0.30 | 0.46 | 0.14 | 0.05 | 0.02 |
Concerted practice | 0.29 | 0.24 | 0.39 | 0.29 | 0.42 | – | 1.00 | 0.00 | |
Share of the decisions annulled in the first instance concerted practice | Horizontal per se | No | 0.71 | 0.61 | 0.57 | 0.57 | 0.59 | 0.40 | 0.20 |
Concerted practice | 0.19 | 0.37 | 0.61 | 0.35 | 0.67 | – | 0.00 | 1.00 | |
Share of the decisions finally annulled | Horizontal per se | No | 0.68 | 0.60 | 0.57 | 0.43 | 0.54 | 0.39 | 0.25 |
Concerted practice | 0.14 | 0.33 | 0.58 | 0.35 | 0.67 | – | 0.00 | 1.00 |
Using horizontal agreements illegal per se as a benchmark we understand that FAS decisions for this type of conduct are not error free (see
While we present data for the whole period, an informative comparison of the data for two types of anticompetitive conducts is possible only until 2011 (for annulments — until 2013, taking into account the timing of judicial review), when the number of decisions on concerted practice dramatically drops. During this period, data on the analysis undertaken by competition authorities and correspondence between the type of analysis and legal standard allow us to derive conclusions on FAS approach.
Before 2011 the number of decisions on concerted practice is relatively large — as large as the number of decisions on price-fixing, market-sharing and bid rigging. The number of instances that measures the costs of litigation for the parties is not significantly higher for concerted practice, in contrast to horizontal agreements illegal per se. As for litigation in Russia in general (
At the same time, convicted parties make significant efforts to persuade the judges to apply higher legal standards than Clarification of the Supreme Commercial Court implies. An important indicator is the share of the decisions, which were submitted for the Highest Court for supervisory review. The share of decisions is extremely high both for horizontal agreements illegal per se and concerted practice. However, the targets for challenging the evidence are different in these cases. For the agreements illegal per se the target for critique is the reliability of hard evidence. For concerted practice decisions, the target for critique is low legal standards. Table
Blocks of economic analysis and annulment rates before and after the changes in Law-2006
2008–2010 | 2011–2013 | ||||||||
Number of instances | Share of finally annulled decisions | Obs. | Number of instances | Share of finally annulled decisions | Obs. | ||||
Only nature of the conduct | Horizontal per se | 2.76 | 0.79 | 34 | 2.77 | 0.62* | 13 | ||
Concerted | 3.00 | 0.67 | 12 | 1.75 | 1.00* | 4 | |||
Market boundaries and market structure | Horizontal per se | 3.33 | 0.66 | 3 | 3.29 | 0.71* | 7 | ||
Concerted | 3.44 | 0.44 | 16 | 3.11 | 0.32* | 19 | |||
Market power enhancing effect | Horizontal per se | 3.71 | 0.59** | 17 | 4.56*** | 0.56 | 9 | ||
Concerted | 3.54 | 0.32** | 56 | 3.00*** | 0.44 | 16 | |||
Theory of harm to consumers | Horizontal per se | 2.25 | 1.00 | 4 | 4.67 | 1 | 1 | ||
Concerted | No | No | No | No | No | No | |||
Total | Horizontal per se | 3.03* | 0.74*** | 58 | 3.56 | 0.66* | 32 | ||
Concerted | 3.44* | 0.39*** | 84 | 2.92 | 0.44* | 39 |
Still, during the first period, the annulment rate in concerted practice decisions is, on average, lower than for horizontal agreements illegal per se. As a result, the number of decisions on concerted practice is high enough.
In 2011–2012, the environment surrounding the investigation, and decisions, on concerted practice, changed under the influence of at least three factors. First were changes in Law-2006, which stressed the importance of the effect on competition as a reason for infringement decision. Second were amendments of Law-2006 towards concerted practice, which highlighted the impact of an explicit announcement of further strategy by either party. After these amendments, the causal relationship between announcement and presumed concerted practice became a necessary part of the evidence. Third was accumulated experience of the discussion of the evidence of market power enhancing effects in first instance, appellate and cassation courts. As a result, legal standards expected by courts increased. Table
In terms of modeling the choice of legal standard by reputation-maximizing authority, these changes can be interpreted as two shocks: one is the change of legal standards expected by courts, and another is the increased cost of applying legal standard. It is worth noting that the increase of legal standards applied by commercial courts took place not due to the Clarification of Highest Court, but instead in spite of the content of the special Clarification. During a relatively short period, Russian commercial courts overcame the effect of initially imposed legal standards, which contradicted economic logic and internationally recognized best practice.
For comparative statics, the model predicts either an increase of legal standards applied by competition authorities or a decrease of the number of enforcement targets. Statistics on the number of decisions on concerted practice (Tables
2008 | 2009 | 2010 | 2011 | 2012 | 2013 | 2014 | 2015 | 2016 | 2017 | 2018 | |
Number of investigations | n.a. | n.a. | n.a. | n.a. | 31 | 11 | 8 | 5 | 3 | 5 | 2 |
Number of infringement decisions | n.a. | n.a. | n.a. | n.a. | 23 | 2 | 4 | 4 | 0 | 0 | 0 |
Number of precautions issued by the FAS | n.a. | n.a. | n.a. | n.a. | n.a. | n.a. | 5 | 2 | 4 | 1 | 19 |
Number of appealed infringement decisions | 21 | 46 | 31 | 31 | 12 | 0 | 1 | 1 | 0 | 0 | 0 |
At the same time, Russian competition authorities did not forget about the threat of concerted practice as tacit collusion and/or agreement to be prosecuted without direct evidence. An important observation is the increase in the number of precautions
Overall, since 2012, the approach of the Russian competition authority towards concerted practice does not substantially change. It is still considered as tacit collusion, or implicit agreement, to be proved by the analysis of strategic interaction between companies. There is little or no progress in deciding what evidence meets the burden of proof. The FAS concentrates on specific remedies (for example, the ban on price announcements or price forecasts by sellers or authorities). However, it is not clear if the evidence of forecasts together with contextual market analysis and parallel pricing are sufficient for infringement decision.
Enforcement against concerted practice illustrates the specific path of enforcement development in recent competition jurisdiction, when the increase of legal standards expected by courts, and followed by the competition authority, results in the decrease of the number of enforcement targets instead of the increase of the quality of enforcement. It revealed that, under higher legal standards, the cost of investigation exceeds the effects of enforcement perceived by the competition authority.
Even in the infringement decisions against motor fuel suppliers (2008–2009), which obtained highest political and public opinion support, competition authorities hardly succeeded in persuading judges that infringement decisions were properly proven. Changes in competition law, which identify the concerted practice as a violation of competition law only because of the restriction of competition, makes the proof of market power enhancing effect necessary part of the infringement decisions.
Dissatisfaction with the outcome of judicial review results, first, in the attempts to narrow legal definition, and, second, in the sharp decrease in the number of investigations and decisions towards participants of concerted practice. The introduction of “public announcement” as a necessary evidence of concerted practice in order to avoid legal errors does not help to collect proofs for power-enhancing effect of conduct.
Statistics of the enforcement supports prediction of the model of legal standards in particular cases (group of cases) as a decision of reputation-maximizing authority.
From a legal point of view, Russian experiences demonstrate the difficulty, first, in finding a legal definition of concerted practice as tacit collusion, and, second, in obtaining sufficient evidence to prove tacit collusion.
The evolution of legal definition and standards of evidence for concerted practices in Russia highlights the important role of judicial review in development of legislation and enforcement practice, and, in particular, the importance of first instance courts. Especially interesting is the role of lower (first instance and appellate) courts. Under the enforcement of concerted practice, decisions of lower courts overcome relatively low legal standard established by the decision of the Supreme Commercial Court.
This work was supported by the Basic Research Program at the HSE University. The authors cordially thank Yannis Katsoulacos, Joseph Harrington and all participants of the International Symposium on Imperfect Forms of Collusion (Stellenbosch, 2018) for extensive discussions and fruitful comments. The authors are grateful to the anonymous reviewer for helpful comments and suggestions.