Journal of Antitrust Enforcement

Filho, Arthur Guerra

doi: 10.1093/jaenfo/jnaa043pmid: N/A

Abstract In 2017, Brazil’s president faced impeachment proposals for allegedly receiving money from a businessman in exchange for influencing a decision of the Conselho Administrativo de Defesa Econômica (Cade), the Brazilian competition authority. But, considering that Cade is an independent agency with a reputation for transparency, how may politicians influence antitrust enforcement? By analysing the responses given to a questionnaire of mine by anonymous former Cade commissioners, as well as other reports related to political influence on the agency and records from Congress, Cade, and other institutions, this article will examine the channels of political influence on Cade. First, I examine channels of democratic control, such as the appointment process. Secondly, by discussing those situations in which politicians approach the agency, I observe that, even in the absence of an explicit promise or threat, contact from politicians bears a special weight for Cade’s officers. I argue that, regardless of conclusive evidence that political contact or pressure has actually impacted Cade’s enforcement, public confidence in the agency might be harmed when there is the appearance that it has done so. Therefore, the appropriateness of increased transparency surrounding meetings between elected politicians and Cade’s officers is a matter for consideration. I. INTRODUCTION The year was 1971 and President Richard Nixon was at the White House speaking with Charles Colson, his special counsel, about the negative coverage that the main television broadcasters in the USA (ABC, NBC and CBS) were giving to his administration. During the conversation, Nixon argues that his administration should threaten the broadcasters with an antitrust suit to make the coverage more favourable: If the threat of screwing them is going to help us more with their programming than doing it, then keep the threat … Our gain is more important than the economic gain. We don’t give a goddam about the economic gain. Our game here is solely political … As far as screwing them is concerned, I'm very glad to do it.1 That same year, President Nixon allegedly attempted to settle an antitrust case involving the International Telephone & Telegraph Corporation (ITT) in exchange for campaign finance contributions.2 But Nixon was not the only American president who reportedly attempted to use antitrust enforcement as a bargaining chip to gain resources for political competition: President Lyndon Johnson, for instance, allegedly promised the Department of Justice (DOJ) clearance to a merger in exchange for a newspaper’s political support.3 Attempts to use antitrust enforcement for political purposes are not exclusive to American antitrust history. Such attempts have been made in Europe4 and in various other countries around the world.5 Brazil is no exception, as will be seen throughout this article. This article will discuss how politicians may influence enforcement of antitrust legislation by Cade, the Brazilian competition authority, which is an independent government agency.6 The article is structured as follows. Section II addresses the ‘democratic controls’ on Cade. In order to influence antitrust enforcement, elected politicians do not necessarily need to have direct contact with or exert pressure on Cade’s officers—the Presidency of the Republic and Congress have channels through which to influence antitrust enforcement because Cade’s powers are under democratic control. Subsequently, Section III explores situations in which elected politicians do approach Cade’s officers. I will argue that contact from politicians bears a special weight for Cade’s officers even if the politician does not make any explicit threat or promise. Moreover, there is a discussion in this section of the limitations to transparency on the part of Cade. Section IV will then try, through the analyses of Cade’s landmark cases, to connect political influence with some of the large corporations that would have benefited from it. Since a few large corporate donors have dominated the donations made in Brazilian federal elections,7 we will take a closer look at how politicians have attempted to influence antitrust enforcement of cases of the main donors’ interests. The section will argue that, regardless of conclusive evidence of causality between donations and favourable antitrust enforcement, the appearance of integrity on the part of Cade in its decision-making is crucial to its evolution as an institution and as such should be protected. Method I must say that I had good and bad luck while researching this article. To better understand the political pressures on Cade, I decided to submit an anonymous survey to Cade’s former commissioners and, thus, asked for approval from the King’s College London Law Research Ethics Subcommittee on 27 February 2017, which was granted on 8 May 2017.8 However, just nine days later, on 17 May 2017, a massive quid pro quo corruption scandal broke out, in which Brazil’s president, Michel Temer, and a member of his staff, contacted and arguably influenced Cade’s high-ranking officers.9 The scandal had enormous repercussions within the media; audio recordings of conversations between politicians and Cade’s officers were revealed, and President Michel Temer was charged with quid pro quo corruption.10 On this basis, perhaps there could be no worse time to submit a survey to former commissioners asking about political pressure. However, even after the scandal, and bearing in mind that a willingness to disclose information about such a sensitive topic may be inherently low, one in four respondents answered the survey—10 out of the 39 former commissioners to whom it was sent.11 Respondents to the questionnaire were told that: all data for analysis would be anonymized; people and cases should not be named; and self-identified contributions or those identifying others would be kept confidential. The questions were multiple choice; provided space for comments; and could be disregarded. There were 16 questions in total, divided into two groups. The first group of questions related to political pressure and contact during the appointment process, including questions such as: ‘Did any politician contact you explicitly offering his/her support to your nomination and/or approval in exchange for a certain decision towards a specific antitrust case?’. The second group of questions related to the respondents’ work as commissioners, such as: ‘Have you been contacted by a politician about an antitrust case to be decided by Cade?’; ‘What was/were the reason(s) for the contact?’; ‘Were there any promises or threats?’; ‘In what circumstances do you think it would be improper for a politician to contact a commissioner?’; and ‘Did you consider any of the politicians’ contacts to be improper in any way?’. The aim of the survey was not to draw any conclusive estimate of the frequency of political contact and pressure on Cade, nor to undertake any other kind of statistical analysis, but rather to obtain reports from former commissioners about their experiences while holding office, in order to qualitatively improve discussions of the way that political pressure on Cade might operate in practice. For this reason, I have used the information provided by former commissioners only where appropriate, in order to ground some of the discussion in this article—see, for example, pages 11, 14, 15 and 19. As the survey was designed to protect respondent anonymity owing to the sensitive nature of the topic, the commissioners are identified sequentially by number according to the order in which they appear in the article—as, for example, ‘anonymous former commissioner 1’, and so forth. To further protect their identity, the commissioners’ responses are not quoted verbatim. The responses of the former commissioners to the survey represent one of many sources that are used throughout this article. Also consulted are books containing interviews with former commissioners and high-ranking Cade officers; books and articles concerning the politics of competition law; news articles; legislation; and records of Congress, the judiciary, the Attorney General’s Office (AGU), the Federal Public Prosecution Office (MPF), and Cade. I shall also clarify that when the term commissioner is used throughout this article, this refers to all members of Cade’s board, including the chairman. What is relevant for this article is the fact that the commissioners and chairman jointly compose Cade’s board and together, therefore, have the final say on any of the agency’s cases.12 Also, to improve reading of the article, the pronoun ‘he’ is used for all anonymous former commissioners, despite the fact that they may be male or female.13 Finally, it should be noted that the objective of this article is not, in any way, an analysis of whether alleged potential quid pro quo practices have actually occurred, nor whether a given Cade officer has committed any kind of wrongdoing. A brief introduction to the structure of Brazilian antitrust enforcement Before discussing the channels of political influence, we must briefly look at the structure of Brazilian antitrust enforcement. The main institution of antitrust enforcement in Brazil is the competition agency, Cade. And the judiciary can potentially review the agency’s decisions. Within Cade’s structure, there are two main bodies—the General Superintendence and the tribunal (Figure 1). Cade’s General Superintendence is the gatekeeper: it has the power to investigate and to instruct antitrust cases. If the general superintendent issues an opinion recommending the approval without conditions of a merger and no commissioner puts forward the opinion that the case should be analysed by Cade’s tribunal, then the general superintendent’s decision will legally be considered the final decision of the agency.14 The general superintendent’s power is also significant when it comes to anticompetitive conduct cases, that is, cartel and unilateral anticompetitive conduct cases. Since the deadlines15 to close an anticompetitive conduct case are much longer than those for mergers, the General Superintendence has more discretion to decide when a given anticompetitive case will make it to the competition agency’s tribunal. Until 2012, this power to investigate and to instruct cases concerning anticompetitive conduct was granted to the Secretariat of Economic Law (Secretaria de Direito Econômico (SDE)) of the Ministry of Justice.16 Considering that the head of the SDE did not have tenure and, therefore, could be dismissed at any time by the Minister of Justice, many have questioned how political were the decisions of the SDE.17 Although it is unclear what level of dependence the SDE had from the government in practice,18 the fact that the Secretary did not have formal decision-making autonomy certainly did not enhance its insulation from political influence. Cade’s tribunal, in turn, has the final say on all the agency’s cases—even when the General Superintendence decides to dismiss a merger or conduct case, the commissioners have the power to request a hearing of the case in the tribunal.19 Therefore, the commissioners of the Cade tribunal have the final say on whether to clear a merger, and deliberate on whether to impose fines in cases involving anticompetitive practices. The competition agency’s final decision is dispersed between seven commissioners (including the chairman) and at least three of them are required to deliberate on a case. This decentralised decision-making process makes political influence more difficult, as illustrated by an audio recording revealed during the Temer–Batista scandal, in which the politician Rodrigo Loures said to businessman Joesley Batista that ‘what matters there [at Cade] is the number of votes that are sufficient to be able to approve things’.20 Cade’s decisions can potentially be reviewed by the judiciary, which is less vulnerable to political influence (see Section II). Since the judiciary may change Cade’s decisions, one could ask whether businessmen and politicians really care about the agency. The ratio of Cade’s decisions that are changed by the judiciary is subject to debate.21 But, in any event, firms certainly seem to have an interest in the agency. First, if a company wants to prosecute another firm for unilateral abuse of dominance, it has no option but to a file a case at Cade, since this violation is exclusively established by the Competition Act, and the competition agency has exclusive authority to enforce its provisions (as stipulated in both the former 1994 Act and the current 2011 Act). Likewise, if a merger meets the thresholds for antitrust analyses, the merging firms are obligated to file a case at Cade to obtain legal clearance. Secondly, firms might want to avoid damaging litigation in court following a Cade decision, be it in a merger, cartel, or abuse of dominance case. The competition agency’s decisions are extrajudicial enforcement instruments, as established by Article 93 of the Competition Act 2011 (Article 60 of the Competition Act 1994). As such, according to Caroline Viera’s research, the judiciary has usually ruled that a decision from Cade has a presumption of legitimacy and self-enforceability, and therefore should be changed by the judiciary only when there is unequivocal and concrete evidence of Cade’s illegality.22 Thirdly, companies might want to prevent the rise of other potential lawsuits analogous to a Cade case, such as a criminal investigation in a cartel case and/or a lawsuit concerning potential damages. Although these proceedings are formally independent of Cade’s decision, they may use evidence that has emerged during the agency’s investigation. Finally, firms might want to avoid the negative media coverage that usually results from the report of a contrary investigation and/or decision by Cade. The next section will show evidence that the agency’s enforcement is indeed relevant to corporations. II. THE DEMOCRATIC CONTROLS ON CADE As argued by William Kovacic, no strong competition agency around the globe is likely to be entirely insulated from the political process. As he puts it, [i]t is difficult to imagine that a jurisdiction would give broad powers to a competition agency—for example, to gather business records, to review a wide range of business behaviour, and to impose strong sanctions—without also creating some mechanisms that press the agency to exercise its powers in ways that serve society’s interests.23 Therefore, channels of democratic control will inevitably exist in most jurisdictions—even in the case of competition agencies created as ‘a stand-alone commission’ and if ‘its members have tenured appointments’.24 What then are the channels of democratic control within Brazilian legislation? The most obvious one is law-making. But our main concern here is, once legislation has been set, could politicians influence the way that this legislation is then enforced. From this perspective, perhaps the main channel of democratic control in Brazil is the appointment process. Nevertheless, there are two other important controls: the setting of the budget and human resources of the agency; and the tacit restraint that the Presidency of the Republic and the Congress might impose on Cade’s decision-making, on the basis of their prerogative to extend or to diminish the agency’s power through the approval of a bill, which might be a more subtle way of elections influencing Cade’s decision-making. Elected politicians have also attempted to influence antitrust enforcement through three other unusual methods: President Fernando Henrique Cardoso tried to grant antitrust immunity to the financial sector, by approving an opinion of the Attorney General of Brazil; Cardoso’s Minister of Justice once accepted a hierarchical appeal25 to cancel a decision of Cade; and Congress has threatened to enact a legislative decree to cancel a decision of Cade. However, I will make no further consideration in this section of these instruments because they were only a threat (the legislative decree), or ineffective (the hierarchical appeal), or their constitutionality and legality are subject to a highly controversial judicial discussion (the president’s approval of the Attorney General’s opinion). In any event, these unusual instruments will be discussed in Section IV. The appointment process To address this first channel of influence, I will explain initially which antitrust enforcement positions politicians can appoint; and go on to consider whether empirical evidence sustains the argument that the ideology of Cade’s enforcers affects the enforcement of the legislation. Positions appointed by politicians Elected politicians are empowered to appoint and confirm three crucial positions in antitrust enforcement: (i) Cade’s general superintendent; (ii) Cade’s chairman and the six commissioners, who together compose the tribunal of the competition agency; and (iii) judges of the appellate and superior federal courts (which may be relevant when there is a judicial review of Cade decisions).26 The Presidency of the Republic has the power to appoint Cade’s chairman, commissioners, and general superintendent; and the Senate has the power to confirm the nominations.27 The chairman and the commissioners have a four-year tenure, with no option of renewal (until 2012, tenure was for a two-year period, which could be renewed once); the general superintendent has a two-year term, which can be renewed once. All these officers may only be removed in very specific situations, such as a decision of the Senate upon request of the President of the Republic, or on the basis of an undisputable criminal conviction for intentional crimes, and so forth.28 No chairman, commissioner, or superintendent has been removed since 1994, when the first de facto antitrust legislation was enacted.29 To my knowledge, since 1994 there have been no investigations or reports of political intimidations regarding the potential removal of Cade’s officers. Although the antitrust officers’ mandates are not coincident, each President of the Republic has the opportunity to appoint and the Senate to confirm all the above-mentioned antitrust positions within a four-year time frame, that is, during a legislature. The judiciary is less vulnerable to political influence. All lower court judges are nominated after challenging entrance examinations largely insulated from the influence of political bodies. The political branches have a relative influence on the composition of the appellate federal courts and the Superior Tribunal Court (STJ), which is the highest court of appeal on non-constitutional issues: the President of the Republic nominates a candidate from a list of three, who have been selected by the courts themselves. Subsequently, the Senate approves the President’s choice. Political influence on the composition of the Supreme Court (Supremo Tribunal Federal (STF)), which is the highest court of appeal for cases that discuss constitutional matters, is greater: the President of the Republic is free to choose his nomination for the STF, which must be approved by the Senate. However, as of October 2019, the STF has rarely heard cases involving Cade's decisions. Cade officials’ ideologies affect enforcement It is fair to assume that those appointed to Cade will share or at least have an ideology similar to the administration that appointed them.30 We may suggest that the administration and the nominees will generally tend to agree on how competition law should be enforced, although this does not mean that the government will agree with every decision made by an appointee after taking office. As just outlined, appointments need to be confirmed by the Senate. Since 1994, nobody has been rejected by the Senate—probably because as the executive frequently asks the opinion of senators before making an appointment, they can gauge beforehand whether a person will be approved or not. In 2004, for instance, the Lula administration initially reappointed commissioner Cleveland Teixeira, but then later refused to put the nomination forward for voting. Reportedly, this was to prevent rejection by the Senate, since certain senators were clearly dissatisfied by the decision of Teixeira and other commissioners to block the acquisition of Garoto by Nestlé.31 Likewise, the Temer administration32 and the Bolsonaro administration33 also cancelled nominations because of the need to negotiate with the Senate. In any event, the executive most frequently has had a majority in the Senate, which formed the government coalition.34 Thus, the ideology of the executive most often has overlapped with that of the Senate’s majority.35 Another question is whether ideology affects enforcement.36 Brazil still lacks empirical studies on this topic, but a recent article from Marcio Moran and others statistically analysed 5091 merger control cases scrutinized by Cade. These researchers found that the greater the number of commissioners seated on Cade’s plenary who had been appointed by a centre-left Workers’ Party administration, the more interventionist decisions tended to be; and equally, the greater the number of commissioners seated on Cade’s plenary who had been appointed by a centre-right Brazilian Social Democracy Party (PSDB) administration, then the less interventionist the decisions tended to be.37 Another indication that ideology has an impact on antitrust enforcement in Brazil is the lobbying of big business around Cade nominations. An anonymous former commissioner has reported that large beverage (Ambev), mining (Vale), and food (Nestlé) corporations have lobbied senators against various nominations for Cade’s board.38 Reportedly, Vale even sent a letter to the Minister of Justice arguing against the nomination of Arthur Badin as head of the agency in 2008.39 More recently, the media reported audio recordings in which a senator (Ciro Nogueira40) offered influence on Cade’s appointment process to meat-packing businessman Joesley Batista, who was a large donor for his campaign and was concerned about the composition of Cade.41 Thus, some firms seem to believe that Cade’s composition does impact on antitrust enforcement. Politicians and their parties have also revealed a belief in the impact of the ideology of officers on antitrust enforcement. In 2008, all 16 votes against the nomination of Arthur Badin as head of Cade came from the opposition parties Brazilian Social Democracy Party (PSDB) and the Democrats (DEM), according to senator Agripino Maia, the leader of DEM. Senator Maia argued that Badin was inflexible towards certain big companies, which would be problematic in the prevailing mega-merger era.42 Other democratic controls: Cade resources and the tacit restraint Members of both the executive and legislative branches are responsible for setting Cade’s financial and human resources. The agency’s budget is determined through annual legislation written by the executive and subsequently reviewed, amended, and approved by the legislature.43 Human resources also depend on the approval of the political branches. To date, there are no empirical studies concerning the impact that different budgeting and human resource levels may have on enforcement by Cade. Finally, there is tacit restraint. According to a model proposed by Alexandre Gheventer, Cade’s officers tend to consider how their antitrust enforcement might echo within the political branches because elected politicians might ‘punish’ the agency if they are upset.44 In the same vein, Kovacic argues that, [i]f the agency must submit its budget estimates through an executive branch ministry, the process gives that ministry the ability to reward or punish the competition agency for past behaviour. If the legislature is the final gatekeeper for budget approval, the agency must consider how legislators might take the agency’s behaviour into account in deciding how to vote on the budget.45 According to this model, therefore, Cade might shape its decisions in accordance with the expectations of the executive and the legislature. For instance, around 2016–17, the then commissioner and interim chairman Gilvandro Araújo met with politician Rodrigo Loures, a key member of the Temer administration, to discuss Cade’s budget (Araújo said that he wanted to prevent possible cuts).46 If Cade decided a case in a way that disappointed the Temer administration, the agency could have been punished through cuts to the budget for the following year. Similarly, the agency’s top officers would also be tacitly restrained because Congress and the Presidency have the prerogative to extend or diminish Cade’s power through the approval of a bill. For example, Cade could be granted the power to analyse a given practice or apply a certain sanction—or have that power revoked. Nonetheless, it may be hard to detect this subtle effect of Cade’s officers ‘wanting to remain in the good graces’ of the political branches.47 Moreover, it is questionable to what extent Cade’s officers will apportion significance during antitrust enforcement to potential rewards or punishments offered by either branch. III. PRESSURE EXERTED OR CONTACT BY POLITICIANS WITH CADE In the last section, we saw that there will inevitably exist channels of democratic control, by which politicians may influence antitrust enforcement: politicians select Cade’s top-officers and set the agency’s resources, and Cade’s officers might tacitly be constrained by the politicians’ prerogative to curb the agency’s power and resources. In all of these situations, politicians may influence antitrust enforcement without having to approach Cade’s officers. In this section, we will look at situations in which elected politicians contact or put pressure on Cade’s officers in order to try to influence the agency’s decision-making, be that in general terms or regarding a specific case. In Brazil, the view that antitrust enforcement should be ‘independent’ has gradually gained strength.48 But the degree of independence is always an open question. The social acceptance of a given politician’s approach to Cade will depend on how this approach is made. I will begin this section by analysing the different ways by which a politician can approach Cade—I will argue that we must differentiate mere political ‘contact’ from ‘pressure’. Next, I will claim that there are limitations to the two main safeguards of Cade’s independence—that is, the commissioners’ tenure and the transparency of the agency. Subsequently, I will explore theoretically how elections might have a bearing on the pressure or contact made by politicians on Cade. In the end, however, I will argue that questions remain concerning just how effective political pressure and contact really have been. The difference between pressure and contact In the complexity of everyday life, it is not always possible to draw a clear line between what is aggressive and what is not. The position that a politician occupies within the government or in Congress, the time and place of the approach, the tone of the politician’s voice, the way he looks at the commissioner, and his manners are all characteristics that may indicate a greater or lesser aggressiveness. But the interpretation of such behaviours and circumstances also depends on the personality of the listener, in this case, the commissioner. One commissioner, for example, may view the behaviour of a given politician as significantly aggressive, whereas another might interpret that same contact as completely normal. It is subjective. For this reason, I propose that a line be drawn in the following terms: when a threat or promise is not explicitly made, that constitutes a contact, and not a pressure. Pressure A ‘pressure’ occurs when an elected politician or a member of his team approaches a Cade officer and makes a promise or a threat. For example, politicians may promise that they will: (i) work to provide more funding and/or human resources to Cade; (ii) give political support for the approval of a competition bill or for a specific provision of the competition bill that will benefit or be in the interests of the agency; and/or (iii) give political support for the reappointment process49 or for the appointment for another public office job. And politicians may threaten that they will: (i) draft legislation to curb Cade’s powers or even a decision made by Cade; (ii) work to cut the agency’s funding and/or human resources; (iii) work against a commissioner’s reappointment or appointment for another public office job; and/or (iv) open a parliamentary inquiry committee,50 or produce parliamentary hearings about a case. Some of these pressures have occurred in practice. With respect to Cade’s decision on the Nestlé–Garoto merger, commissioners were publicly threatened with the opening of a parliamentary inquiry committee; and a threat of a legislative decree to cancel Cade’s decision.51 Also, an anonymous former Cade commissioner has reported that there were cases in which commissioners were pressured by politicians through an explicit threat to make legislation to restrain the agency’s powers; an explicit threat to accept a hierarchical appeal; and with explicit promises of political support for the approval of a competition bill (or a specific provision of the competition bill) that would benefit or was in the interests of the competition authority.52 Contact By contrast, ‘contact’ occurs in cases where no such promises or threats are made—for example, when a politician just wants to express his point of view regarding a certain antitrust case. This differentiation between pressure and contact may be inferred from a report made by former Cade chairman João Rodas, when he said that during his four years in office (two and a half years during centre-right Cardoso’s second term and one and a half during centre-left Lula’s first term), he had ‘several contacts from the minister of justice and other bodies of the government, but not one single pressure on Cade’s decisions’.53 Former commissioner Luiz Prado also seems to assume that there is a difference between pressure and contact when he says that ‘to hear the opinion of a politician or a member of his team is a duty of the commissioner’, and that this form of communication ‘does not characterise pressure’.54 Anonymous reports from former commissioners corroborate the view that a contact is not necessarily a pressure. One said that all the contacts he had with politicians were completely normal and that it is part of the duties of Cade’s board to listen to all versions and views about the cases.55 Another has said that politicians pleaded for adequate antitrust solutions in all contacts.56 Similarly, one former commissioner said that he was contacted by politicians or lobbyists in about 10 cases, and only in one did he consider the contact improper.57 Another anonymous former commissioner reported that politicians most frequently contacted him to ask for information about the progress of a given case.58 Therefore, a contact made by a politician may be natural: he might have a legitimate argument to make. Moreover, a contact made by a politician is often inevitable: for instance, a member of Cade’s board, usually the chairman,59 will necessarily be in contact with the executive and/or Congress to discuss the agency’s budget, and it is reasonable to assume that a politician might seize the opportunity to ask for information about the development of a case. And, finally, a contact made by a politician is often even salutary: it may be helpful for Cade to hear new arguments and opinions, from different perspectives. Nonetheless, even in the situations where no threat or promise is made, contact made by an elected politician always has a subtle special feature: his communication with the member of Cade’s board is made with tacit knowledge that he has the authority to make decisions that may improve or diminish Cade’s power. In other words, even if the elected politician says nothing in relation to a promise or a threat, it remains tacit the fact that he has the power to work to increase or decrease Cade’s power and resources. The mere fact that the elected politician does have this power may lead Cade’s officers to listen to him more carefully than they would in the case of contact made by a non-politician. Anonymous former Cade commissioners have reported situations that illustrate this subtle special feature. One has said that a deputy contacted him and initially brought legal arguments about a case. However, realizing that his arguments did not convince the Cade commissioner, the congressman used his position to voice non-technical arguments, but without making explicit threats or promises. That being said, the congressman’s behaviour during the meeting did leave the impression on the former commissioner of a certain abuse of power.60 Another anonymous former commissioner has reported that, although the deputy (who was also a minister of the Presidency of the Republic at the time) made no promises to him nor threatened him, he felt that such things would have happened had there been disagreement between them.61 Coincidentally, the politician and the commissioner happened to share the same view about the case, but the antitrust officer felt that there would have been consequences had they disagreed. In addition, contact by an elected politician might also be considered special because of its potential representative aspect. Cade’s officers may consider that the opinion of the elected politician on a particular case represents the opinion of the people from his constituency, who might be impacted by the agency’s decision. Therefore, because of the tacit power of the elected politician to increase or decrease Cade’s power and/or because of its symbolic democratic element, a contact by an elected politician is a special one, and this may subtly influence Cade’s antitrust enforcement. The limitations of Cade’s transparency and reappointment ban Several former commissioners have argued that their tenure62 and Cade’s transparency63 enhance the agency’s autonomy from political pressure and contact. But a few words must be said about the limitations of Cade’s transparency and the vulnerability that some commissioners might face even after the Competition Act 2011 removed the provision which enabled their reappointment. Transparency Cade has a reputation for being a transparent agency, especially vis-à-vis other Brazilian agencies, such as the ones designed to exclusively regulate certain economic sectors (eg the telecommunications sector). Cade’s trial sessions are open to the public and audio recordings of trial sessions are available online alongside written decisions. In addition, the diary of Cade’s high-ranking officers is posted online. These features of transparency unquestionably assist in the avoidance of political pressure. However, Cade’s transparency does have limitations. The Access to Public Information Act 2011 enables Cade’s chairman to keep sensitive information (for instance, a sensitive meeting with a politician) confidential for 15 years and allows commissioners to maintain the confidentiality of sensitive information for five years.64 As the Public Information Act came into force only in 2012, at the time of writing (ie October 2019) we still do not have much information about the extent to which Cade’s high-ranking officers have used these confidentiality provisions. But in 2013, Cade reportedly issued an official note stating that the occurrence of a meeting will not be disclosed if the disclosure might ‘harm an ongoing investigation’.65 Moreover, meetings may be informal and out of office, such as a coffee between friends. For instance, the media reported that, in 2013, Cade’s chairman, Vinícius Carvalho, met in a café with Gilberto Carvalho, then a key minister of the Rousseff administration.66 However, the meeting—which, according to the media, occurred on 7 October 2013—was not published on Cade’s website.67 This is not to say that Vinícius Carvalho acted wrongly, but simply indicates that there might be limits to Cade’s transparency. The removal of the reappointment option Until 2012, Brazilian legislation allowed for the reappointment of commissioners for a second term, which could constitute a channel for political pressure: a politician could promise a commissioner political support for their reappointment or could threaten to work against it.68 Moreover, the possibility of reappointment could have a more subtle effect: even without any explicit pressure, it could foster in the commissioner a desire to please the Presidency (who appoints) and the Senate (who approves), so tempting him to alter his attitude towards a certain case.69 The new Competition Act 2011 removed the option for reappointment and extended tenure from two to four years. As noted by former commissioner Lúcia Salgado, the removal of the reappointment option helps ‘to prevent any feeling of pressure from the commissioner who wanted to be reappointed, to accommodate, to relativise his positions on a case that is in the interest of the government or parts of the government’.70 Or, as former commissioner Renault Castro has put it, removal of the reappointment option helps to minimize ‘the risk of the commissioner who wants to be reappointed to relativise his or her understanding in any case that is in the interest of the political power [ie elected politicians]’.71 However, it should be noted that the general superintendent, a central figure in the antitrust enforcement system, still has a two-year term, which can still be renewed once. Under the same rationale, it could be argued therefore that the general superintendent’s reappointment option increases his vulnerability to political pressure or contact. In addition, although the Competition Act 2011 forbids reappointment for the same position, it does not forbid appointment to other positions within Cade’s structure. For instance, a Cade commissioner who wants to be appointed as the agency’s superintendent will still face the risk of becoming vulnerable to succumb to political pressure or contact. Commissioner and interim chairman Araújo apparently faced this risk when he was contacted by politician Loures, from the Temer administration, to discuss his permanent nomination for the position of head of Cade; and later received from Loures a request regarding anantitrust case (see Section IV for a description of the case).72 Although, as of October 2019, there is absolutely no evidence of any form of illegality involving Araújo,73 the point here is that this could potentially represent a vulnerable situation for some officers. The Temer–Batista scandal also shows that some of Cade’s officers might have frequent contact with members of the executive to discuss, for example, potential nominees to the agency’s top positions. Araújo, as interim chairman of Cade, and Eduardo Frade, as the agency’s superintendent, had several contacts with Rodrigo Loures to discuss the applicants for Cade’s top positions.74 In this context, if a Cade officer wants to have influence on the future composition of the agency, there is a risk that he will be in a vulnerable situation if the politician has a request about a given antitrust case. I am not claiming that Araújo and Frade did feel vulnerable—merely that a given officer could potentially feel vulnerable in this situation. Therefore, even after the Competition Act 2011 removed the provision that enabled the reappointment of commissioners, they may still find themselves in vulnerable situations because the legislation enables the appointment of commissioners to other positions within the agency. Equally worryingly, the Competition Act 2011 still enables the reappointment of the general superintendent. Finally, a high-ranking Cade officer might also feel vulnerable if he has frequent contact with politicians to discuss the structure of the agency. How elections might affect the pressure or contact made by politicians on Cade There are at least three reasons why elections might influence political pressure on or contact with Cade. The first relates to the fact that elections determine who will occupy political offices. Different electoral winners may have different views on whether to approach Cade or not and, if they do make an approach, the content of their arguments might also differ. For instance, say a given presidential election was affected by war chest corruption75; and that this furthered the victory of a pro-business candidate who is sceptical of antitrust laws. It is likely that if he decided to pressure or contact Cade about a given case, then he would argue that Cade should adopt a non-interventionist approach. From this perspective, the fact that some politicians might be more sceptical about the importance of Cade’s independence than others may influence the selection of Cade’s officers. As they have the power to select Cade’s high-ranking officers, the Presidency and the Senate may select those people more likely to be responsive to what politicians have to say. In other words, the executive and the Senate have an opportunity to choose those individuals who are likely to be open to political contact; and to ‘filter out candidates who seem certain to ignore’ it.76 In Brazil, this may be illustrated by the Temer–Batista scandal, when the politician Rodrigo Loures said to businessman Batista that the government should give a ‘mission’ to Cade’s interim chairman, Araújo (who was appointed by the previous administration), to see if he ‘responds well’. If Araújo responded well, Loures argued, the government should then appoint him as Cade’s permanent chairman.77 In other words, Loures apparently said that the government should test the interim chairman to see if he would be receptive to government contact. Secondly, elections might affect the political pressure placed on or contact with Cade if a donation gives the donor privileged access to the elected politician. After being lobbied by the donor, the politician may pressure or contact the agency because he was convinced ideologically by the argument (ie convinced that the argument has merit). Thus, the competitive advantage of big donors in lobbying may impact political pressure on or contact with Cade. Or, the third reason is that the elected politician might surrender to the corporate lobby or perhaps simply approach Cade in the absence of any lobbying because he wants to have the business as a frequent donor in future elections. Borrowing from the rent-seeking model described by Kovacic in relation to the Federal Trade Commission (FTC),78 we may argue that politicians might pressure or contact Cade in exchange for campaign finance contributions—in Brazil, the Temer–Batista scandal shows that influence on Cade’s enforcement may generate resources for politicians, as will be seen in Section IV. The uncertain relevance of political pressure or contact on Cade The frequency of political contact with or pressure on Cade remains unknown.79 In any event, even if we assume that political contact and pressure are rare,80 there is still reason to worry as those corporations that may benefit from it will have an advantage over their competitors. If, for instance, the political pressure or contact occurs just in a few cases involving large companies, it would be disturbing if this leads to a more favourable antitrust enforcement for them. Under these circumstances, antitrust enforcement would be, in the end, a competitive advantage for large companies. Nonetheless, it is unknown just how effective political pressure and contact really are. It is one thing to say that commissioners have been pressured or contacted by politicians. It is quite another to say that those commissioners have succumbed to such pressure or contact. Even if a commissioner believes that a given enforcement of an antitrust case would lead to negative political consequences,81 it is still questionable whether the commissioner would view these factors as of much significance in his decision-making.82 It is hard to tell when a commissioner has taken pressure or contact into consideration. It is extremely unlikely that a commissioner who succumbs to such pressure or contact will admit to it publicly. Due to this lack of willingness to disclose, Salop speculates (while discussing antitrust enforcement in the USA) that the ‘incidence of successful political pressure may be understated’.83 As argued by him, a commissioner who succumbs to pressure or contact will probably ‘have a strong incentive to keep it secret to avoid being tarred as someone who can be controlled’.84 The questionnaire I sent to former commissioners is no exception: any willingness to disclose self-surrender was low. Even though the questionnaire was designed to protect respondent anonymity and the confidentiality of the responses, all respondents were informed that no method can guarantee 100 per cent anonymization and confidentiality and, therefore, the absolute absence of risks in this sense.85 As a consequence, the survey sent to former commissioners had the limitation of not overcoming the low willingness to disclose self-surrender to political pressure. It is equally difficult to detect the subtle effect of pressure or contact on those Cade officers who, although not intending to succumb consciously or explicitly, ended up being affected by it implicitly. Therefore, it is difficult for a legal scholar to estimate just how effective political pressure on and contact with antitrust enforcement truly has been. Perhaps an economist, a political scientist or an investigative journalist might have better tools with which to detect the actual responses by the competition authorities to such political pressures or contacts.86 At the time of writing, there is not enough evidence to affirm, with a strong suggestion of causality, whether political pressure on or contact with Cade has indeed been effective. IV. THE APPEARANCE OF POLITICAL INFLUENCE IN CASES INVOLVING BIG DONORS Despite the observations made in the last section, we should keep in mind that the appearance of integrity regarding the competition authority’s decision-making is crucial. Even in the absence of conclusive evidence87 about the effectiveness of political contact or pressure, public confidence on the impartiality of the antitrust enforcement institution will be undermined if people perceive that elected politicians are using antitrust to punish enemies or reward friends. While examining antitrust enforcement institutions in the USA, Jonathan Baker observes that: ‘even if agency decisions are free from direct presidential influence, the concern that they might not be undermines confidence that enforcement actions serve the public interest and undercuts political support for antitrust institutions and norms’.88 Public confidence is key for the development of competition authorities in all countries. With the current global crises of liberal constitutional democracies,89 the protection of antitrust enforcement institutions against public distrust gains in importance.90 But different countries have different levels of distrust. It is one thing to say that there is a risk of erosion of the public confidence in the impartiality of antitrust enforcement in a country such as the USA, where antitrust law is a centenary institution and a historical element of the country’s capitalism; perceptions of corruption are in a low to moderate level; and liberal constitutional democracy is solidly consolidated. But it is a whole other thing to say that for a country such as Brazil, where de facto antitrust law is just 25-year-old and is thus a relatively new element of the country’s capitalism; perceptions of corruption are among the highest in the world; and liberal constitutional democracy is just 31-year-old. The fact that Brazil has had since 2019 a president, Jair Bolsonaro, who likes to challenge the institutions of the country’s liberal constitutional democracy, adds to the concern regarding Brazil’s institutional fragility.91 There are no comparative public opinion surveys concerning the trust that citizens may have in the competition authorities of their countries. Reference is not made here to surveys of the opinions of the antitrust community (ie expert opinion)—such as the Global Competition Review ranking, which is partially based on such expert opinion. When political influence is considered, we are less concerned about the perceptions of the antitrust community and more concerned about the influence that non-expert people (politicians), elected by thousands and sometimes millions of non-expert people, have on the competition authority. But we can look at other data, such as the results of a cross-national survey, which compiled the responses of more than 43,000 citizens in the Americas in 2016–17.92 From the 22 countries surveyed, Brazilians were the ones who expressed the highest perception that corruption is widespread—83.4 per cent of Brazilians perceived that ‘more than half or all politicians are corrupt’ in the country.93 In the USA, by contrast, 45.3 per cent of the population had the same opinion, which gives the country the third best result in this item of the survey. Trust in the judiciary also seem to be dramatically different among the countries. Brazil, again, had the worst indicator among the 29 countries surveyed for this item—89.9 per cent of Brazilians had little or no confidence that the Judiciary would punish the guilty of a robbery or assault.94 In the USA, by contrast, 35.5 per cent of Americans held the same opinion, ranking the country once again as the third best result. There are other comparative public opinion surveys corroborating the well-known and long-established distrust of Brazilians in public institutions.95 Unsurprisingly, the one indicator that we have about (a fraction of the) public opinion towards the competition authority shows similar results. In the survey Global Competitiveness Report of the 2017 World Economic Forum in Davos, thousands of business executives were asked: ‘In your country, how effective are anti-monopoly policies at ensuring fair competition?’.96 For this, Brazil ranked 55th among 137 countries—while the USA ranked third. To provide a developing country parameter, Chile ranked 35th. Therefore, this suggests there is room for improvement in Cade’s reputation from the perspective of the Brazilian business community. This context—the current global crises of liberal constitutional democracies and the fragility of the Brazilian one—indicates that we should take special care about public confidence in the impartiality of Cade’s enforcement. If there exist cases in Brazil that, despite the lack of conclusive evidence, give the perception that corporations or economic sectors might be obtaining favourable antitrust enforcement because of political influence, then this may undermine people’s confidence in the impartiality of Cade’s enforcement and may lead to setbacks, such as reductions in the agency’s funding and human resources, or even the revocation of Cade’s power to analyse certain practices and apply significant sanctions. This section will analyse some of Cade’s landmark cases in order to explore the connection between political influence and certain large corporations and/or economic sectors that would have benefited from it. Perhaps we, academics, will never draw a decisive conclusion about whether antitrust enforcement did, in fact, favour corporations unfairly in the cases outlined below; however, it is sufficiently worrying in itself if reasonable people might think that it did so. Law enforcement must be rooted in confidence; and confidence is undermined when reasonable people—including those beyond academia and the antitrust community—suspect that enforcement was biased.97 For example, in the Temer–Batista scandal (described later in this section): when the media report widely that the President of Brazil is facing criminal investigation and possible impeachment for sending a member of his cabinet to influence a Cade decision in exchange for money, and Cade reportedly acts in accordance with the President’s wish, then we might never know whether the president did, in fact, influence Cade’s decision-making—but some people will have reason to believe that he did so. And this might be sufficient to undermine people’s confidence in the impartiality of Cade’s enforcement. Considering that a few large corporate donors have dominated donations in Brazilian federal elections,98 it may be interesting to look at how politicians have attempted to influence antitrust enforcement concerning these large donors.99 As attempts to influence antitrust enforcement frequently involves not just a single company but also a whole sector, these will be the focus of attention here.100 The top five donors in the 2014 federal elections were the following: (i) construction, engineering, and urban infrastructure; (ii) agribusiness; (iii) finance, insurance, and consortia; (iv) beverages; and (v) mineral extraction, steel, metallurgy, and metal products.101 Apparently, these sectors have been large donors at least since 1994.102 I will try not to overclaim and, as such, I will not draw any conclusions about causation: there is the chance that a given politician may have acted in favour of a donor not because of campaign finances but because of his view of the public interest, or else to enhance his popularity (since some antitrust cases are reported on widely by the media). Moreover, it is difficult to measure just how effective were the politicians’ acts in favour of their donors when it comes to democratic control (such as an appointment), and particularly with respect to pressure on or contact with Cade enforcers, as seen in the last section. Instead, the aim here is to ask: does the evidence as outlined undermine people’s capacity to believe that the attempts made by politicians to influence antitrust enforcement did not favour their donors in some way or else generate a significant risk of doing so?103 My argument is that such a capacity on the part of people was undermined. In summary, I aim to demonstrate three things here: big donors have been concerned with antitrust enforcement104; some politicians may have incentives to follow donor wishes; and that the cases below undermine some people’s confidence in the impartiality of Cade’s enforcement. As seen in the last section, there are limitations to Cade’s transparency (eg political contact in the President Temer–Batista scandal was not registered in the agency’s public schedule) and politicians’ contacts bear a special weight for Cade enforcers (even in the absence of any explicit threat or promise). If this context is correct, then it might be sensible to discuss the appropriateness of increased transparency surrounding meetings between elected politicians and Cade’s officers in order to enhance the appearance of integrity in its decision-making.105 Mineral extraction, steel, metallurgy, and metal products After the enactment of the Competition Act 1994, which gave birth to the first effective Cade, the acquisition of the steelworks Pains by Gerdau was perhaps the agency's first controversial case.106 Gerdau, a Brazilian provider of steel-related products worldwide, was one of the largest donors to Cardoso’s successful presidential campaign in 1994.107 In 1995, Cade vetoed the deal and ordered Gerdau to discontinue the acquisition of Pains because it would result in too great concentration on the flat-steel market. Although the Competition Act established that Cade’s decisions could not be reviewed within the administrative sphere (ie by any other organ of the executive branch), Gerdau filed a hierarchical appeal to Cardoso’s Minister of Justice, Nelson Jobim. Surprisingly, Jobim accepted the appeal and suspended the competition agency’s decision.108 Subsequently, however, Cade sent a letter to the Ministry of Justice stating that the agency did not recognize the minister’s competence to review its decisions and, as a consequence, would forward the case’s docket to the Federal Public Prosecutor’s Office in order for the decision to be executed, rather than to the Ministry of Justice, as requested.109 In 1996, Cade’s plenary had a new composition (ie new commissioners), and the agreement was made with the Gerdau Group that they could buy Pains, but on two conditions: the transportation company Transpains would have to be sold and the hierarchical appeal against the agency dropped.110 The crisis between Cade and the Ministry of Justice reportedly led to the formation during the Cardoso administration of an inter-ministerial group: the purpose of this group was to identify potential changes in the legislation that would reduce Cade’s powers, particularly in relation to mergers, and these would be transferred instead to the Secretariat of Economic Law of the Ministry of Justice and to the president of the republic himself. According to the press, the Cardoso administration was displeased with the excessive interventionism by Cade.111 Media account reported that the government even drafted a provisional measure (ie temporary legislation that can be enacted by the president of the republic without congressional approval) on the subject, although this was never issued.112 Nevertheless, there were also reports of a reaction against Jobim’s attempt to cancel Cade’s decision. Miguel Rosseto, a Workers’ Party deputy, reportedly petitioned the Brazilian Federal Public Prosecution Office noting that Gerdau was the main donor to the 1994 campaign for governor of Antonio Britto, who was Jobim’s friend and colleague at the Brazilian Democratic Movement Party in Rio Grande do Sul.113 Beverages After the Gerdau–Pains case, Cade faced another significant challenge in 2000: the merger between Brahma and Antarctica, which resulted in the Brazilian multinational Ambev.114 The merger seemed to have the political support of Brazil’s then-president, Fernando Henrique Cardoso, who reportedly said to the press that Brazil should have a multinational beverage company to compete internationally, considering that there was a global trend of mega-mergers.115 Ministers of the Cardoso administration and the leader of Cardoso’s Brazilian Social Democracy Party reportedly made calls to the head of Cade at the time, Gesner Oliveira.116 For some, Cade’s decision on the merger was considered ‘weak’ because it did not impose strong conditions for its clearance.117 Former commissioner Ruy Santacruz, who was on Cade’s board at the time, considers Cade’s decision on the Ambev merger left the feeling that it is always possible to approve a large merger when there is political and media pressure.118 This was not the first antitrust case involving Brahma in which the Cardoso administration had contacted Cade—it had reportedly occurred previously in 1997. Although former commissioner Renault Castro did not report any explicit threat or promise, he felt pressured by the government and by politicians in the Brahma–Miller case.119 We may question whether this was indeed pressure since apparently there was no explicit threat nor promise. But the aggressiveness of one of the contacts is less easy to dispute: on the weekend before the trial set to re-examine the case, a minister of the Cardoso administration telephoned Castro’s residence requesting ‘a special treatment’ (and the commissioner replied that he would give the same treatment as to all cases).120 After Cardoso left the government, Ambev was one of the companies that financed his think tank, the Fernando Henrique Cardoso Foundation.121 Moreover, Ambev has subsequently been a top donor in Brazilian federal elections,122 providing donations to the main parties, including Cardoso’s Brazilian Social Democratic Party.123 Financial, insurance, and consortia In 2001, antitrust enforcement in the financial sector came into focus. Reportedly triggered by the Central Bank (which was, in turn, allegedly influenced by commercial banks),124 the Brazilian Attorney General, as head of the Advocacia-Geral da União (AGU), published an opinion stating that the Central Bank had exclusive authority to regulate competition in the financial sector.125 Subsequently, Brazil’s President Fernando Henrique Cardoso approved the AGU’s opinion giving it binding legal force, so benefiting the financial sector—which was the top donor to Cardoso’s 1994 and 1998 electoral campaigns for president.126 According to former commissioner Celso Campilongo, the government had made it clear that it did not want Cade to have jurisdiction over the banking market.127 In other words, the government attempted to grant to the financial sector an antitrust immunity. When Cade was about to hear the first banking sector case after presidential approval of the AGU’s opinion,128 Cardoso’s Minister of Justice apparently hinted to commissioner Campilongo that a decision against the government’s wishes could give rise to political reactions.129 Although Campilongo and other commissioners of Cade have challenged the constitutionality and legality of the presidential approval of the AGU’s opinion, the antitrust immunity for the financial sector since then has been the object of a judicial dispute, which is now being analysed by the Supreme Court (STF), as of October 2019.130 Construction, engineering, and urban infrastructure sector Perhaps the first controversial antitrust case in the construction sector emerged in 2014, when Operation Car Wash began and revealed cartelization practice within the sector, as part of a massive quid pro quo corruption scandal.131 The cartel between the largest Brazilian construction corporations in public bids conducted by Petrobras (Brazil’s national oil company) was the one which most affected democracy in the country.132 Although there is certainly no evidence from which to draw any causation as of October 2019, it is disturbing that the same government coalition parties (led by the Workers’ Party), who benefited from the scheme by receiving bribes in the form of political donations, were also those in charge of the maintenance of Cade’s structure—and Cade did not discover the cartel. Nonetheless, the unfolding of Operation Car Wash revealed evidence that the construction sector may have had an impact on antitrust bill n 3937/2004, which resulted in the Competition Act 2011. Cláudio Melo, who was a director of the Odebrecht corporation (the leader of the cartel), declared in a leniency agreement (struck with the Federal Public Prosecution Office in the criminal sphere) that Senator Romero Juca had intervened in the bill to favour Odebrecht, in exchange for campaign finance.133 However, as of October 2019, there has been no further explanation of Melo’s declaration and it has yet to be confirmed by the investigation. Moreover, there is the possibility that Odebrecht also had links with two other congressmen who were involved in drafting the bill. Senator Francisco Dornelles134 introduced an amendment that gave immunity for consortia formed for public bids; and when the bill returned to the Chamber of Deputies, Deputy Pedro Eugênio135 accepted this part of Dornelles’ amendment.136 The amendment would have been in the interests of construction companies, as they often need to arrange themselves into consortia in order to compete for public works.137 Although it is still an ongoing investigation, it should be noted that Dornelles and Eugênio are now being investigated for possibly having received illegal electoral contributions from Odebrecht.138 Nevertheless, while it might be tempting to claim that Odebrecht had an impact on the legislative process of the Competition Act 2011, this cannot be determined from the evidence provided by the ongoing investigations as of October 2019. Agribusiness In 2017, another quid pro quo scandal shocked Brazil. This involved Joesley Batista, who is the controller of JBS, the world’s largest meat producer and the top donor to the 2014 federal elections; also Michel Temer, the President of Brazil, and a key member of his staff, Deputy Rodrigo Loures; and members of Cade. In a leniency agreement struck with the Federal Public Prosecution Office in the criminal law sphere, Batista declared that Temer and Loures had agreed to deliver a Cade decision in exchange for money. More specifically, Temer and Loures would make Cade issue a preventive measure to stop Petrobras from practising alleged anticompetitive practices against a company in Batista’s group, Empresa Produtora de Energia (EPE).139 In return, the businessman offered the politicians 5 per cent of the profit that his company would make as a result of Cade’s preventive measure. Batista first met Temer in a non-registered, out-of-office hours appointment, when the businessman revealed, among other topics, his concern about the appointment of the new head of Cade.140 At the end of the conversation, Temer said to Batista that ‘the best way’ to talk to him was through Loures. Subsequently, Batista met Loures and offered the quid pro quo agreement, and Loures accepted the deal not only in his but also in Temer’s name, according to Batista’s perception.141 During a meeting with the businessman, Loures telephoned the interim head of Cade, Araújo, and asked him to take a ‘sympathetic’ look at EPE’s request concerning the preventive measure. The Cade officer replied: ‘OK’.142 According to Batista, Loures’ contact with Araújo had a positive effect: after the call, members of Cade allegedly warned Petrobras’ lawyers that the agency could impose severe sanctions on the company.143 Arguably, this led Petrobras to enter into an agreement with EPE, under terms that interested Batista—Raquel Landim, a CNN Brazil senior journalist who recently wrote a book about the case for which she interviewed hundreds of people, corroborates Batista’s assertion that Cade’s officers contacted Petrobras, leading to the outcome desired by Batista.144 Due to the new deal with Petrobras, Batista’s corporation was able to regain profits and so the businessman began to pay the promised bribe to Loures.145 Cade denies any wrongdoing.146 As of October 2019, there is no conclusive evidence to confirm that members of Cade ‘warned’ the Petrobras’ lawyers. V. CONCLUSION I have discussed several ways in which politicians may influence antitrust enforcement. First, political influence on Cade may occur through ‘democratic controls’: the appointment process, resources setting, and the tacit restraint that the agency’s officers may experience regarding the political bodies. Or, secondly, it may arise through contact made with or pressure exerted on the agency by elected politicians. There is significant evidence of the impact that appointments have on Cade enforcement. There are records of large Brazilian corporations lobbying elected politicians on Cade nominations, which indicates that those firms believe that the agency’s composition matters. Politicians and their parties also have the same belief, as some reports indicate. And there is statistical evidence suggesting that the beliefs of these firms and politicians are justified: empirical research by Moran and others on more than 5091 merger cases found that enforcement varied according to the party that made the nomination. Political influence through appointments, however, seems less worrying. Although the degree of independence of the competition authorities remains relatively controversial,147 perhaps it is to be expected that enforcers will have different interpretations of the law; and believe that it is justifiable a shift in antitrust enforcement, at least to a sensible degree, according to the government entitled to make the appointment. But it is more worrying if a given corporation obtains favoured antitrust enforcement as a result of good relations with elected politicians—this would constitute a corruption of Cade’s decision-making process (not in the sense of bribes being accepted, but in terms of distorting the enforcers’ implementation of antitrust technique). In this sense, it should be noted that large donors have great ascendancy over the legal channels of political influence on Cade. First, large campaign donors have had an interest in antitrust enforcement, as outlined in Section IV, and are likely to continue to do so. And it is probable that businessmen will always find a way to legally fund politicians’ campaigns—currently, they can donate up to 10 per cent of the gross income earned in the year preceding an election. Secondly, some politicians may have the incentive to follow the wishes of a business donor, as the President Temer–Batista scandal shows—this case involved alleged bribes, but it could equally well have involved other resources, such as legal campaign finance. Thirdly, as seen in Section III, there will inevitably be frequent contact between one of the members of Cade, usually its head, and politicians in order to discuss matters such as the agency’s budget. And politicians who contact Cade may make comments or requests concerning a specific ongoing antitrust case. According to anonymous reports from former commissioners questioned for this article, even in the absence of any pressure (an explicit threat or promise), the suggestions of elected politicians are likely to carry particular weight for members of the agency. It remains unclear just what the effects of politicians’ pressures on or contacts with Cade really have been. But that does not mean that there are no problems with them. Perhaps we will never reach a conclusion as to whether antitrust enforcement did, in fact, favour donors unfairly in the landmark cases discussed in Section IV. However, if reasonable people think it did or that there is the risk that this will occur in the future, then it is already sufficiently worrying. For instance, when the media reported widely that the President of Brazil was facing criminal investigation and impeachment proposals for trying to influence a Cade decision in exchange for money, and the competition authority reportedly acted in accordance with the president’s request, we may never know if the president did, in fact, influence Cade’s decision-making—but some people will have reasons to believe he did do so. And this is enough to undermine people’s confidence in the impartiality of Cade’s enforcement. The appearance of undue political influence becomes even stronger when the contact with Cade is non-registered but eventually is revealed to the public—as it happened in the President Temer–Batista scandal. Although Cade has a well-deserved reputation for being more transparent than other Brazilian agencies, there are significant limitations to its transparency, as outlined in Section III. The problem with there being less confidence in the impartiality of Cade’s enforcement is that it may result in setbacks, such as reduced allocations to the agency’s budget and human resources. Therefore, it seems sensible to consider how to protect against this fragility, such as the appropriateness of the introduction of greater transparency with respect to meetings between elected politicians and Cade’s officers, to improve the appearance of integrity of the agency’s decision-making. Justice must be done, and be seen to be done.148 Figure 1. Open in new tabDownload slide The structure of Brazilian antitrust enforcement. Figure 1. Open in new tabDownload slide The structure of Brazilian antitrust enforcement. This work was funded by a CAPES–Brazil scholarship (BEX 1646-15/0). I would like to express my gratitude to Keith D. Ewing and Chris Townley for their support, advice and comments on the PhD thesis from which this article is derived, and also to Mariana C. Tavares, Bill Kovacic, Caio Pereira Neto, Tim Kuhner, Graeme Orr, Ana Carvalho, Eduardo Ribeiro, Fernando Marques, Luiz Prado, Rishi Gulati and Murilo Lubambo for general comments about the implementation of this research. I extend my gratitude to all anonymous former commissioners of Cade who answered the questionnaire for this research. Finally, I am grateful to the anonymous reviewers for valuable comments on an earlier version of this article. All errors are my own. I confirm that I practised competition law in Brazil from 2007 to 2014, when I represented companies of the cement, chemical, food and beverage, oil and gas, pharmaceutical, telecommunications, and transportation sectors. Footnotes 1 Walter Pincus and George Lardner Jr, ‘Nixon Hoped Antitrust Threat Would Sway Network Coverage’, The Washington Post (1 December 1997) <http://www.washingtonpost.com/wp-srv/national/longterm/nixon/120197tapes.htm> accessed 14 July 2019. See comments about the case in Ariel Ezrachi, ‘Sponge’ (2017) 5 Journal of Antitrust Enforcement 49, 56. 2 George Lardner Jr, ‘On Tape, Nixon Outlines 1971 “deal” to Settle Antitrust Case Against ITT’, The Washington Post (4 January 1997) <https://www.washingtonpost.com/archive/politics/1997/01/04/on-tape-nixon-outlines-1971-deal-to-settle-antitrust-case-against-itt/246628a9-8abf-47f3-80ec-379569e0f350/?utm_term=.2a517172ea4d> accessed 14 July 2019. 3 For a brief discussion of this case and other records of political pressure on American antitrust enforcement, see Steven C Salop, ‘What Consensus? Why Ideology and Elections Still Matter to Antitrust’ (2014) 79 Antitrust Law Journal 601, 637. For a comprehensive analysis of the influence of American presidents on the DOJ, see James F Rill and Stacy L Turner, ‘Presidents Practicing Antitrust: Where to Draw the Line?’ (2014) 79 Antitrust Law Journal 577. See also Jonathan B Baker, The Antitrust Paradigm: Restoring a Competitive Economy (Harvard UP 2019) 53. 4 See, for example, European Commission, Commission Staff Working Document, ‘Impact Assessment—Accompanying the Document Proposal for a Directive of the European Parliament and of the Council to Empower the Competition Authorities of the Member States to Be More Effective Enforcers and to Ensure the Proper Functioning of the Internal Market’ SWD(2017) 114, 26 <http://ec.europa.eu/competition/antitrust/impact_assessment_report_en.pdf> accessed 14 July 2019. 5 ‘Journal of Antitrust Enforcement Agency Effectiveness Study’ (2016) 4 Journal of Antitrust Enforcement 229, 260. 6 Artigo 4 da Lei nº 12529, de 30 de novembro de 2011, DOU 01/11/2011 e 02/12/2011. 7 On average only 0.4% of Brazilian companies used to make donations. Bruno Carazza, ‘Interesses Econômicos, Representação Política e Produção Legislativa no Brasil sob a Ótica do Financiamento de Campanhas Eleitorais’ (Tese de Doutorado em Direito, UFMG, 2016) 49 <http://www.bibliotecadigital.ufmg.br/dspace/bitstream/handle/1843/BUOS-ASNHA2/bruno_carazza___tese___vers_o_para_defesa.pdf?sequence=1> accessed 14 July 2019. In the 2010 general elections, only 191 companies (1% of the donors) accounted for 61% of the total amount donated. STF, ADI 4650/DF, Rel Min Luiz Fux, DJ 04/03/2016, Inteiro Teor do Acórdão 50. 8 REC Reference Number: HR-16/17-2546. 9 ‘Dono da JBS grava conversa com Michel Temer, diz O Globo’ (G1/Jornal Nacional, 17 May 2017) <http://g1.globo.com/jornal-nacional/noticia/2017/05/dono-da-jbs-grava-conversa-com-michel-temer-diz-o-globo.html> accessed 14 July 2019. 10 The decision on the accusation is pending as of October 2019. See Section IV for a discussion of the case. 11 I have attempted to contact (by e-mail) all former chairmen and commissioners who were appointed and approved to Cade between 1994 and July 2012. The following four former commissioners have passed away: Carlos Carvalho, Leonidas Xausa, Miguel Barrionuevo, and Luiz Schuartz. Moreover, it was not possible to obtain e-mail or other contact information for the following five former commissioners: Marcelo Soares, Edgard Rosa, Antonio Silva, Paulo Pinheiro, and Mércio Felsky. All of the remaining 39 former commissioners were contacted. 12 All members have the same voting rights but, in the case of a draw, the chairman has an extra vote (ie he votes twice); however, only rarely is this necessary. 13 Unfortunately, female representation on Cade’s board has adhered to the established patriarchal cliché: of the 39 former commissioners, I was able to contact, only five were women. 14 Lei nº 12529/2011, artigo 65, inciso II. 15 ibid, artigos 46 and 88, inciso II, parágrafo 2; e 88, parágrafo 9. 16 Artigo 14 da Lei nº 8884, de 11 de junho de 1994, DOU 13/6/1994. 17 For instance, Neide Mallard, former commissioner of Cade, said that she declined an invitation to become the head of the SDE because she ‘knew the difficulties of effectively enforcing anticompetitive practices when there is no hierarchical independence and autonomy to make decisions’. In Pedro Dutra (ed), Conversando com o CADE (Singular 2009) 41. All speeches quoted in this article that were made in Portuguese have been translated to English by me. 18 The lack of formal decision-making autonomy prevented the SDE from investigating certain matters: for example, the Secretariat refused to investigate a case in the banking sector, making the claim that, as a body of direct public administration, it was bound by an opinion of the Attorney General’s Office which gave antitrust immunity to that particular economic sector—Processo Administrativo n° 08700.003070/2010-14 (see Section IV). Somewhat ambiguously, however, in a few cases the SDE has issued an opinion in conflict with that of the government. For example, in the merger between largest Brazilian brewery companies Brahma and Antarctica, although the government was apparently in favour for a non-interventionist clearance of the merger (see Section IV), SDE issued an opinion advocating quite strong intervention—Ato de Concentração n° 08012.005846/1999-12. For reports concerning situations when, despite the institutional vulnerability of SDE, the Minister of Justice assured the autonomy of the Secretariat in practice, see Arthur Badin in Pedro Zanotta (ed), Conversando com o CADE…e com a SDE (Singular 2017) 49; Barbara Rosenberg in Dutra, ibid 208. As noted by Kovacic, ‘a variety of informal customs, norms, and habits can either increase or decrease the amount of independence that formal organisational structures and operating procedures might indicate’. William E Kovacic, ‘Competition Agencies, Independence, and the Political Process’ in Josef Drexl, Wolfgang Kerber and Rupprecht Podszun (eds), Competition Policy and the Economic Approach: Foundations and Limitations (Edward Elgar 2011) 297. The level of SDE’s independence might have varied according to just who occupied the positions of head of the Secretariat and of the Ministry of Justice. 19 Lei nº 12529/2011, artigo 65, inciso II (with respect to merger cases). Lei nº 12529/2011, artigo 67, parágrafo 1 (regarding anticompetitive conducts). When an administrative proceeding is concluded with respect to anticompetitive conduct, and the General Superintendence of Cade decides to dismiss it, the case is automatically sent to the analysis of Cade’s tribunal. Lei nº 12529/2011, artigo 13, inciso VII. Under the former Competition Act 1994, all mergers and conduct cases were automatically sent to the analysis of Cade’s tribunal, once the SDE had issued its opinion. 20 Julia Affonso and Luiz Vassallo, ‘Aliado de Temer sugeriu a Joesley dar “missão” para testar presidente do Cade’ O Estado de S. Paulo (27 May 2017) <http://politica.estadao.com.br/blogs/fausto-macedo/aliado-de-temer-sugeriu-a-joesley-dar-missao-para-testar-presidente-do-cade/> accessed 14 July 2019. 21 For example, after analysing judicial verdicts on Cade decisions taken between 1999 and 2007, Caroline Vieira concluded that most decisions of the agency were upheld by the judiciary. ‘CADE X Judiciário: estudo sobre a revisão judicial das decisões do CADE no contexto da globalização’ (Dissertação de Mestrado em Direito, PUC/SP, 2008). By contrast, another study has found mixed results on the judiciary’s verdicts about Cade’s decisions that were taken between 1994 and 2000. Fabrício Almeida (ed), Revisão judicial das decisões do CADE: pesquisa empírica e aplicada sobre os casos julgados pelos TRFs, STJ e STF (Fórum 2011) (as cited in Andréia Henriques, ‘Para Badin, Justiça é "loteria" ao definir decisões do Cade’ DCI (8 November 2010) <https://www.dci.com.br/servicos/para-badin-justica-e-loteria-ao-definir-decis-es-do-cade-1.292938> accessed 14 July 2019). 22 Vieira, ibid 158–59. 23 Kovacic (n 18) 299. 24 ibid 299. 25 The hierarchical appeal is a general legal proceeding to be used within the federal executive branch to reverse an administrative decision. Since 1994, it gradually consolidated the view that hierarchical appeals do not apply to decisions by Cade, which can be reviewed only by the judiciary. 26 In Brazil, cartels are not only an administrative offence but also a crime, such cases being heard by the judiciary. In practice, however, its criminal enforcement has been much less intense than antitrust enforcement through Cade. 27 Lei nº 12529/2011, artigo 6. In the former Competition Act 1994, the appointment process was established in art 4. Lei nº 8884/1994. 28 Lei nº 12529/2011, artigos 8 e 13, parágrafo 3. Lei nº 8884/1994, artigo 6. 29 For an account of the evolution of the Brazilian antitrust legislation, see Francisco Todorov and Marcelo T Filho, ‘History of Competition Policy in Brazil: 1930–2010’ (2012) 57 The Antitrust Bulletin 207. 30 I use ideology here as ‘the political and economic philosophy and experience that frames one’s preferred enforcement policy and antitrust legal standards’. Salop (n 3) 603. 31 Ato de Concentração n° 08012.001697/2002-89. See Juliano Basile, ‘Cade está sem quórum para julgar criação da Brasken’ Valor Econômico (25 November 2004) <https://www2.senado.leg.br/bdsf/bitstream/handle/id/458315/noticia.htm?sequence=1> accessed 14 July 2019. 32 Lorenna Rodrigues and Eduardo Rodrigues, ‘Pressionado, Temer muda indicação para superintendência do Cade’ O Estado de S. Paulo (12 September 2017) <http://economia.estadao.com.br/noticias/geral,pressionado-temer-muda-indicacao-para-superintendencia-do-cade,70001987717> accessed 14 July 2019. 33 Julio Wiziack, ‘Bolsonaro cede a Alcolumbre indicações no Cade às vésperas de votações-chave no Senado’ Folha de S.Paulo (6 August 2019) <https://www1.folha.uol.com.br/poder/2019/08/bolsonaro-cede-a-alcolumbre-indicacoes-no-cade-as-vesperas-de-votacoes-chave-no-senado.shtml> accessed 11 October 2019. 34 As of October 2019, the Bolsonaro administration has been an exception (it does not have a majority) and has to intensively negotiate every nomination with the Senate. See, for example, Lorenna Rodrigues and Mateus Vargas, ‘Bolsonaro avalia nomes de Pedro Florencio, Fernando Meneguin e Marcos Geich no Cade’ O Estado de S.Paulo (9 August 2019) <https://politica.estadao.com.br/noticias/geral,bolsonaro-avalia-nomes-de-pedro-florencio-fernando-meneguin-e-marcos-geich-no-cade,70002960844> accessed 11 October 2019. 35 The degree of ideological homogeneity between coalition parties may vary. 36 As in other jurisdictions, the former Competition Act (1994) and the current one (2011) were promulgated with provisions that have abstract terms. Thus, enforcers have had broad discretion in terms of interpretation of the legislation. 37 MR Moran, A Seth and AA Fischmann, ‘Does Antitrust Commission Structure Bias Merger Reviews? The Case of Brazil’ (2016) Discussion Paper—INSEAD. There have been some differences in enforcement between administrations in the USA—Baker (n 3) 48. 38 Anonymous former commissioner 1. 39 Juliano Basile, Raquel Ulhoa and Paulo Lyra, ‘Genro garante que indicação de Badin ao Cade será mantida’ Valor Econômico (29 July 2008) <https://www.pressreader.com/brazil/valor-econ%C3%B4mico/20080729/281659660818991> accessed 14 July 2019. 40 From the right-wing Progressive Party. 41 Hugo Marques, Thiago Bronzatto and Laryssa Borges, ‘Exclusivo: Em áudio, Joesley promote repassar dinheiro para Ciro Nogueira’ Revista Veja (6 October 2017) <https://veja.abril.com.br/politica/em-audio-joesley-promete-repassar-dinheiro-para-ciro-nogueira/> accessed 14 July 2019. 42 Juliano Basile, ‘Arthur Badin é aprovado para presidência do Cade’ Valor Econômico (5 November 2008) A10. Notwithstanding the votes against his nomination, Badin received the support of the majority of the Senate. The approval of his nomination by the Senate took unusual four and a half months of political negotiation. ‘Senado finalmente aprova Badin’ Revista Veja (4 November 2008) <https://veja.abril.com.br/blog/radar/cade-senado-finalmente-aprova-badin-2/> accessed 14 July 2019. 43 For an overview of the process, see Câmara dos Deputados, ‘Entendendo o Orçamento’ <http://www2.camara.leg.br/orcamento-da-uniao/leis-orcamentarias/entenda/cartilha/cartilha.pdf> accessed 14 July 2019. 44 Alexandre Gheventer, Autonomia versus Controle: Origens do novo marco regulatório antitruste na América Latina e seus efeitos sobre a democracia (Editora UFMG 2005) 179. 45 Kovacic (n 18) 301. 46 As reported by Araújo himself. Departamento de Polícia Federal, Inquérito Policial nº 4483/DF (DPF, IP nº 4483/DF) fls 476. 47 As argued by Salop with respect to the American antitrust enforcement institutions. Salop (n 3) 637. 48 See Pedro Navarrete, ‘As Origens do Sistema Brasileiro de Defesa da Concorrência: o CADE (1962–1994)’ (Dissertação de Mestrado em Economia, UFRJ 2013). 49 Until 2012, when the Competition Act 1994 was in force. 50 Known in Brazil as Comissão Parlamentar de Inquérito (CPI). Artigo 58, parágrafo 3, da Constituição da República Federativa do Brasil, Senado Federal, DOU 5/10/1988. 51 Julianna Sofia, ‘Cade diz que poderá reavaliar caso Garoto’ Folha de S. Paulo (12 February 2004) <http://www1.folha.uol.com.br/fsp/dinheiro/fi1202200423.htm> accessed 14 July 2019. 52 Anonymous former commissioner 1. 53 João Rodas in Dutra (n 17) 136. 54 Luiz Prado in Dutra, ibid 261. 55 Anonymous former commissioner 2. 56 Anonymous former commissioner 3. 57 Anonymous former commissioner 4. He claimed that the contact was improper because the politician mentioned that a parliamentary inquiry committee could be opened about the case. 58 Anonymous former commissioner 1. 59 Former commissioner Luis Schuartz reported that Cade’s chairman is usually the one who has the most political contact. In Dutra (n 17) 269. Reports from several other former commissioners have confirmed this dynamic. For instance, an anonymous former commissioner (who is not Schuartz, since he was not heard by the survey due to his passing) reported that he had a meeting with a senator about a case, but the chairman of Cade—who was also at the meeting—was the one person first contacted by the politician (anonymous former commissioner 5). In the same vein, former interim chairman Araújo reported several contacts with Loures (a politician) in the Temer–Batista case. DPF, IP nº 4483/DF (n 46) fls 479–80. 60 Anonymous former commissioner 6. 61 Anonymous former commissioner 7. 62 See, for example, Cleveland Teixeira in Dutra (n 17) 217; Luis Rigato in 240; Roberto Pfeiffer in 178; Ricardo Cueva in 250; Elizabeth Farina in 230. 63 See, for example, Arthur Barrionuevo in Dutra, ibid 91; Thompson Andrade in Dutra, ibid 167; Elizabeth Farina in Dutra, ibid 230. 64 Artigo 24, parágrafo 1, incisos II e III, e artigo 27, incisos II e III, da Lei nº 12527, de 18 de novembro de 2011, DOU 18/11/2011. According to art 23, the security of the society or of the state is indispensable and, therefore, any information should be kept confidential if disclosure of it may: risk an ongoing investigation or inspection, related to the prevention or repression of infractions; present a significant risk to the country's financial, economic, or monetary stability; or jeopardize the security of institutions. 65 ‘Registros comprovam reuniões entre presidente do Cade e petista’ Revista Veja (5 November 2013) <https://veja.abril.com.br/brasil/registros-comprovam-reunioes-entre-presidente-do-cade-e-petista/> accessed 14 July 2019. 66 ‘Conselho amigo’ Revista Veja (12 October 2013) <https://veja.abril.com.br/blog/radar/conselho-amigo-2/> accessed 14 July 2019. 67 Cade, ‘Agenda 2013, Presidente Vinícius Marques de Carvalho’ (2016) <http://www.cade.gov.br/acesso-a-informacao/institucional/agendas/agendas-arquivos-pdf/2013/agenda-presidente-vinicius-marques-2013.pdf> accessed 14 July 2019. Mr Carvalho was being criticized by the opposition for allegedly implementing a partisan antitrust enforcement in favour of the Workers’ Party. Carolina Gonçalves, ‘Deputados querem que Gilberto Carvalho esclareça conduta do presidente do Cade’ Agência Brasil (11 December 2013) <http://www.ebc.com.br/noticias/politica/2013/12/deputados-querem-que-gilberto-carvalho-esclareca-conduta-do-presidente-do> accessed 14 July 2019. 68 Under the Competition Act 1994, the commissioner had a mandate of two years, which could be renewed once. Lei n° 8884/1994, artigo 4, parágrafo 1. 69 Kovacic (n 18) 300. 70 Lucia Salgado in Dutra (n 17) 65. 71 Renault Castro in Dutra, ibid 74. 72 DPF, IP nº 4483/DF (n 46) fls 475–77. 73 Procuradoria-Geral da República, IP n° 4483/DF, manifestação n° 162973/2017 (PGR, IP nº 4483/DF) 19. 74 DPF, IP nº 4483/DF (n 46) fls 479–80. 75 That is, when the volume of wealthy donations is so overwhelming that it distorts electoral competition, favouring parties and candidates aligned with business interests. See the High Court of Australia (the highest court in the Australian judicial system) in McCloy v New South Wales [2015] HCA 34, 47, [38]. The decision takes into account not only Australian but also American and European cases on campaign finance. 76 Kovacic (n 18) 299–300. Even if a politician does appoint an antitrust officer, there are antitrust issues that might arise following the appointment, which could potentially lead to disagreement between them. Moreover, politicians may sometimes change their mind about what they want. 77 Affonso and Vassallo (n 20). 78 Kovacic (n 18) 294. 79 According to some reports, political pressure on or contact with Cade can be quite rare—for example, two former commissioners said that they were never contacted by politicians (anonymous former commissioners 8 and 9). By contrast, another former commissioner reported that he was contacted by politicians or lobbyists in around 200 different cases (anonymous former commissioner 10). Commissioners may have sat on Cade’s board for a period of two years (eg those commissioners that served Cade for just one term before the introduction of the Competition Act 2011) up to around seven years (eg Vinícius Carvalho, who worked three years as a commissioner and then four years as Cade’s chairman). 80 Perhaps political contact is rarer because the antitrust officers were effectively screened during the appointment process to reflect the politicians’ preferences. 81 For example, a significant impact on the agency’s budget. 82 Lawrence Frankel makes this argument in relation to merger enforcement in the USA. Lawrence M Frankel, ‘The Flawed Institutional Design of US Merger Review: Stacking the Deck Against Enforcement’ (2008) 2008 Utah Law Review 159, 185–86. 83 Salop (n 3) 637. 84 ibid. 85 The respondents were informed that all methods (online survey, physical mail, and so on) do have weaknesses. I opted to use an online platform designed for academic research, which adopts safeguards in order to protect the anonymity and confidentiality of responses made. 86 ibid. Public choice scholars have used statistical studies to research the impact of such political pressure or contact on the American antitrust authorities. See, for example, Roger L Faith, Donald R Leavens and Robert D Tollison, ‘Antitrust Pork Barrel’ (1982) 25 The Journal of Law and Economics 329; Malcolm B Coate, Richard S Higgins and Fred S McChesney, ‘Bureaucracy and Politics in FTC Merger Challenges’ (1990) 33 The Journal of Law and Economics 463. The public choice approach has been criticized for making the assumption that elected politicians and antitrust officers behave essentially on the basis of self-interest, without pursuing the public interest. See, for example, Albert A Foer, ‘The Politics of Antitrust in the United States: Public Choice and Public Choices’ (2000) 62 University of Pittsburgh Law Review 475; Robert A Katzmann, Regulatory Bureaucracy: The Federal Trade Commission and Antitrust Policy (MIT Press 1980). Similarly, public choice scholarship has been criticized for ignoring that members of competition agencies might have incentives to not to succumb to political pressure. See, for example, Russell Pittman, ‘The Causes and Consequences of Antitrust’ (1997) 12 Review of Industrial Organization 447. 87 For the purposes of this article, I use the term ‘conclusive’ to mean that there is strong evidence of causality. One could argue that the term ‘conclusive’ is inaccurate—when it comes to investigating the influence of a politician over a commissioner, evidence will hardly ever be conclusive in the sense of giving total certainty of causality. For example, even if an audio recording existed in which a given commissioner states explicitly that he would change his decision based entirely on a request made by a politician, it may be that the commissioner had already decided to do this prior to talking to the politician and reported otherwise only to gain potential political advantages. 88 Baker (n 3) 57. For a discussion concerning the political support of antitrust law in the USA, see chs 2, 3 and 10 of the book (202–09). 89 See, for example, Aziz Huq and Tom Ginsburg, ‘How to Lose a Constitutional Democracy’ (2018) 65 UCLA Law Review 78. 90 William E Kovacic, ‘The CMA in the 2020s: A Dynamic Regulator for a Dynamic Environment’ (speech delivered at Policy Exchange, 25 February 2020) <https://www.gov.uk/government/speeches/the-cma-in-the-2020s-a-dynamic-regulator-for-a-dynamic-environment> accessed 28 February 2020. 91 See, for example, Antonio Maués, ‘Bolsonaro’s First Year: Trying to Erode Democracy’ (International Journal of Constitutional Law Blog, 1 February 2020) <http://www.iconnectblog.com/2020/02/bolsonaros-first-year-trying-to-erode-democracy/> accessed 27 February 2020. 92 Mollie J Cohen, Noam Lupu and Elizabeth Zechmeister (eds), The Political Culture of Democracy in the Americas, 2016/17: A Comparative Study of Democracy and Governance (August 2017) <https://www.vanderbilt.edu/lapop/ab2016/AB2016-17_Comparative_Report_English_V2_FINAL_090117_W.pdf> accessed 5 March 2020. 93 ibid 60. 94 ibid 91. 95 For instance, in a survey of thousands of citizens of 18 Latin American countries in 2015, Brazilians were those who expressed the lowest rate of trust in government. See Latinobarómetro, ‘2015 Database’ (2015) <http://www.latinobarometro.org> accessed 22 January 2020. 96 See Klaus Schwab and others, ‘The Global Competitiveness Report 2017-18’ (World Economic Forum, 2018). This was the latest edition to ask the question about anti-monopoly policies. 97 The original structure of this sentence is found in the remarks of Lord Denning MR regarding judges’ appearance of impartiality in Metropolitan Properties Co (FGC) Ltd v Lannon [1969] 1 QB 577 599. 98 See n 7. Although in 2015, the STF banned corporate donations in the ADI 4650/DF (n 7), businessmen can still donate as individuals up to 10% of their gross income earned in the year preceding an election. Artigo 23, parágrafo 1 da Lei nº 9504, de 30 de setembro de 1997, DOU 01/10/1997 (Elections Act 1997), atualizado pela Lei nº 13165 de 29 de setembro de 2015, DOU 26/11/2015. 99 The results from Thomas Philippon’s statistical study suggest that lobbying and campaign contributions caused weaker antitrust enforcement in the USA. T Philippon, The Great Reversal: How America Gave Up on Free Markets (Harvard UP 2019). 100 I am not arguing that the attempts to influence Cade’s enforcement have only occurred in cases involving these top donors: the acquisition of Garoto by Nestlé, one of the most controversial cases in the agency’s history, shows that other economic sectors have experienced this phenomenon. 101 FGV/DAPP, ‘Mosaico Eleitoral’ (2015) <http://dapp.fgv.br/transparencia-politica/mosaico/> accessed 14 February 2018. 102 For the campaign finance data on the 1994 and 1998 federal elections, see David Samuels, ‘Money, Elections, and Democracy in Brazil’ (2001) 43 Latin American Politics and Society 36. For the presence of those top donors as Lula’s top funders in 2006, see Silvio Navarro and Fábio Zanini, ‘Setor bancário deu maior doação à campanha de Lula’ Folha de S. Paulo (29 November 2006) <http://www1.folha.uol.com.br/folha/brasil/ult96u87203.shtml> accessed 14 July 2019. 103 The structure of this question was inspired by Lawrence Lessig, Republic, Lost: How Money Corrupts Congress—and a Plan to Stop It (Twelve 2011). 104 As it will be seen, in two of these sectors (construction and financial services), the concern was dispersed among the key players, as the cases relate to cartel practices and an attempt for antitrust immunity; while in the other three sectors, the concern was concentrated on the leading company, with cases relating to mergers and abuse of dominant position. Contrary to other competition authorities, such as the British Competition and Markets Authority, Cade does not list priority sectors that it will focus on in its annual plan. 105 High courts of different countries have emphasized the need to protect people’s confidence in the institutions of liberal constitutional democracy. For example, the Brazilian Supreme Court ruled that the mere appearance of dependence on the part of politicians on corporate contributions, or of distortion of the electoral process due to an overwhelming amount of corporate contribution, will constitute sufficient justification for campaign finance reform on the basis of an imperative to protect people’s trust in the integrity of elections. ADI 4650/DF (n 7). In the same vein, see Supreme Court of Canada’s Harper v Canada (AG) (2004) SCC 33, [2004] 1 SCR 827; and the High Court of Australia in McCloy (n 75). For the need to protect elections against appearance of quid pro quo corruption, see Supreme Court of the USA in Buckley v Valeo (1976) 424 US 1, 26 and 27. For the need to protect the judiciary against appearance of impartiality, see footnotes 97 and 148 in this article. 106 Ato de Concentração n° 16/1994. 107 ‘Área de siderurgia ajuda a eleger 20’ Folha de S. Paulo (8 October 1995) <http://www1.folha.uol.com.br/fsp/1995/10/08/caderno_especial/15.html> accessed 14 July 2019. 108 Alexandre Gheventer, ‘A Dimensão Política da Regulação Antitruste’ (2004) 18 Boletim Latino-Americano de Concorrência 26. 109 Mallard in Dutra (n 17) 47. 110 Gheventer (n 108) 26. 111 ibid. 112 Fernando Rodrigues, ‘FHC quer superpoder com lei antitruste’ Folha de S. Paulo (5 March 1996) <http://www1.folha.uol.com.br/fsp/1996/3/05/dinheiro/9.html> accessed 14 July 2019. For a discussion about the government’s reaction with respect to the provisional measure, see Carlos Alberto Bello, Autonomia Frustrada: O CADE e o Poder Econômico (Boitempo 2005) 149. 113 Bello, ibid 146. 114 Ato de Concentração n° 08012005846/1999-12. 115 Jornal do Brasil (5 July 1999) (as cited in Gheventer (n 108) 27 and 28). 116 Ruy Santacruz in Dutra (n 17) 103. 117 See, for example, Salgado in Dutra, ibid 61 and 62, who did not vote in the decision. 118 Santacruz in Dutra, ibid 104. 119 Ato de Concentração n° 58/1995. Castro in Dutra, ibid 71. 120 Castro in Dutra, ibid. 121 Daniel Bramatti, ‘Estatal doou R$500 mil a instituto de FHC’ Terra magazine (17 January 2007) <http://terramagazine.terra.com.br/interna/0,,OI1352138-EI6578,00.html> accessed 14 July 2019. 122 See, for example, Mariângela Gallucci, ‘Campanha de Marina declara R$24 mi em doações’ O Estado de S. Paulo (1 November 2010) <http://politica.estadao.com.br/noticias/geral,campanha-de-marina-declara-r-24-mi-em-doacoes,633369> accessed 14 July 2019; José Toledo, Lucas Maia and Rodrigo Burgarelli, ‘As 10 empresas que mais doaram em 2014 ajudam a eleger 70% da Câmara’ O Estado de S. Paulo (8 November 2014) <http://politica.estadao.com.br/noticias/geral,as-10-empresas-que-mais-doaram-em-2014-ajudam-a-eleger-70-da-camara,1589802> accessed 14 July 2019. 123 Transparência Brasil, ‘Às Claras — Quem financia quem nas eleições’ <http://www.asclaras.org.br/@index.php> accessed 14 July 2019. This database uses official data from the Superior Electoral Tribunal. 124 Celso Campilongo in Dutra (n 17) 153. For Cláudio Considera and Paulo Corrêa, the banking industry was interested ‘in keeping competition policy under Central Bank responsibility’ at that moment because Febraban (the association of banks) was being ‘prosecuted by SDE for fixing the interest rate’. ‘The Political Economy of Antitrust in Brazil: From Price Control to Competition Policy’ (2002) 28 Annual Proceedings of the Fordham Corporate Law Institute 533. 125 Parecer AGU/LA-01/2001 (Anexo ao Parecer GM-020), Processo n 00001.006908/2000-25 <http://www.agu.gov.br/atos/detalhe/8413> accessed 14 July 2019. 126 ‘Bancos lideram doações para campanha de FHC’ Folha de S. Paulo (26 November 1998) <http://www1.folha.uol.com.br/fsp/brasil/fc26119803.htm> accessed 14 July 2019. 127 Campilongo in Dutra (n 17) 154. 128 Ato de Concentração n° 08012.006762/2000-09. 129 Celso Campilongo, ‘Antitruste no setor bancário e de meios de pagamento’ (22° Seminário Internacional de Defesa da Concorrência do IBRAC, 12 December 2016) 31:25 <https://youtu.be/KUXaTcLSDRg> accessed 14 July 2019. 130 STF, RE 664189. 131 Due to the amount of money that circulated (at least £2.2 billion), some have wondered if the scheme has been so far the ‘biggest corruption scandal in history’. See Jonathan Watts, ‘Operation Car Wash: Is This the Biggest Corruption Scandal in History?’ The Guardian (1 June 2017) <https://www.theguardian.com/world/2017/jun/01/brazil-operation-car-wash-is-this-the-biggest-corruption-scandal-in-history> accessed 14 July 2019. For econometric and statistical methods used by the Brazilian Federal Police to prove cartelization, see Regis Signor and others, ‘It Is Not Collusion Unless You Get Caught: The Case of “Operation Car Wash” and Unearthing of a Cartel’ (2019) 7 Journal of Antitrust Enforcement 177. 132 Processo Administrativo n° 08700.002086/2015-14. To enable the workability of the cartel, the construction companies corrupted not only high-ranking employees of Petrobras—who were responsible for the public bids—but also the politicians that gave political support for the appointment and permanence of these employees. See Justiça Federal, 13ª Vara da subseção Judiciária de Curitiba, Sentença, Ação Penal nº 5036528-23.2015.4.04.7000/PR, j. 08/03/2016, parágrafos 399 e 1028. The unleashing of the Car Wash cartel unfolded other antitrust investigations concerning other alleged construction cartels involving the same large corporations, such as on the construction of facilities for the 2014 FIFA World Cup (Inquérito Administrativo n° 08700.006630/2016-88), and so on. 133 MPF (Ministério Público Federal), Delação Premiada, Cláudio Melo Filho. If Melo Filho’s declaration eventually gets confirmed by other pieces of evidence, a potential free-rider problem will have to be explained: while Melo Filho is referring only to Odebrecht’s donation and lobbying, other construction might have benefited from it. 134 From the right-wing Progressive Party. 135 From the centre-left Workers’ Party. 136 Câmara dos Deputados, Parecer Pedro Eugênio, 54 <http://www.camara.gov.br/proposicoesWeb/prop_mostrarintegra?codteor=928019&filename=PSS+1+PL393704+%3D%3E+PL+3937/2004> accessed 14 July 2019. 137 See, for example, Inquérito Administrativo n° 08700.004468/2017-44 and Processo Administrativo n° 08700.003241/2017-81. 138 Fernando Rodrigues, ‘Documentos da Odebrecht listam mais de 200 políticos e valores recebidos’ UOL Noticias (23 March 2016) <https://fernandorodrigues.blogosfera.uol.com.br/2016/03/23/documentos-da-odebrecht-listam-mais-de-200-politicos-e-valores-recebidos/> accessed 14 July 2019. 139 Inquérito Administrativo n° 08700.009007/2015-04. Petrobras has state monopoly in the gas supply market. The case discusses price discrimination and refusal to contract. 140 See ‘Leia na íntegra a conversa entre o presidente Temer e Joesley Batista’ Folha de S. Paulo (19 May 2017) <http://www1.folha.uol.com.br/poder/2017/05/1885414-leia-na-integra-a-conversa-entre-o-presidente-temer-e-joesley-batista.shtml> accessed 14 July 2019. 141 MPF (Ministério Público Federal), Delação Premiada, Joesley Batista, Anexo 9. 142 DPF, IP nº 4483/DF (n 46) fls 858–59. 143 PGR, IP nº 4483/DF (n 73) 19. 144 Why Not (Editora Intrínseca 2019) 275. 145 PGR, IP nº 4483/DF (n 73) 19. 146 Cade, ‘Esclarecimentos do Cade sobre o termo delação de Ricardo Saud’ (20 May 2017) <http://www.cade.gov.br/noticias/esclarecimentos-do-cade-sobre-o-termo-delacao-de-ricardo-saud> accessed 14 July 2019. 147 For example, for a discussion concerning the degree of independence of European national competition authorities, see ch 3 at Christopher Townley, A Framework for European Competition Law: Co-ordinated Diversity (Hart Publishing 2018). 148 The original structure of this sentence is found in the remarks of Lord Chief Justice Hewart about judges’ appearance of impartiality in R v Sussex Justices; Ex parte McCarthy [1924] 1 KB 256. © The Author(s) 2020. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: [email protected] This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Lin, Ping; Ross, Thomas W

doi: 10.1093/jaenfo/jnaa039pmid: N/A

Abstract After years of debate, Hong Kong’s new competition law, the Competition Ordinance (CO), took effect in December 2015. Laying out rules to support competitive markets and creating the institutions to administer and enforce those rules, the CO is a modern competition law in many respects, following many best-practices and respecting recent learning in competition economics. This article argues, however, that—at least from an economist’s perspective—in its drafting a series of decisions were made that weaken the law. None is that unusual or critical on its own, however collectively they leave the law less powerful than competition enthusiasts might desire in a modern market economy. We discuss the implications of these decisions and go on to consider some other more unique aspects of the law that might need reconsideration at some point. Finally, we document and discuss the early activities of the Competition Commission of Hong Kong. We conclude that Hong Kong is off to a good start with its new law and its enforcement but that several reforms have the potential to bring a more robust competition policy regime. I. INTRODUCTION After several years of debates, studies, proposals, and revisions, in 2012 the Legislative Council of Hong Kong finally adopted a cross-sectoral competition law, the Competition Ordinance (CO).1 The CO took effect in December 2015. With language clearly influenced by similar statutes in Singapore, Europe, and Australia, the CO reflects, in most respects, a modern approach to competition law, one that integrates accumulated learning in competition economics. In the years since the CO was passed into law, a Competition Commission has been constituted and has launched into an ambitious agenda to first advocate for competition; educate business and consumers; to issue significant guidance documents; and, more recently, to begin active enforcement. A Competition Tribunal has also been created to hear matters brought to it by the Commission. The Tribunal has recently begun hearing its first cases, issuing two important decisions. Modern competition law is a form of economic regulation. It seeks to preserve the competitiveness of markets principally because we believe that competition drives efficiency. And efficiency leads to greater wealth for individuals and societies. As positive and welcome as this first step undoubtedly is, and while the CO generally follows many of the best-practices that have been informed by competition economics, in our view this first version should not be the final word. There are many debatable, to an economist, elements in the CO that merit further thought, possibly leading to legislative or other policy changes over time. The purpose of this article is to offer an economist’s perspective on the new law and the first years of its implementation.2 We feel that this perspective is particularly timely, given indications that there may soon be a review of Hong Kong’s competition policy regime.3 We hope that our views may usefully contribute to this process. We begin by providing some background on the development of the new law in Section II. Section III then briefly summarizes the basic elements of the CO as adopted in 2012. This is meant to be a broad and brief overview and not a detailed, clause-by-clause, review. A close examination of the various components of the CO, in our view, reveals a missed opportunity to embrace as robust and effective a competition regime as economists might like. Section IV makes this case, discussing a number of elements in the CO that are not as strong on certain threats to competition as they might have been. Some additional interesting features of the CO are discussed in Section V. Section VI then briefly reviews the early activity of the Hong Kong Competition Commission. Finally, Section VII concludes, summarizing areas of concern that we feel merit additional attention. II. SOME BACKGROUND ON COMPETITION POLICY IN HONG KONG4 The road to Hong Kong’s new competition regime began many years earlier. It started with the introduction of competition elements to sectoral regulation: sections on abuse of dominance were added to the Telecommunications Ordinance in 2000 with an additional section allowing for the review of mergers in the sector added in 2004;5 and sections on anticompetitive conduct and abuse of dominance were introduced to the Broadcasting Ordinance in 2000.6 In July 2005, the Hong Kong Special Administrative Region (SAR) government set up the Competition Policy Review Committee (CPRC) to evaluate the effectiveness of Hong Kong’s existing competition policy, drawing on international experience, and to examine the need to introduce a comprehensive and cross-sector law on fair competition. A year later, in June 2006, the CPRC released a report (Competition Policy Review Committee, 2006)7 which recommended that a comprehensive competition law be introduced in Hong Kong to guard against anti-competitive conduct that would have an adverse effect on economic efficiency and free trade in Hong Kong. According to this report, the new legislation should apply to all sectors, rather than targeting only some sectors of the economy. In November 2006, following the recommendation of the CPRC, the government issued a public consultation document, entitled ‘Promoting Competition – Maintaining our Economic Drive’ and invited responses from the general public. Respondents included political parties, private companies, academics, the Hong Kong Consumer Council, and trade organizations. While the majority of the respondents supported the introduction of a cross-sector competition law in Hong Kong, there were concerns in the business sector, especially among some small and medium-sized enterprises (SMEs), at the possible effects a competition law might have on their business operations. For example, members of the Federation of Hong Kong Industries ‘are concerned that enacting such a law would impose hefty compliance burden on SMEs to the detriment of their competitiveness vis-à-vis large corporations’.8 After an extended period of consultation and debate, the CO was enacted in June 2012. Getting to this point was not easy for competition law supporters: as noted, there was considerable unease in some quarters in Hong Kong—particularly in the SME sector—about regulatory overreach. Several compromises were necessary to move forward, but the result was a modern competition law, at least with respect to horizontal agreements and abuse of dominance.9 It is interesting to speculate about what finally allowed the proponents of a general competition law for Hong Kong to overcome the entrenched resistance.10 While a more detailed analysis of the battles for the CO is beyond the scope of this article, we speculate that a couple of factors were important. First, Singapore—a country of similar economic development and size, and another leading rich country ‘outlier’ not in the competition policy world—adopted its own competition law in 2004. Singapore’s experience has been widely viewed as successful. Secondly, as noted, a series of compromises were found (discussed further below) to address some of the most extreme concerns of opponents. III. THE COMPETITION ORDINANCE OF 2012: BASIC ELEMENTS The key prohibitions of the CO take the form of two ‘Conduct Rules’ modelled on European antitrust law: (i) the First Conduct Rule (‘FCR’) prohibits agreements and concerted practices that restrict competition; and (ii) the Second Conduct Rule (‘SCR’) prohibits an undertaking with substantial market power from abusing that power by engaging in conduct that restricts competition. The CO also contains a ‘Merger Rule’ that—at this time—only applies to the telecommunications sector. The First Conduct Rule The First Conduct Rule seeks to prohibit arrangements between market participants (whether they are competitors or not) which prevent, restrict or distort competition in Hong Kong. For example, it prohibits competitors from cooperating on key parameters of competition such as price, output or bidding strategies. The key provision here is contained in section 6(1) of the Ordinance: ‘An undertaking must not (a) make or give effect to an agreement; (b) engage in a concerted practice; or (c) as a member of an association of undertakings, make or give effect to a decision of the association, if the object or effect of the agreement, concerted practice or decision is to prevent, restrict or distort competition in Hong Kong.’ While there is not a lot of guidance provided in the CO about the full scope of agreements (or concerted practices etc.) that could be subject to the FCR, it does create a subset that it refers to as involving ‘Serious Anti-competitive Conduct’. It defines (in section 2) the following as being serious anti-competitive conduct violating the FCR: (i) fixing, maintaining, increasing or controlling the price for the supply of goods or services; (ii) allocating sales, territories, customers or markets for the production or supply of goods or services; (iii) fixing, maintaining controlling preventing, limiting or eliminating the production or supply of goods or services; and (iv) bid-rigging. Appropriately, these are the typical targets of provisions aimed at collusion, striking particularly at the so-called ‘hard-core cartels’. The Guidelines on the First Conduct Rule, issued by the Hong Kong Competition Commission, tell us a little more about the potential scope of the FCR provisions, pointing out (at Paragraph 1.5) that the section certainly applies to vertical arrangements as well as those between competitors as there is no such limitation in the FCR. The Guidelines also include some hypothetical examples, providing insights as to the set of arrangements to which the Commission sees the potential application of the FCR. Building from these examples and experience with similar provisions in other jurisdictions we can see the FCR being applied in cases including: Cartels (ie agreeing with your competitors to fix prices, share markets, rig bids or restrict output); Exchange of information; Activities of trade associations and industry bodies; Joint Ventures; Vertical price restrictions; Exclusive distribution and exclusive customer allocation; The CO provides for some exemptions and exclusions to the FCR (see section 9(1)) and some of these come up in our discussion below. Exclusions can be provided for agreements that enhance overall economic efficiency (under schedule 1). Exemptions can be allowed for public policy reasons (section 31) or to comply with international obligations (section 32). Importantly, there is a broad exclusion for statutory bodies (under section 3). The Second Conduct Rule The Second Conduct Rule targets businesses with a ‘substantial degree of market power’ that abuse that power with a view to protecting or increasing their position of power and profits. Certain conduct (see examples provided below) engaged in by businesses with substantial market power can have the object or effect of excluding competitors from the market, thereby limiting choices available to consumers. The specific language of section 21(1) provides that: ‘An undertaking that has a substantial degree of market power in a market must not abuse that power by engaging in conduct that has its object or effect the prevention, restriction or distortion of competition in Hong Kong.’ The CO explains, in section 21(2) that conduct ‘may, in particular, constitute such an abuse if it involves: (a) predatory behaviour towards competitors; or (b) limiting production, markets or technical development to the prejudice of consumers.’ The Commission’s Guidelines on the Second Conduct Rule provide a few more examples of what might constitute abusive behaviour, offering the following non-exhaustive list at Paragraph 5.1: predatory pricing tying and bundling margin squeeze conduct refusals to deal; and exclusive dealing. Significantly, some of these items, notably (b) and (e) involve agreements, raising the possibility that some arrangements could be attacked under either the FCR or the SCR, though the latter approach would require that one firm have a substantial degree of market power. Interestingly, the Commission seems to have taken a clear position that resale price maintenance (another kind of vertical agreement) will be attacked only under the FCR, presumably even if forced on small distributors by large and powerful manufacturers.11 The Merger Rule Mergers that have or are likely to have the effect of substantially lessening competition in Hong Kong are prohibited under the Ordinance. While laying out a modern approach to merger review as practiced in other advanced jurisdictions, the scope of application of the Merger Rule is currently limited to mergers relating to undertakings directly or indirectly holding carrier licences issued under the Telecommunications Ordinance. In this sense, they represent an updating of the merger provisions that already existed in the Telecommunications Ordinance. Some of the inconsistencies in the approach to mergers here compared to the conduct rules (see Section V below) could be simply due to the fact that they were adapted from different sources: the conduct rules drawing heavily on the European law and experience and the merger regime adapted from the sectoral legislation which had its own roots. The New Competition Institutions As becomes important for some of the issues we discuss below, the CO provides for a prosecutorial or adversarial enforcement regime. To do this, it created two important new institutions. The Hong Kong Competition Commission (hereafter HKCC) is established in Part 9 of the CO. Its functions are laid out in section 130: The Commission has the following functions— (a) to investigate conduct that may contravene the competition rules and enforce the provisions of this Ordinance; (b) to promote public understanding of the value of competition and how this Ordinance promotes competition; (c) to promote the adoption by undertakings carrying on business in Hong Kong of appropriate internal controls and risk management systems, to ensure their compliance with this Ordinance; (d) to advise the Government on competition matters in Hong Kong and outside Hong Kong; (e) to conduct market studies into matters affecting competition in markets in Hong Kong; and (f) to promote research into and the development of skills in relation to the legal, economic and policy aspects of competition law in Hong Kong. The first appointments to the HKCC were made in April 2013 by the Hong Kong SAR Government, and the first chief executive officer was named in July 2014. To hear the cases brought by the HKCC, Part 10 of the CO created a specialized Competition Tribunal (hereafter Tribunal) as a superior court of record. This is not a lay panel, or even a mixed lay/judicial body as some competition adjudicative bodies are. Section 135(1) provides that: ‘The Tribunal consists of the judges of the Court of First Instance appointed in accordance with section 6 of the High Court Ordinance (Cap 4), by virtue of their appointments as such judges.’12 Section 142(1) lays out the jurisdiction of the Tribunal: The Tribunal has jurisdiction to hear and determine— (a) applications made by the Commission with regard to alleged contraventions, or alleged involvements in contraventions, of the competition rules; (b) applications for the review of reviewable determinations; (c) private actions in respect of contraventions, or involvements in contraventions, of the conduct rules; (d) allegations of contraventions, or involvements in contraventions, of the conduct rules raised as a defence; (e) applications for the disposal of property; (f) applications for the enforcement of commitments; and (g) any matter related to a matter referred to in paragraph (a), (b), (c), (d), (e) or (f) if the matters arise out of the same or substantially the same facts. In fulfilling its responsibilities, the Tribunal can issue a variety of types of orders and can determine the appropriate level of penalties and/or remedies in matters brought before it. The appointments of the first President and Deputy President of the Tribunal took effect on 1 August 2013. IV. A ‘SOFT’ BEGINNING? As noted, among developed economies at least, Hong Kong is a relatively late arrival to the modern world of competition policy and the path to its first law was neither easy nor was its success pre-ordained. Given the resistance to be overcome in securing passage, it is perhaps not surprising that there are elements in the new law that make it ‘softer’ or weaker, in terms of its ability to detect, prosecute, and punish anticompetitive actions, than it might have been. In drafting a modern competition law there are many decisions that need to be made on the various elements—decisions that can range in their effect, in the process making the law tougher or softer. In a large number of cases the drafters of Hong Kong’s CO chose softer options. While taken individually, each is not unlike choices made in certain other jurisdictions; collectively they make the CO a less powerful protector of competition than the laws of many other countries. To be clear, we do not make this point as a criticism of those drafting the law or of the CO itself: such compromises may have been necessary to build enough support to have a law passed at all. However, we do suggest that, as competition policy evolves in Hong Kong, there is a list of elements that may need revisiting to provide for a more robust competition policy regime. In this section we survey a number of the key elements that we view as softer, discuss how they could have been designed differently and compare them to the approaches taken in other jurisdictions. We begin with elements that cut across the various substantive areas of the law, then focus on elements specific to particular areas such as cartels, abuse of dominance and mergers. Our point here is not to review the individual details of elements that have been examined carefully in earlier work, for example by Kwok,13 rather it is to highlight the pattern that emerges from seeing them collectively. At least from an economist’s perspective, the product is a competition law with less bite than might have been desired. Crosscutting Elements Here we wish to draw attention to four aspects of the new law that cut across the various substantive areas and which, in each case, represent approaches that seem to be softer than they could have otherwise been. First, the CO adopts a ‘judicial (or prosecutorial) model of enforcement’, rather than an administrative one.14 Under a judicial model, the adjudication function is separated from the investigative and prosecutorial functions. In Hong Kong, the HKCC has the authority and duty to investigate potential violations of the CO but, should it wish to challenge agreements or conduct, it must do so before the separate Competition Tribunal. This separation of functions is typically viewed as favourable to respondents who get—it is argued—a truly independent third party review of their case. While administrative structures, such as those of the European Commission or the US Federal Trade Commission, in which competition authorities also render decisions usually allow for case decisions to be made by parties different from the case team, the concern has often been expressed that this separation of tasks does not provide the same level of independence.15 Of course, Hong Kong is far from alone adopting a judicial model, for example matters handled by the Antitrust Division of the US Department of Justice and all cases pursued by the Canadian Competition Bureau proceed through judicial processes.16 Secondly, relative to those of other jurisdictions, the ‘maximum level of fines’ for undertakings violating the FCR and SCR is quite low—up to only 10 per cent of turnover in Hong Kong for a maximum of three years for a single contravention (section 93).17 This contrasts with the approach in other jurisdictions that do not limit the number of years (eg USA and Canada) and/or base fines (with a similar percentage rate) on global turnover (eg European Commission, France, Norway). There is a serious question here whether fines of such a level will adequately deter anticompetitive behaviour by, in particular, multinational companies operating in Hong Kong.18 The case for prosecutions targeting individuals could be different. While it is not clear how often the HKCC will pursue penalties for individuals, it is interesting to note that here the CO provides that the Tribunal ‘may order that person to pay to the Government a pecuniary penalty of any amount it considers appropriate’. (section 93(1)).19 Thirdly, the CO provides a very broad ‘exemption for statutory bodies’ (section 3). Such entities will only be brought under the law if explicitly added to a list by the Chief Executive. Out of a large number of bodies that, to at least some extent, engage in commercial activity and could possibly engage in anticompetitive activity, a very small number have (to this point) been put on this list.20 It is common for a new authority to have to devote considerable resources to advocacy within government—persuading other branches of government of the value of competition. This broad exemption will undoubtedly make the HKCC’s task more difficult and means that the coverage of the law is more limited than it might have been. Other jurisdictions, such as the EU have taken a much more case-by-case approach to the question of which statutory bodies should be exempt from competition law.21 Finally, when there is a concern that competition in some market may be compromised but it is not clear what the specific issues or causes may be, many competition laws allow for authorities to undertake ‘market studies’. These can be wide-ranging examinations of the structure and performance of markets, as well as the behaviour of market participants.22 There have been, broadly speaking, three types of market study or market investigation powers granted to competition authorities. The most limited arise when the authority is authorized to conduct the studies but not to compel participation by industry players.23 Other authorities can compel participation, but cannot take any action directly as a result of the study’s findings—a formal case must be initiated to do that.24 In still other jurisdictions, the authority can launch studies, compel participation and take some actions based on the study’s findings.25 While the CO does authorize the conducting of market studies (in section 130(e)), it is with the weakest of powers described here—the Commission cannot compel participation.26 It goes without saying that this limits significantly the Commission’s ability to get at the kind of inside information it may need to accurately assess the state of competition in a market and the causes of any deficiencies.27 Horizontal Agreements The CO does not clearly distinguish between vertical and horizontal agreements in its First Conduct Rule prohibitions, which may be somewhat unfortunate given the significantly different potential for harm the two types of agreements present. This is, however, not an uncommon approach, and the important distinctions may come to be drawn through jurisprudence.28 With respect to the treatment of agreements between competitors—horizontal agreements—we wish to note five features of the law that render it softer than comparator laws. First, while international best practices certainly endorses taking a rule of reason or effects-based approach to most competition matters, there has emerged a relatively strong consensus on the value of per se approaches to naked cartel behaviour. That is, horizontal agreements that are simply about fixing prices, allocating markets and/or cutting output are typically condemned without the need for expensive hearings about effects. As noted, the CO does prohibit anticompetitive agreements, including cartel agreements, if their object or effect is to harm competition. It may have well been the intent of the drafters of the CO—as has emerged under the European Commission working with a similarly worded provision likely the model for this section of the CO—to capture naked collusion on a per se basis by finding such agreements to have an anticompetitive object and presenting an almost impossibly high hurdle for the establishment of offsetting efficiencies. It remains to be seen, of course, whether this is an interpretation the Competition Tribunal will give the CO. It is also the case that even parties to an agreement to fix price have access to an efficiency defence in Schedule 1 (section 1). Should it transpire that the Competition Tribunal is willing to give a hearing to efficiency arguments even in cases of relatively straightforward collusion, the law in Hong Kong will have clearly moved away from standard best-practice.29 Secondly, the CO is ‘not criminal law’—violators are not subject to criminal liability and punishments. While it is generally viewed as desirable to adopt administrative or civil law approaches to most competition matters given the complexity of many cases and the associated difficulty of establishing effects with great certainty, criminal sanctions are available in a number of jurisdictions (eg Canada and the USA) for the most egregious offenses—usually just hard-core cartel conduct. It is argued that the threat of criminal punishments provides a more powerful deterrent, and also a greater incentive to make use of leniency programmes that expose cartels.30 The CO—and again Hong Kong is not alone here—does not provide this powerful weapon.31 In making this observation, we are very aware of the fact that there is a substantial debate in the literature regarding whether criminalizing cartel conduct is a good idea. Questions raised include, for example: whether cartels are morally blameworthy enough to justify such punitive treatment; whether criminalizing collusion really enhances deterrence; whether the higher costs of criminal enforcement and procedure are worth any alleged benefits; and about the challenges of getting international cooperation to fight cartels when some countries apply criminal sanctions and others do not.32 Our point here is just that, despite something of an international trend toward criminalizing cartel, Hong Kong chose not to do so.33 Thirdly, the CO provides for a ‘free pass’ for violators of the first conduct rule, provided their agreement does not constitute serious anticompetitive conduct.34 The Commission must do this by issuing a ‘warning notice’ to the alleged offenders, and if the warning notice is complied with it can take no further action. Only if the warning notice is not heeded and the action continues can the Commission take the matter to the Competition Tribunal and seek some sort of punishment. There are a few objections one might raise about such an approach. While it might not be unusual for competition authorities to use prosecutorial discretion on less serious cases and allow violators in less economically important matters to adjust their behaviour without suffering punishing fines, to take even the threat of punishment away from the Commission could greatly reduce its leverage, raising enforcement costs.35 Given the fact that what constitutes serious anticompetitive conduct is not absolutely clear, this could lead to substantial disputes over what constitutes such conduct. Cases that should turn on the effects of certain conduct will instead be determined by whether or not it meets the definition of serious anticompetitive conduct. The requirement that they provide offenders with a free pass may also reduce the willingness of the Commission to pursue First Conduct Rule cases that do not involve serious anticompetitive conduct. Over the long term, this could prove detrimental to the development of a comprehensive competition policy regime in Hong Kong.36 Fourthly, while the CO does authorize private follow-on actions for damages by harmed parties (section 110), it does not allow for such parties to initiate stand-alone actions. While in other jurisdictions in which private competition law enforcement is permitted, many (and possibly most) such actions do follow successful prosecutions by a competition authority, it is also true that stand-alone actions have their place. Thus, the CO approach represents a missed opportunity to supplement the resources available to the Commission with enforcement resources from private parties.37 And it limits victims’ rights to recover damages from harms suffered in cases the Commission, for whatever reason, chooses not to pursue. While it is beyond the scope of this article to review the development of private enforcement of competition law globally, it is certainly the case that an increasing number of jurisdictions are expanding victims’ rights to seek damages from violators of competition law.38 On the other hand, it could be argued that the lack of a class action legal framework in Hong Kong will be the larger barrier to vigorous private enforcement.39 Finally, there is an exemption from the application of the First Conduct Rule in Schedule 1 for ‘agreements of lesser significance’ which are defined to be agreements between parties that have combined turnover of less than HK$200 million. This exemption is not available if the agreements involve serious anti-competitive conduct.40 While it can certainly make sense not to devote significant enforcement resources to matters involving very small firms, many other jurisdictions would deal with this via the exercise of prosecutorial discretion or possibly using consent-type agreements rather than formal proceedings.41 It is a bit troubling to think that the Commission would be powerless to address some anti-competitive agreements or concerted practices that might have significant effects, albeit in smaller markets.42 Abuse of Dominance We focus here on two features of the abuse of dominance provisions, contained in the Second Conduct Rule, which may soften its application relative to what is observed in some other jurisdictions. First, when the Commission feels that a contravention of the Second Conduct Rule has occurred, rather than launching an action before the Competition Tribunal, it may issue an infringement notice offering not to take action if certain conditions are met.43 While it is important to give enforcement authorities the flexibility to resolve less serious matters with consent-style arrangements, the concern would be that the explicit provision of these formal processes could lead them to become practically required. If it becomes difficult for the Commission to support an action under the Second Conduct Rule without first issuing an infringement notice, then parties again—as with the warning notices—will effectively get a free pass.44 Secondly, similar to those under the First Conduct Rule, the CO provides for exemptions from the application of the Second Conduct Rule for ‘Conduct of lesser significance’ meaning conduct engaged in by a firm with less than HK$40 million of turnover in a year.45 We have the same concern here about a statutory exemption (as opposed to the simple application of prosecutorial discretion), protecting behaviour based simply on the size of a firm, irrespective of its relative size or power in a relevant market. In our view, de minimis exemptions have their place, as they provide comfort and guidance for business. However, in contrast to the approach taken in the CO we would argue that: (i) the de minimis thresholds should be expressed in market shares (which are more closely related to market power) than in turnover; and (ii) they are best included in guidelines rather than in legislation. To put them directly into legislation creates at least two potential problems. First, it invites costly and protracted debates about market definition, as respondents push for wider markets that reduce their shares below the thresholds. Secondly, the effects of agreements and behaviours will always be case specific, and there will be the occasional case in which firms with small shares are taking actions or entering agreements that lessen competition.46 The Commission, in our view, should retain the authority to intervene when competition is being threatened. Mergers This is clearly the most obvious example of where the CO lacks the punch of most other modern competition laws. Under the CO, the only mergers subject to review are those in the telecommunication sector, despite the fact that the merger section as drafted is rather general in its approach. This does present the opportunity to add a more complete merger review regime by simply removing the restriction to that one sector; perhaps this was, in fact, contemplated by the drafters.47 A second concern is that this section of the CO is to be enforced by the sectoral regulator.48 One of the advantages of a general purpose competition law and enforcement regime is that it is less vulnerable to regulatory capture than sectoral regulators might be. Delegating merger review to the sector regulator in this case is another example of how the law is softer than it might have been. Finally, the merger provisions allow for a generous consideration of efficiencies. A merger that lowers competition may be allowed if it generates efficiencies that ‘outweigh the adverse effects caused by any lessening of competition in Hong Kong’.49 This language, similar to that found in the Singapore law, might be read to suggest something close to a total surplus approach to a trade-off between competitive harm and efficiencies. While not unique, this is more weight given to efficiencies than is available in most jurisdictions. In contrast to the tougher ‘price’ or ‘consumer surplus’ standards usually associated with the US and Europe, for example, it would permit mergers that harm buyers to some degree if there were sufficient efficiency benefits. We might ask how an authority, charged with preserving and promoting competition within its own jurisdiction, might adapt its practice in the absence of a merger review power. One possibility would involve using other sections of the competition law—here the FCR and SCR—to craft a limited sort of merger regime. Mergers are, after all, agreements—so could a merger be viewed as a harmful agreement under the FCR if it was felt its object or effect was to restrict competition? This would not be unprecedented—many competition authorities with full merger review powers will choose between rules on agreements and rules on mergers when reviewing strategic alliances, for example. Similarly, acquisitions by large firms of small nascent competitors or vertical mergers involving dominant firms that might seem to raise barriers to entry could, one might think, come under review under the SCR. However, these approaches are in fact not possible in Hong Kong: it would appear that the drafters of the CO anticipated this possibility and wanted to block it. Section 4 of schedule 1 expressly prohibits applying the FCR and SCR to mergers.50 If the FCR and SCR cannot be applied to review mergers, removing ex ante control of competition problems, they may be all the more necessary to control anticompetitive agreements and conduct ex post and thereby to discourage mergers with anticompetitive motives. Even FCR and SCR cases with relatively small effects might be worth pursuing aggressively if the actions discourage a significant number of other anticompetitive mergers by, for example, increasing the costs of collusion.51 As a final observation regarding the implications of having no general merger law, we note that Hong Kong is denied a seat at the table when competition authorities from other jurisdictions review international mergers with implications for Hong Kong consumers and businesses. One recent example comes to mind. In 2019 Cathay Pacific Airways took over Hong Kong Express Airways, leaving Cathay Pacific in control of three of the four airlines in Hong Kong. The carrier said the transaction is expected to generate synergies as the businesses of Cathay Pacific and Hong Kong Express are largely complementary. The acquisition was reviewed and approved by the Taiwan Fair Trade Commission (TFTC). The TFTC was of the view that the transition would likely lead to lower prices for Taiwan passengers for flights connecting Taiwan to Hong Kong, Macao, or Shenzhen airports, due to the synergy effects and benefits of scope economy the transaction would bring about.52 While it was not the responsibility of the TFTC to consider the effects on Hong Kong passengers or businesses,53 it is unfortunate that Hong Kong’s interests could not be represented in a review of the transaction.54 V. OTHER INTERESTING FEATURES OF THE NEW LAW In this section, we draw attention to some particular, to us, noteworthy features of the new law and guidelines that did not receive attention above. Though their relationship to the ‘softness’ or ‘toughness’ of the CO may not be so obvious, they present challenges worthy of consideration in any review of the CO. Efficiencies First, with respect to the incorporation of efficiencies in the evaluation of cases, it is striking that the CO applies different standards with respect to the First Conduct Rule, Second Conduct Rule, and for merger review.55 Under the First Conduct Rule’s efficiency exclusions (Schedule 1, section 1), agreements that otherwise might violate the rule will nevertheless be permitted if they generate efficiencies ‘while allowing consumers a fair share of the resulting benefits’. It is not clear what ‘fair share’ represents here, but similar terms in other jurisdictions (eg Europe) have been interpreted to mean that consumers must not be harmed by the agreement. That is, they suggest a consumer surplus or price standard—any upward pressure on price deriving from the agreement must not be greater than the downward pressure coming from falling (marginal) costs.56 Contrast this with the approach to efficiencies under the Second Conduct Rule: there are no statutory provisions for the consideration of efficiencies under abuse of dominance which is unfortunate given that many of the (especially vertical) practices that a dominant firm might employ will have efficiency advantages even as they might harm competition.57 It is possible that the Competition Tribunal (and higher courts) will read in a ‘business justifications’ defence by carefully limiting the application of the ‘object’ test in favour of an ‘effects’ analysis.58 This has happened to some extent in Europe, but it is far from guaranteed to be the outcome in Hong Kong. Finally, as described above, under the limited merger regulations in Schedule 7 there is an exclusion from the merger rule ‘if the economic efficiencies that arise or may arise from a merger outweigh the adverse effects caused by any lessening of competition in Hong Kong’. (section 8). As noted, this language is similar to that found in Singapore’s law where it has generally been interpreted as applying a total surplus test.59 It is not clear what to make of these different approaches—if they indeed turn out to be different as interpretations are provided and decisions rendered. It is at least worth asking why efficiencies might matter more in some cases than in others. To us the most troubling aspect is the lack of an efficiency exclusion for Second Conduct Rule matters, though as noted there might be judicial work-arounds possible here. It might also not seem so unreasonable to have a more generous treatment of efficiencies in merger cases than in cartel cases (admittedly only a subset of First Conduct Rule cases). Putting more weight on efficiencies is essentially like putting more weight on the profits of the firms being investigated. In a merger case we might be more comfortable treating the surplus accruing to the merging firms as socially valuable in a way we might not view the (‘ill-gotten’) gains of cartel members. Competitive Effects Tests In a similar way, the language of the various competitive effects tests varies across the CO. Under the First and Second Conduct Rules, agreements and conduct are problematic if they have as their object or effect the ‘prevention, restriction or distortion of competition in Hong Kong’.60 The related language in the merger rule in Schedule 7 (section 3) is a little different. It prohibits a merger ‘that has, or is likely to have, the effect of substantially lessening competition’.61 Whether the Tribunal and courts will interpret ‘prevent, restrict, distort’ differently from ‘substantially lessen’ is an open and interesting question. One specific aspect of the difference that could matter relates to the absence of the ‘prevent’ term in the merger rule.62 In Canada where the merger provisions contain both terms, they are generally seen to cover different cases, specifically different types of merger counterfactuals. A merger lessens competition when it reduces the level of competition below the level observed immediately before the merger. A merger prevents competition if, but for the merger, competition was expected to increase (possibly because of entry) but the merger prevents this from happening (eg when an incumbent buys the potential entrant).63 Of course, reducing competition in the future is still ‘lessening’ future competition and the Tribunal may have no trouble being forward-looking this way in the right case. However, it is interesting that the prevent language exists in the CO for the First and Second Conduct Rules, which makes its omission in the merger rule somewhat more troubling.64 Object versus Effect in the Second Conduct Rule As noted, under the First Conduct Rule, agreements can be subject to sanction if their ‘object or effect’ is to harm competition. This language and approach would appear to be borrowed from the European model and Article 101 of the Treaty on the Functioning of the European Union (TFEU). However, whereas the Second Conduct Rule has similar ‘object and effect’ language related to abuses of substantial market power, this is not imported from the European abuse of dominance provisions (Article 102 of the TFEU) which instead simply target any abuse of a dominant position. It is not clear why this difference in approaches exists, but one possibility is that there might be a desire to use the ‘object’ test as a way to create per se prohibitions of certain kinds of practices.65 To be fair, the European Commission has, in the past, come close to establishing per se rules for some actions by dominant firms, however the increasing trend now is to look for effects before taking action.66 Issues with the Second Conduct Rule Guidelines There are two issues with respect to the Guidelines produced for the Second Conduct Rule that we highlight here. The first relates to the Guideline’s discussion of ‘substantial market power’ which is required by section 21 (section 1) to pursue a case under the Rule.67 The Guidelines define (at paragraph 3.2) substantial market power this way: ‘Substantial market power can be thought of as the ability profitably to charge prices above competitive levels, or to restrict output or quality below competitive levels, for a sustained period of time.’ (footnotes omitted). To most economists this would function as a standard definition of (any) market power generally, but not ‘substantial’ market power. To be fair, it is not clear how one should define substantial market power; however, it would seem like the definition should be different from the definition of any market power.68 Secondly, there is little attention paid to the so-called ‘cellophane fallacy’ in the market definition section of the Guidelines, which reads much like market definition guidance from merger guidelines. In cases in which firms already possess market power, it has long been recognized that the market definition techniques appropriate to mergers are not adequate. Those techniques ask whether the merging firms are facing enough competition that they could not profitably raise prices if they acted together. In an abuse of dominance or monopolization case however, the firm is already exploiting its market power—knowing that it cannot raise price further does not tell us that it has no market power now. Admittedly, defining markets in such cases can be very difficult, but here the Guidelines could be more helpful by highlighting the challenges to market definition and providing guidance regarding how the Commission will approach this problem. Unfortunately, the Guidelines barely mention the cellophane fallacy, providing only a brief reference to the issue, in footnote 8.69 VI. EARLY ACTIVITY OF THE HONG KONG COMPETITION COMMISSION In this section we briefly review how the Competition Commission has begun its work enforcing the statute.70 Established under Part 9 of the CO, the Competition Commission serves a number of competition policy functions in Hong Kong, listed in section 130 of the CO and in section III above here. In its relatively short life—it was appointed in May 2013 and began recruiting staff at that point—the Commission has been active.71 This activity has been made possible by the provision of substantial government funding. The overall budget provided to the Commission and the number of funded staff positions have also grown over time. In its last annual report, the Commission reported government subventions of approximately HK$105.3 million and employment of a staff of about 57.72 While it is always difficult to compare the levels of funding provided to different competition authorities by their governments, a comparison with Singapore might be of some interest.73 There are many parallels between Singapore and Hong Kong: Singapore’s competition law is also relatively new (the Commission was established in 2005),74 it is also a high income, free market economy in a relatively small Asian jurisdiction. And Singapore’s early experience with competition policy has been widely viewed as successful.75 For fiscal year 2017 (before the Singapore Competition Commission had consumer protection added to its mandate) the Singapore Commission had a total staff of 70 people and a budget of approximately S$17 million.76 With these levels of agency funding both the Singapore (in its last fiscal year before adding consumer protection) and Hong Kong (most recent year) governments are allocating about 0.03 per cent of total government consumption expenditures to their competition authorities. On its establishment, the Commission immediately began work on a set of guidelines (as required by the CO) and to this point has produced an impressively large set of substantial guidance documents:77 (a) Guidelines on the First Conduct Rule (b) Guidelines on the Second Conduct Rule (c) Guidelines on the Merger Rule (d) Guidelines on Complaints (e) Guidelines on Investigations (f) Guidelines on Applications for a Decision under sections 9 and 24 (Exclusions and Exemptions) and Section 15 Block Exemption Orders It has also produced guidance on certain policies: (g) Enforcement policy guidance (f) Leniency policy guidance. The Commission has also become very active in competition advocacy and outreach both within the Hong Kong business community and toward the public at large. The Commission’s creative public advocacy work has included the production of educational videos and television spots.78 A number of the Commission’s advocacy and outreach programmes have garnered international awards from, among other bodies, the International Competition Network. On the enforcement front, the Commission has been active as well. From the Annual Reports we can see where and how frequently they have made enforcement contacts (many based on complaints). A summary is provided in Table 1. Table 1. Enforcement contacts (14 December 2015–31 March 2019): Total = 3607 First Conduct Rule . Second Conduct Rule . Cartel conduct 1010 Tying and bundling 156 Resale price maintenance 411 Exclusive dealing 131 Exchange of 0information 141 Refusal to deal 73 Exclusive dealing 167 Predation 50 Others 430 Others 293 Others General state of competitiona 441 Not related to a Conduct Rule 1106 First Conduct Rule . Second Conduct Rule . Cartel conduct 1010 Tying and bundling 156 Resale price maintenance 411 Exclusive dealing 131 Exchange of 0information 141 Refusal to deal 73 Exclusive dealing 167 Predation 50 Others 430 Others 293 Others General state of competitiona 441 Not related to a Conduct Rule 1106 a The General State of Competition numbers are only available in the 2016/17, 2017/18 and 2018/19 Annual Reports Source: Each case may involve allegations of multiple types of anticompetitive conduct or agreements Open in new tab Table 1. Enforcement contacts (14 December 2015–31 March 2019): Total = 3607 First Conduct Rule . Second Conduct Rule . Cartel conduct 1010 Tying and bundling 156 Resale price maintenance 411 Exclusive dealing 131 Exchange of 0information 141 Refusal to deal 73 Exclusive dealing 167 Predation 50 Others 430 Others 293 Others General state of competitiona 441 Not related to a Conduct Rule 1106 First Conduct Rule . Second Conduct Rule . Cartel conduct 1010 Tying and bundling 156 Resale price maintenance 411 Exclusive dealing 131 Exchange of 0information 141 Refusal to deal 73 Exclusive dealing 167 Predation 50 Others 430 Others 293 Others General state of competitiona 441 Not related to a Conduct Rule 1106 a The General State of Competition numbers are only available in the 2016/17, 2017/18 and 2018/19 Annual Reports Source: Each case may involve allegations of multiple types of anticompetitive conduct or agreements Open in new tab From this table it is clear that concerns have been raised frequently about many types of potentially anticompetitive conduct or agreements. Particularly noteworthy is the large number of contacts related to cartel activity. The matters that reach the stage of initial assessment are similarly dominated by First Conduct Rule (mostly cartel) issues. This is shown in Table 2. Table 2. Initial assessment and investigation cases (14 December 2015–31 March 2019): 230 First Conduct Rule: Cartel conduct 175 First Conduct Rule: Exchange of information 43 First Conduct Rule: Resale price maintenance 41 First Conduct Rule: Others 40 Second Conduct Rule 49 Others 7 First Conduct Rule: Cartel conduct 175 First Conduct Rule: Exchange of information 43 First Conduct Rule: Resale price maintenance 41 First Conduct Rule: Others 40 Second Conduct Rule 49 Others 7 Source: Initial Assessments are commenced in response to Enforcement Contacts and other intelligence gathered by the Commission. For more information on how the Commission conducts Initial Assessments, see paragraphs 3.1–4.4 of the Commission’s Guideline on Investigations. As above, each case may involve allegations of multiple types of anticompetitive conduct or agreements. Open in new tab Table 2. Initial assessment and investigation cases (14 December 2015–31 March 2019): 230 First Conduct Rule: Cartel conduct 175 First Conduct Rule: Exchange of information 43 First Conduct Rule: Resale price maintenance 41 First Conduct Rule: Others 40 Second Conduct Rule 49 Others 7 First Conduct Rule: Cartel conduct 175 First Conduct Rule: Exchange of information 43 First Conduct Rule: Resale price maintenance 41 First Conduct Rule: Others 40 Second Conduct Rule 49 Others 7 Source: Initial Assessments are commenced in response to Enforcement Contacts and other intelligence gathered by the Commission. For more information on how the Commission conducts Initial Assessments, see paragraphs 3.1–4.4 of the Commission’s Guideline on Investigations. As above, each case may involve allegations of multiple types of anticompetitive conduct or agreements. Open in new tab Another interesting feature of this data is the frequency of resale price maintenance (RPM) contacts and assessments. While consistent with a general sentiment that RPM is common in Hong Kong, it does raise concerns should the Commission choose to take many actions here. The concern would be particularly acute should the Commission try and succeed with the Tribunal in having RPM assessed on an ‘object’ basis, possibly making it effectively per se illegal. RPM could then offer tempting easy wins for the Commission, despite the fact that the economic literature has now provided many explanations for how RPM can be efficiency enhancing and international trends seem to be toward a more effects-based approach to RPM cases.79 In the three years since the law took effect in December 2015, the Commission has taken five cases into formal proceedings before the Competition Tribunal. All five involve allegations of serious anticompetitive conduct violating the First Conduct Rule. On 17 May 2019, the Competition Tribunal handed down decisions in two of these cases. While in both cases the Tribunal found violations of the First Conduct Rule, some of its reasoning is problematic in light of concerns raised above. The first case involved bid-rigging for the provision of IT services to the Hong Kong Young Women’s Christian Association80 and the second involved a price-fixing and market sharing agreement related to the provision of renovation services at a public housing site in Hong Kong.81 On the positive side, the decisions make clear that the Tribunal is prepared to find a violation of the First Conduct Rule by object in cases of price fixing, market allocation and bid-rigging. However, the Tribunal did not go all the way toward establishing these practices as per se violations because it was willing to consider, and did consider, efficiency arguments.82 Though it found the efficiencies insufficient in the renovation services case, the fact that the Tribunal has demonstrated a willingness to consider them suggests that cases alleging these most serious of anti-competitive behaviours will be more difficult and costly to pursue.83 We can expect that many respondents will now try to mount some sort of efficiency defences even if weak and we would consider this not consistent with international best practice.84 Compounding this problem, the Tribunal, in its IT decision indicated that the possibility of significant monetary penalties (among other reasons) meant that respondents in First and Second Conduct Rule cases are owed the protections of normal criminal proceedings. Very importantly, this means that the criminal standard of proof (‘beyond a reasonable doubt’) must be applied. While such a high standard may be appropriate (and is observed in other jurisdictions) for individuals facing truly criminal sanctions for offences that are per se in nature, this is also not consistent with international best practice for cases in which effects need to be established.85 It can be very difficult to prove to such a high level, that, for example, certain behaviours of dominant firms are harmful to competition given that many such actions (eg very low pricing) can have benefits as well as costs. We will undoubtedly learn more as the third, fourth, and fifth cases work their way through the process.86 The third and fourth cases are very similar to the second in the nature of the conduct and the industry (ie decorations and renovations in public housing).87 Importantly, the third and fourth cases include charges against individuals and not just companies. The fifth case is somewhat different. On 22 January 2020 the Commission offered and had accepted its first infringement notice, with Nintex Propriety Limited in an alleged bid-rigging scheme involving IT services. Importantly, the notice was part of a leniency agreement—the first leniency agreement—reached with the Commission in exchange for Nintex’s cooperation.88 As infringement notices cannot include fines for participants, one would hope they would not often be used for FCR violations involving serious anticompetitive conduct. However, they make sense as part of a leniency arrangement under which the cooperating party might already qualify for exemption from monetary punishments, as is the case here. Enforcement actions in this matter continue against the other cartel participant.89 Beyond all this important enforcement activity, the Commission has been engaged in a number of non-contested matters, including conducting a market study into petroleum retailing in Hong Kong,90 granting a block exemption for vessel sharing in the liner shipping industry,91 and evaluating exclusion applications regarding the creation of a code of banking practices92 and to permit the collection and sharing of sales data by members of the Hong Kong Pharmaceutical Industry.93 These were all substantial projects. Taken together then, we believe this represents a high level of activity for a new agency. VII. CONCLUSIONS In summary, we have in this article tried to register and support three key points about the new competition policy regime in Hong Kong. First, while the CO is in general a modern competition law, it does appear to be less aggressive in its approach to anticompetitive agreements, practices and mergers than most professional observers may have wished for.94 We concede that a more tentative approach might have been necessary to build support for any kind of competition law at all—Hong Kong is late adopter of modern competition law after all. However, we hope the CO will be seen as just a first step toward a full ‘best-practices’ system. In particular we would urge: (i) development of an expanded general merger review system; (ii) efforts to make naked cartel agreements clearly per se violations while treating other agreements (particularly vertical agreements, including resale price maintenance) under a rule of reason; (iii) adding provisions for stand-alone private actions for damages, at least from collusion; (iv) a higher maximum for fines (particularly for collusive agreements); (v) the end of warning notices; (vi) the Commission be given the power to compel participation in market studies; and (vii) the replacement of the de minimis statutory exemptions laid out in the CO, with the exercise of prosecutorial discretion, to be explained in Competition Commission guidelines.95 Secondly, early cases and policy work under the current law will help shape this version of the CO in important ways and could even address some of our listed suggestions. For example, they will—and already are—clarifying the extent to which the CO is seen to require criminal law processes and where burdens of proof for pieces such as efficiency exemptions reside. It will also be important to see if the list of statutory bodies subject to the CO will be greatly expanded, and what kinds of cases will earn exemptions from the Chief Executive. Finally, we observe that the Competition Commission has quickly become an active enforcement body. It has been handling large numbers of complaints, undertaken ambitious advocacy campaigns and it has now engaged in litigation on five cases. It has also produced, for such a young agency, an extraordinary number of detailed guidelines. The cases being litigated will are serving as important tests for the Commission, just as they are for the new law. Success here and with the next few cases—which we hope will include second conduct rule issues—will do much to further establish the Commission among the community of modern, skilled competition agencies. The authors gratefully acknowledge very valuable discussions with Dennis Beling, Rasul Butt, Thomas Cheng, Carter Chim, Lilla Csorgo, Jindrich Kloub, Kelvin Kwok, Catrina Lam, Tim Lear, Philip Monaghan, Sharon Pang, Steven Parker, Marcus Pollard, Pierre Regibeau, Derek Ritzmann, Brent Synder, Rose Webb, Anna Wu, and Natalie Yeung, though the views expressed here are solely those of the authors and should not be attributed to these good people. Very valuable comments were received from two referees. The authors would also like to recognize the very capable research assistance provided by Jennifer Ng. Footnotes 1 Hong Kong Special Administrative Region (SAR), Laws of Hong Kong, Chapter 619: Competition Ordinance (2012). 2 We recommend Kelvin Hiu Fai Kwok, ‘The New Hong Kong Competition Law: Anomalies and Challenges’ (2014) World Competition 541 for an early, more law-focused, and highly complementary assessment of the CO. 3 See, eg Kanis Leung, ‘Tightening of Hong Kong’s Competition Laws to Cover Mergers on the Horizon, says Competition Commission chairwoman’ South China Morning Post (17 January 2019). <https://www.scmp.com/news/hong-kong/hong-kong-economy/article/2182425/tightening-hong-kongs-competition-laws-cover>. See also Conventus Law (17 January 2019) <http://www.conventuslaw.com/report/the-hong-kong-competition-ordinance-2018-in-review/> accessed 29 June 2020. 4 Our review of some key elements of the history of competition policy here will be very brief and is intended only to set the stage for our evaluation of the Competition Ordinance and its early enforcement. More detailed information about the history can be found in Thomas Cheng, ‘Trade Associations and Cartel Conduct under the New Hong Kong Competition Law Regime’ in Thomas Cheng, Sandra Marco Colino and Burton Ong (ed), Cartels in Asia: Law & Practice (Wolters Kluwer Hong Kong Limited 2015) 295, Sandra Marco Colino, ‘Punishing Cartel Behaviour: Means to Encourage Compliance with the Hong Kong Competition Ordinance’ in Thomas Cheng, Sandra Marco Colino and Burton Ong (ed), Cartels in Asia: Law & Practice (Wolters Kluwer Hong Kong Limited 2015) 315 and Knut Fournier, ‘A New Competition Agency Learns to Deal with SMEs: The Case of the Hong Kong Competition Commission’ in Michael T Schaper and Cassey Lee (eds), Competition Law, Regulation and SMEs in the Asia-Pacific: Understanding the Small Business Perspective (ISEAS–Yusof Ishak Institute 2016) 345. 5 In particular, sections 7K (anticompetitive practices including agreements), 7L and 7N (both related to abuse of dominance) and 7P (mergers). See Edward KY Chen and Ping Lin, ‘Competition Policy under Laissez-Faireism: Market Power and its Treatment in Hong Kong’ (2002) 21(2) Review of Industrial Organization 145 and Thomas K Cheng, ‘A Tale of Two Competition Law Regimes - The Telecom-Sector Competition Regulation in Hong Kong and Singapore’ (2007) 30 World Competition 501 for discussions of competition regulation in the telecom sector in Hong Kong under this sectoral regime and the historical background of competition policy development in Hong Kong. 6 In particular, sections 13 (on concerted practices and agreements) and 14 (abuse of dominance). A critical review of the success of the competition policy under this sectoral approach, with a particular focus on broadcasting, can be found in Thomas K Cheng, ‘Competition Law Enforcement in the Television Broadcasting Sector in Hong Kong: Past Cases and Recent Controversies’ (2010) 33 World Competition 317. Sandra Marco Colino, ‘A History of Competition: The Impact of Antitrust on Hong Kong’s Telecommunications Markets’ (2109) 29(3) Fordham Intellectual Property, Media & Entertainment Law Journal 931, examines the evolving competition rules for telecommunications sector. 7 See, Competition Policy Review Committee, Report on the Review of Hong Kong’s Competition Policy (2006) <https://www.cedb.gov.hk/citb/doc/en/publication/cprc.pdf> accessed 29 June 2020. 8 The Federation further states that: ‘Contrary to the popular belief that introducing a cross-sector competition law would benefit SMEs, there is evidence that the law would provide a convenient avenue for large corporations to sue their smaller counterparts for anti-competition. Since many SMEs cannot afford to pay the huge legal costs involved, not to mention the time and energy required of management in such lawsuits, large corporations could eliminate competitors in the courtrooms without having to compete with them in the market place.’ See Federation of Hong Kong Industries, ‘Public Consultation on the Way Forward for Competition Policy in Hong Kong’ submission to the Hong Kong Economic Development and Labour Bureau (2007). 9 See, eg Ping Lin and Jingjing Zhao, ‘Recent Amendments to Hong Kong’s Competition Bill’ (2012) 1(6) CPI Asia Antitrust Column <https://www.ln.edu.hk/econ/staff/plin/CPI2012asialinzhao.pdf> accessed 29 June 2020. For much more detailed treatments of the very interesting—and counterintuitive to some—approach taken by SMEs in the debates surrounding a new cross-sector competition policy for Hong Kong, see Cheng (n 4) and Fournier (n 4). 10 The irony of Hong Kong not having a competition law when the much less market-oriented economy of Mainland China had one (the Anti Monopoly Law of 2007), was not lost on observers. 11 Hong Kong Competition Commission, ‘Guideline on the First Conduct Rule’ (July 2015) paras 6.71–6.77. For a detailed discussion of the treatment of RPM in Hong Kong, see Ping Lin, ‘Treatment of Resale Price Maintenance in Hong Kong’ (October 2015) Competition Policy International: Antitrust Chronicle <https://www.competitionpolicyinternational.com/treatment-of-resale-price-maintenance-in-hong-kong/> accessed 29 June 2020. 12 While the CO does not put ‘experts’ (competition economists, experienced business people etc.) on the Tribunal, it does provide that the Tribunal may draw assistance from ‘specially qualified assessors’ (schedule 141(1)) though such assessors will not vote on a matter. 13 Kwok (n 2). 14 On the differences between judicial and administrative approaches to competition policy see, eg Frederic Jenny, ‘The Institutional Design of Competition Authorities: Debates and Trends’ in Frederic Jenny and Yannis Katsoulacos (eds), Competition Law Enforcement in the BRICS and in Developing Countries: Legal and Economic Aspects (Springer International Publishing 2016) 1. 15 Of course, most modern competition jurisdictions that use administrative structures do allow appeals from authority decisions to a higher level (and independent) body. Jenny (n 14) reports that the administrative approach dominates among European member states. 16 Jenny (n 14) also includes Australia, Ireland, Austria, and Sweden as using this model. Early discussions about the design of the proposed law in Hong Kong did contemplate using an administrative model. The move to the judicial model was, in part, due to a decision in a securities case: Koon Wing Yee v Insider Trading Tribunal, (2008) 11 HKCFAR 170. It its decision, the Court of Final Appeal asserted that, when substantial punishments were possible, a respondent in Hong Kong had a right (under the Bill of Rights) to some of the protections afforded defendants in criminal proceedings. These protections could (though the court did not speak to this directly) include the right to a hearing before an independent judicial body. 17 This is similar to the limits under the Singapore law. 18 Information about the maximum fines in many other countries can be found in International Competition Network Cartels Working Group, ‘Setting of Fines for Cartels in ICN Jurisdictions’ in Report to the Seventh International Competition Network Annual Conference (Office for Official Publications of the European Communities 2008) <https://www.internationalcompetitionnetwork.org/wp-content/uploads/2018/05/CWG_SettingFines.pdf> accessed 29 June 2020, particularly in the table at pp 35–38. Kwok (n 2) and Colino (n 4) are among other authors who worry that the level of maximum punishments under the CO may be too low. 19 However, this lack of a cap on penalties to individuals contributed to a potentially problematic finding by the Competition Tribunal (discussed below) that could make prosecuting individuals very difficult. 20 At this writing there are six such bodies listed. See the Competition (Application of Provisions) Regulation, (17 April 2015) <https://www.elegislation.gov.hk/hk/cap619A> accessed 29 June 2020. For an excellent discussion of the ‘generosity’ of this blanket exemption of statutory bodies, see Kwok (n 2) 556–59, where the author indicates that 575 statutory bodies qualify for this exemption (at p 597). Related to this, and also discussed by Kwok (n 2) 559–60 are the broad powers allocated to the Chief Executive in Council to grant exemptions from the CO. 21 See, eg pp 87–90 of Richard Whish and David Bailey, Competition law (OUP 2012). 22 On market studies, including examples of different market study powers by country, see, eg OECD (Organization for Economic Cooperation and Development), The Role of Market Studies as a Tool to Promote Competition: Background Note by the Secretariat, Unclassified DAF/COMP/GF(2016)4, <https://one.oecd.org/document/DAF/COMP/GF(2016)4/en/pdf> accessed 29 June 2020. 23 eg Canada and South Korea, OECD, Methodologies for Conducting Market Studies – Summaries of Contributions, Unclassified DAF/COMP/WP3/WD(2017)30, <https://one.oecd.org/document/DAF/COMP/WP3/WD(2017)30/en/pdf> accessed 29 June 2020. 24 eg Australia [OECD, Methodologies for Conducting Market Studies–Note from Australia, Unclassified DAF/COMP/WP3/WD(2017)1 <https://one.oecd.org/document/DAF/COMP/WP3/WD(2017)1/en/pdf> accessed 29 June 2020], Singapore [OECD, Methodologies for Conducting Market Studies -Note by Singapore, Unclassified DAF/COMP/WP3/WD(2017)1 <https://one.oecd.org/document/DAF/COMP/WP3/WD(2017)24/en/pdf> accessed 29 June 2020] and the US Federal Trade Commission [OECD , Methodologies for Conducting Market Studies -Note by the United States, Unclassified DAF/COMP/WP3/WD(2017)19 <https://one.oecd.org/document/DAF/COMP/WP3/WD(2017)19/en/pdf> accessed 29 June 2020]. 25 eg the UK [OECD (n 23)], Mexico [OECD (n 22)] and Iceland [OECD (n 22)]. When actions can be taken directly following the study, the process is more often referred to as a market investigation, rather than a market study. 26 The Commission noted its lack of powers in its report on Hong Kong’s auto-fuel market: ‘It is important to take note that this market study is not conducted as part of an investigation, and therefore the Commission does not have compulsory information gathering powers at its disposal and has to rely heavily on stakeholders’ willingness to provide information and materials gathered from the public domain’ ( at para 1.4 of Hong Kong Competition Commission, ‘Report on Study into Hong Kong’s Auto-fuel Market’ (2017) <https://www.compcomm.hk/en/media/press/files/Full_Report_Auto_fuel_Market_Study_Report_Eng.pdf> accessed 29 June 2020.). 27 One additional aspect, not discussed above, of the CO that suggests a desire for a softer, slower implementation is the unusual (in a statute) requirement that the HKCC issue guidelines: (i) on its interpretation of the conduct rules (section 35 1(a)); (ii) on its interpretation of the block exemption provisions (section 35 1(a)–(b)); (iii) on the procedures it will adopt for investigations (section. 40); and, (iv) on how it will conduct merger reviews (Schedule 7, section 17)—and that the full implementation of the CO waited until after the required guidelines had been issued. In terms of providing greater certainty to business, this may well have been an excellent idea. (Though drafting guidelines before you have case experience to draw from is a challenge.) 28 As has been the case in Europe under the European Commission. 29 As discussed below, this is not a hypothetical possibility. The Competition Tribunal has found violations of the FCR ‘by object’, however, in its second case the Competition Tribunal did admit evidence on efficiencies in a collusion matter: CTEA2/2017 Case name: Competition Commission v W. Hing Construction Company Limited and Others. This case is discussed briefly below. Our point here is simply that it will take additional cases for us to get a good sense of the Tribunal’s tolerance of efficiency arguments in cartel cases. 30 The relatively low levels of maximum fines, discussed above, further limits the attraction of the leniency programme. 31 As discussed below, there is a question, addressed in a recent judgment from the Competition Tribunal, as to what extent criminal procedures might be required in proceedings under the CO, even if the punishments are not criminal. 32 For a nice collection of articles addressing some of these questions see Caron Beaton-Wells and Ariel Ezrachi (eds), Criminalising Cartels: Critical Studies of an International Regulatory Movement (Hart Publishing 2011). Other significant contributions include Mark Furse, Criminal Law of Competition in the UK and in the US: Failure and Success (Edward Elgar Publishing 2012); Andreas Stephan, ‘Why Morality Should Be Excluded from the Cartel Criminalisation Debate’ (2012) 3(2) New Journal of European Criminal Law 127 and Andreas Stephan, ‘Four Key Challenges to the Successful Criminalization of Cartel Laws’ (2014) 2(2) Journal of Antitrust Enforcement 333; Gregory J Werden, ‘Sanctioning Cartel Activity: Let the Punishment Fit the Crime’ (2009) 5(1) European Competition Journal 19; Peter Whelan, ‘A Principled Argument for Personal Criminal Sanctions as Punishment Under EC Cartel Law’ (2007) 4(1) Competition Law Review 7; Peter Whelan, The Criminalization of European Cartel Enforcement: Theoretical, Legal, and Practical Challenges ( OUP 2014);Peter Whelan, Report Examining the Desirability of Introducing Criminal Sanctions for Cartel Activity (Submitted to the Finnish Competition and Consumer Authority) (2014) <https://www.kkv.fi/globalassets/kkv-suomi/ajankohtaista/tiedotteet/2014/whelan-selvitys-27-5-2014.pdf> and Wouter Wils, ‘Is Criminalization of EU Competition Law the Answer?’ (2005) World Competition 117. 33 We do not want to exaggerate the magnitude of this trend: Gregory C Shaffer, Nathaniel H Nesbitt and Weber Waller Spencer, ‘Criminalizing Cartels: A Global Trend?; Comparative Competition Law’ (2011) 12 Sedona Conference Journal 313, also Chapter 12 in John Duns, Arlen Duke and Brendan Sweeney (eds), Research Handbook on Comparative Competition Law (Edward Elgar Publishing 2015) 301 discuss the movement toward criminalization but point out that in some cases criminal prohibitions have not been actively enforced. 34 Calling this a ‘free pass’ is something of an exaggeration we (and others) use for convenience. Of course, a firm receiving a warning notice will have costs associated with its own legal representation and managerial response—and there is a chance that a firm may suffer some reputational damage. 35 In the reforms that created the Canadian Competition Act in 1986, abuse of dominance cases also offered essentially a ‘free pass’ to violators. Many objected to this free pass, see eg comments by the then Commissioner Sheridan Scott in a presentation to the US Department of Justice in 2006: ‘We are on record, supported by others such as the OECD, that a lack of financial consequences for dominant firms found to have abused their position is a significant shortcoming of the current legislation.’ <https://www.justice.gov/atr/abuse-dominance-under-competition-act>. In March of 2009, an amendment to the Act introduced “Administrative Monetary Penalties (AMPs) to this section. 36 Kwok (n 2) 562 makes this point, as well. 37 Interestingly, the draft bill did allow for stand-alone actions, but this was removed in the final bill. See, eg Lin and Zhao (n 9) and Colino (n 4). This would appear to be another example of a compromise designed to address SME concerns that large firms would launch legal actions against them. Again, see Cheng (n 4). 38 eg on how private enforcement has been spreading among European countries see Luigi Pier Parcu, Giorgio Monti and Marco Botta, ‘Chapter 1: Introduction’ in Pier Luigi Parcu, Giorgio Monti and Marco Botta (ed), Private Enforcement of EU Competition Law : The Impact of the Damages Directive (Edward Elgar Publishing 2018) 1 and Christopher H Bovis and Charles M Clarke, ‘Private Enforcement of EU Competition Law’ (2015) 36(1) Liverpool Law Rev 49. 39 Law Reform Commission of Hong Kong, Report: Class Actions (May 2012)advocated an incremental approach to the development of a class action regime for Hong Kong and specifically did list ‘Antitrust/competition cases’ in a list titled ‘Types of cases that might be suitable for class action proceedings’ in Annex 1 (p 275). <https://www.hkreform.gov.hk/en/docs/rclassactions_e.pdf> accessed 29 June 2020. 40 In debates about the appropriate thresholds to put into the law, the Bills Committee was told that the average annual business turnover of small and medium-sized enterprises in Hong Kong was HK$11 million. (See Hong Kong Legislative Council, Bills Committee on Competition Bill, Minutes of the twenty-fourth meeting held on Tuesday, 15 November 2011, LC Paper No CB(1)1427/11-12, Ref: CB1/BC/12/09 (2011) 9.) Hence, this threshold would be enough to exempt groups of up to 18 firms with this average size from actions that could reduce competition in localized markets, eg in many consumer-facing service industries—as long as the actions do not represent serious anti-competitive conduct. 41 To provide greater guidance, this more informal approach to smaller companies can be spelled out in guidelines, as has been done in Singapore. 42 We note, as well, that turnover is not a satisfactory measure of market power. As described in Lin and Zhao (n 9) and Cheng (n 4) these safe harbours were offered as a compromise to build support for the new law in the face of resistance, particularly from SMEs. Thresholds based on turnover do, however, avoid protracted debates about market definition required if thresholds are to be expressed in terms of market shares. As stated in the text, our preference would simply be to express thresholds as part of enforcement guidelines to be flexibly interpreted in ways consistent with the overall objectives of the Ordinance. 43 Infringement notices may also be issued to resolve concerns about violations of the FCR related to serious anti-competitive conduct. In 22 January 2020 the Commission offered and had accepted its first such infringement notice, with Nintex Propriety Limited. The case is discussed briefly below. 44 A late amendment to the bill removed a provision that would have allowed infringement notices to come with fines, but this was removed. See Lin and Zhao (n 9). 45 Schedule 1, section 6. 46 This second point reinforces the first. If the Commission always has the authority to intervene there is less to be gained by respondents trying to game the system by arguing for wider market definitions that reduce their shares below thresholds. Thresholds inserted into legislation further reinforces the importance of the market definition exercise at a time when some experts are urging a movement away from formal market definition in competition cases. See, eg Louis Kaplow, ‘Why (Ever) Define Markets?’ (2010) 124(2) Harvard Law Rev 437. 47 Kwok (n 2) 549–51 discusses some of the arguments made when the initial version of the CO was submitted for debate in 2010. 48 Strictly speaking, the CO grants concurrent jurisdiction to the Communications Authority and the Competition Commission (at section 159), however the Memorandum of Understanding between the two agencies indicates that the Communications Authority will ‘ordinarily take the role of Lead Authority on matters which fall within the concurrent jurisdiction of the Authorities’. 49 Schedule 7, Part 4, section 8. 50 To be clear, using the FCR and/or SCR for merger review in place of a proper merger review process could be problematic. Challenges to this approach would come from the simple and obvious fact that the conduct rules are not designed for merger review—there is now a fairly well-developed and broadly applied (by advanced competition authorities) approach to merger review that is seen as reasonably fair, transparent and predictable. It is less clear what merger review under the FCR or SCR would look like. 51 To be clear here, we are not advocating that the Commission take FCR or SCR cases that are not meritorious under those rules. eg we do not support actions on excessive pricing (not explicitly covered by the SCR in any case, but see Kwok ( n 2). 545 suggesting they cannot be ruled out completely) simply to discourage anticompetitive mergers. 52 See, TFTC Newsletter, No 134, (in Chinese) <https://www.ftc.gov.tw/upload/1081106-2.pdf> accessed 29 January 2020. 53 It is possible that Hong Kong passengers could be affected differently from Taiwanese passengers even on routes between Taiwan and Hong Kong—for example if the market is somewhat segmented (eg due to loyalty programmes) such that Hong Kong passengers tend to use Hong Kong-based airlines while Taiwanese passengers tend to use Taiwan-based airlines. As the transaction combined Hong Kong airlines, it may have more effect on their prices and services than on those of the Taiwanese airlines. Add to this concern for routes served by the Hong Kong airlines but not protected by competition from Taiwanese carriers and it is clear that Hong Kong passengers may have been harmed even if those from Taiwan are not. 54 Brent Snyder, CEO of the Hong Kong Competition Commission said that the transaction ‘constitutes a merger and places the transaction beyond the scope of the Ordinance. That said, the Commission is actively engaging with the government to encourage a competition assessment of the acquisition to the extent permissible within the relevant regulatory ambit and also to offer the Commission’s services in that regard’. See, Speech by Brent Snyder, at British Chamber of Commerce and Freshfields Breakfast Briefing: The Hong Kong Competition Ordinance 3rd Anniversary 30 April 2019; <https://www.compcomm.hk/en/media/campaigns_events/completed/files/20190430_BritCham_Freshfields_Breakfast_Briefing_Speech.pdf> accessed 29 January 2020. 55 While striking, Hong Kong is not alone in its different treatment of efficiencies; eg the differences between the treatment of efficiencies under the First and Second Conduct Rule here parallel those in Europe. See OECD, Directorate for Financial and Enterprise Affairs, Competition Committee, Roundtable on the Role of Efficiency Claims in Antitrust Proceedings: Note by the Delegation of the European Union, DAF/COMP/WD(2012)81, <http://www.oecd.org/competition/EfficiencyClaims2012.pdf> accessed 29 June 2020. 56 The exclusion does not apply if the agreement eliminates competition; see Schedule 1, section 1(c). 57 Consider, eg the various efficiency reasons a dominant firm might have for using resale price maintenance. See, eg Chapter 6 in Massimo Motta, Competition Policy: Theory and Practice (CUP 2004),or Frank Mathewson and Ralph Winter, ‘The Law and Economics of Resale Price Maintenance’ (1998) 13(1–2) Review of Industrial Organization 57. 58 It is also possible that, if the Commission wanted to allow the parties to make efficiency claims, a vertical agreement by a dominant firm with another firm could be evaluated under the First Conduct Rule. 59 The Canadian Competition Act (at section 96) contains similar language that has been interpreted to mean something close to a total surplus standard: mergers are exempt if the efficiencies ‘are greater than, and will offset’ the harm to competition. See Thomas W Ross and Ralph A Winter, ‘The Efficiency Defense in Merger Law: Economic Foundations and Recent Canadian Developments’ (2005) 72(2) Antitrust Law Journal 471. 60 This is the exact language from the Second Conduct Rule (section 21 (1)). The language for the First Conduct Rule (section 6 (1)) is virtually identical: an undertaking must not participate in an agreement or concerted practice if ‘the object or effect of the agreement, concerted practice or decision is to prevent, restrict or distort competition in Hong Kong’. 61 As indicated above, this language in the Merger Rule of the CO is exactly the same as that in the relevant section of the Hong Kong Telecommunication Ordinance. Specifically, s 7P(1) of the Telecommunication Ordinance (2012) provides that where there is a change in relation to a carrier licensee, the Telecommunication ‘Authority may conduct such investigation as it considers necessary to enable it to form an opinion as to whether or not the change has, or is likely to have, the effect of substantially lessening competition in a telecommunications market;…’ One possibility is that at the time of introducing the Competition Ordinance, the law-makers inserted the language in the existing Telecommunication Ordinance into the Merger Rule which applies to telecommunications sector only, so as to maintain legal continuity. 62 The Canadian Competition Act’s merger provisions (section 92) seek to block mergers that ‘prevent or lessen competitive substantially’. 63 See Thomas W Ross, ‘Competitive Effects and Efficiencies: The Canadian Supreme Court’s Decision in Tervita’ (2016) 2 Competition Law & Policy Debate 54 for a discussion of an important ‘prevent’ case in Canada that raises questions about the ability of an authority to predict future changes in the market. 64 Note, as well, that the merger test requires that the effect on competition be substantial; a term not included in the competitive effects tests under the conduct rules. As with the treatment of efficiencies discussed above, this may reflect a more permissive attitude toward mergers than the other competition matters. 65 Also, recall that there are no efficiency exclusions under the Second Conduct Rule. 66 See, eg Whish and Bailey (n 21). 67 While not a focus for us here, the requirement of ‘substantial degree of market power’ rather than ‘dominance’ (as under art 102 of the TFEU in Europe) is actually an example of an element that is possibly tougher than it could have been—though in practice, the differences may not be material. The ‘substantial’ test appears to be borrowed from the Australian and New Zealand legislation. 68 The Canadian provisions on Abuse of Dominance do not need to get into this exactly. They apply to firms that ‘substantially or completely control .. a class or species of business’ which the Competition Bureau has interpreted to mean that the firm has market power (but no requirement of substantial market power). The Singapore law requires dominance which is presumably more than just some market power. Their Guidelines define market power the standard way (as above) but then add that dominance requires ‘substantial market power’ (paras 3.3 and 3.4). 69 In contrast, the Singapore Guidelines on Market Definition devote three paragraphs to this issue (5.5–5.7). These paragraphs more clearly lay out the challenge posed by pre-existing market power and suggest some approaches toward dealing with the challenges. 70 For a discussion of the Commission’s first year enforcing the Ordinance, see Ping Lin and Thomas W Ross, ‘First Year of Enforcement of the Competition Ordinance in Hong Kong’ (March 2017) Competition Policy International: Antitrust Chronicle <https://www.competitionpolicyinternational.com/first-year-of-enforcement-of-the-competition-ordinance-in-hong-kong> accessed 29 June 2020. 71 And might have been even more active but for the unfortunate passing of its first Chief Executive Officer, Stanley Wong, in April 2016 after less than two years of service. 72 This is from the 2018/19 Annual Report. Compared with fiscal year 2013/14 this represents growth of over 62% in annual funding with employee numbers almost tripling. 73 Comparisons are difficult for a variety of reasons, including potential differences in the agencies’ mandates (eg is the authority also responsible for consumer protection and/or sectoral regulation) and differences in costs of key inputs like skilled labour. 74 The Singapore law is also heavily based on the European model. Importantly, consumer protection responsibilities were added to the renamed Competition and Consumer Commission of Singapore in 2018. 75 See, eg p 3 of Kala Anandarajah and Dominique Lombardi, ‘Competition Law in Singapore’ (August 2015) Competition Policy International: Antitrust Chronicle <https://www.competitionpolicyinternational.com/competition-law-in-singapore/> accessed 29 June 2020; ‘The CCS has clearly established itself as a serious regulator not to be ignored and established Singapore as one of the leading countries in the implementation and enforcement of competition law.’ 76 These numbers for Singapore are from OECD, Annual Report on Competition Policy Developments in Singapore (2017) Unclassified DAF/COMP/AR(2018)27 <https://one.oecd.org/document/DAF/COMP/AR(2018)27/en/pdf> accessed 29 June 2020. 77 Available from the Commission’s website at: <https://www.compcomm.hk/en/legislation_guidance/guidance/guidance.html> accessed 29 June 2020. 78 They are available at: <https://www.compcomm.hk/en/media/advertisements/tv_video.html> accessed 29 June 2020. 79 To be fair, not every jurisdiction has moved to a more effects-based treatment of RPM—Europe’s move has not been smoothly in that direction, for example. However: with its Leegin decision in 2007, the US Supreme Court adopted an effects-based approach to minimum RPM; in amendments to its Competition Act in 2009, Canada moved from a per se criminal treatment of RPM to an effects based civil review approach; Australia and New Zealand have both allowed for authorizations for RPM if there are efficiencies and an authorization was granted in Australia to Tooltechnic in 2014; and Singapore’s relatively new (2004) law does not have any specific provisions related to RPM. 80 CTEA1/2017 [2019] HKCT 2. Case name: Competition Commission v Nutanix Hong Kong Limited, BT Hong Kong Limited, SiS International Limited, Innovix Distribution Limited (trading as Innovix Distribution) and Tech-21 Systems Limited. 81 CTEA2/2017 Case name: [2019] HKCT 3. Competition Commission v W. Hing Construction Company Limited and Others. The third case, not yet decided is similar in many respects: CTEA1/2018, Case name: Competition Commission v Kam Kwong Engineering Company, Goldfield N&W Construction Company Limited, Pacific View Engineering Limited, Chan Kam Shui and Lam Po Wong. 82 The Respondents’ argued, eg that allocating floors to specific firms under the agreement allowed for efficiencies in materials delivery costs, labour time (not as much time spent moving between floors), and in marketing and cleanup costs. The Tribunal was not persuaded that the requirements for a successful efficiency defence were met. See paras 147–281 of the decision. 83 To be fair, the Tribunal set a fairly high bar (and put the onus on defendants) to establish an efficiency defence. For the defence to stand, four conditions must be met: (i) the agreement must generate efficiencies; (ii) it allows consumers a fair share of the resulting benefit; (iii) it does not impose restrictions that are not indispensable to the attainment of the objectives; and (iv) it does not afford the undertakings concerned the possibility of eliminating competition. (at para 153). 84 There is also a danger that the willingness to consider efficiencies will undo the benefit of being willing to find violations ‘by object’. The advantage of labelling some hard-core conduct like price fixing as a violation by object is that it avoids long drawn-out battles over the presence and magnitudes of effects. However, if efficiencies are to be considered, in cases in which they are found to be substantial they will need to be weighed against the harm to competition. It will be difficult to do this without some measure of the anticompetitive effects of the agreement. 85 This would appear to leave Hong Kong occupying an odd middle ground in competition policy targeting hard-core cartel activity—without the deterrent power of true criminal law (eg jail sentences) that has so helped encourage use of leniency programmes, but with a legal process that gives respondents the protections of a criminal law approach. 86 We should also note that there are appeals underway in the first two cases and a decision on the pecuniary penalties in the second case is still pending. Therefore, there is also more information to come from these cases. 87 The third case is: CTEA1/2018, case name: Competition Commission v Kam Kwong Engineering Company, Goldfield N&W Construction Company Limited, Pacific View Engineering Limited, Chan Kam Shui and Lam Po Wong. The fourth case is: CTEA1/2019, case name: Competition Commission v Fungs E & M Engineering Company Limited, Yee Hing Metal Shop, Cheung Min trading as Accord Construction & Decoration Co, Hing Shing Construction Company, Luen Hop Decoration Engineering Company Limited, Dao Kee Construction Company Limited, Wong Wai Chuen, Wong Fu San and Cheung Yun Kam. 88 See the press release at <https://www.compcomm.hk/en/media/press/files/20200122_ENG_PR_Competition_Commission_takes_IT_cartel_conduct_case_to_Competition_Tribunal.pdf> accessed 29 June 2020. 89 This fifth case is: CTEA1/2020, case name: Competition Commission v Quantr Limited and Cheung Man Kit. 90 Hong Kong Competition Commission (n 26). 91 Block exemption granted August 2017. See, eg <https://www.compcomm.hk/en/enforcement/registers/block_exemption/files/Block_Exemption_Order_and_Guidance_Note_final.pdf > accessed 29 June 2020. 92 Decision denying the application for exclusion issued October 2018. See, eg <https://www.compcomm.hk/en/media/press/files/COMMISSION_DECISION_UNDER_SECTION_11.pdf> accessed 29 June 2020. 93 Decision denying the application issued October 2019. See, eg <https://www.compcomm.hk/en/media/press/files/20191022_Competition_Commission_publishes_Decision_in_relation_to_a_proposed_pharmaceutical_sales_survey_eng.pdf> accessed 29 June 2020. 94 To be clear, other authors such as Kwok (n 2) have come to similar conclusions about some of the issues we raise. Our goal here is to combine views (our own and others) on many elements of the current CO to build a case for a more robust competition law for Hong Kong. 95 If the removal of the de minimis exemptions was to meet with too much resistance, a weaker alternative would be to make the exemptions in terms of market shares, and keep the exempted class very small. © The Author(s) 2020. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: [email protected] This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)