The Institutional Design of Competition Authorities: Debates and Trends

Abstract

The issue of institutional design of competition authorities has attracted increasing interest since the early 2000 but requires further elaboration. This article attempts to fill some gaps by providing a general framework to examine a number of dimensions of this issue under three headings: the goals, the functions and the organization of competition authorities. While there is no unique institutional design which would fit all countries, a number trade-offs should be considered in designing a competition authority. These trade-offs may lead to different designs across countries depending on the local conditions. Ultimately choosing the best possible design for the competition authority given the local conditions is crucial to ensure that the competition authority is most effectively able to discharge its duties.

1 Introduction

The issue of the institutional design of competition authorities has attracted increasing interest since the early 2000 for a variety of reasons.

Prominent among the reasons for which the issue of the institutional design of competition authorities has become an increasingly important topic of discussion is the fact that as competition authorities have become more prominent and powerful in a number of countries they have also become more conscious of the fact that they need to be (and to be seen to be) effective in discharging their duties. Thus in a number of countries, there have been recent changes in the institutional design of competition authorities (for example, in Europe, in Denmark, in the Netherlands, Spain and the United Kingdom) or there are changes contemplated (for example in Australia). Some of those changes have been partly spurred by economic constraints on government in periods of low growth (for example in the Netherlands), partly spurred by the desire to increase the effectiveness of the competition law system (for example in the United Kingdom), partly spurred by the desire to better integrate regulatory policy and competition policy (for example in Spain).

The fast increasing globalization of markets which has characterized the last decade of the twentieth century and the first decade of the twenty-first century has led to an increasing interest both in facilitating international trade and in promoting the convergence of competition law regimes in trading nations. This has led to reflections on what the competition law regimes should converge on. Although this debate has been largely focused on the substance of the competition analysis, it has also touched upon institutional issues. For example, there has been a lively debate on the importance of ensuring that competition authorities throughout the world are in a position to examine speedily transnational merger transactions which fall under their domestic merger control law and therefore on the importance of the adequate funding of competition authorities. Similarly, the question of the independence of competition authorities has been raised as some of the important exporters or foreign investors feared that they would be treated unfavourably by competition agencies in some countries in which such agencies seemed to be dependent on the national government or national dominant firms .

Furthermore, in a world in which a large number of countries have recently adopted a competition law and created a new competition authority (the number of competition authorities in the world has increased by at least 40 over the last 20 years), there has been an increasing demand on the part of developing countries for guidance on the institutional design they should adopt for their newly created competition institution.

Finally, a number of well known competition specialists have produced influential articles about the design of competition authorities. The most prolific and influential of those authors, Bill Kovacic, has had a long standing interest in the issue of institutional design and has long argued in articles and conferences that it was important for competition authorities to devote more attention to the issue of the relationship between the goals of competition law, the effectiveness of the agency and its institutional design.¹ Other prominent competition authors such as Philip Lowe or Eleanor Fox have also contributed to the discussion .²

The scope of what one should consider to be the institutional design of a competition authority is extremely wide as it covers every aspect of the governance of the authority, of its internal organization and of its relationship with the outside (be it the government, parliament, the business community ). Rather than attempting to systematically cover all the bases this article focuses on a limited but significant number of dimensions of the institutional design of competition authorities which have been recently publicly discussed. It builds primarily on a set of OECD Competition Committee Roundtable on institutional design as well as on some OECD roundtables on issues related to specific dimensions of the institutional design of competition authorities which were held over the last 15 years. The OECD Competition committee held its first roundtable on the optimal design of a competition agency in its Global Forum on Competition in February 2003. It held a second roundtable on changes in the institutional design of competition authorities in December 2014 and again in the spring of 2015. In between the OECD Competition Committee held roundtables on the relationship between Competition Authorities and Sectoral Regulators in 2005 and on the Interface between Competition and consumer Policies in 2008. Besides the work of the OECD, the ICN also did work on institutional issues, for example through its Agency Effectiveness Project the results of which were presented In Kyoto during the ICN annual conference in 2008.

This chapter will discuss a number of questions related to the institutional design of competition authorities regrouped under three main themes: the goals of competition authorities, the functions of competition authorities and the organization of competition authorities.

For each theme we will show the diversity of situations prevailing across jurisdictions and explain the main justifications for each institutional design.

A short conclusion will follow.

2 The Goals of Competition Authorities

The question of what are the goals of competition authorities is by no means new and it has been the object of repeated discussions over the last 10 years. It was first discussed in the OECD Competition committee in May 1992. Then it was discussed in the OECD Global Forum on Competition in February, 2003.³ Finally this issue was raised again in a recent debate on institutional changes at OECD in the December 2014.⁴

In 2003, the OECD secretariat note⁵ observed that “the basic objectives of competition authorities were to maintain and encourage the process of competition in order to promote efficient use of resources while protecting the freedom of economic action of various market participants”. It also noted that competition policy was also viewed to achieve or preserve a number of other objectives as well: pluralism, de-centralisation of economic decision-making, preventing abuses of economic power, promoting small business, fairness and equity and other, socio-political values.

Consumer surplus

A lively debate around the goals of competition law took place in the United States in the aftermath of the publication of Robert Bork “the Antitrust Paradox”⁶ in the late seventies. The debate turned around the question of whether the standard for illegality under competition law should be a “consumer welfare” test or a “total welfare” test. However as Herbert Hovenkamp observed⁷: “the volume and complexity of the academic debate on the antitrust welfare definition creates an impression of policy significance that is completely belied by the case law, and largely by government enforcement policy”. Indeed, as J. Kirkwood and R.H Lande found in 2008⁸ and as Hovenkamp observed in 2013 the reality is that US enforcement agencies have consistently follow a consumer welfare standard.

Over time the narrow economic goal of protection of consumer surplus has gained wide acceptability.

Wider economic goals

However, whereas nearly all competition authorities are concerned with the protection of the consumer surplus, there are differences of opinion about whether the protection of consumer surplus is a natural result of competition or an underlying goal of competition law.

Furthermore, among the jurisdictions for which consumer surplus is indeed a goal of competition law, there are differences of opinion between those which consider that consumer surplus is the only goal of competition and those which consider that competition law enforcement may also have other economic goals.

Finally, among the jurisdictions for which consumer surplus is one of the economic goals of competition there are differences between those which consider that economic goals are the sole goals of competition law and those for which competition law may also have social or political goals.

In 2011 the International Competition Network published a document on “Competition Enforcement and Consumer Welfare”.⁹ It recorded the responses to a 2010 ICN survey.¹⁰ Many respondents stated that even if consumer welfare were an important end goal, economic growth in general and total welfare were the more specific goals of competition law.

In countries like Australia, Norway or Swaziland, the goal of competition law is the protection of total welfare rather than consumer welfare.

Thus, in Australia, the ACCC has powers to grant exemption from competition law in certain circumstances, such as where benefits to the public from the anticompetitive conduct outweigh the detriment that the conduct may cause. In assessing benefits to the public the ACCC may have regard to total welfare effects.

In Norway, the goal of the Competition Act is: “(…) to further competition and thereby contribute to the efficient utilization of society’s resources.”

The competition authority of Swaziland also uses a total welfare standard and noted in its response to the 2010 ICN survey that “besides consumers, there are other equally important stakeholders, such as competing businesses, and that this can lead to the importance of ensuring welfare of groups other than consumers”. The strategic goal of the Competition Authority of Swaziland is thus to promote active competition for the public benefit.

In Kenya competition law sometimes seeks to maximize producer and consumer surplus, not consumer surplus alone.

Among the countries that have a broader economic agenda than the strict promotion of consumer surplus, one may also include Germany, Hungary, Iceland, Ireland Korea, Switzerland or Uzbekistan. In Germany, according to a recent draft guideline issued by the Bundeskartellamt, the purpose of merger control is “to protect competition as an effective process,” which the draft guidelines explain “may sometimes coincide with protecting competitors.”¹¹ In Hungary, the goals of the competition law are the maintenance of effective competition and the promotion of efficiencies. The Icelandic Competition Act aims to promote effective competition and thereby increase the efficiency of the factors of production of society. According to the Irish Competition Authority, the primary goal of its work is to ensure competitiveness in the Irish economy, which will ultimately benefits the consumer (although the benefits of this law enforcement activity might not always be immediately clear to consumers). The main goal of Switzerland’s Cartel Act is to prevent the harmful economic or social effects of cartels and other restraints of competition.

Non-economic goals

Besides broader economic goals than the promotion of consumer surplus, a number of competition laws also have social or political goals. These might include, for example, the promotion of employment, regional development, national champions (sometimes couched in terms such as promoting an export-led economy or external competitiveness), national ownership, economic stability, anti-inflation policies, social progress, poverty alleviation, the spread of ownership stakes of historically disadvantaged persons, security interests and the “national” interest. In addition, a number of domestic competition laws in Europe include the Treaty of Rome objective of market integration within the European Union.

As the OECD Secretariat noted in 2011¹²: “The specific objectives behind merger control (…) may differ between jurisdictions”. “(…) For example, protecting local or small and medium size competitors, achieving various socio-economic and socio-political objectives, protecting employment, encouraging enterprise, and achieving various industrial policy objectives including promoting the international competitiveness of the local economy and building strong national firms.”

In the note prepared for the discussion of the objectives of competition law and policy which took place in the OECD Competition Committee in 2003,¹³ the Secretariat offered the view that: “Among OECD countries, there appears to be a shift away from use of competition laws to promote what might be characterised as broad public interest objectives, and use of public-interest based authorisation procedures, exemptions or political over-rides (collectively, “public interest objectives”) in competition laws, that contemplate a consideration of factors which extend well beyond what appear to be the generally accepted “core” competition policy objectives of promoting and protecting the competitive process, and attaining greater economic efficiency (the “ core competition objectives”)”.

In hindsight, this assessment seems to have been overly optimistic. It is true that countries which did not have a public interest provision in their competition law did not add such provisions to their competition law. But it is equally true that (1) a number of countries which had a public interest provision in their competition law did not eliminate them and that (2) a number of developing countries which have since adopted a competition law have included public interest provisions in their law.

Among the countries which had a public interest goals in their competition law and did not eliminate them (even if they use them sparsely) , one can mention Canada. The goals of competition law in Canada are to promote the efficiency and adaptability of the Canadian economy, to expand opportunities for Canadian participation in world markets while at the same time recognizing the role of foreign competition in Canada, and to ensure that small and medium-sized enterprises have an equitable opportunity to participate in the Canadian economy.

Similarly, the Korean competition law goals are a mix of economic and non-economic goals. Article 1 of Korea’s Monopoly Regulation and Fair Trade Act (MRFTA)¹⁴ states that “The purpose of this Act is stimulate the creative initiative of enterprisers, to protect consumers, and to strive for the balanced development of the national economy by promoting fair and free competition through the prevention of the abuse of market dominance and excessive concentration of economic power by enterprisers and through regulation of improper concerted practices and unfair trade practices”.

With respect to the developing countries which have public interest clauses in their law one can mention that the Competition Act of South Africa and that of Namibia have very wide goals that are both economic and non-economic.

The purpose of the South African Competition Act¹⁵ is to promote the efficiency, adaptability and development of the economy; to provide consumers with competitive prices and product choices; to promote employment and advance the social and economic welfare of South Africans; to expand opportunities for South African participation in world markets and recognize the role of foreign competition in the Republic; to ensure that small and medium-sized enterprises have an equitable opportunity to participate in the economy; and to promote a greater spread of ownership, in particular to increase the ownership stakes of historically disadvantaged persons.

Similarly, the Anti-Monopoly Law of China (the “AML”), which took effect in 2008, has a variety of goals including “the protection of fair competition in the market” and “the interests of consumers,” but also “the promotion of the healthy development of the socialist market economy.” Another stated objective of the Chinese AML is to protect the “lawful business operations” of undertakings in industries “controlled by the State-owned economy and concerning the lifeline of national economy and national security.”

The goals of single functions competition law institutions

There are three themes of discussion around the goals of competition law.

The first one is a discussion on why competition authorities’ goal should be to protect consumer surplus; the second is a discussion on why competition authorities should not have other goals besides the protection of consumer welfare and the third one is a discussion about why competition authorities should not have public interest goals.

First, the justification for following such a standard is usually that the balance between consumer welfare losses and the attributable efficiency gains would be too complicated for competition authorities to perform.

Second, the justification for competition authorities having only one (economic) goal is that non-competition policy mechanisms are generally superior for achieving noncompetition policy objectives. In other words, it is considered that restricting competition in order to achieve a broader policy objective, whether economic or not, will have inevitable anti-competition side effects, e.g. granting protected monopoly profit to a firm or firms.

Third the justification for not entrusting competition authorities with “public interest” goals is that broadly specified policy objectives can be ambiguous and as such are subject to “capture” or “hijack” by the politically strongest private interests, usually those of producers or workers. Thus de jure public interest objectives may de facto serve private interests.

Those justifications are not fully satisfactory.

First, with respect to whether competition authorities should follow a consumer surplus standard rather than a total welfare standard, we know that in doing so, competition authorities may not take into consideration efficiency gains which could outweigh the consumer loss due to the increase in price and reduction in output due to the anticompetitive nature of the practice or the transaction. Thus, using the wrong test may entail a social cost. On the other hand assigning a total welfare standard to the competition authority may lead it to err in its judgment because of the complexity of the assessment it has to do. Thus it may also entail a social cost. Assuming that society is risk neutral, the question is then whether the expected cost of the errors due to the adoption of a consumer welfare test (i.e. the probability of such an error multiplied by its cost when it happens) is larger or smaller than the expected cost of errors that would happen if the competition authority had to perform a more complex task by following a total welfare test. This a difficult question and we are not aware of any empirical work which would support the choice of a consumer surplus standard over that of a total welfare standard.

Second with respect to the idea that using a restriction in competition to achieve a broader policy objective will entail a social cost, this argument is convincing from society’s point of view only if the alternative ways to fulfill the broader policy objectives (presumably through another agency) do not entail social costs which larger than the ones incurred if the competition authority restricts competition to fulfill these objectives. If, for example, the agency in charge of fulfilling these other objectives has a higher chance of being captured or if the consequence of its actions is to restrict competition more than what the competition authority would have deemed necessary to fulfill those objectives, it may be that letting another agency fulfill these objectives may end up being more costly to society than entrusting the competition authority with the fulfilment of these objectives.

Finally, with respect to the argument that public interest goals may lead to a capture of the competition authority by private interest, one can argue that the competition authority will not be as easily captured as another agency dedicated to the fulfillment of these public interest clauses would be. Also, one could argue that the competition authority, precisely because it is in charge of promoting competition, will be more restrained in the enforcement of the public interest provisions than other parts of government would be.

Thus the objections to the fact that the competition authority may have to enforce public interest provisions when they enforce competition law are unconvincing because they fail to consider the possible costs of alternative solutions.

For sure, a number of critics of public interest provisions in competition laws would prefer that such provisions did not exist. They rightly point out that the enforcement of such clauses may lessen the intensity of competition and be contradictory with the objectives of competition law and policy. But what they fail to acknowledge is that in a number of countries, particularly developing countries (for example in South Africa or in China), the only alternative is between a competition law containing public interest provisions and no competition law at all. In this second best situation it is arguable whether or not such clauses should be tolerated.

The issue of the goal of competition law must also be considered in relation with the possibility of multiple functions of competition authorities, a question to which we now turn.

3 The Functions of Competition Authorities

The second dimension of institutional design we want to explore is the question of the functions of competition authorities.¹⁶

Competition enforcement and consumer protection

The first sub-question is that of knowing if competition authorities should also be entrusted with consumer protection responsibilities. Over the recent years, quite a number of OECD countries have changed their institutional design from that point of view. For example, basing ourselves on the submission to a recently held roundtable on Changes in Institutional Design at the OECD Competition Committee (December 2014), it appeared that seven countries have merged the competition and the consumer enforcement functions in a single agency since the beginning of the century (Denmark (2010), Finland (2013), Ireland (2014), Italy (2007, 2014), Korea (2006, 2008), Lithuania (2000), Netherlands (2013)). But three jurisdiction have separated consumer protection from competition partly (in the case of the United Kingdom (2013–14)) or completely (in the case of Iceland (2005) and Japan (2009)). Finally, four countries (Brazil (2012), Bulgaria (over the last few years), Estonia (2008), Chinese Taipei (2005)) have considered merging those functions and decided against doing it. Altogether nearly half of the competition authorities of the OECD countries have consumer and competition law enforcement functions whereas the other half do not have a consumer protection function. In some countries, where there are several competition authorities, the picture is even more complex because one agency is a single function competition authority whereas the other one has both a competition and a consumer enforcement function (this is the case in the US where the US FTC has both competition and consumer protection enforcement functions whereas the DoJ is a single function competition authority and in France where the Autorité de la concurrence is a single function competition agency whereas the competition division of the Ministry of economic Affairs (DGCCRF) has both a competition enforcement function (at the local level) and a consumer protection function.

These figures reflect a certain ambivalence about the wisdom of merging the two functions. The arguments in favour of merging the functions and against merging them have been extensively researched in a background paper prepared by Allan Fels and Henry Ergas for the above mentioned discussion of Institutional changes in the OECD Competition Committee.¹⁷

They, first observe that each of the two policies can be used to advance the goals also pursued by the other: “competition policy, by keeping markets effectively competitive, can reduce the work that needs to be done by consumer policy; consumer policy, by enhancing the ability of consumers to exercise choice, can help make markets more effectively competitive and force firms to compete on the merits, thereby supporting the ends of competition policy”. As former FTC Chairman Timothy Muris has said, “The policies that we traditionally identify separately as ‘antitrust’ and ‘consumer protection’ serve the common aim of improving consumer welfare and naturally complement each other.”¹⁸

But Fels and Ergas also note that each policy can create challenges for the other.

They thus note that “ when a market becomes more exposed to competition than it was previously (say, because of the removal of trade barriers or deregulation), the incentives of market participants may change in ways that raise consumer protection concerns” and that in some sectors consumers may have a difficult time coping with the complexities of competition . Examples are numerous and would include the fact that the opening up to competition of a number of sectors in transition economies has led to deceptive practices that required consumer protection, that the introduction of competition into some public utility markets (such as electricity and telecommunications) has given an incentive to firms to lock in consumers so as to avoid losing customers to the competitors. Consumers may have difficulties dealing with complex pricing schemes on service markets (such as in banking) or be exposed to risks in competitive markets when they cannot assess the quality of services (such as on professional services markets).

Equally consumer protection may lessen competition both by imposing constraints on suppliers (such as a ban on comparative advertising or the imposition of regulatory standards) or by promoting transparency which may lead to a weakening of competition.

Fels and Ergas then assess the growing importance of behavioural economics and changes in the extent and functioning of markets on the debate on the relationship between the two sets of policies. They point out that in recent years researches in behavioural economics have explored issues about the inherent limitations on the quality and efficacy of consumer choice. Those studies have important implications for policy design, most obviously of consumer protection measures. It should also be noted that market forces may in some cases be important ways of addressing concerns about the efficacy with which consumers take complex choices, because firms in competitive markets have incentives to offer consumers solutions that allow potential gains from trade to be more fully realised. It remains, however, that there are cases where the two policies should interact and be coordinated (such as for example in professional services, health care).

Altogether according to Ergas and Fels, there are three major advantages to integrating the primary responsibility for competition policy and consumer policy within a single institution. There can be advantages from using those two as instruments that can be flexibly combined and more generally managed within a single portfolio of policy instruments; second there are possible gains from developing and sharing expertise across these two areas (for example to develop an understanding of the interaction between the supply side and the demand side of markets); and third the visibility and understanding of consumer and competition policies may be greater if they are integrated ion the same agency.

Whereas the second and the third advantages are widely recognized, there is less consensus about the usefulness of combining in a flexible way the policy instruments of competition law and consumer protection. For example, FTC Commissioner Maureen Ohlhausen recently stated¹⁹ “In some cases, the FTC has blurred the line between competition and consumer protection—with respect to both the alleged violation and the remedy sought by the agency—to the potential detriment of effective and transparent enforcement in both areas. This blurring of the lines, while in some sense an integration of competition and consumer protection principles, is more accurately viewed as an improper and unhelpful muddying of the two disciplines”.

But integrating consumer and competition policy may also entail costs. There are differences in the nature of the instruments and in the ways in which the policies are implemented²⁰ more limited instruments in the case of competition policy than consumer policy, a large number of smaller cases in consumer protection, a small number of larger cases in competition enforcement, geographically localized policy in consumer protection and centralized policy in competition). These differences may create practical difficulties in the management of consumer protection and competition policy within a single organization.

Furthermore, because of these differences and the fact that consumer policy is inherently less centralized than competition policy, the degree of integration between these policy instruments may be difficult or impossible to achieve.

Finally commentators have mentioned other possible practical difficulties of integrating those functions within an agency such as the potential for one mission to dominate the other to the detriment of the latter, a lack of clarity of purpose of the agency, resulting in diminished support for the agency’s overall mission, the potential for “destructive rivalry” between the competing missions within an agency for prestige, headcount, and budgetary resources²¹

As a result of these conflicting tendencies, Ergas and Fels conclude:

In practice, what appears most important is:

• To ensure that the competition authority has in-house access to the skills involved in the formulation of consumer policy, and at the very least a watching brief with respect to consumer policy, as well as scope to intervene in consumer policy decisions that have material competition implications; and

• That there be within government, an entity that has “whole of government” oversight of consumer protection, and that exercises that oversight in a manner mindful of competition concerns.

It is useful, keeping this approach in mind, to eek the perspective expressed by the competition authorities which have merged the two functions and by those which have decided against such a merger during the OECD roundtable on Institutional changes.

Among the countries which have recently merged the two functions (Denmark 2010, Finland 2013, Ireland 2014, Italy 2014, Korea 2008, Lithuania 2000, Netherlands 2013), the motivation most frequently mentioned in the OECD discussion on institutional changes are the increase in the effectiveness of both policies, the development of synergies between consumer policy and competition policy, and the development of expertise in the understanding of market mechanisms. Furthermore two countries (Ireland and the Netherlands) indicate cost saving as an important determinant of the merger of the functions and one country (Korea) indicates that the objective was to make consumer policy more consistent with competition policy.

Thus the reasons for which competition enforcement and consumer protection have been brought together in those countries are broadly in line with the complementary nature of the two policies outlined by Ergas and Fels and the desire to make both policies more effective.

For example, echoing the assessment of Ergas and Fels, the Irish contribution the OECD Roundtable on Institutional design²² explained that the rationale for “amalgamation” of the two functions in this country rested on the idea “that combining competition enforcement, consumer protection and consumer awareness in one body will build a more effective organisation which is better equipped to foster a pro-competition culture across the economy. An independent authoritative body provides a source of consistent information to business and consumers about their rights, and provides administrative savings and skill enhancement through the pooling of information, skills and expertise. Competition authorities are expert in assessing how firms compete with one another thanks to an internationally accepted toolkit for competition analysis while the enforcement of consumer law brings awareness of problems that arise in business to consumer transactions even in markets that are competitive. In addition, the rapid rise of behavioural economics has given regulators deeper insight into how consumers actually make choices in competitive markets. The experience of deregulation has shown that supply side reform on its own is not sufficient to ensure that all consumers fully benefit from competition as there may be behavioural barriers which prevent consumers from making the best choices for themselves or indeed unfair commercial practices causing consumer harm before and after they buy. The increasing awareness of behavioural issues in competitive markets serves to reinforce the logic of having competition and consumer experts working side-by-side. In newly competitive markets there tends to be gaps in understanding among consumers and this confusion can be exploited by firms. This gap can be bridged by co-ordinating consumer and competition policy”.

In addition, the Irish competition authority considered that there could also be operational advantages from having the competition and consumer functions within the one regulator. It stated “For example, a competition case might raise concerns about consumer harm due to market power but there might be insufficient evidence or constrained resources to bring an enforcement action. Having a single agency overseeing both competition and consumer protection allows the different courses of action to be considered simultaneously”.

The contribution of Poland to the OECD roundtable on changes in institutional design²³ emphasized the particular complementarity of the two policies in transition economies where there is no widespread culture of market economy. It stated: “A consumer perspective in competition enforcement is of particular importance in transition economies, where market liberalisation is often, rightly, a key policy objective as a means of creating foundations for long-term growth and consumer welfare. However, short-term impact on consumers cannot be ignored. A liberalized market must from the start meet consumer expectations with regard to access, choice, price, quality, security and reliability, and must be independently regulated and enforced. From UOKiK’s experience, we often see that such liberalisation aimed at long-term benefits for consumers may result in short-term infringement of consumer rights. This is why we believe that impact assessment accompanying legal regulatory changes needs to include a consumer impact forecast for both the short and the long run. A competition and consumer protection agency is well placed to offer government a balanced view in this respect during the legislative process. It is also well positioned to counteract any short-term negative effects of market liberalization without jeopardizing its long-term benefits.

A practical example would be the electricity markets. Since the opening of residential retail markets in Poland in July 2007 there have been numerous problems with door-to-door selling. In this case, antitrust law is not the solution. This issue should be addressed through other means such as legislation on commercial practices, trade standards etc. Door-to-door selling became a major source of consumer dissatisfaction shortly after the retail market was fully opened up to competition. The bulk of consumer complaints focused on the fact that they were being misled into signing contracts to switch suppliers when they were under the impression that they were only agreeing to approve a visit from a consultant, obtain information or have their meters read. UOKiK is currently conducting a number of proceedings against the most aggressive suppliers. These cases show that market liberalisation may create incentives for unfair, deceptive and unlawful business practices, against which our consumer protection law is the only defence. Similar problems occurred during the liberalisation of the telecom market in the early 2000s. However, actions undertaken by the telecom regulator as well as the competition authority to create a diverse market along with consumer rights enforcement have led to a substantial improvement in the sector, as demonstrated by today’s fierce price and quality competition as well as fewer consumer complaints.”

It seems that the concerns about the difficulty of integrating the two policies have not been a major concern in those countries.

The objective of cost savings invoked by some countries may not be met to the extent that, as Ergas and Fels argue, the integration of both instruments may be quite challenging given their different natures. Yet it seems reasonable to assume that some cost saving can be achieved in the support functions (such as communication, personnel, general administration etc….) when the two instruments are merged in a single institution.

In those countries which merged the consumer protection and the competition authority, different concerns were raised at the time of the merger or shortly after.

In several case there was a negative reaction on the part of consumer representatives about the merger of the functions (for example in Korea) or a concern that either competition law enforcement would come to dominate consumer protection or that “easy” consumer protection cases would crowd out the “more difficult” competition cases. This last consideration was , for example, the reason advanced by the Monash Business Policy Forum in Australia to advocate the separation of the consumer functions from the ACCC. It argued that such a separation was necessary in order to²⁴ “free a potential bias in the present operation (of the ACCC) where consumer protection gets more enforcement work because it is easier law to prosecute”.

The difficulty of prioritization of cases in agencies that have both consumer protection and competition law enforcement functions was also mentioned in the contributions to the OECD Competition Committee Roundtable on Institutional Changes. For example, the contribution from Finland²⁵ illustrated the problem it faced in the following way: “In the field of competition law, the legislation practically obligates the authority to prioritize between investigated cases and also gives the right not to investigate insignificant issues, whereas there are no actual provisions regarding prioritization in consumer affairs. However “The Consumer Ombudsman must be active especially in areas which are particularly significant for consumers or where it can be assumed that problems for consumers would most commonly occur”, but in practice as there is a lack of appropriate provisions regarding prioritising, enforcement has to be targeted at all the areas that are defined as being under the aegis of the Consumer Ombudsman”.

In some other case the agency felt it difficult to merge the different cultures of the consumer protection personnel and of the competition enforcers. Those concerns are understandable in light of the difference in the instruments described by Ergas and Fels.

But other concerns were also expressed with respect to the identification of a common strategy and the structure of the new institution.

Finally, in organizational terms it is worth noting that overall either the functions of consumer protection and competition are separated by law (Finland) or the enforcement of consumer protection, and competition are de facto separated (for example in Denmark and in the Netherlands). Ireland seems to follow a more integrated model than the other countries which have merged consumer protection and competition enforcement into a single body. In most cases, however, a number of support functions are merged, such as communication, policy and legislation, strategy.

The market analysis function is integrated (used both for consumer protection and competition enforcement) in Denmark and the detection function is integrated in the Netherlands.²⁶

Finally, it should be noted that in the countries which have decided to unbundle the consumer protection and competition enforcement functions and to create two separate institutions (Iceland and Japan) , the reason given was to increase the effectiveness of competition policy (in Iceland) and to increase the effectiveness of consumer policy (in Japan). These motives may be explained by the difficulty of agencies having the two functions to find the proper balance between them and to prioritize their enforcement activities. This is suggested, for example, by the contribution from Iceland to the OECD Debate on institutional changes²⁷ which stated: “In the view of the Icelandic Competition Authority (ICA), the move from a multifunctional design towards a single functional one has made competition enforcement and advocacy more effective. The fact that the ICA is “solely” responsible for competition enforcement and advocacy, enables a very clear goal-orientation, which in return facilitates prioritization and makes the Authority well equipped to tackle changes in the economic environment. The institutional design has enabled the ICA to put its weight on the most important tasks at any given time, and by that facilitate quality decisions and active advocacy and guidance. The prerequisite for quality decisions is the ability to attract and maintain high-level expertise. The current institutional design has served as a basis for success in this regard. The ICA has also been able to use its focus and goal orientation to prioritize cases with the aim to improve the length of procedures”.

The second sub-question is that of knowing whether the competition enforcement function should be merged with regulatory functions.

Competition and sectoral regulation

There is a diversity of situations throughout the world with respect to the relationship between competition law enforcement and the enforcement of sectoral regulations.

In Australia, for example, the ACCC has a range of regulatory functions in relation to national infrastructure industries as well as a prices oversight role in some markets where competition is limited. According to the Australian contribution to the OECD debate on Institutional changes²⁸ the regulatory functions of the ACCC include: “assessing access undertakings under the ‘National Access Regime’, which facilitates third party access to certain services provided by means of significant infrastructure facilities; a number of responsibilities regarding the National Broadband Network; supporting the development and operation of efficient water markets in the Murray-Darling Basin; and assessing notifications of price increases in relation to certain services (regional air services, services to airports and airlines, and certain services provided by Australia Post)”.Within this model there is a specificity with regard to the energy market. Under the Competition and consumer Act of 2010, the Australian Energy Regulator is an independent entity staffed and funded through the ACCC’s agency appropriation which has some regulatory functions mostly related to energy markets in eastern and southern Australia and which assist the ACCC with energy-related issues arising under the Consumer and Competition Act, including enforcement, mergers and authorizations. Thus in the field of electricity there are two, closely related, regulators, one of which is the ACCC.

At the other end of the spectrum, in the United Kingdom, the sectoral regulators²⁹ have powers to apply some aspects of competition law in relation to their particular industry sector. ‘Concurrently’ with the Competition and Markets Authority they enforce the prohibitions on anti-competitive agreements and abuse of dominance under Articles 101 and 102 TFEU and the UK national equivalents. They also have the power to make a Phase 1 market study and refer a market for a full Phase 2 market investigation by the CMA Panel. These competition powers are in addition to the sector regulator’s regulatory powers.

In between those two extreme models, a number of countries follow a “division of labor” model between the competition authority and the sectoral regulators. For example, in Portugal³⁰ “the powers to enforce and promote competition rules, to defend consumer’s interests as such, and to regulate markets are entrusted to different bodies: the Portuguese Competition Authority, the Directorate-General for Consumers and National Regulatory Authorities, respectively”.

In the recent years there have been changes in the allocation of regulatory and competition law enforcement powers. In a number of countries some of the regulatory functions were given to the competition authority. Such was the case, for example in Denmark (2009) where the competition authority was given regulatory functions in the water distribution sector, in Estonia (2008) where the competition authority was given regulatory functions in the energy, rail, and telecom sectors, in the Netherlands (2013) where the competition authority became the telecom and post regulator, in Spain (2013) where the competition authority became the airports, audio visual products, energy, rail, post, and telecom regulator or in Lithuania (2009, 2011) where the competition authority became the rail regulator.

Conversely in a few countries there was a movement to separate competition law enforcement from sectoral regulatory functions. Such was the case in Denmark (2010) where the Danish Energy Regulatory Authority was separated from the Competition Authority (at the same time that the consumer protection function was added to the competition authority) . This was also the case in Estonia in 2014 where the competition authority which had been given, in 2008, regulatory functions in the energy, water, heating , post, railway, airport, telecom, lost its regulatory functions in the telecom sector (which it previously shared with a technical regulator). As we shall see below there is also a lively debate in Australia on whether the ACCC should keep its regulatory functions.

The arguments in favour of entrusting competition authorities with regulatory functions are the following:

• First the fact that the competition authority will have a more flexible range of instruments to promote and maintain competition, particularly in newly deregulated sectors.

• Second, the fact that the competition authority may be better able to detect/manage policy or enforcement conflicts (e.g., ensuring that a competition remedy does not conflict with regulatory requirements or vice versa).

• Third, the fact that the pooling of sectoral responsibilities may make the agency more adaptable to changing markets (e.g., where convergence is occurring such as in the information sector).

• Fourth, the fact that there is less risk that the competition authority will be captured than the sectoral regulators because competition authorities deal with a wide variety of markets whereas sectoral regulators always deal with the same, comparatively small, number of regulated firms.

Along those lines, Ergas and Fels, examining the allocation of responsibility for regulating the former public utilities to the competition authority (as was done in Australia and in New Zealand) state: “The advantage of (this) approach, at least in theory, is that it extends the range of instruments that the authority can bring to bear. For example, it may be that the most efficient solution to a particular regulatory problem is to restructure the market in ways that promote competition and then more vigorously enforce the competition rules. By ‘internalizing’ into the same authority the competition and regulatory instruments, the authority may be more inclined to efficiently mix and match problems and instruments, avoiding the ‘silo mentality’ that can compromise good decision-making. At the same time, there may be instances where identifying the efficient regulatory solution requires an analysis of competition impacts, which such an integrated authority may find it easier to undertake”.

Against those possible advantages, there are a number of possible difficulties associated with the merging of regulatory and competition law enforcement responsibilities into a single entity.

A first category of difficulties may accrue from the complexity involved in managing different functions. As we saw when discussing the amalgamation of consumer protection and competition law enforcement, prioritization of cases and the efficient allocation of resources becomes more difficult as the number of different functions of the authority increases.

A second category of difficulty may come from the fact that, as the competition authority accumulates different functions, its support is eroded because it becomes more and more difficult for economic actors and the general public to understand what it does and to assess its quality and its accountability.

A third source of difficulty may be due to the complexity of mixing within the same organization staff members having different cultures and approaches (the ex-ante and prescriptive approach of regulators and the ex-post and legalistic approach of competition enforcers).

A fourth source of difficulty may be due to the loss of competition between sectoral regulators and the competition authority in advocating regulatory changes for the regulated sectors. This loss of competition between regulators may entail a social cost for society.

A fifth source of difficulty may be due to the fact that the goals which should be ascribed to an institution which is both a competition policy enforcer and a sectoral regulator are far from clear.

This last point was made by the Dutch contribution to the OECD debate on institutional changes.³¹ As mentioned earlier, the Netherlands Authority for Consumers and Markets was created on 1 April 2013 through the consolidation of the Netherlands Consumer Authority (CA), the Netherlands Independent Post and Telecommunication Authority (OPTA) and the Netherlands Competition Authority (NMa). In its contribution the competition authority stated: “As ACM sees it, one of the authority’s strengths is its focus on consumers. This focus has not gone un-criticized within the Dutch system, where many commentators argue that ACM should more correctly focus on orderly market processes and on competition in the market, rather on the effects on consumer welfare. ACM’s Establishment Act determines that ACM is to ensure that markets function well, that market processes are orderly and transparent, and that consumers are treated with due care. (…) Eighteen months after the merger, ACM can raise these issues for discussion but cannot, as yet, give experience-based answers to these questions”.

Over and beyond the advantages and difficulties previously mentioned from entrusting sectoral regulatory functions to competition authorities, one should note that if regulatory oversight can be complementary to competition law enforcement (for example both may require a common vision of what the relevant markets are and it is clear that the possibility of effective competition on a regulated market is a function of both the structure of the market and the sectoral regulation applicable to it), the deregulation of a market and the establishment of a competitive market is a fundamentally different function than the protection of competition on a deregulated and structurally competitive market. The opening up of a formerly legally monopolized markets to competition, particularly in sectors where the incumbents are managing essential facilities, requires a number of ex ante decisions of an industrial policy nature to establish the possibility of competition such as: at which rhythm should entrants be allowed (to avoid too much competition among the entrants resulting in an inability for each of them to meaningfully compete with the incumbent); which entrants should be chosen; what should be the interconnection obligations of the incumbent monopolist both quality-wise and from the standpoint of the financial terms; what public policy is necessary to decrease the importance of the bottlenecks and to facilitate the development of infrastructures etc…. To discharge these functions it is not clear what the comparative advantage of a competition authority is.

Thus in a number of European countries (France, for example) the choice was made to entrust competition law enforcement and sectoral regulations to different agencies but to ensure that the two agencies would communicate on questions of mutual interest. Thus , for example, the Autorité de la concurrence in France has the duty when it deals with a competition issue in a regulated sector to ask the opinion the sectoral regulator on the technical issues underlying the competition question it deals with. The opinion of the sectoral regulator is not binding on the competition authority but it is made public and the competition authority must explain in its decision why it departs from the opinion of the sectoral regulator. Likewise when the technical regulator is dealing with a technical issue which may have an impact on competition, it must consult the competition authority on the implications for competition of the question it deals with.

Such arrangements allow each institution to fulfill its function and to have the input of the other institution without having to bear the costs of difficulties attached to multi-task institutions. Such a system requires, however a clear delineation of the responsibilities of each institution as well as a clear and transparent procedure for the exchange of opinions between the competition authority and the sectoral regulators. Yet it is not always the case that such responsibilities are clearly delineated and, for example, during a period in the nineties, Spain had a system where the competition authority and the sectoral regulator were simultaneously competent to deal with a number of issues which caused a certain amount of confusion and dissatisfaction. But when the system is well set up, as it is in France, it can run very smoothly to ensuring the cooperation and the consistency of the sectoral and the competition enforcement approaches.

The optimal arrangement when it comes to whether one should entrust sectoral regulatory functions to the competition authority may differ depending on the size of the country. Indeed smaller countries may have difficulties supporting separate institutions given their public resource constraints and the important weight attributed to possible economies of scope or economies of scale in those countries may tip the balance of advantages and costs in favour of multi-function agencies.

Finally, the choice of having a single function competition agency or a multifunction agency with sectoral regulatory powers or a system of cooperation between agencies may also be determined by the economic history and past experiences of the country in the area of deregulation.

For example, in the case of Australia , Ergas and Fels suggest that the decision to confer responsibility for economic regulation of telecommunications on the ACCC was shaped , among other factors, by the perception that the industry-specific regulator had not been a success and by the unfounded expectation that industry-specific telecommunications regulation would ‘wither away’, as a rapid transition to competition was envisaged. They explain the current debate over whether the sectoral regulation functions of the ACCC should be transferred to another institution by the fact that the historical factors in favour of the multi-function agency are not relevant anymore in Australia.

Similarly the contribution of Spain to the OECD debate on institutional changes³² makes the point that the regulatory model for energy and telecommunications based on specialized regulators had been designed at the beginning of the liberalization process but that, as competition developed in both sectors and as the frontier between telecommunication and the digital economy became more blurred, the need for better coordination among sectoral regulators, on the one hand, and between sectoral regulators and competition law enforcers, on the other hand, required a different and more integrated regulatory system which led in 2013 to the creation of the new Spanish National Authority for Markets and Competition.

4 The Organization of the Competition Authority

A third set of question relates to how , given its assigned functions, the competition authority should be organized.

We will try to address the following questions: what are the respective advantages of the prosecutorial model and the administrative model of competition authorities? Should investigation and adjudication be separated? Should competition authorities have a single commissioner or have a board and, in the latter case, what the function of the board should be? How to ensure the independence of the competition authority be (and what one means by independence)? What are the ways to organize the funding of competition authorities? A final question will deal with the management of its resources by the competition authority (recruitment of staff, prioritization of cases, organization of the work between lawyers and economists).

Administrative versus prosecutorial model

In a prosecutorial model, the competition authority prosecutes the cases that it brings in an adversarial proceeding in a courtroom. In such a model the court is the decision maker and not the competition authority. This is, for example the case in the Us (for the Antitrust Division of the Department of Justice), Australia, Canada and Ireland as well as in some European countries (in Austria and in Sweden).

In an administrative model, the competition authority investigates and adjudicates cases. This model is the dominant model among European member states. It is also the model followed by the FTC in the United States and by a large number of countries throughout the world. The administrative model has itself two variants: the variant in which the authority’s decision are appealable to a general court (such as for example in France or in the EU) and the model in which the competition authority’s decisions are appealable to a specialized court (such as in Portugal or in the United Kingdom or in Mexico under the new law of 2013).

The question of whether a prosecutorial model is preferable to an administrative model was hotly debated recently both in the United Kingdom and in Switzerland, two countries which ultimately decided to stick with the administrative model. There was also some discussion along those lines in Germany.

The perceived legal advantage of a prosecutorial model is, first and foremost, the fact that the impartiality of the proceedings is better protected through the separation of investigation and adjudication in a judicial context than in administrative proceedings were those functions are combined in a single entity.

In Europe, however, this argument has not been successful. As mentioned by Slater, Thomas and Waelbroeck³³ “Traditionally, the view is taken that, it is sufficient for Commission decisions in antitrust cases to be subject to review by the Community courts and particularly by the Court of First Instance (“the CFI”), even if the Commission itself is not an “independent and impartial tribunal” under Article 6 ECHR”. The same view applies in the Member states which are (unlike the European Union) signatories of the European Convention on Human Rights. This view is based on the European Court of Human Rights Le Compte, Van Leuven and De Meyere v Belgium judgment,³⁴ in which the European Court stated that: “Whilst Article 6 par. 1 (art. 6-1) embodies the “right to a court” (….), it nevertheless does not oblige the Contracting States to submit “contestations” (disputes) over “civil rights and obligations” to a procedure conducted at each of its stages before “tribunals” meeting the Article’s various requirements. Demands of flexibility and efficiency, which are fully compatible with the protection of human rights, may justify the prior intervention of administrative or professional bodies and, a fortiori, of judicial bodies which do not satisfy the said requirements in every respect; the legal tradition of many member States of the Council of Europe may be invoked in support of such a system”.^{35 , 36}

Similarly in Member States which have signed the European Convention of Human Rights, the Le Compte, Van Leuven and De Meyere v. Belgium, judgment of the ECHR is seen as the basis on which administrative agencies even when they do not meet the standards of an independent and impartial tribunal are not considered to breach the right of parties to a fair trial provided that their decisions can be appealed to such an independent and impartial tribunal.

A second possible advantage of a prosecutorial model is the economic equivalent of the legal advantage previously discussed: a prosecutorial system to avoid the confirmation bias which is likely to characterize the administrative proceedings of a competition authority which acts as investigator and adjudicator.

The possibility of a bias is consistent with behavioural economics and has some empirical support.

Behavioural economics suggests that a confirmation bias occurs when people filter out potentially useful facts and opinions that don’t coincide with their preconceived notions. Behavioural economics holds that such bias affects perceptions and decision making in all aspects of our lives and can cause us to make less-than-optimal choices. Thus, if a competition authority is both a prosecutor and an adjudicator, it may be tempted to confirm and justify as an adjudicator its decisions to prosecute by finding the parties it has decided to investigate guilty of a violation or engaged in an anticompetitive transaction.

This theory has found support in a small number of empirical studies. For example, two economists analyzed the decisions of the US FTC when the FTC sat in appeal of its own administrative law judge decisions following previous FTC decisions challenging mergers and referring them to the FTC administrative law judge for adjudication (see footnote 36). They found that the appeal was much less likely to be successful when the FTC commissioners sitting in appeal were the same as the commissioners who had originally opposed the merger and more likely to be successful when the commissioners sitting in appeal were different from the commissioners who had originally objected to the mergers.

A third advantage of the prosecutorial model is held to reside in the fact that the judicial decision process is (often) more transparent than the administrative process and therefore more credible. This argument was invoked in the United Kingdom by those who were in favour of switching to a prosecutorial model during the discussions that led to the creation of the Competition Market Authority. Indeed, there were complaints about what parties and their counsels perceived to be the lack of transparency of the OFT decision making process and the impossibility to either know who made decisions or to be heard by the decisions makers.

However one should note that those arguments are somewhat inconclusive in the sense that one could conceive of an administrative model in which the prosecution and the adjudication would be separate and done by different staff members and in which the decision making would be transparent.

Thus even if one accepts the usefulness of the separation of investigation and adjudication and of the transparency of the process, it does not follow that the administrative model is necessarily flawed.

Finally, it is sometimes argued that the number of appeals would be lower if the courts rather than the competition authorities made the decisions and that this would save time and money in the enforcement system. This assertion, however is called into question by the fact that judicial proceedings can drag on for a long time. For example, the Annual Report of the Austrian Federal Competition Authority 2011 stated that “the proceedings before the Cartel Court often drag on for years without there being comprehensible reasons for their excessive length”. It mentions cases brought in 2004, 2007 and 2009 but not resolved by the end of 2011. In the case of Sweden, the court proceedings in the TeliaSonera abuse of dominance case started in December 2004 and the Stockholm City Court made a decision on 2 December 2011. Thus the court proceedings in this last case took 7 years (minus 2 years because of a reference to the ECJ).

The prosecutorial model is also frequently considered to have some drawbacks compared to the administrative system.

First, the Courts hearing the competition cases are often not specialized unlike competition authorities with the result that they are less likely to understand the underlying economic issues. The reason is that generalist judges for whom competition cases represent but a small minority of the cases on their dockets, have less incentive to invest their time in learning the intricacies of the economic underpinnings of competition law than specialized judges for whom competition cases represent a large portion (or the entirety) of their caseload. Thus the quality of the lower level decisions in the prosecutorial model may be an issue unless the relevant court is specialized.

Second, because competition cases are often seen as more complex and involving more work for the judges than the other cases coming to the court, because of the difficulty to understand the underlying economic issues, they may not be given a high priority by the courts resulting in delays in the court proceedings.

Altogether a number of the arguments offered in favour or against the administrative model or in favour or against the prosecutorial model appear not decisive. The only decisive advantage that the prosecutorial model offers is that it guarantees a separation between investigation and adjudication, something which is not guaranteed to the same extent in the administrative model.

Separation between adjudication and investigation

As mentioned earlier, within the administrative model several sub-models can exist reflecting varying degrees of separation between investigation and adjudication.

As the EC study on institutional design of competition authorities suggest³⁷ that two main configurations can be distinguished within the administrative model: “the first involves a functional separation between the investigative and decision-making activities of the single administrative institution whereby the inquiry is carried out by investigation services and the final decision is adopted by a board/college/council of this administrative institution. For example, in France and Spain a full functional separation between investigative and decision-making bodies has been set up, where their respective competences are carried out independently from one another. The second configuration follows a more unitary structure and does not have different bodies carrying out different steps in the procedure although there may be different divisions (e.g. a Competition department and a Legal department) inside these authorities that deal with separate aspects of the same case”.

The functional separation of adjudication and investigation is widely considered to have a number of advantages for the competition law enforcement process and to improve the quality of decisions.

The first benefit of the separation of investigation and adjudication is the possibility to avoid mistake by having “a second set of eyes” reviewing the evidence and the proposed qualifications.

This is, for example, largely why, previous to the creation of the CMA in the United Kingdom, merger enforcement was split between two institutions, the Office of Fair Trading and the Competition Commission. The OFT reviewed information relating to merger situations and, where necessary, referred any relevant mergers to the Competition Commission for further investigation if it is felt that the merger was likely to lead to a substantial lessening of competition within any market for goods or services in the UK. This system has now been abandoned with the creation of the Competition and Market Authority.

However within the CMA a separation has been kept between investigation and decision making in the context of the CMA for mergers reaching what was once referred to as a Phase 2 level of enquiries. “Phase 2” merger and market decisions must be made by a group drawn from a separately appointed panel of experts (the Panel) who are not CMA staff. The investigatory teams in the two phases are also largely different.

A second possible benefit from the separation of investigation from decision making within the administrative model is that more information is likely to be provided to the decision maker when the decision-maker is independent of both the investigator and the defense. Indeed in such a case neither party has an incentive to hide information.

A third advantage of the separation between investigation and decision within the competition authority making lies in the fact that the authority is perceived to be more respectful of due process and therefore more legitimate. For example, during the revision of the Mexican competition law which led to the creation of the Federal Economic Competition Commission (Cofece) and of the Federal Institute of Telecommunications (Instituto Federal de Telecomunicaciones) (IFT) in July 2013, the separation of investigation and adjudication in both institutions was seen as necessary in order to guarantee impartiality and objectivity of the competition authorities. Thus, the reform provided for a separation between the authority in charge of the investigation and the authority in charge of the resolution (both within each institution).

A fourth, and may be the most important benefit from the standpoint of the quality of the decision making process is the fact that the separation of investigation and decision making limits (somewhat) the risk of confirmation bias whereby an authority having invested a large amount of resources to bring a case against a firm or a set of firms has a natural tendency to legitimize its past efforts by finding the investigated firms in violation of the competition law.

If there are thus clear potential benefits of separating investigation and adjudication, some have questioned the importance of those benefits and others have pointed out that there are nevertheless costs and inefficiencies involved in keeping the two functions apart.

Belgium is, for example a country which underwent a change of institutional design of its competition authority in 2013. The contribution of Belgium to the Competition committee roundtable on institutional design³⁸ states that the newly established Belgian Competition Authority (BCA) is an autonomous authority with its own legal personality. It is managed by a board of four members appointed by the Government with a mandate of 6 years: the president, the competition prosecutor general (auditeur général), the chief economist and the general counsel. Formal cases are opened by the prosecutor general after hearing the chief economist. Investigations are managed by the prosecutor general who appoints one member of the investigation service in charge of the daily management of the case and one as ‘peer reviewer’. These three officers can together decide as auditorat (1) to bring the case before the Competition College, or (2) to settle the case, or (3) to drop the case. The Competition Colleges consist of the president and two assessors designated in alphabetic order from a list of twenty. They hear and decide the cases brought for them by the auditorat (or the parties who wish to appeal a decision of the auditorat to drop a complaint.

An interesting comment by the President of the BCA (who was the director general for competition of the Authority before the institutional change) is found in the Belgian contribution to OECD.³⁹ It states: “ I wish to reiterate, however, what was said in earlier surveys of the OECD and the IMF. As Director General in the previous authority in which I was member of the Board of the ministry of economic affairs, I never experienced the slightest restriction of the independence of the agency. This of course also held true for the Competition Council that was an administrative tribunal whereby the authority only act in formal infringement cases as the prosecutor”.

In administrative models in which a board independent of the investigatory team is called upon to make the final decision on a merger or on an antitrust violation three risks exist.

The first risk is that the board may lack means to monitor that quality and/or the quantity of the work done by the investigatory body of the same agency. The lack of possible feedback from the decision board to the investigatory arm of the competition authority may lead to a suboptimal use of resources and/or an ineffective process if the two parts of the administrative agency do not share the same vision of the goals of the institution.

First there is the risk that the decision-makers, because they have not participated in the investigation, may not know or understand as well as the investigatory team the implications of the results of the investigation. This could happen in some very fact intensive cases where the analysis requires the ability to put into context numerous elements revealed by the investigation. Even if the decision makers read the investigatory file they may not have as intimate a knowledge or understanding of this file as the investigators themselves who have spent many months painstakingly putting together its elements.

Alternatively, in some institution the decision-makers benefit from a second investigatory team, usually separate from the team which conducted the initial investigation.

The risk in that case is the opposite. Indeed the risk exists that there may be unnecessary duplication between the two successive investigations. This was, for example, frequently a complaint voiced by the business community when, previous to the creation of the CMA, merging businesses were faced by requests for information , first, by the Office of Fair trading and, second, when there was a reference to the Competition Commission by the staff of the Commission. The business community clearly felt that these repeated requests, often on the same points imposed on them an unnecessary cost. As mentioned earlier the CMA has since tried to alleviate the problem by including in the second phase investigatory team some members of the original investigatory team. But the trade-off between ensuring the independence of the investigatory and the adjudicative processes and making sure that the adjudicators are not entirely dependent on the information provided by the primary investigators is a tricky one.

The problem of the separation of the adjudication from the investigation is particularly acute in civil law systems when the decision maker is a court and when the court does not have separate investigatory powers or independent means of investigation. In such cases, the court may be in fact very dependent on the evidence and the economic interpretation of this evidence proposed by the prosecuting entity even if the defendants lawyers try to provide the court with whatever evidence would exculpate their clients. The court often cannot ask for additional or different investigations and the defendants do not have the powers of investigation of the prosecutor.

Thus the separation of investigation and adjudication may come with two costs: the cost of duplication and/or the cost of a diminution of the relevance or the quality of the investigation for the adjudicator which partly offset some of the benefits of the system. The author’s personal experience in the French system suggests another illustration of this phenomenon. In the case of France, within the Autorité de la concurrence, there is a strict separation between the investigatory team (under the leadership of the Rapporteur Général) and the board under the leadership of the President of the Authority. As a result it is quite difficult for the board to deal with an unsatisfactory or slightly incomplete investigation. The board is limited by the charges which have been notified in the statement of objection. When it feels that a charge is missing , it has the power to send back the case to the investigatory team but is necessarily limited in its ability to say exactly what it considers is missing in the investigation for fear of jeopardizing the independence of the investigation and the adjudication. The investigators understandably consider that having a case sent back to the investigatory team by the board is one of the worst outcomes for them as it clearly establishes that the initial investigation was insufficient. This has two effects on the investigators; first they try to do as good a job as possible and this tends to ensure a high level of quality of the investigations. But, second, it also leads the investigators to send to firms statements of objections with charges which go far beyond what would be reasonable and justified. The reason is that the investigators, not knowing always what the board would consider to be sufficient evidence of a violation would rather notify violations that do not have merits which the board will not retain in its final decision than not include in their statement of objection questionable charges of violations of the law because in the latter case, they run the risk of having the board send back the case for further investigation. This is not costless for investigated business as they have to answer numerous dubious charges of violations, even in cases where it is clear that the case officer is not fully convinced of the validity of his reasoning. Furthermore dubious charges of competition law violations are often all the more difficult to fight that their legal or factual or analytical basis is very shallow.

A third possible drawback of the separation between investigation and adjudication was much discussed in the United Kingdom during the lengthy process which eventually led to the creation of the CMA. One of the questions discussed at some point in the process was whether a prosecutorial model should be preferred to an administrative model. The OFT staff was rather alarmed by the prospect of a prosecutorial model which would imply they (or their successors) would not be in charge of making decisions on case. The argument they put forth was that moving to a system were the investigative staff would not make decisions would discourage the most talented staff who would see their area of responsibility restricted to the investigation. The argument was made that the separation of investigation and adjudication could have the effect of lowering the quality of the investigatory staff of the competition authority.

It is fair to say that no empirical evidence was presented to back such an assertion. Many competition authorities such as, for example, the US FTC or the French competition authority have a long tradition of separation of investigation and adjudication and top quality staff. So it is entirely possible that the argument was nothing more than a corporatist argument by the OFT staff to retain its status.

Some of the afore-mentioned costs or risks of the separation between investigation and adjudication may explain why the European Commission chose an opposite path to that of a number of competition authorities. Several decades ago the functions of investigation and adjudication were separate within the Directorate in charge of competition then called DG4. The service in charge of investigation would eventually turn over its files to the service in charge of decision making and the two services were independent of each other. This organization has given way to a different one were the services of the Directorate General for Competition which are in charge of the investigation are also the drafters of the draft decision (after consultation with other interested services and, in particular, the legal service) under the supervision of the Commission and the Commissioner’s cabinet. This decision is eventually examined by the Commission. Over the years, this model has drawn criticism for three main reasons: the lack of separation between investigation and adjudication; the fact that the parties rarely if ever meet the decision makers and the fact that the Commission itself will mostly be former ministers.

Finally, it is worth mentioning that South Africa went from an integrated administrative model to a prosecutorial model in 1999 with the competition agency bringing cases to the Competition Tribunal and the decisions of the Competition Tribunal being reviewed by the appeal court. This move, which was followed by a number of African countries, ensures the independence of the investigation and the adjudication. However the South African delegation to the OECD debate on changes in institutional designs of competition authorities considered that it was too early to say whether the model adopted in 1999 has worked out or not. According to the delegation of South Africa, some of the likely advantages of the new model are a better respect of due process, more independence in decision making , a diminished likelihood that the competition authority might be captured, a lesser probability of corruption and a more rigorous decision making process. Against these potential benefits some of the drawbacks of the new system may be, first the fact that cases may take longer to be disposed and the fact that the cost of running two institutions is higher for the government than in an integrated system. In addition, the South African delegation noted that for such a bifurcated system to work well, it is necessary that there be sufficient expertise in the adjudicative body.

The organizational structure of competition authorities: collegial board versus single commissioner

There is a variety of organizational designs of competition authorities. The competition authority may consist of a single Commissioner (as in Canada or in the Antitrust division of the Justice department) or be a collegial body (such as the US FTC) . If there is a collegial body it may be an administrative board (such as in the CMA in the United Kingdom) or a board making decisions on cases (such as in the case of France). Finally the board members may be full time or part time members.⁴⁰

The collegial body model with a decision making board allows for the possibility of board members having different areas of expertise (for example law, economics, business) and thus may seem to be more legitimate to make competition decisions which require a combination of skills. The single decision maker may have fewer skills than a collegial body and therefore a more limited ability to take into consideration all the relevant elements which should be considered in a given case.

The fact that a competition authority has a collegial body may also seem to be likely to make its capture (by business interests or by government) more difficult than if the competition authority is a single individual. Indeed it would seem that as the number of commissioners increases the likelihood that they are all captured by the same interest diminishes.

Finally the collegial model may, if the appointment of the board members are staggered, lead to more stability in the jurisprudence whereas a change of a single commissioner at the end of his term may lead to wide swings in the way the competition maw develops.

There are, however, possible downsides of the collegial body model.

First, the decision making process may be slower in a collegial body than when a single individual is responsible for the decision.

Second, the decisions of the authority may be less consistent in collegial bodies if there are disagreements among the members than the decisions made by a single head of agency.

Third, when there are disagreement between the board members, the competition‘s decisions may be criticized publicly by the members of the board which were in the minority and this may contribute to a loss of trust in the institution.⁴¹

Independence of competition authorities

The issue of the independence of competition authorities has been increasingly important in the debate about the optimal design of a competition authority and has several dimensions: structural independence, operational independence, organizational independence and financial independence. We will deal with organizational and financial independence later on and the current development will be devoted to the issue of operational independence.

Twenty years ago most delegates to the OECD competition committee were representative of ministries. Nowadays they are mostly representatives of independent administrative agencies which are not directly part of the executive. It is now considered to be appropriate for a competition authority to be insulated from government interferences in its law enforcement activity. Equally, conventional wisdom also suggests that competition authorities should be independent of business interests to protect the integrity of their law enforcement activity.

Yet as a recent EC study noted,⁴² even in Europe “a number of NCAs are formally assigned to, or come under the responsibility of, a minister or ministry. Moreover, some NCAs may in principle be subject to general supervision or to general instructions by the executive branch or parliament although, such supervision may not have been exercised in practice, or at least not recently”. The degree of independence from government that a competition authority enjoys may also vary with the subject matter it deals with. Thus, for example, the above quoted EC document stated “In a number of Member States a specific form of government intervention exists in merger cases. It usually means that the government or competent minister may intervene on public interest grounds after the NCA has analysed the merger's impact on consumers and businesses. In one Member State, the Prime Minister may declare a merger to be of state interest and, as a consequence, exempt from competition scrutiny by the NCA”.

It is assumed that the more independent the competition authority is structurally and operationally, the less likely it is that it will be under pressure to start investigations or to decide on cases for reasons extraneous to the logic of competition. This is important because competition laws are written in such a way that competition authorities often have a wide discretion when it comes both to the prioritization of cases and the competitive assessment they make in each case. They can thus relatively easily hide motives unrelated to competition or consumer benefit, if they have such motives, in decisions which look formally fairly reasonable by manipulating market definition, a finding of market power, concerns about a vertical restriction or an abuse of dominance or the counterfactual to assess the impact of a merger on competition .

Because it is thus not easy to detect departure from an economically justified interpretation of competition laws, it is particularly important that competition authorities be as sheltered as possible from the risk of capture by the executive branch of government or by business interests . Thus the first benefit expected from independence of the competition authority is a better quality of decisions and an implementation of competition law more in line with economic analysis and legal principles. If independence does not guarantee the competence of the competition authority, at least it limits the risks that illegitimate goals will interfere with the decision making process.

A second benefit, linked to the first one may be a greater consistency of decisions. Indeed if one assumes that the executive is tempted to intervene in some cases where it may have a particular interest, those cases may end up being treated by the competition authority differently than other cases. A similar situation may accrue if the competition authority is captured by business interests. The cases were those business interests are involved may be treated more leniently than other cases.

Consistency in the treatment of cases may contribute to build trust of the stakeholders in the institution and make it more respected by the firms.

A third possible benefit of the structural and operational independence of the competition authority is the fact that it limits the incentive of economic agents to lobby the authority since this lobbying is less likely to be successful, thus freeing resources which can be put to a better use for society.

Thus as Bill Kovacic and Marc Wineman put it⁴³ “Implicit or explicit in many discussions of independence are conditions that we believe represent a sensible core domain of decisions that are shielded from political interference. The most important of these is the exercise of law enforcement authority which can lead to the imposition of significant sanctions upon juridical persons and natural persons. The political branches of government ought not to be able to (a) dictate, by rule or by custom, which entities an agency investigates; (b) determine whether the agency will prosecute such parties; or (c) influence how specific disputes will be resolved, including the choice of punishments for alleged wrongdoers.(…) These conditions assume greater importance as the severity of the agency’s power to punish increases.”

If there are advantages to the structural and operational independence of the competition authority vis à vis the government in its enforcement function, there may be also be trade-offs between the protection of the integrity of the enforcement activity of the competition authority and its ability to advocate. Indeed, the more independent of the executive is the competition authority the less effective its advocacy is likely to be.

This trade off , and the necessity to be effective in advocacy, is the justification often given by the Korean competition authority to justify the fact that the head of the Korean competition authority participates in the Cabinet meetings. As the KFTC has remarked in many occasions, this allows the competition authority to be informed in a timely fashion of any governmental plans which may have a negative impact on competition and to advocate against them.

There may also be a trade-off between accountability and independence. Accountability, if it takes the form, for example, of a close scrutiny by the executive or the legislative branch of the ways in which the competition authority has discharged its functions and/or has allocated its revenues, may limit the ability of the competition authority to act in the way it considers most appropriate for fear of displeasing the bodies which are reviewing its activities (and sometimes deciding on its budget). But at the same time there cannot be independence (and the ability for the competition authority to impose strong sanctions) without accountability.

Kovacic and Wineman ask what safeguards are most appropriate to guarantee the level of operational independence needed for competition authorities. They suggest the following list:

• Legal commands or customs that impede the head of state, government ministries, or the legislature from taking direct or indirect steps to shape broad policy or to determine how the agency exercises its power to prosecute cases or adopt secondary legislation.

• An absence of judicial review of agency decisions, or requirements that courts abide by a highly deferential standard of oversight.

• The absence of, or severe limits upon, the ability of citizens, nongovernment bodies, or commercial entities to influence the agency’s agenda or to monitor its operations by having access to the agency’s records or by participating in its activities.

• Sources of funding that do not depend upon the exercise of discretion by the head of state, executive ministries, or the legislature.

However they recognize that meeting those formal conditions could be problematic from the point of view of the accountability of the competition authority and they argue that a more limited set of conditions which allow political institutions to offer guidance or recommendations to the competition authority about large issues of policy but would prevent them from dictating or blocking a specific decision may be more realistic.

The situation in which the competition authority is part of a ministerial department is generally considered to be least consistent with the requirement of independence. Indeed in such a model, the head of the agency generally can be dismissed at will by the executive.

More consistent with the necessity to ensure the independence of the competition authority is the situation where the Competition agency is an administrative body which is outside the ministerial structure if its members are appointed for a fixed term and cannot be removed from office except for cause.⁴⁴

Mexico is one of the countries which underwent a significant institutional change of its competition authority in 2013. Two agencies with full constitutional autonomy—responsible for competition matters were created: the Federal Telecommunications Institute (Instituto Federal de Telecomunicaciones—IFT), for broadcasting and telecommunications, and the Federal Economic Competition Commission (Comisión Federal de Competencia Económica—COFECE), for all other sectors. COFECE replaced the former Comision Federal de la Competencia , a body with technical autonomy which formed part of the Ministry of Economy.

Cofece was given special institutional characteristics and new powers to effectively promote and protect the competition process. This autonomy includes features such as: (a) distinct legal personality and self-patrimony; (b) full independence in the decision-making process; (c) budgetary autonomy, (d) the power to enact rules regarding administrative organization, (e) the power to enact implementing regulation; (f) the power to file a constitutional recourse before the Supreme Court of Justice in case the Federation violates or affects its authority.

With respect to accountability, COFECE is obliged to submit reports to the Executive and the Federal Congress, and is subject to various accountability and transparency mechanisms and an Internal Comptroller, appointed by the Chamber of Deputies, has to oversee the application of the COFECE’s budget as well as the conduct of public officers regarding administrative responsibilities.

Besides the formalistic safeguards mentioned previously several set of provisions can protect the independence of the competition authority.

First, the conditions of the appointment of its members, the rules governing the conflicts of interest of the board members as well as the time limits put on their mandates can contribute to the independence of the competition authority. From that standpoint, it seems preferable that the chair and the Commissioners be appointed for a period sufficiently long so that they can acquire the basic skills necessary to deal with the interface and law and economics. But it is equally important that the mandate of the members of the board not be renewable for another mandate at the end of its mandate. Indeed , if board members can be re-appointed, the possibility exist that they will eventually become concerned with their chances of re-appointment and will start adapting their judgement to what they believe would please the authority in charge of reappointing them, thus foregoing deliberately their independence. The staggering of the appointment of the board members may from that point of view have the added advantage of making the appointment of any commissioner less important for the balance of the institutions.

Over an beyond the formal rules, it is widely acknowledged that greater transparency in operations can, in general, increase the agency’s perceived legitimacy and can be a useful barrier against government or business encroachments. Transparency can be achieved through a diversity of technical means such as press releases, the publication by the Commission of guidance papers, the publication of well-written and well-argued decisions etc…. The more transparent the process of decision making by the Competition authority the more visible would be the result of an undue influence and the less likely it is that the competition authority will let itself be influenced by outside forces.

It should, however, be noted that excessive transparency may have some drawbacks. The experience of Cade in Brazil in its early years provides an interesting example. The Cade board had the duty to deliberate in public. This meant that each of the commissioners had to announce publicly its vote and what justified it. There were allegations that this excessive transparency led to the fact that each Cade commissioner in the deliberation was more interested in what the press would report about their position than in establishing a real dialogue with the other commissioners and that this may have affected both the quality of decision making and the independence of judgment of the Cade Commissioners.

Closely related to the discussion on the independence of competition authorities, there is a debate on whether Ministers or Governments should be allowed to give a strategic steer to the Competition authorities.

In its contribution to the OECD Competition Committee debate on institutional changes⁴⁵ the EC Commission states: “The majority of NCAs are not subject to supervision by another state body. However, a number of NCAs are formally assigned to, or come under the responsibility of, a minister or ministry. Moreover, some NCAs may in principle be subject to general supervision or to general instructions by the executive branch or parliament although, such supervision may not have been exercised in practice, or at least not recently. In addition, the degree of supervision differs and may range from guiding and co-ordinating the NCA’s activities or outlining the NCA’s activities without intervening or deciding on individual cases or on the actual application of the law, to giving instructions regarding the general application of the law or regarding budgetary issues or general policy matters which is also directed to other governmental institutions. In a number of Member States, the minister may instruct the NCA, for example, to carry out sector inquiries or competition studies or analyses, which the NCA cannot otherwise initiate itself, but without, however, directing the outcome. The minister may also instruct the NCA to investigate a particular case or examine the need for interim measures”.

The debate on the appropriateness of strategic steers of competition authorities can be illustrated by the recent divergent evolutions of the UK and Portuguese competition laws.

A new feature in the recent landscape changes in the UK is the requirement for the Government to provide the CMA with a strategic steer. The UK delegation to the debate on institutional changes held at OECD states⁴⁶ “Whilst at the time of the reforms certain concerns were raised that such a statement risked weakening the CMA’s perceived independence, the Steer is a public document setting out the Government’s high-level aims and expectations for the CMA in an open and transparent way”. And “Whilst the CMA has regard to the Steer and remains accountable to the Government for its performance assessed by reference to the Performance Framework, its decision-making remains fully independent from Government”. Thus according to the Delegation from the United Kingdom, the ability of the government to give the CMA a strategic steer is a minor concession to the request for the accountability of the institution and the fact that the steer is public should be seen as a protection against undue secret pressures. Incidentally this is not the only element of accountability of the Competition Authority in the case of the United Kingdom. The CMA is also subject to a performance framework which is agreed with the government . The Performance Framework sets out the performance the Government expects from the CMA and describes how the CMA will fulfil the performance reporting requirements of the Act. The CMA Board is accountable for the success of the CMA as a whole and the delivery of the objectives set out in the Performance Framework. Thus, the CMA must report annually on a number of benchmarks, including: the delivery of a target of direct financial benefits to consumers of at least ten times its relevant costs to the taxpayer (measured over a rolling 3-year period); the ratio of direct financial benefits to consumers and costs for its principal tools and, finally, its assessment of wider benefits of its work, for example on growth, business and consumer confidence in markets, compliance with competition law and deterrence of anticompetitive behaviour.

Not everybody takes a positive view of the United Kingdom situation in which the UK Governmùent can give a strategic steer to the CMA. For example the business community (The Business and Industry Advisory Committee of OECD) stated in its contribution to the debate on institutional changes⁴⁷: “In the U.K., for example, in the CMA’s infancy, there are questions regarding its independence from government. When the CMA was formed, the government outlined a non-binding ministerial statement of strategic priorities for the CMA, or a “steer”, which essentially outlined how the government thought the new body would fit within its broader economic policies. Further, the CMA possesses broad new investigative powers regarding issues of ‘public interest’, such as national security, and the government recently called upon it to intervene in the energy and financial services sectors. While the CMA has emphatically stated that it will make its own decisions on which markets to investigate, there are nevertheless questions regarding its independence from government”.

The situation in the United Kingdom is in sharp contrast with the situation in Portugal. The previous Portuguese Competition Act’s bylaws of 2003 stated that the independence of the competition authority in the performance of its duties was “without prejudice to the guidelines on competition policy set out by the Government (…) or to the acts subject to ministerial oversight” (article 4). The need to comply with Government competition guidelines was perceived by some as lessening the competition authority’s independence. This provision has now been removed from the Bylaws, which state instead that the Competition authority “is not subject to governmental supervision” and that “the Government cannot make recommendations or issue directives to the Board on the priorities to be adopted by the Portuguese Competition Authority in carrying out its mission” (Article 40(1) of the PCA’s Bylaws). The law explicitly excludes, therefore, the possibility of external interferences with the activities of the Competition Authority.

To sum up, it is fair to say that competition authorities which can receive a strategic steer (such as the CMA in the United Kingdom, but also the ACCC in Australia) from their government would probably be happier if this possibility did not exist as they are faced with an unsatisfactory alternative : either they disregard the steer they have received and they risk being isolated from other policy makers or they follow the steer at the risk of being perceived as subservient to the government.

Performance indicators are a politically more neutral way of ensuring that the Competition Authority is accountable. However they can have drawbacks and lead the competition authority astray if they are simplistic or badly conceived. For example, it seems clear that the preference of the OFT for advocacy and negotiation over enforcement came from the fact that the assumed direct effect of settlement or the publication of guidelines always seemed to be much more important than the effect of an enforcement decision. Similarly, the OFT was not keen on small enforcement cases and only accepted to take very large cases that could have an influence throughout the entire industry rather than on some firms both because the assumed effect of those cases applied industry wide and because there were some scale economies in the investigation of such cases. However, what was crucially lacking in the methodology used to assess the effect of the actions of the OFT was the deterrent effect of enforcement (which depends both on the probability of detection and the severity of the sanction); furthermore, taking only very few and very large enforcement case, some of which failed at the appellate level, was increasing the risk that OFT would be perceived as an ineffectual enforcer.

Funding of competition authorities

Competition authorities have often argued that securing sufficient funding was a necessity and that lack of proper funding could jeopardize the quality and the integrity of the decision making process and ultimately of the competition law enforcement process.

Very little will be said in this article on the amount of funding necessary for a competition authority. The heterogeneity of competition authorities for reasons discussed in this article with respect to their role, their scope of activity, the legal context in which they operate, the size of the countries over which they have jurisdiction, the level of market development of the economy they oversee, the importance assigned to market competition etc… makes it exceedingly difficult to compare budget allocations for competition authorities across countries and to come up with a conclusion on an appropriate benchmark for the funding of a competition authority.

The best resourced competition authorities (having a budget above or roughly equal to US$90 million) are the US Department of Justice Antitrust Division, the US Federal Trade Commission, the Korean Fair trade Commission, the EU Competition Division and the Japanese Fair Trade commission of Japan.

Non European competition authorities with a mid-range budget (between US$15 and 30 millions) are found in Israel, Turkey, South Africa, or Canada.

Within Europe, the competition authorities of the United Kingdom, Sweden, Germany, France , Italy and Spain have budgets upward of US$20 Million (US$24 million for France and nearly US$75 million for the United kingdom). The competition authorities of Norway, Denmark, and Greece have a budget of US$10–15 million. Hungary, Poland, Ireland, Portugal, Belgium, the Czech Republic have budgets between US$5 and 10 million. Finally, the competition authorities of four countries Cyprus, Austria, the Slovak Republic, Lithuania and Latvia have budgets between US$1 and 3 million and a number of smaller countries have budgets inferior to US$1 million (Slovenia, Malta, Estonia) .

In Latin America Mexico has the best resourced competition authority with a budget of US$30,000,000. In three countries Brazil, Chile, Ecuador, the competition authority seems to have annual resources comprised between US$9 and 15 million. In three countries Argentina, Columbia, El Salvador, the competition authority has an annual budget comprised between US$2 and 3 million and in four countries Honduras, Uruguay, Costa Rica, Nicaragua the annual budget of the competition authority is comprised between US$900,000 (in the case of Honduras) and US$300,000 (in the case of Uruguay).

As mentioned earlier those figures are not easily comparable, among other reasons, as the scope of activity of the competition authorities varies widely from one country to the next and as the economic sizes of countries are themselves very different.

In its peer reviews the OECD has taken the habit of comparing the resources allocated to the competition authority with the resources allocated to other regulators such as the telecom or the electricity or the media regulators. Even though those comparisons are not without their problems, they show that in a number of countries, the competition authority is rather poorly resourced compared to the technical regulators without such difference being clearly justified.

There are, however, at least three issues worth raising regarding the financing of competition authorities. The first issue is that of the independence of the funding from the case selection and from the decision making of the competition authority. The second issue is that of the possibility for the competition authority to be self-funded. The third issue is that of the budgetary autonomy of the authority in the spending of its budget.

The first question is crucial and the source of the funding of the competition authority can be problematic in two types of circumstances. First, if the government or parliament have the ability to “punish” the competition authority for either pursuing cases that they do not like or for findings that they disagree with, by modifying the budget allocated to the competition authority, the independence of judgment of the competition authority (i.e. its ability to decide to pursue cases uniquely on the legal and economic merits of those cases) will be in jeopardy.⁴⁸

This could, for example, explain partly the findings of some researchers investigating the US FTC merger challenges. It is worth remembering that the US Federal Trade Commission has to go through a Congressional reauthorization process at regular intervals through the Senate Committee on Commerce Science and Transportation and the House Committee on Energy and Commerce. Furthermore, the appropriation committees of the House and Senate appropriate funds for the agency on an annual basis. Coate, Higgins and Mc Chesney⁴⁹ studied the FTC challenges to mergers during the period 1982–1986. They suggested that political pressure on the FTC to investigate more merger cases came in two different varieties: the call by Congress on FTC commissioners and staff to defend their antitrust record (hearings during which individual congressmen and senators can even advocate particular enforcement actions) and the budget process whereby “antitrust agencies budgets are allocated according to the enforcement zeal shown”. Analyzing the enforcement record of the FTC they find that to challenge mergers the FTC relies on a number of the (economic) criteria mentioned in the merger guidelines (such as the Herfindhal-Hisrschman Index, the ease of entry, and the likelihood of collusion) but also on the “the desires of politicians to stop mergers”. They conclude “ As we show in statistical tests, greater political pressure does cause the FTC to challenge more mergers”.

Second, if the competition authority has a financial stake in the outcome of the cases it investigates, which is the case, for example, if it directly benefits from the fines it imposes, then there is the possibility of “moral hazard” and the decisions of the competition authority may not be based exclusively on the legal and economic merits of the cases it handles. This is why it is generally considered to be inappropriate for a competition authority to be directly funded by the fines it imposes on competition law violators. It is thus generally held that the sanctions meted out by the competition authorities should go to the general budget rather than to the specific budget of the competition authority. Not all countries abide by this principle. For example, the competition authorities of Portugal, Bulgaria and Peru are partly financed by the fines they impose on competition violators. Portugal and Bulgaria can use (for funding their budget) up to respectively 40 and 25 % of fines. The Hungarian competition authority is allowed to use (indirectly) 5 % of its fines for funding conferences and external research projects.

Whereas, it is generally the case that competition authorities are not the direct recipient of the fines that they impose, there is, however, a more subtle way in which competition authority indirectly benefit from the aggregate amount of sanctions they impose. It is indeed very tempting for competition authorities who are competing with other administrative agencies for the allocation of public funds to make the point that they contribute, through their fining policy, sometimes mightily to the resources of the state and that therefore their efforts should be appropriately rewarded by an adequate budget.

However, some authors have suggested that allowing the competition authorities to benefit from some of the fines they impose on violators may, in some circumstances, in fact contribute to correcting their natural bias toward under-enforcement. Pierluigi Sabbatini⁵⁰ suggest two reasons for the suboptimal level of fines meted out by European competition authorities. First, to appear to be successful, competition authorities want to ensure a high success rate before appeal courts and therefore have a high incentive to favour relatively low fines (so as to reduce the incentive to appeal), commitment decisions (which will not be appealable) or simple cases involving firms with large turnover (and fines that appear high in absolute value, thus ensuring the visibility of the enforcement action of the competition authority, but are low in percentage terms, thus ensuring the acceptability of the fines). Second, Sabattini suggests that competition authority officials may have an incentive to act in a way that benefits their future careers prospects and their opportunities once they will have left the competition authority. Some of those staff are likely to join an economic consultancy firm or a law firm or academia after the end their career at the competition agency. This would lead them to want to open many cases and follow a high sanction strategy. But commissioners or board members of competition authorities, who are usually political appointees, may in some cases be more interested in pursuing a career in government after they term ends at the competition authority. In order to have a higher chance of being appointed to another government job or in a political position, their technical skill at deterring competition violators may be less important than their skill at mediating between different interest groups. This would lead them to overlook the deterrent effect of punishment and to focus more on the adequacy of sanctions with respect to the harm done to the immediate victims of the antitrust behaviours examined.

Assuming on this basis that competition authorities (at least in the European context) tend to be myopic (or to under enforce as is claimed by Connor), Sabbatini goes on to suggest that this bias could be partly eliminated, first, if the competition authority were entitled to retain a limited portion of the sanctions which have been confirmed at the appellate level, second if the sums retained by the competition authorities would come in addition to (and not in substitution to) their governmental funding and, third if these sums could only be used for purposes specifically designed to incentivize the staff to bring about successful fining cases (i.e. fining cases upheld on appeal or not appealed) and to improve the quality of decisions. For example, such sums could be used to provide additional incentives for the staff to successfully bring cases of sanctions, to fund temporary personnel such as the members of the chief economist staff or to finance conferences or studies on competition policy.

In conclusion of this point there may be a complex set of trade-offs between the quality of the enforcement of the competition authority and the independence of its budget from the sanctions it metes out.

The second issue related to funding is that of the possibility for the competition authority to be self-funded.

As was clear from the previous development, the funding of the competition authority through the budgetary process cannot immunize the competition authority against more or less subtle pressures to adapt its performance to the expectations of those (government or parliament) who control its budgetary process. For that reason, the allocation of funds to the competition authority through the budgetary process leaves the possibility that the competition authority will behave strategically in dealing with its cases in order to maximize its chances of seeing its budget increase.

Self-funding of the competition authority can in some circumstances avoid this problem. By self funding we mean that the competition authority benefits from resources which are both independent of its record and of the budgetary process.

A good example of a self-funding mechanism for the competition authority is given by the case of the Turkish Competition Authority. Article 39 of the Turkish Competition Act provides that the revenues of the Authority are made, first, of funds to be allocated to the Authority in the budget of the Ministry of Customs and Trade, second, of a tax of 4/10000 on the capital of all newly established incorporated limited liability partnerships or of the increase in capital of already established partnerships and third, of publications or other revenues of the Authority. Since its creation, in 1997, the Turkish Competition Authority (which had a budget of US$27,000,000 in 2014) has relied entirely on the tax on limited liability partnerships to fund its budget and has never needed to benefit from tax revenues of the Ministry. Thus the Turkish competition authority has avoided the implicit or explicit bargaining associated with the budget process and the volume of its resources, provided by the shareholders of limited liability partnerships rather than by the taxpayers, is independent of its enforcement decisions.

This type of self-funding mechanism which allow the funding of a regulator by the regulated firms rather than by the taxpayers can have a great appeal in countries where the government is cash strapped but one of the risks faced is that this formula be replicated to fund a number of other regulators in which case the transaction costs on the formation of limited liability firms can become quite significant leading firms to incorporate themselves outside of the country.

In its contribution to the 2014 roundtable on Changes in Institutional Design,⁵¹ the Italian Competition Authority explained that it had adopted a new funding system since 2012 (effective in 2013) which is somewhat similar to the Turkish system. Law Decree no. 1 of 24 January 2012 introduced a mandatory contribution for companies incorporated in Italy whose turnover exceeds the threshold of 50 million euro. This contribution replaces previous financial resources of the Italian Competition Authority, i.e., public budget and merger fees. The contribution stated: “As a result of the new funding system, the Authority no longer needs to engage in negotiations with the Government every year to secure its financial resources, thus being reinforced in its independence”.

Another slightly different example of a (partly) self-funding system is that of the Portuguese Competition Authority which was modified in 2014 following the adoption of a Framework Law setting the main principles for the functioning of regulatory bodies in Portugal under the Economic Adjustment Programme.⁵² One of the key structural reforms set out by the Programme has been to ensure the independence and adequate financing of sectoral regulators and of the Competition Authority.

According to Article 35 of the Portuguese competition law the Portuguese Competition Authority, first, may charge fees for the services it provides, second, receives 40 % of the proceeds of fines imposed for competition law violations, third, receives revenues from a tax imposed on a large number of sectoral regulators.⁵³ The amount of the tax is fixed yearly by the Minister of Finance for each sectoral regulator and must be comprised between 5.5 and 7 % of the own resources of the regulator (excluding the product of fines or sanctions it has imposed) and , finally, if necessary, may receive a budget allocation from the Ministry of the economy.

Another way to provide self-funding for the competition authority is to allow it to charge for services rendered. In particular in a number of countries, the competition authorities when they review mergers can charge a filing fee.⁵⁴ The question of whether a competition authority is authorized to charge a filing fee for merger review depends partly on whether the prevailing view is that the cost of the service should be borne by the merging firms who benefit directly from the merger control service or whether one considers that it is a public service that should be funded through taxpayers.

In countries where it is considered that the cost of the merger control should be borne by its direct users, several institutional designs are possible.

First, the filing fees can go to the general budget and not to the budget of the competition authority itself. Alternatively, the filing fees can go directly to the budget of the competition authority.

Second, the filing fee may be a flat fee,⁵⁵ a function of the size of the transaction,⁵⁶ a function of the services rendered by the competition authority⁵⁷ or a function of the complexity of the competitive analysis the merger entails.⁵⁸ In this last case the fee can vary depending on whether a phase II investigation is undertaken by the competition authority.

A flat fee may be easily predictable but seen as unfair as the parties to smaller transactions are subsidizing the parties to larger transactions. Fees based on the size of the transaction may lead to uncertainties or controversies about the computation of the size of transactions. Fees based on the costs incurred by the Competition Authority are more likely to allow the competition authority to recover the full cost of merger control but may be difficult to predict for the parties and may be disproportionately high in sectors which have not previously been studied by the competition authority. Finally, when fees are based on the complexity of the transaction, there is the risk of a moral hazard problem as the competition authority may have an incentive to artificially open unjustified second phase investigations.⁵⁹

When the fees are supposed to cover the cost of the merger reviews, one of the difficulties is to determine the average cost of such reviews. As the ICN study states: “Setting the level of filing fees to cover the cost of merger review is not an exact science, as the number of annual merger notifications can vary significantly and the measure of the cost of reviewing any particular transaction varies depending on what costs are taken into account, such as whether fixed costs are considered. Full cost recovery is not always a practical or desirable policy goal. For example, the UK Competition Commission’s costs of reviewing twelve mergers referred in 2002 and 2003 varied from £262,000 to £524,000 per case. A UK Consultation Paper on merger fees published in 2004⁶⁰ noted that “a fee based on even the lowest of these costs would seriously impact the economic rationale of some mergers.” ” Similarly, a discussion paper published by the New Zealand Ministry of Economic Development in 2004⁶¹ on the pros and cons of increasing the merger control filing fees levied for the Commission’s services emphasized that merger control “fell well short of full cost recovery” and that “the fees levied did not even begin to cover the cost incurred by the Commission in processing applications.”

In some countries, particularly in developing countries where the cost of reviewing the merger may be substantially lower than the filing fee, the explicit goal of merger review filing fees is to finance, if possible, the overall budget of the competition authority, or to contribute to the budget of the competition authority over and beyond the cost of processing the merger applications.

A particularly telling (and ultimately unsuccessful) example of a merger control filing fee mechanism which was designed to allow the cross subsidization of the other activities of a Competition Authority was the Comesa⁶² merger control set up. Initially the filing fees were set at an extremely high level for regional mergers (namely, “where” both the acquiring firm and target firm or either the acquiring firm or target firm operate in two or more [COMESA] Member States). The Assessment Guidelines published in 2013 stated that that the filing fee was equal to the lower of (1) COM$ 500,000 (approx. 389,166 €) or (2) the higher of 0.5 % of the parties’ combined annual turnover or value of assets in the COMESA Common Market Area. The law provided for the establishment of a turnover or asset threshold below which merging firms did not have to notify their mergers but this threshold was initially set at $ COM 0; furthermore the competition authority gave an extensive interpretation of the concepts of “operation in the member states” and an extensive interpretation of the concept of “control” for the purpose of defining when a merger has occurred. Finally, under the initial guidelines published in, it was not clear that where the merger control applied at the Comesa level, national merger controls did not apply.

The result was that very few mergers were notified to the Comesa Competition Commission. The number of mergers notified to the Comesa Competition Commission averaged about one a month for its first year of operation and there are allegations that a number of notifiable mergers were not notified as many merging firms considered that the cost of filing was prohibitive given their very low level of activity in the Comesa region.

In October 2014, revised guidelines were published by the Comesa Competition Authority which established a positive threshold of assets or turnover (US$5 million) for the target company unless each of the merging parties generated two thirds of their annual turnover within the same COMESA member state. This means that mergers which have no nexus in the region will not have to be notified to the Comesa Competition Commission but that for the notifiable mergers, the merging parties continue to face in Comesa one of the highest merger filing fees in the world. According to sources in the Comesa competition commission, the number of notified mergers has not increased since the revision of the guidelines

Thus there is clearly a limit on the possibility to cross subsidize the general activity of the competition authority through the merger control fees. Raising those fees to very high levels compared to the levels in other jurisdictions may either discourage mergers (irrespective of whether they are anticompetitive or not) or push merging parties to bypass the notification procedures. In both cases the revenue of the competition authority will be less than hoped for and the merger fee setting policy will lead to efficiency losses.

Finally, it is worth pointing out that if self-funding mechanisms, properly implemented, can enhance the budgetary independence of competition authorities, they make the funding of these Authorities independent of their level of activity. This may, in certain circumstances (for example when there is a decrease in the number of economic mergers due to a downturn in economic activity), be problematic for the Competition Authority which may be lacking the resources necessary to discharge its duties in the short term. This explains why in most cases, as we saw, the laws provide for different forms of financing of the competition authority (such as a combination between possible budgetary entitlement and a self-funding mechanism).

Management of its resources by the competition authority

The effective management of its resources by a competition authority requires that, on the one hand, it be able to choose the most qualified personnel, given its financial constraint, and, on the other hand, that it uses its resources in the most effective way given its goals. With respect to this second point two issues can be discussed: the issue of prioritization of cases and the issue of the working methods of the competition authority.

It is not always true that competition authorities are at liberty to choose their staff members through an open, transparent and competitive process. Particularly in countries which transition from a system of price regulation to a system of market competition or in which several formerly independent regulators are merged within the competition authority or in countries in which the competition authority is a department of an economic ministry, competition authorities may face constraints in terms of recruitment of their staff and only be allowed to recruit from the pool of employees of the former price regulators or sectoral regulators or of the ministry they belong to. Similarly competition authorities may not always be at liberty to determine the compensation or the status of their employees and may be bound by the general rules applicable to civil servants.

For example, one can compare two different situations.

When the Turkish competition authority was created in 1997, it benefited from a substantial budget, the possibility to recruit freely its staff and the ability to set to a large extent the level of compensation of its employees. As a result it was able to attract a very large number of quality applicants with law or economic degrees who preferred to work for the competition authority rather than to work for other parts of the civil service or even in the private sector.

In France in 2008, there was a wide ranging reorganization of the competition law enforcement system. Up to that point, the Directorate for competition, consumer affairs and fraud of the Ministry of economic affairs (which was a distant successor of the price division of the Ministry) had a subdivision in charge of the basic fact gathering in competition cases as well as the powers necessary to do the dawn raids to gather evidence. The Competition Authority for its part, because it did not have the resources to do the basic investigation, relied largely on the preliminary investigations of the Ministry and limited itself to the analysis and the adjudication of cases. The separation between basic investigations and analysis of the cases and the fact that in some cases the basic investigation of the Ministry was considered by the Competition Authority as insufficient to serve as a basis for a competition case while in some other cases where the Competition Authority requested further investigation of a case, the Ministry did not come up sufficiently rapidly or adequately with the requested elements inevitably led to frictions between the Competition Authority and the Ministry. Thus in 2008, the competition authority, was given the means to do the preliminary investigations of the cases and its powers were enlarged. The Competition Authority was allowed to hire 60 new staff members to discharge its enlarged duties but was constrained in its possibilities of recruitment by the fact that the former investigators of the competition, consumer protection and fraud division of the Ministry had either to change job within the Ministry or to exercise a right to request their transfer to the Competition authority (48 of them did). It turned out that some of these former investigators were not particularly fit to become competition case handlers in a framework were investigation and analysis were integrated and had to be retooled at a non-negligible cost by the Competition Authority.

There are at least two possible problems associated with constraints imposed on the recruitment of staff members by the competition authorities.

First, it is often the case that governments moving toward a more important role for the market economy, tend to reduce their regulatory functions and to expand the surveillance function of the competition authority. Thus there is a problem of reallocating civil servants which have some economic experience to different tasks. However, what is not always perceived is that the handling of competition cases requires legal and economic analytical skills which are different from those useful in regulation whether this regulation is done by economic ministries or by independent sectoral regulators. Indeed economic regulation is mostly an a priori administrative intervention which is designed to ensure a pre-determined economic performance of the regulated firms or industries, whereas competition law enforcement is an a posteriori judgement on whether a violation of the competition law resulting in a less intense or less free competition on the market deserves sanctioning. Thus the civil servants experienced in economic regulation may not be the best prepared people to handle competition cases. Furthermore when they have had a long experience in regulation, they may have less flexibility to adapt themselves to a new task. For example, former staff members of price divisions in economic ministries, may, in handling competition cases, focus on whether investigated firms are able to justify their prices by their cost rather than on what would be the pricing mechanism if the market was competitive. Former staff members of sectoral regulators handling competition cases may have normative views about what the level of efficiency of the firms should be rather than focus on whether it can be established that their behavior investigated qualifies as violation of the competition law; they may also more prone to recommend behavioural remedies than would be justified from the competition standpoint because of their regulatory culture.

Altogether constraints on the ability of competition authorities to choose their staff outside established ministries or sectoral regulators may involve significant costs for the competition authority in adapting the skills of these staff to their purpose.

Second, in many countries, and in particular in developing countries, the scale of salaries of civil servants with a legal or economic background is quite a bit lower than the scale of revenues that people with the same level of education and skills could get in the private sector. Imposing on a competition authority that its staff should be paid according to the pay scale of the civil service has two opposite effects. On the one hand it lowers the budgetary expenses of the competition authority compared to what those expenses would be if its staff was paid competitive wages. On the other hand, however, it contributes to the importance of the turnover of staff personnel in competition authorities and to make competition authorities less able to discharge their duty than they would be if they could pay higher salaries and had a more stable personnel.

The high level of turnover of the staff of young competition authorities, particularly in developing countries, is fueled by a combination of factors. As mentioned earlier, the turnover in the staff of young competition authorities is, prima facie, due to the difference in potential revenues between the private and the public sector for legally or economically trained staffs. But this difference in potential revenues is itself due to three factors. First, because competition procedures, unlike many other economic administrative procedures, take the form of contradictory legal proceedings, there is a great demand in the private sector for skilled competition law experts with a legal or economic background to help defendants present their cases. Second, the combination of legal and microeconomic skills necessary to argue against the allegations of a competition authorities are quite specific because competition authorities generally follow a rule of reason approach to competition law enforcement which means that an intimate knowledge of the thinking of the competition authority is often as useful to assess how to mount a defence as the examination of legal precedents. This means that former staff members of the authority who have an intimate knowledge of the functioning of the competition authority may be more valuable than equally competent outsiders. Third, as competition authorities become more experienced at handling cases and at imposing fines, the stakes for firms which are alleged to have violated competition law increase and the so does the private demand of law firms or economic experts consulting firms for the services of former staff members of competition authorities.

Whatever its level of resources and its ability to retain its staff, the competition authority has to decide how it is going to allocate its limited resources.

In most countries merger control is an ex ante authorization process which includes an obligation on the parties to notify reviewable mergers and the competition Authority has no possibility of to prioritize its activity in this area as it has to make a decision for each notified merger on whether the merger can go through or is approved under conditions or is blocked on competition grounds.

However, when it comes both to antitrust enforcement and to the advocacy function of competition authorities, the competition authority may have some discretion to decide on its priorities.

In some countries the competition authority has a very limited ability to allocate its resources because it is required to act on all of the cases for which it gets a reference.

An example is provided by the French competition authority which has an obligation to publish an appealable decision in each case for which it gets a reference. For cases which are without merit whatsoever, either because the issue raised is outside the scope of the competition law or because there is no evidence of anticompetitive practice, the Competition Authority can decide to throw out the case without investigation it but must publish an appealable decision explaining that the reference was misguided. But for cases which have merits but that it would not have chosen to investigate, for example because the social cost involved in pursuing those cases is disproportionate compared to the social benefits of the case (including its potential deterrent effect), the Competition Authority must investigate to avoid the possibility (which has occurred in several occasions) that the Court of appeal will overturn its decision to dismiss the case as unjustified.

The reason for which the French Competition Authority has this obligation to investigate all cases is that, because it is not a court but merely an administrative (independent) body, there was a concern that it should not be given the ability to pick and choose the cases it wanted to investigate. Thus there was a conscious decision of preserving fairness to the detriment of effectiveness. However, to the extent that its resources allow it to initiate cases over and beyond the cases for which it gets a reference, the French Competition Authority can and does choose its priorities.

In some other countries, where the competition authority did not have much freedom to choose its antitrust cases, a reform of the procedure increased its degree of liberty. Such was the case, for example, in Greece. The Greece’s Economic Adjustment Programme, included a variety of fiscal measures, as well as structural reforms aimed at enhancing the overall competitiveness of the Greek economy. Acknowledging the central role of the Hellenic Competition Authority in the efforts to strengthen the functioning of the Greek economy, the Economic Adjustment Programme also included a competition-related component providing for the revision of the Greek Competition Act. Among other objectives, the proposed revision of the Greek Competition Act aimed at consolidating the deterrent and overall systemic effect of its enforcement action, focusing notably on procedural efficiency and independence. One of the weaknesses which had been identified as hindering the effectiveness of the Competition Authority was the fact that it did not have a margin of discretion in setting its own strategic objectives and priorities. A new Competition Act (2011) enhanced the Hellenic Competition Authority’s ability to set strategic goals and to prioritize important cases, with a view to increasing the systemic effect of its enforcement action. The Greek contribution to the OECD debate on changes in institutional design of competition authorities states⁶³ “The Hellenic Competition Commission adopted—pursuant to Art. 14 (2) of the Law—an internal management tool in the form of a “Point System” for the investigation of cases by the Directorate-General (decision no. 539/VII/2012). In particular, the Directorate-General shall now investigate pending cases according to their ranking on the basis of a point system, which essentially exemplifies and quantifies the priority criteria set out in the abovementioned decision. The Point System aims at enhancing the efficiency of investigations, the focus being on important cases with increased estimated impact on the functioning of effective competition and/or overall systemic effect, while promoting a more coherent and targeted policy of prioritizing pending cases, whereas it provides for the possibility of rejecting complaints that get a low priority ranking”.

Similarly, in 1996, Hungary went from a system of no discretion to a system of discretion for the competition authority to choose the cases it pursues.

In most countries where the Competition Authorities have the power to select the cases it wants to investigate, they have established a system of prioritization. The criteria for prioritization and the process of prioritization, however, vary from country to country as is documented by the ICN.⁶⁴

Some countries prioritize markets where they believe they are the most likely to find competition law violations, other countries prioritize sectors that have been recently liberalized or sectors which are prone to market failures. Another group of countries prioritize sectors because of their significance which may be judged by the importance of the turnover of firms in the sector, the impact of the investigation, or the likely deterrence effect of the investigation. Still other countries adopt a principled approach to prioritization setting a number of criteria to be considered simultaneously such as the effect on consumers, the strategic significance of the sector, the likelihood of a successful outcome, the resources to be mobilized for the case etc…

Countries also differ on the extent to which they make their prioritization criteria public and the extent to which they justify the reasons why they refuse to investigate certain cases. Whereas in some countries the prioritization criteria are publicly discussed by the competition authority with stakeholders, in other countries they result from internal consultations and are not always known by firms.

The ICN document on prioritization stated several reasons why agencies might want to communicate their prioritization criteria externally and why they might be reticent to do so.

Reasons to communicate prioritization criteria to the general public include the desire to increase the transparency of the work of the agency, the desire to build a strong relationship with different stakeholders, the desire to build a strong network of partners of the agency in order to promote a competition culture or the desire to encourage complaints that reflect the priorities of the competition authority.

On the other hand, competition authorities may be reluctant to publish their prioritization criteria or to discuss them with stakeholders for fear of giving an incentive to the firms in the targeted sectors to conceal the proofs of their violations of the law or for fear of raising the expectations of stakeholders with respect to the delivery of enforcement decisions in the sectors prioritized or for fear of disclosing confidential information or for fear of prompting challenges of their prioritization criteria. For example, the Greek contribution to the OECD debate on Changes in institutional design⁶⁵ stated: “The Point System is intended solely for internal use (such that the ranking of each individual case at the investigation phase is not made public, nor notified to the complainant), in an apparent attempt to immunize the system from litigation”.

Whereas there is no question that prioritization is a useful tool of management to help competition authorities to focus on the achievement of their strategic goals, too strict a prioritization process can lead to rigidities in the agenda setting of the competition authority to the detriment of the agility necessary to deal with new or important challenges as they arise. Thus a regular reassessment of the prioritization criteria and an occasional departure from those criteria may be necessary to make the competition authority sufficiently responsive to its environment.

Related to the last point, it is worth noting that a number of requests for international cooperation on antitrust cases seem to be denied by the requested authority on the basis of the fact that the request clashes with the priorities of the requested authority. Indeed, unless international cooperation is itself one of the criteria of prioritization, there is the possibility that the move toward more extensive prioritization processes by competition authorities may lead to a decrease in the flexibility needed to accommodate requests for help from foreign agencies.

Assuming that the resources are adequate in quantity and quality and that the prioritization processes of competition authorities has helped them select the most relevant cases given their strategic goals, one last question is whether there are working methods to ensure that legal staff and the economic staff of the Competition Commission work most efficiently to deliver high quality investigations at the least cost for the Authority.

There does not seem to be a unique solution to this problem as there is a trade-off between two possible desirable objectives: first, the objective of integrating the legal and economic approaches to ensure that the economic case of the competition authority will stand in court and that the legal case make economic sense and, second the objective of not allowing compromise which may ruin the integrity of each approach at the investigative stage.

The practices of competition authorities seem to vary significantly on the organization of their staff for working on cases.

For example during the debate at OECD on Changes in Institutional Designs the competition authority of the Netherlands,⁶⁶ discussing its handling of mergers, explained that “there was a concerted effort within the ACM to set-up case teams that bring together the knowledge necessary to deal with the case at hand and to harmonize procedures and working methods”. It further stated : “This flexibility is crucial. What emerges from an analysis of these initial structural issues is that designing an authority is not—or at least not only—about the structure. To a far greater extent, it is all about people, about how they work and how they work together”.

But during the same debate the US FTC contribution explained that the competition lawyers and the competition economists within the FTC worked mostly separately from each other on cases. It stated: “At all critical points of a competition investigation, including the decision to issue compulsory process, to begin adjudicative procedures, or to accept a consent decree, the lawyers and economists write separate recommendation memoranda and submit them to the decision-makers through their own Bureau management. When appropriate, Bureau of Economics and Competition managers write memoranda presenting their own views. Before the matter reaches the Commission for decision, the Director of the Bureau of Competition convenes a meeting to evaluate the matter in which the staff economists and managers from both Bureaus participate. Both sets of memoranda are provided to the Commission and representatives of both Bureaus present their views at Commission meetings”.

5 Conclusion

Altogether several conclusions emerge from the preceding analysis.

First, institution building is an art rather than a science when it comes to designing competition authorities. In many countries, both developing and developed, the institutional design of competition authorities is repeatedly modified over time. Given the number and the complexity of the trade-offs involved, it is only natural that countries need to experiment before they settle on a set of institutional characteristics for the competition authority. In that perspective, establishing a process to regularly review the adequacy of the institutional design of the competition authority is likely to be useful, particularly in the first few years after the creation of a competition authority. As Philip Lowe, the Director General of the Directorate-General for Competition at the European Commission stated in 2008⁶⁷ “ In order to fulfil their role effectively (competition policy) institutions must constantly assess and reassess their mission, objectives, structures, processes and performances. It is only through realizing and adapting to changes in their environment and through carrying out the corresponding improvement that their competences, powers, budget and ultimately existence can be justified before a wider public”.

Second, clearly there is no ideal model that fits all situations. Thus, the best way to go about designing a competition authority is for a country to start by considering the issues which seem to be most relevant to its domestic circumstances and the trade-offs they have to make. For example, in countries in which the main problem is the fact that its judiciary is weak, ill-informed of non-legal matters and slow, it may be necessary not to choose a prosecutorial model but to choose the administrative model even if a prosecutorial model offers superior guarantees in terms of separation of investigation and adjudication. If, on the other hand, the main concern is that the civil service or the political class are corrupt, it may be useful to consider the adoption of various measures guaranteeing the transparency of the appointment process of the members of the Competition Authority and the independence of the status of the members of the competition authority as well as the transparency and the integrity of its decision-making process even if there is a risk that a very independent institution may carry less weight in terms of advocacy. If the main issue faced by the country is the fact that the development of market competition is likely to be challenged because the economic power is concentrated in the hands of a few operators (whether they are chaebols in Korea or oligarchs in Russia), the political backing of the competition authority becomes important to ensure that the market player understand that the government is serious about the promotion of competition and to ensure that the competition authority is not outmaneuvered by the holders of privileged situations. In such cases, it may be that the power of the institution (and its proximity to the ventral government) may become more important than its independence to bring about the hoped for changes. If in a country economic and legal expertise on competition are scarce, the country may have no choice but to settle for a simpler law, a lower cost approach to competition with per se violations and/or a smaller and less costly competition authority without separation between investigation and adjudication etc….

Third, it should be clear that the institutional design of a competition authority will have an influence on the way the competition authority will discharge its duties. Competition authorities which must meet performance indicators will naturally tend to prefer advocacy to enforcement because deterrence of enforcement actions is usually very difficult to measure. Competition authorities which have regulatory functions in certain sectors are likely to have different goals and also to use a more regulatory approach (through the use of behavioural remedies) than competition authorities which are only responsible for competition. Competition authorities which do not have adequate funding will deliver lower quality decisions or advice than more richly funded competition authorities either because they will have insufficient staff at their disposal or because their most qualified staff will be regularly poached by the private sector and they will face a much higher level of turnover of their staff. Those differences in performances or in the way competition authorities discharge their duties are inevitable because competition authorities are set in different legal, political, sociological or economic environment.

What is important however, and what may not have been sufficiently emphasized in the past, is that each country chooses the best possible institutional design given the constraints it faces. There has been a great deal of efforts in the past 20 years to promote substantial convergence among competition authorities and this effort has largely succeeded in bringing about a common understanding of what competition analysis entails and to develop best practices. But, there may have been insufficient thoughts about the institutional design of competition authorities and about the interaction between the institutional design of the authorities and their ability to successfully implement these good practices. Many questions raised by officials in charge of establishing or upgrading competition authorities have been answered through various programs of technical assistance on the basis of the experience of more established competition authorities rather than on the basis of a careful analysis of the local circumstances and of the particularly important trade-offs faced by the countries requesting this assistance. The result is a certain frustration of competition officials in new or small agencies in number of developing countries which are facing great difficulties in emulating the more established competition agencies because while they operate in a very different context and with different constraints, their institution is not well adapted to their environment. In other words, the long term effectiveness of competition authorities is not only dependent on the substantive quality of the economic analysis they perform or of the way they are managed but also dependent on the relevance of their institutional set up in the countries in which they are established.