In the absence of global antitrust law: looking to “ bricks and mortar ” institutions and agency networks

Purpose – This paper aims to draw attention to a broad range of experimental institutional initiatives which operatein theabsence of a globalantitrustregime. The purposeof this paper is to offer food for thought to scholars inother ﬁ elds of internationaltrade lawfacing challengesfromdivergent nationalregimes. Design/methodology/approach – Taking inspiration from political science literature on institutions, this paper crafts a broad analytical lens which captures various organisational forms (including networks), codes (including soft law) and culture (including epistemic communities). The strength and shortcomings of traditional “ bricks and mortar ” institutions such as the European Union (EU) and General Agreement Tariffs and Trade/World Trade Organisation are ﬁ rst examined. Then, the innovative global network of International CompetitionNetwork(ICN) is analysed. Findings – It highlights the value of the global antitrust epistemic community in providing a conducive environment for extensive recourse to “ soft law ” . Examples from the EU and the ICN include measures which ﬁ nd expression in enforcement tools and networks. These initiatives can be seen as experimental responses to thechallengesof divergentnational antitrust regimes. Research limitations/implications – Itis desktop researchrather than empirical ﬁ eld work. Practical implications – To raise awareness outside the antitrust scholarly community of the variety of experimental institutional initiatives which have evolved, often on a soft law basis, in response to the challenges experienced by national enforcement agencies and businesses operating in the absence of a global antitrustregime. Originality/value – It offers some personal re ﬂ ections on the ICN from the author ’ s experience as a non-governmental advisor. It draws attention to the ICN ’ s underappreciated range of educational materials which are freely available on its website to everyone. It submits that the ICN template offers interesting ideas for other ﬁ elds of international trade law where a global regime is unrealisable. The ICN is a voluntary virtual network of agencies collaborating to agree ways to reduce clashes among national regimes. Its goal of voluntary convergence is portrayed as standardisation rather than as absolute congruence. Even if standardisation of norms/processes is too ambitious a goal in other ﬁ elds of international trade law, the ICN model still offers inspiration as an epistemic community within an inclusive and dynamic forum for encouragingdebateandcreatinga culture of learningopportunitieswhere familiarity and trustis fostered.


Introduction
The ongoing absence of a global antitrust (or competition law) regime is significant because of its potential implications for the competitiveness of the globalised marketplace.The lack of a global regime with uniform norms and processes creates challenges for antitrust enforcement agencies which apply different national/regional norms and processes.The concern is that disjointed enforcement is an inefficient way to deal with cross border anticompetitive practices which harm consumers.Moreover, having to comply with divergent substantive norms and enforcement processes creates inefficiencies for businesses that trade across borders.In practice, national (or regional) antitrust is applied extra-territorially.This has been described as a default response of "unilateral jurisdiction" (Gerber, 2010, p. 5).Taking the view that unilateral jurisdiction is an unsatisfactory solution, this paper looks at alternative responses involving institutions (broadly understood).It traverses a broad landscape which starts with traditional 'bricks and mortar' institutions and culminates with a voluntary virtual global network of antitrust agencies.
The European Union (EU) regime is examined first because it exemplifies the most convergent supranational antitrust model.Nonetheless, as a regional regime, it exhibits limitations on the global stage.Thus, attention turns to more globally rooted institutional endeavours.In order to present a sharper contrast, the analysis confines itself to examining two radically different global institutional arrangements.The analysis of General Agreement Tariffs and Trade (GATT)/World Trade Organisation (WTO) considers its (relatively short lived) direct contribution to global antitrust and takes the view it deserves credit for stimulating interest in innovative institutional arrangements such as the International Competition Network (ICN).The ICN is a virtual global network of antitrust enforcement agencies designed to facilitate intensive collaboration among its members with the aim of, at least, reducing friction and, more ambitiously, achieving more standardisation across divergent local regimes.
This paper draws attention to selected modes and fruits of experimentation within a trusted (or, at least, familiar) international epistemic community when a harmonised global regime is not readily realisable.It offers an exploration and critical consideration of institutional initiatives which commenced operating in an environment of divergent national regimes.Its appraisal of their operation highlights incidences of innovation and emphasises their recourse to "soft law" measures which incline against divergence.By focussing on the evolution, motivation and operation of a wide span of institutional models, this paper shines a light on a variety of responses to divergence within the antitrust field.Drawing on the author's experience as a non-governmental adviser (NGA) to the ICN, this paper offers some personal reflections.It concludes by suggesting that the network model (even if not as full blown as the ICN) offers food for thought for other fields of international trade law which encounter challenges in the context of persistently diverse national regimes.

Antitrust in the global marketplace
Antitrust law is enforced locally along national (or, in regional) territorial lines.However, national antitrust regimes, on account of the globalised nature of markets, may impact antitrust agencies and businesses that are located outside the regimes' own territory.
Antitrust enforcement agencies faced with transnational anti-competitive activities (e.g.cross-border cartels) must, in practice, seek to engage in bilateral/multilateral interactions with enforcement agencies operating in other jurisdictions whose substantive norms and enforcement toolkits may diverge significantly from each other.The difficulties faced by agencies operating in different political environments have been recognised in the observation that "rising populist concerns and differences in competition laws, increase tensions among competition agencies and the risk of divergent approaches to enforcement" (Pham and Pecman, 2019, p. 22).
Businesses may be adversely affected by the lack of global antitrust law, for example, in the form of incurring additional costs when complying with varying antitrust rules in multiple jurisdictions.A prevalent illustration is where parties to a proposed merger are obliged to notify their transaction to national antitrust agencies in different jurisdictions with different practical requirements (e.g.documentation) and, sometimes, different substantive tests.
The scale of the problem has been exacerbated by the proliferation of national antitrust laws, especially in recent decades.The reasons for the relative explosion of antitrust laws since the 1990s include: [. . .] the transition from central planning to markets in multiple countries; the evolving widespread desire to introduce greater competition in such markets consistent with retained scope for governmental review; increased cross border merger and acquisition activities; growing concern about international cartels and other anti-competitive conduct with extraterritorial effect, among other factors.Moreover, in trade relations between a number of countries, barriers to market access were increasingly identified as stemming, in part, from private restraints, which had perhaps become more apparent as governmental barriers were reduced (Janow and Rill, 2011, p. 22).
Currently, antitrust laws are found in more than 120 jurisdictions.That number gives a sense of the geographical scale of the context and the spread of antitrust regimes across the globe.
The "unilateral jurisdiction" approach may be the most common solution in practice, but it cannot be regarded as a perfect substitute for a global regime.Some of its drawbacks identified by Gerber include the reality that the national laws were not designed to operate extra-territorially, have limited capacity to deter anti-competitive conduct on global markets and encourage jurisdictional conflicts without providing an effective means to resolve them (Gerber, 2010, p. 5).Relying on national antitrust law in international scenarios, as he elaborates, results in the legal systems with sufficient economic leverage or political power to enforce their laws outside their borders (which in practice means the USA and, to a lesser extent, the EU) provide and enforce transnational competition law rules (Gerber, 2010, p. 4).
The international enforcement landscape has been described as a "heterogenous reality" which is best viewed "as a cluster of islands of domestic enforcement, differing in size, economic power and political interests" (Ezrachi, 2012, p. 5).The fragmentation of the field, especially in the context of enforcement, has been noted (Maher, 2015).
Unilateralism creates "a murky, haphazard and uncertain patchwork of norms, interests, institutions, and procedures that does not provide a predictable framework for economic decision making on global markets [. ..]" (Gerber, 2010, p. 5).These are the shortcomings of the "unilateral jurisdiction" approach which provide the setting for and prompt this paper's interest in examining alternative institutional approaches in the absence of a global antitrust regime.

Institutions
This paper's choice to focus on institutions is underpinned by Gerber's assertion that while competition is an abstract idea which refers to a process of economic exchange, it is institutions which make competition possible and shape its form and intensity (Gerber, 2010, p. 2).

Definitions
Understandings of the concept of "institutions" vary.Some interesting definitions are found in the literature from so-called Schools of Institutionalism in the field of Political Science.Institutions are defined by North (1991, p. 97) as the "rules of the game in a society, or more formally, [. ..] the humanly devised constraints that shape human interaction."Wilks defined institutions as a "series of organisations, enactments and interests, all bound together by mutually understood ideas and by shared normative assumptions" (Wilks, 1996, p. 2).For Thelen and Steinmo, institutions show "how [factors] relate to one another by drawing attention to the way political situations are structured" (Thelen and Steinmo, 2021, p. 12).Hall regards the "formal rules, compliance procedures and standard operating procedures that structure the relationship" between individuals in various units of policy and economy as institutions (Hall, 1986, p. 19).The foregoing sample of definitions evidences the considerable flexibility around the concept of institutions.This paper draws particular inspiration from the approach of March and Olsen (1989, p. 22) which recognises institutions in the shape of "routines, procedure, conventions, roles, strategies, organizational forms and technologies [. ..] beliefs, paradigms, codes, cultures and knowledge."Consequently, the institutional lens adopted in this paper is not confined to organisations and acknowledges the distinction drawn by Selznick (1957).

Antitrust institutions
Typically, national antitrust institutions are conferred with enforcement powers in relation to prohibiting anti-competitive unilateral, bilateral and/or multilateral conduct of businesses.Antitrust institutions in some jurisdictions have additional specific competence to approve or prohibit mergers.Nevertheless, there is significant variety in the national antitrust institutional architectures established throughout the globe (Maher and Papadopoulos, 2012).
Such is the variation that it is difficult to articulate a comprehensive categorisation of antitrust enforcement models in a simple format.The difficulty stems from differences in their allocation of enforcement competences among particular units within the whole enforcement institutional framework.Crucial differences may exist as regards which entity has competence to investigate suspected violations; to make determinations of infringements; to issue binding orders (which may include financial sanctions) and to adjudicate on challenges to administrative determinations.
Fox and Trebilcock offer a classification of antitrust enforcement which comprises three basis models.The first is the "bifurcated judicial model" (where the competition agency goes to court for enforcement); the second is the "bifurcated agency/tribunal model" (where the competition agency goes to a specialised tribunal for enforcement); and their third is the "integrated agency model" (where a unit within the competition agency makes the first level adjudication) (Fox and Trebilcock, 2013, p. 5).The "bifurcated agency model" is found in Canada, South Africa and Chile.Japan and China offer examples of the "integrated agency model."Some jurisdictions contain two types of models.For example, in the USA, the Federal Trade Commission typifies the "integrated agency" model and the Antitrust Division of the Department of Justice (DOJ) exemplifies the "bifurcated judicial" model.Yet, as Fox and Trebilcock concede, even with the nine jurisdictions selected for their study, the three categories do not completely capture the range of antitrust enforcement models.When creating a model some jurisdictions draw elements from different models.For example, India combines elements of the "bifurcated agency" model and "integrated agency" model while Australia and New Zealand combine elements of all three models.
Measuring the performance of competition law institutions is a complex but important exercise.Kovacic offers a perspective which is interesting for the attention it pays to the broader societal significance of the performance of antitrust institutions.In his view, antitrust agencies, by fulfilling their duties capably, directly influence not just the outcome and effectiveness of the substantive antitrust rules, but, additionally, contribute to the confidence of the public in public governance and thereby strengthen the legitimacy of public administration (Kovacic, 2009).
Several norms of antitrust institutional performance across both inquisitorial and adversarial legal systems are articulated in Fox and Trebilcock's study.These include "timeliness of decision-making, predictability, expertise, transparency, independence, efficiency and effectiveness, and accountability" (2013, p. 9).The primary problems they encountered, across several systems, included: excessive delays; lack of predictability; lack of consistency; and a lack of legal and economic expertise.The latter being especially likely in "younger and resource-starved jurisdictions, and also in small economies without a critical mass of competition cases" (2013, p. 9).In some jurisdictions, they observed a "lack of reasoned decision-making, lack of publication of decisions, and lack of independence from political interference at some or all stages-investigative, enforcement and adjudicative.Political interference ranges from rare to pervasive.Institutional arrangements that tend to ensure public accountability of the agency on a regular basis are generally recognised as a virtue" (2013, p. 9).Interestingly, they conclude that the "strengths and weakness in performance do not divide neatly by basis regime type" (2013, p. 9).
The foregoing conveys a flavour of the variety in the design of national antitrust institutions and their performances.The discussion next moves to considering concrete examples of antitrust institutions which operate beyond one national territory.It deals with the EU competition law regime before considering more globally grounded models.

European Union
The EU competition law regime merits examination because it illustrates supranational, transnational and regional dimensions of enforcement.It is a uniquely developed model.Supranational EU competition been described as the first supranational policy of the EU (McGowan and Wilks, 1995).It, of course, operates within the context of the EU (previously EC) law regime whose convergent development owes so much to the seminal EU law judgments delivered by the European Court of Justice (ECJ) (as opposed to legislative measures which require approval from, inter alia, the Council of Ministers).Key EU law judgments include those which established that EU law constitutes "a new legal order" [1]; the doctrine of supremacy (that EU law prevails over conflicting national laws) [2]; and the doctrine of direct effect (which allows certain EU laws to enter directly into the national legal systems and be relied on before national courts) [3].While these doctrines on the interface between national and supranational substantive norms are highly significant the particular focus of article requires that greater attention be paid to the antitrust institutional design.
Achieving consensus from different states on the design of structures for enforcing competition law beyond the confines of one national jurisdiction is an ambitious venture.For this reason, it is instructive to reflect on the founding and evolution of the EU competition law institutional framework.That the initial six Member States of the then European Community did not have convergent pre-existing competition law regimes made it more difficult for them to reach consensus on the format of the legal model, including the design of the institutional structure (Deringer, 1963).
In the face of differences of view, the belief that placing powers to execute the competition laws in the hands of the European Commission would minimize Member States' political interference with enforcement prevailed (Laudati, 1996).The Commission, as the most centralised institution, was seen as the best placed institution to establish the application of EU competition rules across the Union and, thereby, promote market integration by prohibiting private barriers to be erected (Dabbah, 2010).
While the text of the main antitrust substantive provisions were set out in the Treaty of Rome 1957 [4], the legal framework for their enforcement came later in 1962 in the form of Reg.17/62 which was replaced by Reg.1/2003.Reg.1/2003 (the so-called "Modernisation Regulation") came into effect in May 2004 and introduced some new provisions as well as re-enacting some of the provisions previously contained in Reg.17/62.A so-called "one-stop" EU merger regime came into existence only when Member States, after decades of discussions, agreed to the first Merger Regulation in the late 1980s [5].The interesting point to appreciate is the extent of difficulties, even within the closely integrated environment of the EU legal order, to achieving political consensus from a group of states on antitrust measures which extend beyond a national territory.
The antitrust enforcement model created by the EU Regulations may be regarded as an "integrated administrative model."The Commission enjoys remarkable enforcement powers in relation to "undertakings" (interpreted broadly to capture any entity engaged in economic activity) suspected of infringing either of the two main substantive prohibitions contained in Art 101 TFEU and/or Art 102 TFEU.Extensive powers of investigation (in the sense of fact finding) were conferred on the Commission.It may request or require information to be supplied to it by undertakings suspected of having committed an infringement [6].It is entitled, without seeking judicial authorisation, to carry out necessary inspections of undertakings' premises [7].It may copy documents, seal documents or premises and "ask any representative or member of staff [. ..] for explanations on facts or documents relating to the subject matter and purpose of the inspection and to record the answers" [8].Inspections, often known as "dawn raids," may occur without warning and simultaneously at different venues, often in more than one Member States (typically with the assistance of the national competition authority [NCA]).Since May 2004, Commission has power to undertake inspections at private personal premises (typically homes and/or vehicles) if it obtains any judicial authorisation as required by national law of the Member State [9].Following the fact-finding stage, the Commission may decide to close the case file.Otherwise, it may open a formal procedure (which has been termed the "inter partes" stage) where submissions are made and, if the parties wish, an oral hearing may be held.On completion of the "inter partes" stage, the Commission may determine whether any infringement of the substantive prohibition has occurred [10].
In addition to competence to make infringement determinations, the Commission may issue binding order(s).This potential impact of its orders is deep as they may be either behavioural (such as "cease and desist") and/or structural (such as divestiture) [11].The most striking competence of the Commission, as an administrative entity, is relation to imposing fines on undertakings [12].Fines for intentional or negligent infringements of the substantive prohibitions shall not exceed "ten per cent of its total turnover in the preceding business year" [13].When fixing the amount of the fine, regard must be paid to "both the gravity and the duration of the infringement" [14].Fines, up to a maximum of one per cent of turnover, may be imposed for procedural-type infringements, for example supplying incorrect or misleading information [15].Although fining decisions are expressed in the Regulation to "not be of a criminal law nature" [16] this is a highly contentious issue from the perspective of rights (Wils, 2011;Barbier de la Serre and Lagathu, 2018).
This paper next draws particular attention to the measures in the Commission's enforcement toolkit which are not based entirely on black letter formal instruments emanating from the legislative process involving the Council of Ministers/European Parliament.Over the years, the toolkit has been supplemented by various initiatives or policies emanating from the Commission.They may be regarded as "soft law"a genre whose definition and significance is contested, especially in the context of EU law (Cosma and Whish, 2003;Trubek et al., 2006;Stefan, 2013;Terpan, 2015).As formally non-binding measures these are of direct interest to this paper's ambition of highlighting innovative and unconventional institutional responses.
Starting in the 1990s, the Commission began to operate a leniency policy grounded in Notices [17].Following some amendments, the Leniency Programme has become, in practice, one of the Commission's most valued tools for dealing with suspected cartels.Under the Leniency Programme, the Commission may grant a fine remission of up to 100%.Thus, unsurprisingly the programme has sparked some debate (Martyniszyn, 2015).Cartel settlements is another example of a powerful enforcement tool based on a nonbinding Notice.This mechanism has been used by the Commission since 2008 to conclude its investigations into a cartel [18].The Commission may, at its discretion, invite parties to reach a settlement with it after it has indicated its proposed conclusion (of an infringement) and likely level of fine.Under this procedure, the parties must make a final settlement submission in which they admit their liability; that they do not envisage requesting either access to the file or to an oral hearing and the maximum fine.
For several years, the Commission was prepared, in certain instances, to accept so-called "undertakings" (essentially promises as to future conduct) without a firm legal basis being in place.This practice was problematic where the undertaking was breached.The situation was rectified by the inclusion of express power in Reg.1/2003 to accept "commitments' in cases where the Commission was otherwise intending to adopt an infringement decision (Dunne, 2010) [19].The Regulation makes provision for instances in which the commitment is not observed.The commitment mechanism has become quite prevalent in recent years.
It is well understood that the Commission enjoys remarkable powers to directly enforce competition law.What is, perhaps, less widely appreciated, is the extent to which some highly potent enforcement powers exercised by this administrative body do not stem from formally binding measures agreed by the EU legislative process.
The judicial dimension completes the EU competition law enforcement architecture (Andreangeli, 2018;Nazzini, 2012).For decades, the ECJ was the only court established within the EC system.In 1989, the Court of First Instance (CFI) was established to operate as a court of review of certain decisions (including competition law) of the Commission.Its judgments could be appealed, on a point of law, to the ECJ.More recently, the judicial institutions were renamed.The CFI became the General Court (GC), the ECJ is the Court of Justice (CJ).Both courts together constitute the Court of Justice of the European Union and are often referred to as the EU Courts.For undertakings seeking to challenge decisions of the Commission, there are two relevant treaty articles.Firstly, there is a general basis for review of acts of EU institutions set out in Art 263 TFEU which allows for their annulment.That article permits challenge to any "measure the legal effects of which are binding on, and capable of affecting the interests of the applicant by bringing about a distinct change in his legal position" [20].While four grounds of review are expressed in Art 263, they are "in a sense all really encompassed in the third one, the infringement of the Treaties or of any rule of law relating to their application" (Jones et al., 2019, p. 986).It is important to emphasise that the Art 263 mechanism does not involve an appellate rehearing, but the GC may look at the facts to ascertain whether Commission's factual basis was correct or sufficient and whether it discharged the burden of proof.The GC can annul the Commission's decision where it decides that the wrong conclusions were drawn by the Commission from the facts.The second relevant mechanism for undertakings seeking to make a challenge is contained in Art 261 TFEU.It grants the EU Courts "unlimited jurisdiction to review" the Commission's decisions on fines which, in effect, means that fines may be cancelled, reduced or increased.Thus, in relation to fines, the Court can substitute its judgement for the decisions of the Commission.This power contrasts with its more limited jurisdiction under Art 263 TFEU where the Court is confined to reviewing legality and may only annul.

European Union mergers
As noted earlier, a standalone merger regime (EUMR) was instituted in the 1980s.It was designed as a "one stop" shop to deal with the particular exigencies of acquisition and merger type transactions.Pursuant to the EUMR, "concentrations" with an EU dimension must be notified to the Commission for pre-implementation assessment [21].Importantly for parties, the Regulation stipulates a time frame within which the Commission must conclude its Phase I assessment.If that assessment leads the Commission to the view that the transaction raises serious doubts, a Phase II investigation is initiated.Following that phase, the Commission decides whether to declare the transaction to be either incompatible with the common market or to be (unconditionally or conditionally) compatible with the common market.A right of appeal lies from the Commission's decision to the GC, and, further to the CJ.
The foregoing analysis of the supranational dimension of the EU model detailed the extensive enforcement competences of the Commission.Its competences, as an administrative body, to enforce the main prohibitions of EU competition law are extraordinary.It enjoys power to investigate, make determinations of infringements and impose remedies (including fines).In addition, it runs key policies such as leniency, settlements and commitments.It plays the role in controlling mergers with an EU dimension.Also, it pursues fundamental policymaking functions.It provided the driving force for several instruments which aim to improve more harmonious enforcement at national level-for example the Modernisation Regulation, the Damages Directive and ECN þ Directive.
Other regional initiatives, for example, Association of South East Asian Nations have drawn inspiration from the EU model (Ong, 2018).However, to date, these have not succeeded in achieving the same level of supranational operation.This reality must be acknowledged as it indicates the obstacles to achieving the consent required for supranational legal model.That said, the EU model remains a relevant source of inspiration when regard is had to the European Competition Network (ECN).The ECN merits consideration as an example of innovative structured relations among antitrust enforcers at national and regional level.

European Competition Network
Comprising the Commission and NCAs, the ECN has been described as a "fascinating experiment" (Wilks, 2005a(Wilks, , 2005b, p. 133), p. 133).Given this paper's interest in the origins of imaginative non-traditional instruments, it is apt to highlight that the ECN's emergence came, initially, in a non-binding notice whose first paragraph referred to the "creation and maintenance of a common competition culture in Europe" [22].
The legal basis and operation of the ECN was considerably fortified by Reg 1/2003 which came into effect on May 1, 2004.It performs an extraordinary role, especially in terms of decentralising the enforcement of EU competition law to national level (Cengiz, 2010;Maher, 2009;Monti, 2014).Reg.1/2003 obliged each Member State to designate NCAs and to confer on them specified powers to effectively enforce EU competition law.It further provides rules which determine the allocation of cases among NCAs (according to a test of being "best placed" which may, depending on the circumstances, may be one or more NCAs or the Commission) [23].There is express provision requiring NCAs and the Commission to apply the EU competition rules in "close cooperation" [24].A NCA must inform the Commission 30 days in advance of issuing certain decisions.
It is important to emphasise that the ECN functions not only in a horizontal sense (among two or more NCAs) but also incorporates a rather vertical dimension (between the Commission and NCAs).Its operations occur in a number of fora including annual high-level meetings among heads of NCA, plenary sessions, working groups and sector specific sub-groups.The ECN produced a voluntary model template for leniency programmes.Notably, it is an output which was not required or necessitated by the founding regulation but is really an innovative response to diverse national approaches to mergers.
To fully appreciate the cohesiveness of the ECN and its capacity to achieve innovative voluntary collaborative responses, it is helpful to allude to the concept of an epistemic community.An episteme has been defined as "a network of professionals with recognized expertise and competence in a particular policy domain and the authoritative claim to policy relevant knowledge within that domain" (Haas, 1992, p. 3).The power of an EU competition law episteme, according to political scientists van Waarden and Drahos (2002), provided the most persuasive explanation as to why some Member States voluntarily converged the substantive terms of their national competition law provisions.The antitrust episteme includes, at least, academics and practitioners in the fields of competition law and competition economics.The influential involvement of EU and national officials has been portrayed in terms of "'technocrats' or policy-makers who have made their careers in competition law, embrace a commitment to competition as a benign process, and share a similar world view" and noted they are "extremely influential in this technical world where political interests are either unclear or not articulated" (Wilks, 2005a(Wilks, , 2005b, p. 136), p. 136).The existence of a broadly based EU competition law epistemic community, undoubtedly, has a bearing on the functioning of the ECN.Some analyses of the ECN have noted certain common understandings among the member agencies around their approaches to enforcement and goals of competition law and that these are based essentially on trust and peer-esteem.Maher and Stefan remarked on how "through meetings, interactions and exchanges, the members acquire institutional and cultural information about each other [. ..] [and] become socialized into facilitating cooperation among each other [. ..]" (2010, p. 185).
By the same token, criticism has been expressed about the ECN.These include concerns about its consistency and accountability (Maher and Stefan, 2010, pp. 189-192).Commentators have remarked that while the ECN's "clublike qualities" may ensure better decision-making this comes at "the expense of a transparent forum for deliberation.And opacity brings the risk of domination, whether by the Commission or another NCA, dictating enforcement priorities, thus NCA's work, for the European Union's competition policy" (Chalmers et al., 2019, p. 907;Wilks, 2005a, 2005b, Agency Escape).This author, too, recognises the challenges for a network in achieving balanced interactions among members.This may present as a concern given the ECN's membership includes not just NCAs but the Commissionthe supranational entity bearing legal responsibility to ensure the effective enforcement of EU competition law by NCAs.Yet, it would be misleading to conclude that the NCAs inevitably and invariably fall in with the Commission.Notably, in June 2021, a joint paper produced by the Heads of the NCAs proposes a particular approach to the Digital Markets Act [25].
Clearly, the EU competition law institutional landscape is an intricate and complex institutional arrangement.For this paper, its most striking aspect is its institutional evolution through hard and soft law measures.Recourse to the latter can be seen as a strategic response to situations where divergences among national regimes appeared to be incapable of easy resolution through traditional instruments which require political consensus.Experimentation is a hallmark of many of the EU measures (comprising enforcement toolkit and also the ECN) which emerged long after the initial (and very powerful) competences expressed in the 1962 Regulation.Yet, irrespective of its completeness or sophistication, the EU model exhibits an inherent shortcoming as a regional regime.The significance of this limitation becomes most apparent in the face of a misalignment with an external antitrust regime.

Regional
When the EU regime encounters an external antitrust regime (e.g. the USA) difficulties can arise.Their resolution entails resorting to extra-territorial jurisdiction.The extra-territorial jurisdiction of EU competition law is the governed by the "effects doctrine."Under this doctrine, EU competition law is applicable to undertakings whose market conduct may have an actual or potential effect on competition within the EU.This test captures undertakings that are based outside the EU where their conduct is implemented with potential anti-competitive effect within the EU [26].
Difficulties have arisen, even where, as Janow and Rill remark, it was accepted, at least in the abstract, that the USA and EU each enjoyed jurisdiction to review (and impose conditions on) offshore mergers which produced domestic effects (Janow and Rill, 2011, p. 26).Indeed, the potential for clashes between regional regimes is not limited to mergers but readily extends to other enforcement actions, for example, against large businesses in the digital field.These markets offer a fertile ground for potential disagreement as agencies must deal with complex rapidly developing issues (for example two-sided markets, network effects and algorithms).Added to this, in some instances, is the recognised challenge of different regimes' incompatible ideological underpinnings, which can result in tensions between different antitrust enforcement agencies (Capobianco and Nyeso, 2017).
To illustrate this point, it is apt to return to the field of mergers to note one real example of how differences between the EU and the USA prompted the initiation of more structured interaction among antitrust enforcers.While not frequent, "some sharp political flare-ups" in the context of certain "high profile and acrimonious cases" (e.g. the acquisition of McDonnell Douglas by Boeing and the proposed acquisition of Honeywell by GE) acted as "something of a wake-up call to antitrust officials in the US and in Europe that additional steps were needed to ensure that there was sufficient confidence between enforcement agencies to handle the inevitable suspicion that could arise in the body politic as to the motivation behind an offshore merger prohibition" (Janow and Rill, 2011, p. 27).In the wake of this episode, bilateral consultative processes involving information exchanges were developed for the USA and the EU.
It is true that, from a global perspective, bilateral arrangements (even if they are more ambitious than information exchanges) are inherently limited in their reach.In that light what might be achievable by more globally grounded institutional arrangements must be considered.

Global
Two types of global institutional arrangements are next analysed with a view to unpicking not only their operations within the antitrust realm, but, also, their origins.They will be presented as representing contrasting points on a spectrum of possible institutional configurations.Attention is paid to the GATT/WTO before turning to the ICN.This order of presentation allows the article to offer a crisp juxtaposition between a conventional "bricks and mortar" organisation and an unconventional virtual construction.

General Agreement Tariffs and Trade/World Trade Organisation
In the 1990s, antitrust began to emerge as a topic of interest for GATT.In 1993, the "International Antitrust Working Group" presented a Draft International Antitrust Code to the Director General of GATT (Fikentscher, 1996).In 1994, a "Group of Experts" was appointed by the European Commissioner for Competition.Notably, its discussions considered the drafting of "an international competition code to be superimposed on national laws, including the establishment of a single authority responsible for its implementation" but this was not viewed as "a realistic short-or medium-term option" (Mundt, 2019, p. 9).Instead, the Group recommended pursuing a two-prong approach comprising "a deepening of bilateral agreements" and a "plurilateral framework which should in the first instance, include a group of core disciplines and core countries and a structure for dispute resolution" (Mundt, 2019, p. 9).Its report in 1995 proposed the creation of a global competition code with binding common principles which states would apply within the context of the WTO [27].December 1996 marks a key moment at the Singapore WTO Ministerial when the EU initiated the Working Group on the Interaction between Trade and Competition Policy (WGTCP).Its mission was to study issues raised by Member States regarding interactions between trade and competition policy (including anticompetitive practices) with a view to identifying any areas which merited further consideration within the WTO framework [28].
Although some might have preferred a more far-reaching mandate for the WGTCP, it was recognised that there was a limit to what was achievable due to "deep disagreement among WTO members about the appropriateness of making rules on competition at a multilateral level in the WTO" (Mundt, 2019, p. 9).Some jurisdictions (including the USA) were sceptical about the effectiveness of the WTO in this area.This lack of enthusiasm may be attributable, somewhat, to WTO rules on governmental restraints on trade and also to "the lack of antitrust expertise in that trade body" (Janow and Rill, 2011, p. 26).By contrast, the European Commission saw the WTO as the appropriate locus and proposed four areas of work for the WGTCP.The first area was to: [. ..] examine the feasibility of a commitment by all WTO members to adopt domestic competition laws and enforcement systems.The second area was the identification of common principles on the substance of competition law.The third area concerned the creation of an instrument of cooperation between competition authorities.Finally, the working group should think of ways of adapting the dispute resolution mechanism to the area of competition law and policy (Mundt, 2019, p. 10).
The WGTCP issued a number of reports between 1997 and 2003.These included a feasibility analysis of a binding WTO framework supporting effective national competition policies.
A declaration supporting the group's work was issued by the WTO Doha conference in 2001.That Ministerial Declaration focussed on the working group's work on clarifying the core principles (including transparency, non-discrimination and procedural fairness); provisions on hard-core cartels; modalities for voluntary cooperation; and "support for progressive reinforcement of competition institutions in developing countries through capacity building" [29].Although it was decided that negotiations on antitrust should begin after the Ministerial Session in Cancun, the "explicit consensus" to so do was never obtained (Esteva Mossa, 2011).Consequently, although competition was one of the "Singapore issues," antitrust did not feature on the agenda at Cancun.The working group was suspended in 2003.By July 2004, the General Council of WTO had decided that interaction between trade and competition policy would no longer form part of the work programme.The net result was that antitrust was no longer an actively pursued item on the WTO agenda.
At first glance the WTO chapter may be regarded as an unsuccessful venture.However, its existence should not be written off as meaningless.According to Gerber, the WTO experience framed the subsequent discussion of global competition law and caused many to abandon the idea of multilateral agreement for protecting competition and to seek solutions in greater convergence among competition law systems and in bilateral and regional agreements (Gerber, 1999, p. 104).Sokol suggests that the WTO became "a direct cause for experimentation across both international and regional soft law institutions" (Sokol, 2011).For this author, the most striking example of institutional experimentation is provided by the ICN.

International competition network
This paper proposes that the ICN offers if not a template, then, at least some ideas, for other areas of international trade law experiencing challenges stemming from diverse national regimes.In that light, it is essential to convey the novelty of the ICN's design and operation by considering its origins.The ICN was inaugurated in 2001 with a membership of antitrust agencies (including the European Commission) drawn from 14 jurisdictions and has grown rapidly since (Fox, 2009).

Origins
The next account emphasises how the ICN emerged as a consciously distinctive venture with peculiar institutional characteristics in contrast to existing institutions.Some actors were less than content to leave global antitrust issues wholly in the hands of institutions such as WTO and Organisation for Economic Co-operation and Development (OECD).The broad mandates of such institutions, in the opinion of some, rendered them unsuitable to address adequately the practical enforcement matters caused by the internationalisation of antitrust.
The seeds of the ICN extend back as far as November 1997 when the International Competition Policy Advisory Committee (ICPAC) was set up in the USA.Its mandate was to advise on "the new tools, tasks and concepts that will be needed to address the competition issues that are just arising on the horizon of the global economy" [30].Intentionally designed to be non-partisan, ICPAC's membership comprised ten experts drawn from legal, academic and business fields.Its final report in February 2000 recommended the establishment of the Global Competition Initiative (GCI) as a venue for consultation and deliberation over antitrust by government officials, private firms and non-governmental organisations [31].
Three aspects of its Report merit highlighting as they seem, in retrospect, to anticipate the establishment of the ICN.The first is the name of "Initiative."This term was preferred in place of using terminology like "agency" or "organisation" which might have "suggested a resource intensive undertaking with implied performance that was thought to distract discussion from its purpose to its structural and organisational features" (Janow and Rill, 2011, p. 29).The second aspect is the recommendation that the GCI be established as a virtual set-up with minimal staff; be operated by its members and involve non-governmental advisors.The third striking aspect concerns its recommendations on convergence.The Report recommended that the GCI be "directed towards greater convergence of competition law and analysis, common understanding, and common culture" [32].The founding of the GCI was shaped by the hope that nations might "usefully explore areas of cooperation in the field of competition policy and facilitate further convergence and harmonisation" [33].
The GCI Initiative was presented to the OECD Global Forum in New York in 2001.Its distinctive and peculiar mode of operation was made clear when the Assistant Attorney General Antitrust Division, US DOJ elaborated that it would neither be a "bricks and mortar" organisation with a permanent secretariat nor deal with trade issues and, instead, it "would be all antitrust, all the time" [34].Soon after, the GCI was renamed as the ICN and the Director General of Competition Directorate stated that the "ICN will not be just another bricks and mortar international organisation.It will have a kind of virtual structure without a permanent secretariat, flexibly organised around its projects, guided by a steering-group" [35].
Two of the key individuals who led ICPAC have reflected on the unusual combination of global circumstances and individuals that created the momentum to establish the ICN.After noting that rarely can any enterprise of such magnitude stem from one individual or country, they record how after "important initial endorsement from the US and the European Community, officials and experts from around the world weighed in and gave the ICN the time and attention to give fruit to this enduring initiative" (Janow and Rill, 2011, p. 22).Others have similarly acknowledged the close cooperation of the competition agencies on both sides of the Atlantic that immensely facilitated the ICN's creation [36].

Set up and style of operation
The ICN eschewed the traditional international institutional architecture.Its design entails a peculiar setup and mode of operation.
One intentional differentiation clearly manifests itself in the ICN's expansive attitude to eligibility for membership.It takes an inclusive approach which allows membership by agencies located throughout the globe.That the OECD (notwithstanding its considerable engagement with less developed nations) confined full participation to countries from the developed world served as an "important reason for the creation of the ICN, which readily made membership available to all jurisdictions with a competition law and a mechanism for its enforcement" (Hollman and Kovacic, 2011, p. 63).Another inclusive aspect of the ICN's model is the involvement of NGAs from a range of professional and disciplinary backgrounds to contribute to working groups and workshops and to attend, following invitation, the annual conference.
The differentiation of the ICN extends beyond its membership eligibility criteria.It is clearly evident in its distinctive mode of operation which can be traced back to ICPAC's final report (discussed above).The ICN has been described as a "form of voluntary multinational collaboration that commentators have identified as a promising way to facilitate international ordering amid the global decentralization and diversification of economic regulation" (Hollman and Kovacic, 2011, p. 52).The ICN's style of operation is in line with the "creation of private ordering to address many of the most pressing international antitrust issues" which has been facilitated by trust-based repeat relationships through soft law mediated through international institutions (Sokol, 2011, p. 154).Soft law initiatives are key planks of the ICN's operation in the direction of addressing divergences among antitrust regimes.This is noteworthy because there is considerable contention in the literature as regards the topics of divergence and convergence across antitrust regimes (Doern and Wilks, 1996;Gal, 2009;Fox and Trebilcock, 2013;Gerber, 2010;Lucey, 2017;Rodger and Lucey, 2018).

Soft law and convergence
The ICN has been described as one of the "bridging or connecting" mechanisms which, while they lack rule making or decision making power, are "becoming generators of standards of soft law that may harden into world law" (Fox and Trebilcock, 2013, p. 11).Soft law, in this context, entails the use of benchmarking of general practices and that permits "flexibility for jurisdictions to adhere to benchmarked better practices based its own unique legal, political and economic background" (Sokol, 2011, p. 153).The ICN's "paramount goal" has been described as facilitating "convergence on superior approaches, concerning the substance, procedure and administration of competition law" (Hollman and Kovacic, 2011, p. 52).The expectation is "that if competition systems around the world opt-in to superior techniques, they will achieve greater progress towards dismantling competitive restraints within single jurisdictions and across borders" (Hollman and Kovacic, 2011, p. 53).
Convergence within the ICN has been portrayed as the broad acceptance of standards which cover substantive doctrine, procedures and competition agency administration.Such standardisation, it is argued by Hollman and Kovacic, can yield benefits, firstly, in terms of improvements in performance and effectiveness of antitrust regimes and, secondly, by reducing costs.Improvements in the performance of a national regime bring about wider improvements in the "effectiveness of competition policy as a global endeavour, by increasing the capacity of competition agencies as a group, through individual initiative and cross border cooperation, to deter harmful business conduct" (2011, p. 55).To illustrate their argument that the negative effects of poorer practices are not confined to the national jurisdiction, but, may injure economic performance elsewhere they suggest that where country that "applies an inferior approach is economically significant, companies doing business in global or regional trade may feel compelled to conform their practices to satisfy the demands of the single jurisdiction" (2011, p. 53).Thus, they argue that the quality of the individual national antitrust systems is a matter of keen interest for "the large community of nations" (2011, p. 53).The second suggested advantage of standardisation lies in the reduction of unnecessary costs.Such costs imposed on businesses includes ones "caused by subjecting mergers to multiple individual national reviews, where each involves idiosyncratic reporting requirements or where notification obligations sweep in transactions with little connection to commerce within a jurisdiction" (2011, p. 53).If the review process is standardised (e.g. by allowing a common form to report a proposed deal to numerous authorities) then "convergence can reduce the cost of commerce without diminishing the quality of regulatory oversight" (2012, p. 53).
The ICN engages in projects that seek to increase understanding of individual competition systems; consensus about "superior practice"; and encourage individual jurisdictions to opt-in to superior techniques (Hollman and Kovacic, 2011, p. 53).The ICN's style of operation, in practice, is an intentionally participative one which brings competition agencies together in working groups, workshops and annual meetings.Esteva Mosso emphasises that the non-binding nature of the ICN "series of practical recommendations and other tools such as best practices, work books and other analytical frameworks" result in a "common approach [which] has become in many areas of competition law the worldwide accepted standard, to which established agencies try to converge and emerging ones get inspiration to develop theirs" (2011, p. 165).
The field of mergers offers an excellent illustration of the ICN's approach.In 2002, it issued "Guiding Principles for Merger Notification and Review" (which contained eight principles for procedural convergence).By 2011, it had issued 13 Recommended Practices for Merger Notification and Review Procedures and, for substantive alignment, three Recommended Practices for Merger Analysis.In addition, ICN designed supplementary practical products to assist agencies deal with mergers such as its Merger Guidelines workbook (which contains practical checklists); its Handbook on Investigative Techniques for Merger Review and also, the Merger Remedies Report.The products are supplemented by the Merger Template (providing an overview of a member's national rules) and "brought to life in the course of merger workshops and working group meetings" (Esteva Mossa, 2011, p. 168).While the foregoing illustration deals with mergers, for completeness, it should be noted that the ICN is active in several other areas of antitrust such as cartels, unilateral conduct, agency effectiveness and advocacy.
In sum, a common thread throughout the ICN work is its pursuit of standardisation by developing outputs which are practical measures expressed in soft law which build on consensus achieved through repeat interactions among its member antitrust agencies and NGAs.It places considerable value on creating opportunities for close and continuing collaboration among agencies which likely reflects its appreciation of the importance of trust among stakeholders.
Having presented the ICN momentum towards standardisation, for balance, it is essential to acknowledge important limitations around the pursuit of convergence among divergent antitrust regimes.Several enduring sources of differences among antitrust regimes have been recognised.These include variations "in the economic conditions, history, legal process (e.g.civil law versus common law), and political science of individual jurisdictions" (Hollman and Kovacic, 2011, pp. 54-55).According to Hollman and Kovacic, the ICN "does not anticipate the establishment of identical policies and enforcement mechanisms across the world's competition policy systems" (2011, p. 55).They summarise the argument against "absolute congruence" as being that to "insist upon full uniformity across systems, or await unanimous approval before any single system undertook an innovation, would rob competition policy of a valuable source of continuing renewal and vitality" (2011, p. 55).Interestingly, they identify the value of "the useful innovation that comes from decentralization experimentation" (2011, p. 55).Where competition rules in one jurisdiction are based on standards that are appropriate for that jurisdiction (in light of the status and history of its markets, companies, economy, etc.) it is recognised that the same standards may not be appropriate for a different jurisdiction and that raises the need for what has been termed "informed divergence" (Fingleton, 2011, p. 183).Fingleton argues that "informed divergence" assumes particular importance in the context of moving from a bi-polar to a multi-polar world with the development of the economies of China, other Asian countries, South America and Africa (2011, p. 198).

Reflections on International Competition Network
Some reflections on the ICN next follow.It is important to emphasise that these are personal reflections and subjective observations which stem from this author's association with the ICN since 2015 as an NGA in annual conferences and working groups.
The ICN's inclusive approach to membership attracted agencies from more than 120 jurisdictions in less than twenty years.That pace of expansion is remarkable as it indicates significant level of voluntary "buy-in" from around the globe.That said, the membership does not include the China antitrust agency.Nonetheless, some NGAs who work in China in universities and in private practice play a part in the ICN.That additional inclusivity which is made possible by the NGA dimension is arguably an underappreciated strength of the ICN.NGAs are drawn from a wide base.Ex-officio Head of agencies may become NGAs and their continued involvement can bring value to institutional memory and experience.The inclusion of NGAs from a wider pool brings additional expertise and a more external perspective to agency work as well as broadening stakeholder engagement.Yet, over including NGAs may bring difficulties.In this regard, a concern might arise about a perception of privileged access to agencies by NGAs, especially ones connected to private legal practice, economic consultancy, think tanks whose clients or employers (in the case of in-house counsel) may face official interaction (investigation/merger notification) with antitrust agencies.While some sessions at the annual conference are closed to NGAs they may avail themselves of opportunities to engage with agencies during breakout sessions and in social events.NGAs are nominated by agencies rather than centrally by the ICN.The NGA toolkit produced by the ICN offers guidance to agencies.
The question of resources required to operate any entity can be challenging.Although the ICN need not finance a secretariat/bureaucracy or a building, it needs resourcing.The burden of organising its work (annual conferences, working groups) falls on agencies.Inevitably, some agencies are better equipped than others to contribute resources.This reality means that not all agencies are equally well placed to participate, for example, by hosting the annual conference.That said, there are less meaningful and demanding leadership opportunities in the shape of hosting workshop.Moreover, leadership of working groups is sensibly structured to comprise three agencies whose tenure is staggered over three years.
The ICN's relationship with other global entities raises some interesting issues about interaction and duplication.While the ICN was designed differently to traditional global "brick and mortar" institutions, this author does not wish to convey that it operates in hostile opposition to them.The ICN and the OECD published their first joint report on international cooperation in January 2021 which proposed new areas of future work [37].
Finally, it seems to this author that the range of educational materials on the ICN website available not only to members but to anyone deserves more widespread recognition.Some of the materials are created by the ICN Working Groups.In addition to their reports there is an extensive suite of videos whose content range from practical topics (e.g. on conducting investigations) to more theoretical topics (e.g. on the historical origins of antitrust).Moreover, the Document Library contains a repository of market studies and reports on product and service markets conducted by its member agencies.While this initiative was likely motivated by a desire to assist new or developing agencies which lacked sufficient resources to design a report from scratch it is a central archive of material for students and scholars of competition law and, as such, deserves greater publicity.A Work Product Catalog is available [38].

Conclusion
The absence of global antitrust regime provides the backdrop to this paper.Starting from the position that the default unilateral jurisdiction approach is not wholly satisfactory, this paper explored selected institutions which are operating in the space left unoccupied by a global antitrust regime.By adopting a broad interpretation of institutions (inspired by political science approaches) this paper crafted a wide analytical lens which allowed consideration of a wide span of forms (including networks), codes (including soft law) and culture (including epistemic communities).
The analysis of the EU model emphasised its enforcement of antitrust within a supranational legal order, a transnational governance network and a region.The EU, after a decades-long drive towards substantive convergence and, more recently, against procedural divergence, constitutes the most cohesive approach to antitrust enforcement beyond one national jurisdiction.Its innovative ECN has been described as "a 'thick' network in that it is established by law, all its members enforce the same legal norms, albeit in different national contexts and they can share confidential information" (Maher and Stefan, 2010, p. 178).Given these peculiar foundational characteristics it is unrealistic to expect the EU model will be readily adopted as a global template.
For this reason, attention then turned to considering other institutional arrangements.An established global traditional organisation was examined before considering an innovative global network of agencies.
The analysis of the GATT/WTO highlighted that, notwithstanding reports from inclusively constituted Groups which offered thoughtful proposals for progressing antitrust within the WTO, the lapse of necessary political support hampered its antitrust activities.
Consequently, although the WTO enjoys significant enforcement competences in the shape of dispute resolution and issuing rulings, its operation in the field of antitrust has diminished significantly.Yet, it left a valuable legacy by prompting more experimental institutional initiatives.
The ICN's intentionally distinct design as an innovative construction of antitrust agencies was emphasised by analysing its roots, membership criteria and its peculiar operation.The ICN operation centres on realisable practical "work products" achieved through consensus among antitrust agencies in repeated collaborative interactions.This productive dimension is seen by some as being essential in order to both maintain the ICN's credibility and, moreover, to justify the substantial efforts agencies devote to it (Esteva Mossa, 2011, p. 171).The ICN's goal of voluntary convergence was portrayed in terms of standardisation rather than as harmonisation/absolute congruence.The acknowledgement and acceptance of divergence among ICN member regimes was highlighted because it reveals an understanding not only of the need to respect national differences but, also, of the value of divergence as a stimulant of experimentation and innovation.
This article suggests that the agency network model offers food for thought beyond the antitrust field.There are other fields of international trade law where directly enforcing common global norms and processes by a conventional global organisation is not readily realisable (perhaps for lack of adequate political support).The ICN experiment demonstrates how a voluntary global network of agencies (assisted by invited experts) may collaborate on a trust basis through "soft law" actions with the aim of agreeing practical outputs which reduce, in practice, clashes among regimes which are not (and, sometimes should not be) harmonised.If standardising norms and processes is too ambitious an objective for some other fields of international trade law, other aspects of the ICN model may offer inspiration.In such instances, it is submitted that it is worth considering ways to fostering an epistemic community comprising not only agencies/enforcers but, additionally, carefully selected experts from academia and professional practice.The collation and active sharing of relevant studies and reports should be an achievable first step.
The antitrust experience over the decades shows the viability of an inclusive forum for debate and creating a culture of mutual learning which provides a fertile ground for experimentation.The EU's supranational regime and its ECN began as experimental ventures.Experimentation, too, epitomises the ICN.The lesson offered by the analysis in this paper is that experimentation undertaken within a transnational epistemic community can play a useful role where there is not the political consensus required to establish a uniform global regime with direct enforcement competences.
4. The EEC (established by the Treaty of Rome 1957) became the EC which ceased to exist when the Treaty of Lisbon came into effect on 1 December 2009 and was succeeded by the European Union.The Treaty on the Functioning of the European Union (TFEU) contains the competition rules which were previously expressed in the Treaty of Rome.The two main substantive prohibitions are currently contained in Art 101 TFEU and Art 102 TFEU. 5. Reg.4064/89 [1989] OJ L395/1.The current Merger Regulation (EUMR) is Reg.139/2004, [2004] OJ L24/1.