Remedies: Structural, Behavioral, Both, or None? Enforcement Trade-Offs for Mergers and Antitrust

This note reviews the rationale behind both structural and behavioral remedies, not only in merger but also in antitrust cases. The cost-benefit analysis for the enforcement of remedies depends on the case examined, and general presumptions on the relevance of one or another type of remedies and even on their very opportunity might easily be misguided. This calls for a consistent theory of competitive relief, in order to properly frame the enforcement of remedies.

By Andreea Cosnita Langlais¹

 

I. INTRODUCTION

Remedies, broadly defined as measures used to restore competition by removing a competitive concern or putting an end to an infringement, are yet again topical.

2019 was a busy year from this point of view: there were calls in the U.S. to break up the Big Tech companies and fix competition by imposing spin off remedies,² while in Europe the ECN+ Directive granted Member State competition authorities the same power that the European Commission has held since Council Regulation 1/2003 to impose commitments, i.e. structural or behavioral remedies as an alternative to penalties, in order to quickly restore effective competition and ensure the proper functioning of the European internal market.³

On the enforcement side, the U.S. Department of Justice gave up its more favorable view of conduct remedies for mergers as promoted by the 2011 Remedies Guide, and decided to switch to its previous version focusing on divestitures,⁴ while the new chairman of the Federal Trade Commission intends to limit the use of remedies in general.⁵ By contrast, the French Inspectorate-General for Finance recommended, in a recent report on European competition policy, that the European Commission’s recourse to behavioral remedies be facilitated, or that they be at least placed on an equal footing with structural remedies.⁶

In an attempt to understand regulators’ evolving stance on remedy enforcement, this paper briefly reviews the rationale behind both structural and behavioral fixes, not only in merger but also in antitrust cases. Rather than provide definite answers, this discussion will raise some questions.

 

II. MERGER REMEDIES: TOO MUCH OF A GOOD THING?

Despite recurrent doubts as to their effectiveness, remedies are frequently used to solve competitive concerns raised by mergers.⁷ All merger remedies are meant to preserve competition on the market(s) affected by the proposed transaction, and in so doing they help prevent potential anticompetitive conduct post-merger. The type of remedy used will, however, depend on the type of merger, and also on the agency dealing with the case.⁸

Basically, structural remedies⁹ address competitive concerns arising from the change in the structure of the market or of the value chain. They typically enable the entry of a new competitor or the strengthening of an existing player by transferring property rights. By contrast, behavioral remedies¹⁰ are supposed to (temporarily) limit the property rights and hence conduct of the merging firms, so as to allow competitors, customers and suppliers to react to the structural modification of the market or of the value chain triggered by the merger. Hence, agencies tend to prefer structural measures for horizontal mergers, and behavioral restrictions for vertical ones.¹¹

Divestitures are most often used to solve horizontal concerns due to overlaps between merging parties.¹² They may not always be feasible in the case of vertical mergers, because they can endanger the very efficiency gains expected from the deal, i.e. those arising from the elimination of double marginalization. But that does not mean that conduct remedies will necessarily perform better. To implement behavioral remedies, the agency should list all possible circumstances relevant to the potential conduct raising competitive concerns, which generates informational problems for the design of conduct remedies. The agency must also engage in regulatory-like oversight to ensure compliance, because merging firms have incentives to evade behavioral constraints.

Although structural remedies are viewed as more drastic, because they are irreversible, they are also generally less costly to enforce, because they can normally be executed more quickly, and do not require continuous monitoring over long periods of time once they have in fact been implemented (unlike conduct remedies). However, they can easily fail too. If the buyer of divested assets is a new entrant, this firm may lack the necessary information and experience to manage to stay in the market long enough to effectively compete against the insiders and drive prices down. And if the buyer of divested assets is an outsider, this firm has every incentive to maximize the anticompetitive effect of the merger. In fact, the buyer and the sellers of divested assets for a horizontal merger have a common interest to limit competition between them,¹³ which goes against the agency’s objective in imposing a fix for the horizontal overlaps created by the merger. Furthermore, the new industry structure resulting from the asset transfer may be (even more) conducive to collusion,¹⁴ due to more symmetric asset distribution or the creation of multimarket contacts.

All in all, it shouldn’t come as a surprise that the few existing studies on the effectiveness of merger remedies have concluded that they often fail to restore and preserve competition.¹⁵ And given that merger prohibitions are rare, one cannot but wonder whether merger remedies may actually be used much too often.

To answer this question, one should examine the opportunity to enforce merger remedies. It turns out few papers do so, but they all point to the possibility of lowering welfare when clearing mergers conditional on remedies.¹⁶ This basically suggests that the current trend in merger enforcement – to clear as many mergers as possible, with or without remedies – is consistent with the belief that type I errors, i.e. wrongful prohibitions, are much more costly than type II errors, i.e. wrongful clearances. More research is needed to clarify whether this might – or might not – be true.

 

III. REMEDIES IN ANTITRUST CASES: STRUCTURAL (IN)CONSISTENCY?

Antitrust violations are typically punished by fines. These are easy to execute, but will fail to restore competition on the market in the absence of injunctions to cease the violation, and may even provide incentives to increase profits through further unlawful conduct. Hence effective antitrust enforcement relies heavily on antitrust remedies. Contrary to merger remedies, which need to fix the likely anticompetitive effects of the deal beforehand, antitrust remedies must stop the specific abusive conduct undertaken by firms to restore competition on the market.

In principle, both types of remedies can achieve this, but, in reality, structural relief for antitrust violations is rare. Out of more than 400 proceedings initiated by the U.S. Department of Justice under Section 2 of the Sherman Act and surveyed in 2004,¹⁷ some 22 percent involved structural remedies. In Europe, the adoption of structural remedies for abuses of a dominant position has only been possible since 2003, once antitrust remedies were formally introduced into EU law in 2003.¹⁸ The European Commission has hardly ever used structural antitrust remedies, and then only as of 2008.¹⁹

A general reason for favoring structural over behavioral remedies in antitrust stems from the practical difficulty of splitting up an integrated firm, which operates effectively as a single entity, without provoking efficiency losses. It is far from obvious how to do that to a firm that has expanded through internal growth, or at least more complex than divesting assets from an entity resulting from external growth.

Often, putting an end to abusive unilateral behavior does not require drastic remedies, and an injunction to stop may suffice.²⁰ The remaining question is whether that is enough to curb the anticompetitive behavior in order to preserve competition, especially when the incentives for the antitrust violation stem from the structure of the relevant market. In merger cases, it is the future market structure that is likely to trigger anticompetitive behavior, and competition authorities do not hesitate to enforce divestitures in order to limit this risk. In antitrust cases, agencies are not quite ready to do the same, although the anticompetitive conduct is not potential, but has already materialized.²¹ Thus, a dominant firm may keep all its business and assets even after an antitrust violation, but a merger generating some efficiency gains while increasing or strengthening a dominant position will not benefit from the same opportunity.²²

Arguably, some competition agencies do acknowledge the possibility that structural relief can be appropriate to address the lack of effective competition on the market.²³ The expectation that in unilateral abuse cases the buyer of divested assets is more likely to put them to profitable use and compete more fiercely is probably related to the absence of a problem typically associated with divestitures for horizontal mergers – namely that the buyer – and merger-outsider – is not an ally of competition, but instead will gain from maximizing the merger’s anticompetitive effect.

Finally, most antitrust remedies, regardless of their nature, are the result of a negotiated settlement between the agency and the firm: the former forgoes a formal infringement decision to obtain a quicker fix for its concerns, while the latter avoids the sanction and gets a reduced fine.²⁴ Although both competition authorities and companies may prefer resolving a unilateral violation through negotiated commitments, concerns have been raised regarding the use of this solution as opposed to full-fledged infringement decisions. First, without the deterrent effect of the sanction, many fear the possible increase in the incentive for companies to engage in anticompetitive practices.²⁵ In addition, the case-bycase approach taken by agencies when negotiating antitrust remedies provides little guidance to parties trying to establish what types of behavior are lawful, given that commitments or consent decrees identify only conduct raising competitive concerns. To put it differently, formal prohibitions of unilateral abusive conduct will create legal precedent, and thus provide guidance for the assessment of future cases. This informational value of formal decisions to sanction anticompetitive practices needs to be factored in alongside with the deterrent effect they exert whenever the social cost and benefit of antitrust remedies is being assessed.²⁶

 

IV. CARTEL REMEDIES: A MISSED OPPORTUNITY IN ANTITRUST ENFORCEMENT?

Despite the increasing reliance on market screening to uncover them and monetary sanctions to punish them, it would be hard to argue that enough is being done against cartels. Current enforcement, even allowing for non-monetary sanctions and well-designed leniency programs, is still not deterrent enough.

So, why not contemplate additional instruments, such as structural remedies? Arguably, it would be a new tool in the agencies’ arsenal.²⁷ But given that cartel cases are the epitome of anticompetitive conduct triggered by market structure, Harrington (2018)²⁸ makes a truly compelling case for the effectiveness of structural relief for collusion.

Contrary to merger cases, for which anticompetitive effects, in particular coordinated effects, will not have not materialized at the time of the ex ante enforcement decision, a proven cartel infringement reveals beyond doubt that the market is prone to collusion. Collusion could then start again, either in explicit or tacit form.²⁹ A change in the distribution of market shares through asset divestitures would address recidivism (by creating a maverick for instance) while also preventing tacit collusion (by making explicit communication and agreement necessary).

Fine tuning would be easier than in the case of merger remedies, as potentially all cartel members could be required to divest, and given that a cartel does not generate cost savings, there is no risk of writing them off. A mandatory asset transfer could even turn out to be a more effective deterrent than the (necessarily capped) fines,³⁰ since, although akin to a monetary penalty for the firms, the liquidity constraint for the latter would be much less relevant. Finally, and to the extent that the resulting change in market structure will boost competition and thereby lead to lower industry prices, a structural remedy might even provide more extensive compensatory relief to buyers than “mere” damages (which are often available only to direct customers).

On the downside, structural remedies are certainly costlier to administer than monetary sanctions. And they may fail to correct the very problem they are supposed to address – making the market less prone to collusion – depending on the distribution of total industry capacity³¹ and the identity of the buyer (a new entrant? a maverick already on the market? other cartel members?).

Notwithstanding these drawbacks, asset divestitures wouldn’t make things worse for cartel enforcement, and Harrington (2018) rightfully stresses that the benefits of adding this instrument on top of the others would be highest when fines are relatively low, civil claims for damages are hard to bring, and non-monetary sanctions are unavailable.

 

V. SOME FINAL REMARKS

All in all, this note argues that the cost-benefit analysis of the enforcement of remedies will depend – a lot – on the particular case, and therefore general presumptions on the relevance of one or another type of remedy and even on their appropriateness might easily be misguided.

For instance, it might be worth reconsidering the extensive recourse to remedies for both mergers and unilateral anticompetitive conduct cases. In contrast, one might wonder why, so long as the incriminated behavior clearly stems from a faulty market structure, mandatory asset sales are never used against cartels?

In the end, this calls for a consistent theory of competitive relief, in order to properly frame the enforcement of remedies. For the time being we have plenty of theories of harm, but a sound theory of how to remedy harm would come in handy.

---

1 University Paris Nanterre and EconomiX-CNRS.

2 See https://www.vox.com/policy-and-politics/2019/3/8/18256192/elizabeth-warren-medium-google-amazon-facebook or https://www.nytimes.com/2019/05/09/opinion/sunday/chris-hughes-facebook-zuckerberg.html for instance.

3 Directive (EU) 2019/1 of the European Parliament and of the European Council of 11 December 2018.

4 Makin Delrahim, ’Modernizing the Merger Review Process’, Remarks at the Global Antitrust Forum, September 2018.

5 Joseph Simons, “Completed Initial Questionnaire,” Senate Commerce Committee, February 2018.

6 French Inspectorate-General for Finance (Inspection générale des finances), La politique de la concurrence et les intérêts stratégiques de l’UE [Competition policy and the strategic interests of the EU], April 2019.

7 In the U.S., on average two merger cases get litigated out of the roughly 50 investigations every year – Kwoka, John E. (2018) Reviving Merger Control: A Comprehensive Plan for Reforming Policy and Practice, available at SSRN: https://ssrn.com/abstract=3332641. In the EU, only 30 mergers have been banned by the European Commission as of February 2020 since the coming into force of the EU Merger Regulation in 1990 – https://ec.europa.eu/competition/mergers/statistics.pdf.

8 Among national competition authorities in Europe, the German Bundeskartellamt has not cleared any merger conditional on behavioral remedies since 2011, whereas the French Competition Agency made 22 such decisions; out of all merger remedies, the French agency uses behavioral measures in 36 percent of cases, whereas the Commission does so in less than 20 percent of cases – see Autorité de la concurrence. Les engagements comportementaux – Behavioral remedies, La documentation française, February 25, 2020, at https://www.autoritedelaconcurrence.fr/sites/default/files/2020-01/eng_comportementaux_final_en.pdf.

9 These can take the form of mandatory sales (of tangible assets such as subsidiaries, stores, plants, warehouses, branches, or intangible assets such as brands and operating licenses), termination of franchise agreements, non-acquisition of an asset initially included in the scope of the deal, severing ties with a competitor, or the divestiture of minority shareholdings.

10 For instance, licensing a brand to a competitor, providing access to a key, possibly essential infrastructure (network good or service, technology, patent, intellectual property rights), temporarily changing contractual clauses, prohibition of bundling services or products or of product range discounts, giving up certain customers, or preventing the passing of information between structures by putting up “vertical firewalls.”

11 Behavioral remedies must ensure a level playing field for competitors that need to either buy or use key assets, inputs or technologies owned by the merging firms, which usually happens when the merged entity is vertically integrated and there’s a risk of foreclosing the access to the upstream or downstream markets to existing or potential competitors.

12 However, conduct remedies can equally fix horizontal effects in certain cases. The 2020 report on behavioral remedies by the French competition authority (see footnote 8 above) provides examples of mergers in the regional daily press sector, for which behavioral commitments in the shape of “Chinese walls” were used (i.e. provisions guaranteeing the separation of the operational, administrative and managerial activities of the two entities belonging to the same group) in order to fix the risk of horizontal effects in terms of content homogenization and ensuing quality and diversity loss for readers.

13 See Farrell, Joseph (2003) Negotiation and Merger Remedies: Some Problems, in François Lévêque & Howard Shelanski, eds. Merger Remedies in American and European Union Competition Law. Cheltenham UK: Edward Elgar.

14 See Compte, Olivier, Jenny, Frédéric & Rey, Patrick (2002) Capacity Constraints, Mergers and Collusion. European Economic Review 46, or Vasconcelos, Helder (2005) Tacit Collusion, Cost Asymmetries, and Mergers. Rand Journal of Economics 36.

15 There are currently three such studies: two commissioned by the Federal Trade Commission in 1999 (at https://www.ftc.gov/sites/default/files/documents/reports/study-commissions-divestiture-process/divestiture_0.pdf) and 2017 (at https://www.ftc.gov/system/files/documents/reports/ftcs-merger-remedies-2006-2012-report-bureaus-competition-economics/p143100_ftc_merger_remedies_2006-2012.pdf) respectively, and one by the European Commission, DG Comp, in 2005 (at https://ec.europa.eu/competition/mergers/legislation/remedies_study.pdf). Looking at merger remedies in the U.S., Kwoka (2015) finds that mergers cleared with asset divestitures raised prices by 5.6 percent, and those with behavioral remedies by more than 13 percent on average – see Kwoka, John (2015) Mergers, Merger Control and Remedies: A Retrospective Analysis of U.S. Policy, MIT Press.

16 For instance, an asset sale to a new entrant might prevent this firm from opening its own brand – see Cabral, Louis (2003) Horizontal Mergers with Free-Entry: Why Cost Efficiency May Be a Weak Defense and Asset Sales a Poor Remedy, International Journal of Industrial Organization 21(5). But asset reallocation to outsiders already on the market can also lower consumer welfare if the merger does not generate sufficiently synergies – see Vergé, Thibaud (2010) Horizontal Mergers, Structural Remedies and Consumer Welfare in a Cournot Oligopoly with Assets, The Journal of Industrial Economics 58(4). Merger remedies may equally hinder the deterrence of the most welfare-detrimental mergers, which may then force the competition authority to investigate mergers more often – see Cosnita-Langlais, Andreea & Sørgard, Lars (2018) Enforcement and Deterrence in Merger Control: The Case of Merger Remedies, Review of Law and Economics 14(3).

17 See Crandall, Robert W. & Elzinga, Kenneth G. (2004) Injunctive Relief in Sherman Act Monopolization Cases. Kirkwood, J. (Ed.) Antitrust Law and Economics (Research in Law and Economics, Vol. 21), Emerald Group Publishing Limited, Bingley.

18 Council Regulation (EC) No. 1/2003 on the implementation of the rules on competition laid down in Arts 81 and 82 of the Treaty, OJ L 1, 4 Jan. 2003.

19 Actually, this has only concerned the European energy and gas market so far, and upon proposition of remedies by the firms themselves – see Case COMP/39.388, German Electricity Wholesale Market and COMP/39.389, German Electricity Balancing Market (26 Nov. 2008), OJ 2009/ C 36/ 8, Case COMP/39.402, RWE Gas Foreclosure (18 Mar. 2009), OJ 2009/C 133/10, and Case COMP/39.315, ENI, OJ 2010/C 352/09.

20 Actually, in the EU, structural remedies can only be adopted in antitrust cases as a last resort, i.e. provided no equally effective behavioral remedy is available, or if such a solution were more burdensome than the structural remedy considered.

21 In European antitrust this reluctance to target a firm’s dominant position in order to contain its anticompetitive behavior is most likely related to the fact that it’s not the dominant position in itself that is prohibited, but only its abuse. For instance, the French competition agency has never accepted structural remedies for unilateral anticompetitive conduct, explicitly arguing every time that the competitive concern addressed did not stem from the market structure (i.e. dominant position), but from the firm’s deviant conduct – see Autorité de la concurrence. Les engagements comportementaux – Behavioral remedies, La documentation française, February 25, 2020, at https://www.autoritedelaconcurrence.fr/sites/default/files/2020-01/eng_comportementaux_final_en.pdf.

22 Wang, Wei (2011) Structural Remedies in EU Antitrust and Merger Control. World Competition 34(4).

23 This is the case in UK, where the CMA has the power to initiate a market inquiry resulting in divestitures if an ‘adverse effect on competition’ is found – see Competition and Markets Authority, Market Studies and Market Investigations: A Supplemental Guidance on the CMA’s Approach (2014, rev. 2017). A similar provision exists in non-metropolitan French regions, where, since Lurel Act of 2012, the French Competition Agency can enforce ‘structural injunctions’ to tackle competition concerns raised by a dominant position on the final retail market – see Lurel V. (2013) Le retour de l’Etat régulateur en outre-mer. Concurrences, n° 2-2013.

24 Since the application of the Council Regulation No 1/2003 making room for commitments, the European Commission has heavily relied on such decisions: more than 70 percent of the abuse-of-dominance cases were resolved with commitments – see Mariniello, M. (2014) “Commitments or prohibition? The EU antitrust dilemma.” Bruegel Policy Brief 2014/01. The same trend is noted in the U.S., where over the last 30 years the FTC and DOJ have “resolved nearly their entire civil enforcement docket by consent decree” – see Wright, J. D. & Ginsburg, D. H. (2018) The economic analysis of antitrust consents. European Journal of Law and Economics 46. Currently, the negotiated settlement of antitrust investigations is common in both OECD and non-OECD countries – Secretariat, Organization for Economic Co-operation and Development, Commitment Decisions in Antitrust Cases (Mar. 30, 2016).

25 Choné Philippe, Souam Saïd & Vialfont Arnold (2014) On the optimal use of commitment decisions under European competition law. International Review of Law and Economics 37.

26 Cosnita-Langlais, Andreea & Tropeano, Jean-Philippe (2020) Litigation and the Informational Value of Precedent: An Application to Antitrust Commitments. Mimeo.

27 A recent real-life example is nonetheless provided by the Brazilian competition agency, which convicted a cement cartel for price-fixing and market-sharing in 2014. In addition to corporate and individual fines, the authority imposed the divestiture of cement and concrete plants in order to reverse the structural market changes that the cartel had implemented to secure the stability of the agreement. See press release of CADE at http://en.cade.gov.br/press-releases/cade-fines-cement-cartel-in-blr-3-1-billion.

28 Joseph E. Harington, Jr. (2018) A proposal for a structural remedy for illegal collusion. Antitrust Law Journal 82(1).

29 See William E. Kovacic et al. (2007) Lessons for Competition Policy from the Vitamins Cartel, in The Political Economy of Antitrust 149, Vivek Ghosal & Johan Stennek eds.

30 Joseph E. Harrington, Jr. (2017) The deterrence of collusion by a structural remedy. Economics Letters 160.

31 See footnote 14 above.