Convergence and Divergence in International Antitrust

Christine Varney’s speeches as the Assistant Attorney General, such as that in Fordham last year, and yesterday’s speech in Paris (Co-ordinated Remedies: Convergence, Co-operation and the Role of Transparency) have struck a very constructive and conciliatory tone on the issue of convergence.  Convergence between antitrust authorities, and having a coherent global system is clearly important to her, and she rightly points out that consistency will benefit both agencies and businesses.  She also appreciates the limits to convergence, noting that any agency action has to be taken in the context of their own legal systems, and with a view to protect the consumers within that agency’s geographic area.

In merger control, convergence works extremely well, particularly when parties time their notifications to the different agencies so that the agencies can move forward on a case together, discussing their thinking, and co-ordinating their information gathering.  There are still some disagreements, but complete convergence is almost certainly an unattainable goal – given a complex set of facts, reasonable people can disagree on the appropriate analysis.

In recent years, more concern has been expressed about (the lack of) convergence in respect of unilateral conduct.  Given the DOJ’s section 2 report published last year, disavowed by several FTC Commissioners, then disavowed by Christine Varney, and withdrawn by the DOJ this year, transatlantic convergence is not the only problem to be addressed.  There are profound differences of view within the US legal system and legal community as to the appropriate section 2 liability standards.

And these are divisions within one single legal system.

When you look at reducing divisions across different legal systems, the problem becomes far more complex.  Although EU and US antitrust law shares the same objectives, there are differences in the practical enforcement of EU and US antitrust law that substantially limit the value of direct comparisons between the laws.  No helpful discussion of similarities and differences between the two laws can take place without a recognition of the impact of these practical differences.  Bill Kovacic, then FTC Chairman and now FTC Commissioner has famously addressed this.

In roughly the past 30 years, judicial fears that the US style of private rights of action – with mandatory treble damages, asymmetric shifting of costs, broad rights of discovery, class actions, and jury trials – excessively deter legitimate conduct have spurred a dramatic retrenchment of antitrust liability standards. This is most evident in the progression toward more lenient treatment of dominant firm conduct…

… Had the US private rights of action been more constrained (for example, by making treble damages discretionary rather than mandatory), my prediction is that US doctrine for abuse of dominance would more closely resemble existing EU standards. The persistent inclination of US courts to raise liability standards to offset perceived excesses of private rights creates what could turn out to be a permanent fissure between the EU and the US approaches to dominant firm conduct and other forms of business behavior.

William E. Kovacic. Chairman. U.S. Federal Trade Commission. Bates White Fifth Annual Antitrust Conference. Washington, D.C.. June 2, 2008, citing William E. Kovacic, The Intellectual DNA of Modern U.S. Competition Law for Dominant Firm Conduct: The Chicago/Harvard Double Helix, 1 COLUM. BUS. L. REV. 1-80 (2007)

Evidence for US concern at the costs of antitrust enforcement are easy to find: just to mention some recent Supreme Court cases,

  • Justice Scalia in Trinko referred to the ” sometimes considerable disadvantages” of antitrust;
  • Justice Souter in Twombly indicated that in setting the appropriate standards for pleadings sufficient to carry a case through to the discovery stage, it was important to recognise that discovery can be expensive;
  • Justice Breyer in Credit Suisse, citing Hovenkamp, that entrusting antitrust enforcement of securities issues to “nonexpert judges” and “nonexpert juries” will lead to multiple inconsistent results.

Most recently, the oral arguments in the Supreme Court’s American Needle case also show the same concerns.  The legal issue at this stage of the case is relatively straightforward: is the NFL a single entity for the purpose of the antitrust laws (as if it is, section 1 of the Sherman Act would not apply)?  The vast majority of the discussion, however, was not on this question; rather it focussed on the potential costs to the NFL if it were to be dragged into court repeatedly.  The Justices seemed to want to find the solution that would minimise costs to the NFL, rather than one which would answer the direct question in front of them.

Under the EU system, whether a sports league is a single entity would depend on issues of ownership and control, not on the issue of whether one solution or another would lead to disproportionate costs.

I do not mean to suggest that the US approach is wrong and the EU approach is right: any legal system and any set of laws has to take account the potential costs to society of setting the laws in a particular way.  The point is rather that the costs in the US and the costs in the EU are markedly different.

Even if we look only at the issue of treble damages, the difference is stark.  Private litigation scarcely exists in Europe, and even if it occurs, and succeeds, only single damages are available.  If the options discussed in the European Commission’s White Paper on Damages are brought forward, there will still only be single damages for most antitrust violations, with a possibility of double damages for hardcore cartel violations only.

In the US, the Clayton Act provides for treble damages for any antitrust violation – whether overt or covert, clearly illegal or borderline.  It’s inevitable in such a system that concerns at over-deterrence outweigh those of under-deterrence.  A rational business would be cautious about approaching the border between legal and illegal activity; in areas where the border depends on detailed economic analysis, the business might reasonably decide that the risk of treble damages was too great, and pull back from perfectly legal behaviour.  Against this background, setting high liability standards is understandable, whether it is that the business practice should make “no economic sense”, or whether the harm should be substantially greater than any potential benefit, or any of the other means by which liability can be reduced.

Under the EU system, businesses face a different risk assessment.  If a business pushes at the border of legal activity, it may face a fine from the public authority, but private action leading to damages are rare, and even if successful lead to single damages only.  And public authorities rarely take enforcement action against activities that are borderline.

These differences explain many of the similarities and differences between the EU and US antitrust rules.  Convergence on merger control is very strong: there are rarely disagreements on the substantive tests to be applied – even though the formal legal tests are worded differently; there are – rare – disagreements on individual cases, but as mentioned above, that is likely inevitable. This convergence reflects the fact that the overall enforcement system is not markedly different.  Similarly, there is no great disagreement on cartel enforcement; there are no risks of over-deterrence, and cartels do not require detailed assessment to judge their legality.

There is divergence – and I think with decisions such as Trinko, increasing divergence – in unilateral conduct.  I do not believe that the differences in the systems discussed above are the sole reasons for this: it’s fairly clear that there are differences of view as to how best to maximise consumer welfare, that go beyond the problem of enforcement costs; other issues such as the US’s politicisation of the judiciary and of public enforcement also need to be taken into account.  Nevertheless, the enforcement cost issue is real, and any discussion of convergence has to take into account these genuinely divergent enforcement systems.