Antitrust writing awards
Free access to over 100 leading articles
Vote for the best antitrust Writing!
Previous business/academic article Next business/academic article

Cartel leniency programs: Some caveats

Mark Leddy, Concurrences, N° 3-2011, n°37333

See Mark Leddy's resume

Vote for this articleHelp

* Average
** Interesting
*** Good
**** Excellent
***** Must receive an Award!

Readers vote is closed since March 1st, 2012. Readers’ vote has nominated 2 articles for each of the Awards. This short list has been communicated to the Board, with the articles nominated by the Steering Committees. The Board will decide on the award-winning articles on March 27, at the Awards ceremony to take place in DC. See vote results online here.

It seems indisputable that the U.S. DOJ amnesty program has been a remarkable law enforcement success. The number of criminal indictments has significantly increased over the years, monetary penalties have been enormous, and many U.S. and non-U.S. citizens have gone to prison. The U.S. amnesty program has drawn the attention of competition agencies around the globe and more than fifty countries now have some form of leniency program. And, perhaps most importantly, the enforcement actions generated by the leniency programs in the U.S. and elsewhere presumably have had a substantial deterrent effect on cartel behavior, a clear societal benefit.

But there are several important caveats to this success story. These are summarized below in the hope of generating further discussion about how enforcement agencies might improve cartel enforcement generally and leniency programs in particular. Such a discussion would be especially timely as the globalization of cartel enforcement – now frequently in connection with a leniency program – continues.

The first caveat is the “rush to leniency.” Leniency programs create a powerful incentive to report price fixing. But too often the pull of this incentive trumps careful factual analysis and sound judgment, leading to a significant waste of government and private resources. In the U.S., given the severity of criminal antitrust penalties and the benefits of the U.S. amnesty program – immunity from fines, amnesty for employees, and substantially reduced civil liability – some seek amnesty (or what amounts to an advisory opinion) even when the relevant conduct is not hard core price fixing. To its credit, DOJ has recognized this problem and taken steps to address it, though the problem has not been eliminated in the U.S. and remains a serious problem outside of the U.S.

Sometimes lawyers simply lack sufficient experience to determine whether particular conduct is in fact a collective competitive constraint that warrants the label “cartel.” Whether out of inexperience or hyper-risk aversion, they advise and even push a client to seek leniency when the conduct, fairly examined, is not “hard-core” collusion. And once a client seeks leniency in one jurisdiction, a domino effect is often created in which counsel in each other affected country understandably asks, “Well, if you’re going to seek leniency in the EC, maybe we should seek leniency everywhere you do business: in the U.S., Korea, Australia, South Africa, Brazil, Bulgaria, etc., so that someone else doesn’t get to those agencies first.” The result is a mass amnesty application based on the original misguided advice.

Our firm has been involved in a number of cases in which investigations have lasted for years and the regulators have closed the matters with no allegations against anyone. This is a huge waste of both private and public resources. By way of example, in one case we advised a client that the business practice at issue (a lengthy non-compete agreement in a legitimate multi-country business transaction) was probably not even illegal, much less a U.S. criminal violation. But separate European counsel advised that the European Commission might conclude that the practice violated Article 101 and that amnesty should be sought before the other party to the contract sought amnesty. The confused and risk-averse client ultimately decided to seek leniency in the EC and then in more than ten additional jurisdictions, including the U.S. Fifteen months later, after the fortunate intervention by an enlightened enforcement agency staff, there were no charges brought anywhere and the matter was dropped. Nonetheless, this initial miscalculation resulted in very significant legal expenses and great anxiety and distraction for the client.

These issues are not easy problems to solve, and avoiding the unwarranted rush to leniency is a challenge. The stakes are high. These circumstances require lawyers with sufficient experience and confidence to advise clients that it is not necessarily in their best interests to seek amnesty in what, to an experienced counsel’s eye, is at worst an ambiguous situation that enforcement agencies ultimately will not view as hard core collusion. But this is what lawyers are supposed to do – collect the relevant facts, evaluate them, and advise clients about their legal consequences. Again, making these judgments can be tricky even in jurisdictions with mature enforcement agencies and is all the more difficult in countries without well-established antitrust regimes. But that doesn’t justify some of the “rushes to leniency” that we’ve experienced.

One further note on this issue: the “marker” system of some agencies is a useful tool in reducing the harm from an unwarranted “rush to leniency” as it allows for a relatively thorough internal investigation and the withdrawal of the “marker” if the evidence uncovered does not amount to egregious anticompetitive conduct. When invoked, it should be used not only to confirm the possible existence of a violation but also to examine thoroughly whether there is a cartel violation in the first place.

The second caveat to the success of leniency programs is what we call “the stretch,” i.e., the tendency of counsel and the client to shape recalled “facts” in a way that advances their goal of leniency. This temptation is a powerful one because the reward is so significant and because there is always the risk that a competitor might likewise “stretch” employees’ recollections so that it can win the amnesty or leniency “prize.”

This phenomenon is enhanced by what Wouter Wils of the EC has called “confirmation bias.” [1] Our interpretation of confirmation bias is the uncritical acceptance by a prosecutor or regulator of the worst possible interpretation of any potentially inculpatory statement from any source, whatever the context or the incentives of that source might be. The most vague and speculative hearsay statement is at times accepted without question because it “confirms” the regulator’s innate suspicions and feeds the hope for yet another headline-making prosecution. Together, the powerful incentive to stretch and the willingness to accept these statements as “facts” risk serious distortion of what actually happened.

Inculpatory statements must be viewed in context. Amnesty and leniency are after all a bargain – the more “added value” a company or individual is perceived as giving, the greater the benefit in terms of a significant monetary discount or immunity for an individual. Given these benefits, there is a tendency by some companies and their counsel to stretch or alter recollections to justify corporate leniency discounts or to secure immunity for an individual in circumstances where that immunity or leniency might otherwise not be available. Every counsel wants to obtain leniency for his or her client and amid the give and take of witness interviews can be sorely tempted to mold the client’s recollections to give the government what it wants to hear. These are not (to our knowledge) situations in which lawyers or their clients deliberately lie to the agencies. But it is clear to anyone involved in these cases that there is a tendency for some companies or individuals, when reconstructing conversations that are one, five, or even more years old, and in a strenuous effort to obtain leniency, to characterize behavior that is ambiguous or at least ill-remembered as an “agreement” or exchange of information “that must have had an effect on price.”

It is not easy for prosecutors to detect situations in which companies or individual clients are stretching the truth in order to obtain immunity or a leniency discount. But unlike their counterparts in many if not most other jurisdictions, at least experienced U.S. prosecutors who must prove their case “beyond a reasonable doubt” to a judge or jury understand this phenomenon and the need for vigilance about the necessity of these after-the-fact characterizations. An administrative agency like the EC, however, makes findings without having to carry a burden of proof before an independent fact finder and without witnesses being subject to cross-examination by a magistrate or by the accused. [2] Some have criticized these administrative systems, which can impose harsh penalties, for not providing sufficient judicial review, which only encourages lawyers and clients to “stretch” for leniency.

But in the U.S., the “stretch” and the “rush to leniency” problems are exacerbated by the “amnesty plus” program, where companies that have not received amnesty for a particular product under investigation can provide the government with evidence of a violation in another product and secure both amnesty in the second product and a significant fine discount on the first product. These incentives to race for amnesty and to “stretch” the facts in this situation are especially strong.

The increase in prosecutions generated by leniency programs has of course led to huge fines in many cases. A third caveat in the leniency context is the wooden and inequitable application of sentencing guidelines. This is a controversial topic but overall there is simply too much reliance on sentencing formulas both in the U.S. and elsewhere.

Lawyers like to say that every case is different. It is a cliché but it’s true. Not every price-fixing case is as bad a violation as the worst of the cases. Not every case is as egregious as the vitamins case, with regular meetings, monitoring mechanisms, and a clear intent to raise prices collectively. Some cases involve primarily an information exchange with no real “meeting of the minds” or only a loose, unmonitored “understanding” that prices should rise by some unspecified amount. Differentiating the truly serious and effective cartels from “loose talk” is again a challenge but is sound law enforcement and simple equity, the bedrock of law enforcement. And yet less extreme cases are often treated or punished just like the very serious cases.

The U.S. Sentencing Guidelines illustrate this problem with their starting fine calculation based on 20% of “affected commerce”, i.e., 20% of the defendant’s total sales of the product at issue for the period in question. The EU similarly bases its fines on turnover. We would suggest that these formulas be treated like HHIs in the merger context – starting points for discussion that create a rebuttable presumption.

Being a prosecutor is a difficult job; judgments about the relative severity of conduct are not simple. But by treating every case similarly, there is genuine inequity and a real potential for undermining respect for the law enforcement process. This is a problem that can and should be fixed. Instead, the problem is getting worse as jurisdictions around the world adopt more or less strict revenue based fine structures.

The importance of these issues has increased with the globalization of leniency programs. As new regimes develop, many countries’ agencies do not have sufficient experience to deal with the sometimes complex factual and legal analysis involved. The remedy is of course more experience, particularly the invaluable learning that comes from losing contested cases. (Recall the significant merger reforms instituted by the EC after it lost several extremely weak cases at the CFI.) Without the benefit of this experience, few of these new prosecutors will be able to step back and shed their “confirmation bias” and objectively evaluate a complex situation. But such lessons in cartel cases are rare in many jurisdictions because administrative processes lack effective judicial review.

With new regimes ill-equipped to address these problems themselves, the relatively more mature agencies like the DOJ and the EC should reflect on these questions and on possible solutions, whether through the ICN or otherwise. This could help to ensure that the benefits of these leniency programs are secure around the world while reducing the incidence and severity of the problems described in this article.

Mark Leddy is grateful to his colleague, Elaine Ewing, for her invaluable assistance in preparing this paper and to other Cleary Gottlieb colleagues for their valued insights on the issues addressed in this paper. He alone, however, is responsible for its content.


[1] See Wouter P.J. Wils, The Combination of the Investigative and Prosecutorial Function and the Adjudicative Function in EC Antitrust Enforcement: A Legal and Economic Analysis, 27 World Competition 201, 216 (2004).

[2] See generally Ian S. Forrester, Due Process in EC Competition Cases: A Distinguished Institution with Flawed Procedures, 34 E.L. REV. 817 (2009).

© 2012 - Institute of Competition Law Download our brochure