The possible undermining effect of damages actions on leniency programs has been hotly debated. The concern is that the prospect of damages claims may discourage colluding firms from applying for leniency, since the leniency program only shields them from public fines, not from civil damages.
Civil damages may contribute to the goal of preventing cartels by increasing the expected costs of starting a cartel. However, civil damages may not enhance antitrust deterrence if colluding firms believe it to be unlikely that competition authorities will detect their cartel. For leniency programs to put cartel members in a prisoners’ dilemma, confessing must be more attractive than staying quiet. If civil damages are substantial, leniency may not sufficiently improve a colluding firms’ position as compared to their non-reporting co-conspirators, and hence their incentive to apply for leniency will decrease, together with the overall odds of cartel detection.
This note, available here, discusses the ambivalent effect of antitrust damages actions on deterrence. It considers how fines and damages compare for leniency applicants and non-cooperating firms in Europe and the US. It also explores avenues to ensure that the incentives to apply for leniency are retained in an environment of increased private enforcement.
Section II looks at how public enforcement seeks to deter anticompetitive practices.
Classical deterrence theory suggests that a violation can be deterred if the expected sanction, consisting of the value of the sanction and the probability that it will be imposed, exceeds the expected gain from the violation. In order to induce firms to refrain from cartel behaviour, the expected illegal profit from colluding must be lower than the expected sanction, given by the anticipated fine multiplied by the probability of the unlawful conduct being discovered and sanctioned.
Leniency seems to increase the probability of cartels being uncovered. In effect, the primary way in which cartels are unearthed is through leniency applications. In the U.S., over 90 percent of penalties imposed by the Department of Justice for cartel violations since 1996 are linked to investigations supported by leniency applicants. A 2013 EU study found that out of 57 cartels for which a full decision was published since 2000, 53 were discovered through a leniency application. Moreover, with the introduction of leniency programs in the EU and the U.S., the number of cartels detected increased considerably.
While leniency programs are often credited with the successful detection and prosecution of cartels, it cannot be excluded that the increase in the number of cartels detected may also be due to an increase in cartel activity. This would reflect the fact that leniency programs could have an ambivalent effect on deterrence. On the one hand, leniency policies put cartel members in a prisoner’s dilemma, making collusion harder to sustain; on the other, the reduced fines in leniency programs may raise the expected benefit from continuing to collude.
Section III looks at how private enforcement may impact deterrence.
In principle, civil damages increase the expected costs of collusion, thereby contributing to deterrence. However, private claims may also dilute incentives to apply for leniency, since leniency applicants are exempted from paying a fine but still have to pay damages. Given the central role that leniency programs play in discovering cartels, civil damages may reduce the deterrence effects of competition policy.
The U.S. found a remedy against civil damages dissuading potential leniency applicants by limiting the extent to which leniency applicants are liable for compensation – instead of being jointly and severally liable for treble damages, leniency applicants that cooperate with claimants will see their liability limited to actual damages incurred, and will be exempt from joint and several liability. The EU, on the other hand, sticks to a strict compensation principle whereby leniency applicants will be liable for the damages they caused in the same way as other cartelists. EU law protects leniency merely by limiting access for claimants to leniency documents and other sensitive information – thereby ensuring that leniency applicants are not in a worse position in civil claims than non-cooperating conspirators are –, and by exempting immunity recipients from joint and several liability. Thus, the advantage for immunity recipients in private enforcement is considerably smaller in the EU than in the U.S.
Section IV looks at options to ensure the optimal articulation of private enforcement and leniency.
A solution to maintain leniency incentives while promoting private enforcement is to create a gap between the civil liability of immunity recipients and that of other firms. One such approach can be found in the US, which regime favours leniency applicants by reducing their expected civil liability when compared to co-cartelists. Another alternative, which used to be in place in Hungary, would be to limit immunity recipients’ civil liability to a “last resort” mechanism that would exempt the immunity recipient from civil liability unless injured parties cannot obtain compensation from the other cartel participants. Finally, individual sanctions such as prison terms could be introduced or increased. This would add to the deterrent effect of competition policy in a way unrelated to the overall monetary liability that firms found guilty of a competition infringement would face.
This is a concise analysis of the relationship between private enforcement and leniency – which is often depicted as a tension between mechanisms that increase deterrence by either enhancing the size of the penalty or increasing the odds of detection.
It provides a good intro to the topic, and I am sympathetic to the argument that one can optimise the benefits of public and private enforcement by limiting the liability of leniency applicants on both spheres by comparison to other cartelists. Having said this, the argument would benefit from actual evidence demonstrating that increased private enforcement is having detrimental effects on leniency and that such measures are appropriate. This paper is not the place to do this, of course, but I find that the existing literature on the topic is more limited than it could be, and that this is an area ripe for both empirical and more analytical research – such as the one in the paper below.