The efficiency of a leniency programme hinges essentially upon whether it provides maximal incentives for being the first to report the cartel and on whether the conditions for qualifying up front for conditional immunity are transparent, predictable and of a nature that would enable a rapid assessment and decision by the authority. The incentives depend not only on the level of sanctions – since fines lower than the profitability of the cartel leave firms indifferent to report – but also on the authority’s credibility in detecting and prosecuting cartels through ex officio investigations. As to transparency and predictability of conditions and expediency of conditional immunity decision-making by the authority, these sustain the full cooperation of the applicant thus ensuring effectiveness of further investigative phases.
The EU adopted its Notice on Leniency from fines2 for reporting secret hardcore cartels in 1996 and subsequently revised the Notice in 20023 and 20064. Unlike the 1996 Notice, subsequent leniency programmes grant conditional immunity upfront to the first qualifying undertaking that reports an undetected cartel and sets out specific bands of reductions of fines for the subsequent reporting undertakings. The recent developments in the design of the Leniency Notice reflect a number of considerations aimed at enhancing clarity and transparency, to increase the incentives to report and to create a race to report5. The revisions of the 1996 Notice were also adapted to reflect more closely the different circumstances which trigger reporting of cartels. A significant proportion of cartels are reported when the business most involved in the cartel is acquired. Unsurprisingly, therefore, the waves in leniency applications have coincided quite closely with merger waves. Where the acquirer is not an existing or recent member of the same cartel, the information threshold that the applicant needs to supply to obtain conditional immunity may necessarily be less than that sufficient to establish an infringement of anti-cartel law. The EU leniency notice then foresees a grant of conditional immunity when an applicant disclosing its participation in the cartel also submits sufficient (incriminating and other) information for the Commission to take the view that it can carry out unannounced but targeted inspections or when it submits evidence enabling the Commission to establish an infringement. A much wider population of cartels have thus become potentially detectable including those cartels where the merger transactions may actually have occurred in a different jurisdiction.
Once an unannounced inspection has taken place, the race to be the second in line and apply for a reduction of fines hoping to make it to the highest reduction band available is the inevitable consequence of a one-way option that confronts the remaining cartel members. Similarly, given the design of the leniency programme an identical logic of ‘a race to the authority’ applies for cartel investigations initiated ex officio: once an inspection has taken place the least cost option for each cartel member is to supply significant added value in terms of evidence in order to at least secure a fine reduction. While the leniency notice is a significant and effective instrument in the detection of cartels, own-initiated cartel investigations account in recent years for an average of 20 per cent of the Commission’s decisions.
The constitutive elements of a leniency programme necessarily have to correspond to the investigative and sanctioning powers of the authority. They should also reflect the investigative procedure, the subsequent steps leading to a prohibition decision with sanctions and eventually to an appeal. Thus in the EU, Commission fining decisions are intended to deter the undertaking involved: this means that they are addressed not just to the business subsidiary involved in the cartel, but also to the parent company that exercises a decisive influence on it, so that both are held jointly and severally liable for the infringement. By the same token the beneficiary of any leniency granted is also the applying undertaking. Furthermore, since most fining decisions are appealed, the leniency programme has to be crafted in a way that generates from applicants solid evidence, (often incriminating documents contemporaneous to the infringement) or enables the Commission to seize such evidence in the course of unannounced targeted inspections subsequent to an immunity application. Many commentators argue that a leniency programme such as that of the EU mainly helps uncover cartels that are terminated rather than ongoing. Factually this is not correct, since the Commission has often been able to establish infringements lasting until the very inspection date. In any case, from the point of view of enforcement and of instilling serious deterrence, it is of some importance to prosecute both recently terminated and ongoing cartels; by so doing increasingly greater proportions of cartels uncovered by the leniency programmes will be, as the empirical evidence suggests, ongoing.
Clearly the design of a leniency programme would benefit substantially from the existence of a viable theory of cartel formation. While there is a lot of academic literature on the subject, it is far from definitive. In the absence of conclusive views, a critical element for the design is what we can learn from when cartels have actually been reported. One can distinguish a number of events which seem to determine cartel reporting: first, as mentioned above, the due diligence undertaken at the time of a merger or an acquisition; then, situations such as when inspections of a company’s competitors have taken place concerning a related product, or when the company has been fined and has introduced a serious compliance programme or when an ongoing cartel is facing internal destabilisation. Sources of instability vary greatly, but the empirically notable ones are when the cartel is facing competitive entry on a large scale especially in a cyclical sector or on the occasion of hostile mergers among cartel members.
In the context of monetary sanctions for cartel infringements, the level of fines is crucial in defining the absolute and relative significance, in relation to sales affected by the cartel, of the incentive to report. Unless the fines are set at a level that deters, the positive incentives to report may be outweighed by the benefits of the cartel, even after allowance for private damages claims. The liability for fines under the rules on competition in the EU falls on the companies within the undertaking with which the infringing party forms a single economic unit. The 1998 fining guidelines not only increased fines for cartel infringements but make explicit that account would be taken of the overall size representing the capacity to damage competition, of the undertaking concerned. These elements not only translate as serious incentives to report (particularly for recidivist undertakings), but they also provide a powerful impetus for the parental holding to introduce and maintain robust compliance programmes.
Indeed the uptake of vigorous anti-cartel programmes within companies is a useful indicator of the cartel detection rate with which the authority is credited and the deterrent impact of fines. In fact such programmes could borrow substantially from the carrot and stick design of leniency programmes so that even if they fail to prevent participation in a cartel, there are at least benefits for the undertaking from being the first to report.
One may ask what the appropriate measure is to assess the success of leniency programmes. The conventional view is to compare statistics on the number of applications, the number of follow-up applications reporting other infringements and so on. However, statistics are not comparable to the extent that they are programme specific and particular to the legal and institutional context to which they belong. A comparison between leniency programmes can be made on the basis of more revealing parameters. In this regard, account must also be taken of the influence of the programme on other enforcement agencies and the degree of convergence it stimulates on its main features. An example of such convergence is the ECN Model Leniency Programme, adopted on 29 September 2006 by the heads of the EU competition authorities. Certainly, the most relevant parameter should be how often the programme meets its actual objective: its score in successful prosecutions.
1 The content of this article does not necessarily reflect the official position of the European Communities. Responsibility for the information and views expressed lies entirely with the author.
2 OJ C 207, 18.7.1996, p. 4.
3 OJ C 45, 19.2.2002, p. 3-5.
4 OJ C 298, 8.12.2006, p. 17-22.
5 For a complete overview, see ‘Commission adopts revised Leniency Notice to reward companies that report hard-core cartels’, by Sari Suurnäkki and María Luisa Tierno Centella in the Competition Policy Newsletter - No 1 - Spring 2007 pages 7-15.
