The primary objective of this study, available here, is to analyse how national European courts have assessed cartel overcharges. In addition, it also provides figures on the development of cartel damages actions in Europe (how many cases were decided, in which countries, with which outcomes, etc.). It was completed with the help of lawyers, law professors, economists, national competition authorities and national judges from 30 European countries.
Now in its fourth edition, this study shows that national courts in Europe have handed down judgments in at least 239 cartel damages actions in 13 countries, relating to more than 63 cartels. In these judgments, courts have given many insights into how to assess cartel overcharges.
Section I describes the methodology followed.
The process employed for this research involved four steps. The cases were identified; copies of judgments were gathered; using a recent automatic translation service, judgments were translated into English; their content was then analysed.
To identify cases, contributors were asked whether they were aware of relevant cases in their jurisdictions, and to assist with their analysis. In total, 140 lawyers, law professors and economists, 26 NCAs and 44 judges contributed to this study. You can see the list of people involved in the article – it is quite impressive.
Section II provides general figures on the cases gathered.
A “case” means an action for damages with one or several plaintiffs alleging that a cartel caused an overcharge in which a court handed down at least one judgment on the merits. By 2019, 239 cartel damages actions had been identified. This reflects a significant increase from 2018 – the author identified 104 judgments handed down between July 2018 and August 2019. The 239 cases up to August 2019 include 59 cases in which damages were awarded, 86 cases in which liability was established, and 93 cases that resulted in dismissals. A noticeable trend all across Europe is that the number of cases is growing rapidly. There were 52 cases judged for the first time in 2018, compared to 24 in 2017.
Cases came from a large number of different countries: Germany (122 cases), France (46 cases), Spain (33 cases), Hungary (seven cases), the Netherlands (six cases), Italy (five cases), Belgium and Finland (four cases each), Austria and Denmark (three cases each), Greece, Poland and the United Kingdom (two cases each). The small number of cases from the UK does not reveal the true size of London as a competition litigation forum. Many cartel damages actions have been brought before the Competition Appeal Tribunal or the High Court of Justice of England and Wales but most actions were settled before any judgment on the merits.
It is remarkable how much more common follow-on claims are than stand-alone cases. Of the 239 cases identified, 57% followed an infringement decision made by a national competition authority, 40% followed a European Commission decision, and only 2% were stand-alone actions. There is also one case in which the infringement decision was adopted by a regional competition authority. Another interesting fact is that many claimants come from a broadly defined public sector. Publicly owned companies (53 cases), local authorities (49 cases), and central governments (8 cases) initiated 45% of all claims. This may be related to the fact, that in a relatively high proportion of cases (47%), the allegedly affected purchases resulted from tendering processes. [On a personal note, I believe a similar pattern can be observed in Japan].
Section III analyses the award of damages and the reasons why cases were dismissed.
Cartel damages have been awarded in 59 cases. They come from France (23 cases), Spain (21 cases), Germany (eight cases – German courts very often establish liability but do not quantify damages), Denmark (three cases), the Netherlands (two cases) Austria, Finland, Italy and the UK (one case each). The range of overcharges in these cases as a percentage of affected prices varies between 1% and 49%. At the same time, 93 out of the 239 cartel damages actions have been dismissed, and the main reason for this (c. 47% of cases) was a failure to establish loss.
Section IV focuses on how courts have assessed cartel overcharges.
Courts have now been exposed to all major types of methods for calculating damages described in the Practical Guide for estimating damages issued by the European Commission in 2013. In the 59 judgments awarding damages, damages were quantified with the following methods:
(i) Comparison over time (also called “before-and-after”): 31 cases.
(ii) Comparison with an unaffected market (also called “yardstick”): four cases.
(iii) Cost-based and financial methods: nine cases.
(iv) Other methods: 19 cases (this involves relying on presumed rates of overcharge for anticompetitive conduct found in contractual clauses, or estimating rates of overcharge by reference to statistics found in the Commission’s practical guidelines).
For the most common method (i.e. before-and-after), in over 90% of cases the reference price was that which could be observed after the infringement had ended. There appear to be two reasons for this choice. First, it is in general relatively clear when an infringement ceased, but it is sometimes less certain when it started. Secondly, when the infringement lasted a long time, finding data on prices that prevailed before the infringement began can be a challenge.
It is somewhat remarkable that there have been no cases using regression analysis (i.e. “econometrics”) or simulation models. However, the results of regression models were used in a case in which no overcharge was found. In another recent judgment, the Paris Administrative Court of Appeal took into consideration an econometric study in order to reach the conclusion that there was harm, and appointed an expert to determine its amount with precision.
Section IV also deals with passing on and interest.
The data makes it clear that it is quite difficult for passing on defences to succeed. Passing on was raised as a defence in 126 cases, i.e. 53% of all cartel damages actions covered by this study. However, in only four cases did courts find that all or part of an overcharge was passed on by claimants to their customers. This may have to do with rules, particularly in Germany and France, requiring a defendant to show that the claimant suffered no loss at all from the infringement, which sets quite a high bar.
Regarding interest, infringement decisions were adopted on average 8.4 years after the date of purchases; also on average, the first civil judgment is handed down 4.2 years later. The average total time from the moment the harm potentially occurred to the first judgment granting damages is therefore 12.6 years. Situations range, however, from less than 3 years from the date of purchase to the award of damages, to more than 20 years. This long duration explains why prejudgment interest is often an important topic in competition claims.
Finally, section V presents highlights from some recent cases.
This includes a German decision where the court found that umbrella effects were not established; the German Supreme Court decision that removed (or seriously qualified, depending on your perspective) the presumption of harm flowing from a cartel which I discussed a few weeks ago; a number of cases where the presumption of overcharge did not operate because claimants failed to prove that they had purchased the cartelised product; and many, many decisions regarding the trucks’ cartel, with an interesting focus on the evidence required to demonstrate that a claimant was a victim.
I have been in awe of Jean-Francois’ work for some time – this is the fourth time he publishes this paper, which is constantly updated. This is the best source I have come across on private enforcement in Europe (including official documents). The only thing I can do is recommend this paper to you all and hope that there is a new edition next year – even as the number of cases increases exponentially – while pointing out discrete issues that struck me as I read the paper.
First, there seems to be an inversely proportional relationship between being a leading forum for cartel claims and actually awarding damages. You will remember a paper on the Netherlands that I reviewed a few weeks ago found 11 follow-on claims in Netherlands between 2012 and 2015 alone, out of six different sanctioned cartels. Yet it seems that there only six final decisions on cartel claims have even been adopted there. A similar relationship seems to hold in the UK, where final decisions are quite few. It would seem to follow that the most sophisticated jurisdictions devote a significant amount of time and work to procedural issues, upon which the parties can see the writing on the wall and settle. This, however, is not the case in Germany – except that there too, once liability is established, most cases seem to settle. Interestingly, this means that courts in these countries manage to avoid calculating damages, which is notoriously difficult – while courts in Spain are currently issuing a flood on judgments on the estimated harm suffered by victims as a result of the trucks’ cartel (mostly with decisions covering very few trucks).
Second, from the numbers it seems that there is stand-alone cartel private enforcement is virtually non-existing in Europe. This contradicts the information contained in a 2015 book on private enforcement by Lianos, Nebia and Davies which found that stand alone claims are actually much more common than follow on claims. This can perhaps be explained by the fact that this piece focuses on cartels, whereas it is likely that the 2015 numbers concern abuse cases and competition claims in the context of contractual disputes.
Third, given the absence of stand-alone claims, it is slightly odd that 21 cases have been dismissed for absence of wrongdoing. I think that this means that the courts dismissed a case because the defendant was not an addressee of the infringement decision, i.e. they were not part of the relevant undertaking, but if so this could have been made clearer. If this is the case, I wonder whether such cases will be decided differently following the Skanska judgment. Of course, other reasons will also apply here – Jean-Francois has told me that some of these cases concern lawyers suing non-cartel participants regarding umbrella pricing.
Lastly, I am curious about the data on passing on. The author mentions four instances where it was accepted, whereas until last year he had only identified a single case. However, other papers have identified more cases where overcharge was accepted. This also does not really cohere with the finding that mitigation and passing on were the cause for dismissal in 22 cases. This is probably explained by: (i) many cases where passing on was invoked did not relate to cartels but to other infringements, such as those involving bank cards; (ii) prior to the implementation of the Damages Directive, in a number of jurisdictions the claimant had to prove that he or she could not pass on the overcharge (France, Germany, Poland), and failed to provide such proof.