Court in substantial doubt when Posten lost first-of-a-kind Norwegian follow-on claim case against truck manufacturers
The trend seen in Europe over the last decade of launching follow-on claims in anti-trust cases has now also reached Norway. Such claims are seen as a well-trodden path for victims of cartel infringements to be compensated for their economic loss. Companies found guilty of anti-trust infringements on the other side face an increased financial exposure, as they risk not only huge fines, but also having to compensate their customers’ loss.
Follow-on claims are often funded by third parties, something which is also seen as a possibility in Norway, although Posten did not seek external funding.
The truck cartel case has led to an array of follow-on claim cases throughout Europe
On 19 July 2016 the European Commission fined truck manufacturers Volvo/Renault, Daimler, Iveco, and DAF a combined fine of 2.9 billion euro for infringing TFEU Article 101 and the EEA Agreement Article 53 by coordinating on the setting of “gross list” prices for medium and heavy trucks in the EEA, the timing for introduction of emission technologies and passing on the cost for such technologies to the customers. MAN was granted full leniency for revealing the cartel to the Commission. The other manufacturers got their fines reduced according to the settlement procedure. Scania chose not to settle and was later fined EUR 880 million. On 2 February 2022 the General Court dismissed the appeal by Scania and upheld the fine imposed by the Commission. Scania has appealed this judgment to the European Court of Justice.
The “gross list” prices, which the manufacturers had coordinated, relates to the factory price of trucks and works as the basis for price negotiations with the buyers. The infringement was ongoing from 1997 to 2011.
Following this, Posten, which is the incumbent Norwegian postal operator and a provider of logistics services under the brand Bring, filed a claim against the truck manufacturers, arguing that it as well as its subsidiaries in Sweden and Slovakia had been overcharged when purchasing trucks from the defendants during the infringement period. The claim against Iveco, Stellantis and CNH Industrial was subsequently settled out of court.
The Norwegian dairy producer Tine has also launched a separate claims case against the truck manufacturers on similar grounds. This case has been put on hold awaiting the outcome of other cases in the truck cartel complex.
There are several other ongoing follow-on claims in the truck cartel case in other countries, including Germany, France, the Netherlands, Spain, and the UK.
In the UK, the Competition Appeal Tribunal (“CAT”) issued a decision on 7 February 2023 in follow-on claims from Royal Mail and three companies in the BT Group against the truck manufacturer DAF. The CAT found that each company was overcharged a total of 5% for trucks bought from DAF in the infringement period, amounting to GBP 17.5 million in total (Link).
Several procedural issues of principal nature have been clarified ahead of the material claim case
Since follow-on claims is a relatively new phenomenon in Norway, there were several procedural questions of principal nature that needed to be clarified ahead of the material claims case.
The non-Norwegian truck manufacturers disputed Norwegian courts’ jurisdiction. Posten attempted to establish jurisdiction by including the Norwegian company Volvo Norge AS in the lawsuit as an “anchor”, on grounds that foreign parties can be included on basis of the Lugano Convention article 6 no. 1 if the claims are “closely connected”. The truck manufacturers argued that Volvo Norge AS was not part of the Commission’s decision, and the claims therefore not closely connected. This question went to the Supreme Court, which concluded that Volvo Norge AS not being part of the Commission’s decision was not an obstacle for relying on the Lugano Convention article 6 no. 1, and that the claims were indeed closely connected. Norwegian courts thus had jurisdiction over the claims against both Volvo Norge AS and the foreign-based defendants.
A dispute also arose about whether the truck manufacturers were required to disclose the Statement of Objections (“SO”) from the Commission, as requested by Posten. The truck manufacturers argued that the SO contained information subject to confidentiality and claimed that being required to disclose the SO would undermine the leniency institute. Borgarting Court of Appeal concluded that the SO must be disclosed but accepted that business secrets were to be redacted.
Oslo District Court’s judgment – key takeaways
As the truck manufacturers admitted that the participation in the price cartel established grounds for liability on their part, the only substantial question for the court to decide upon was whether Posten had proved that the cartel had caused an economic loss.
Cartel participation not sufficient in itself
As price coordination is considered a restriction of competition by object, it does not require that anti-competitive effect is proved. Therefore, the court had no effects analysis from the Commission to lean on in its deliberations.
The court found that the coordination of gross list prices in itself was insufficient to prove an economic loss for Posten as a customer of the cartelists. The court is thereby rejecting a presumption for economic loss following a price coordination and states this must be assessed case-by-case according to Norwegian tort law.
Likely link between gross list price coordination and actual transaction prices not sufficient
To a large extend, the court based its judgment on the settlement decision when concluding that the gross list prices affected the transaction prices paid by the truck customers. This was not contradicted by the explanations given by witnesses representing the truck manufacturers. The Commission had concluded that the gross list prices affected the price paid by the import companies and the dealer companies, which again affected the retail transaction prices. This finding is strengthened by the fact that truck manufacturers sometimes sell directly to dealers and even to customers.
Based on the above, the court concluded that it is likely that the infringement could have affected transaction prices paid by truck customers. The court does not, however, find this sufficient for imposing liability on the defendants in Posten’s case. Based on the evidence provided by the parties, the court concluded that it was necessary to conduct an empirical analysis of the transaction prices in the “but for” scenario to reveal if Posten had suffered an actual loss.
It is interesting, and to some extent surprising, that the court does not accept that a likely link between the infringement and transaction prices is sufficient to establish a loss on Posten’s part. The legal test, as pointed out by the court, is whether it is more likely than not (i.e. more than a 50% probability) that the infringement led to a loss for Posten. Yet the court seems to conclude that it is necessary to do an empirical analysis proving a quantifiable loss for any loss to be considered sufficiently proved. In other cases than antitrust cases, courts will usually be inclined to establish causality if it is likely that some sort of loss must have occurred, and then use its discretion to determine the quantum of the loss.
The court rejected the empirical analyses of all parties
Both Posten and the truck manufacturers agreed that econometric regression analysis is the best tool to determine the most likely contrafactual price, where the loss is calculated based on the difference between the prices actually paid by Posten and prices in the contrafactual scenario determined by the analysis.
All parties provided econometric evidence, prepared and presented by economic experts over 11 court days. The experts all relied on a “during-after” regression analysis of the price effects. While the analysis of Posten’s experts showed an overcharge at about 25%, the defendants’ expert witnesses presented analyses showing zero or minimal effects.
The court starts its assessment of the econometric analyses by highlighting its struggles in evaluating the presented evidence. The court also notes that the respective experts reluctance to accept criticism from their fellow economists does not particularly induce confidence, especially considering the complexity of the subject matter.
The court goes on to assess the econometric analyses presented by the truck manufacturers’ experts. Based on somewhat different specifications of the models and different data sets, all experts on the manufacturers’ side concluded that the analyses showed that Posten had not been overcharged during the infringement period.
Posten’s economic experts argued that the manufacturers’ analyses were flawed due to the choice of explanatory variables in the regression models. Posten’s experts argued that all model specifications included endogenous variables, i.e. variables affected by the cartel behaviour and the truck manufacturers commercial decisions, making the estimated price effect biased and unreliable and with a risk of providing “self-fulfilling” results. The court sided with Posten’s experts and under doubt dismissed the estimates presented by the truck manufacturers’ experts.
Posten’s experts based their regression analysis on GDP and CPI as main explanatory variables, finding a 25% overcharge. These variables do not pose the same problem of endogeneity, as they are independent of the cartel infringement and the commercial decisions of the truck makers. While the court found that this model specification made sense, it agreed with criticism from the experts appointed by the defendants about the model’s lack of robustness. First, the court raises doubt about the robustness, as Posten’s experts did not present model specifications excluding GDP and CPI but including alternative explanatory variables such as producer price indexes, transport volume and oil prices. Moreover, the court refers to an analysis prepared by Volvo’s experts using the model specification of Posten’s experts on Volvo’s (larger) data set, showing no overcharge. Based on this, the court found the conclusion of Posten’s experts unreliable, also dismissing this analysis under considerable doubt.
The court thereby ended up with no empirical analysis to lean on. Despite the court finding several other indicators of an overcharge, it concluded that an actual overcharge had not been established in the absence of a reliable regression analysis. As the court found that the burden of proof lies on Posten, Posten had thus failed to establish that the illegal cartel activities resulted in an overcharge.
This conclusion raises several questions both on how such cases are argued and presented to the court, but also on the court’s reasoning in the absence of reliable econometric evidence. While the court appears to argue that a robust econometric analysis is required to establish a loss in this case, such evidence is normally not required under Norwegian tort law. The court’s emphasis of econometric evidence in this case may have been impacted by the parties’ heavy reliance of such evidence in the court proceedings, while other indicators of loss appear to be given less weight. In light of the court’s reasoning, it seems that the court proceedings emerged as a battle ground for economic experts bringing focus to technical aspects of econometric modelling, in the end levelling each other out. Therefore, it may be important also to focus on other indicators that could shed light on the actual effects of the cartel activities on transaction prices.
In addition, the court admitted that it struggled to fully comprehend the technical aspects of the presented regression analyses and arguments forwarded by the economic experts. Norway does not have specialist courts for civil claims. Obviously, it may be challenging to communicate advanced econometric models to a generalist judge with likely limited experience in both econometrics and competition law in general. This could suggest that the courts in follow-on claims should be sat with expert lay judges. In this case, the parties did not request that the court should sit with expert lay judges.
The court rejected a reversed burden of proof
The burden of proof normally lies with the claimant. Posten argued that the truck manufacturers had acted reprehensible to a degree that they should bear the burden of proof that no overcharge had occurred but was not heard by the court on this view.
Directive 2014/104/EU on antitrust damages actions Article 17 (2) states that it shall be presumed that cartel infringements cause harm. Furthermore, the defendant shall have the right to rebut this presumption. This has been interpreted as a reversed burden of proof in follow-on claim cases following cartel infringements.
The court, however, recalls that the damage directive is not (yet) implemented in Norway. This indicates that until Norway actively implements this directive, the burden of proof will still lie with the claimant. Awaiting implementation of the directive in Norway, this may lessen the chances for successful follow-on claims in Norway compared to EU member states.
The court adds that the directive nevertheless would be non-applicable in this case, as it entered into force after the infringement ended.
Significant degree of uncertainty in the judgment indicates that the last word has not been said
As stated above, the court concluded that none of the economic experts presented sufficiently reliable analyses. Even if the court found it probable that the gross list prices affected the actual transaction prices, this was obviously not considered sufficient without a reliable empirical analysis demonstrating this effect. In the absence of reliable empirical analyses, the court seemed to have limited evidence to prove an effect on transaction prices and gave the benefit of the doubt to the truck manufacturers.
This uncertainty expressed by the court throughout the judgment clearly suggests that an appeal by Posten is likely, and that we have not seen the end of this case. Expert evidence and empirical analyses are likely to become important also in an appeal case, but the claimant should probably attempt to avoid the appeal case becoming dependent on a battle between the economic experts, through inter alia bringing focus to other indicators of overcharge. Moreover, a key takeaway is that the claimants in cases where econometric evidence plays a key role should strongly consider requesting expert lay judges in order for the court to be better suited to comprehend and assess the presented analyses.