Stellantis Auto SAS v Autoliv AB [2024] EWCA Civ 609 – The Court of Appeal dismisses an appeal against an order of the Competition Appeal Tribunal requiring two defendant groups to use a single joint expert at trial in relation to the issue of overcharge

The background to the matter is set out in detail in our briefing of 29 November 2023. To recap, the Claimants (manufacturers of motor cars) have brought a stand-alone claim against Autoliv and ZF, suppliers of occupant safety systems (namely, seat belts, airbags and steering wheels, together, "OSS Products"). While the supply of OSS Products were the subject of two Commission Decisions (OSS1 and OSS2), those decisions did not concern supplies to the Claimants. The Claimants therefore advance their case on two bases. First, they contend that Autoliv and ZF formed a cartel or were part of a cartel in respect of the sale of OSS Products to the Claimants. Second, they have a follow-on 'umbrella' claim from the Commission decisions OSS1 and OSS2, alleging that the impact of the cartels identified by the Commission was to reduce competition and increase prices in the market generally.

The CAT granted permission for a single expert in the field of competition economics to be instructed jointly by the (then) three Defendant groups[1] (the "First Decision"). This limitation was opposed by Autoliv and ZF. The CAT explained its decision as being made principally on the basis of case management considerations, namely, that the use of a single joint expert would help streamline the expert process. The Tribunal was also of the view – perhaps contrary to the understanding of many practitioners – that there is no established practice in the Tribunal to the effect that defendants in cartel cases should be permitted to use separate experts. While the Tribunal acknowledged that ordering the Defendants to use a single joint expert may give rise to potential conflicts of interest, in the instant case, the Tribunal was of the view that the conflicts identified were theoretical.

Popplewell LJ granted ZF and Autoliv permission to appeal the First Decision to the Court of Appeal.

Prior to the hearing of the appeal, on 15 April 2024, Autoliv and ZF made an application to the CAT asking the CAT to revisit the First Decision and requesting that permission be granted for separate experts for Autoliv and ZF relating to the issue of overcharge (the "Second Decision"). Autoliv and ZF cited three material changes since the First Decision, namely: (i) the grant of permission to appeal; (ii) the claim against the third Defendant Group having been withdrawn; and (iii) the service of the Claimant's expert report, which, they contended, required the Tribunal to revisit the question of whether a conflict of interest had arisen as between Autoliv and ZF that necessitated the use of separate experts. The first two reasons were disposed of in short order on the basis that the Tribunal did not consider it appropriate to second guess the outcome of the appeal, nor did it consider the withdrawal of the claim against the third Defendant Group as a reason for increasing the number of expert witnesses.

As regards the third change, it is helpful to first consider the basis on which the Claimant's expert had prepared his expert evidence, which purports to show, by way of a regression analysis, that prices charged for OSS Products were elevated during the period of the alleged cartels. That analysis will be based on data from supplies to the Claimants and would not seek to distinguish between the OSS1 and OSS2 cartel periods or the activities of the individual Defendant groups. Accordingly, while the regression analysis would seek to isolate the overcharge from other factors which influence price, it cannot identify the cause of the overcharge, including, for example, whether it was caused by a specific cartel, an umbrella effect, or the prices of a certain Defendant group. With this in mind, the Tribunal disagreed that the Claimant's expert report led to any conflict being identified. In particular, the Tribunal found that the fact ZF was not a party to the OSS1 decision did not give rise to a conflict in this context as the (Claimant's) proposed expert evidence would not be capable of informing the Tribunal as to whether the overcharge arose from different cartels or the activities of different cartelists. The Tribunal re-iterated its position (as set out in the First Decision), that the quality of justice will be impacted if the Tribunal is faced with the task of having to resolve different methodologies using different datasets from different Defendants. 

The Court of Appeal (Birss LJ giving the leading judgment, with which Arnold LJ and Sir Geoffrey Vos agreed) was required to address two questions. The first was whether the CAT was right to hold that there was no relevant conflict of interest in relation to Autoliv and ZF's expert evidence. The second question was which principles should be applied when a Court or Tribunal is considering ordering a single joint expert.

Overarching principles

In the first instance, it is helpful to examine the Court's approach to the relevant overarching principles.

The Court agreed with the parties that there was "no relevant difference" between the rules set out in the CPR and the principles applicable to expert evidence in the CAT (as set out in the CAT Rules 2015 (the "CAT Rules") and Guide to Proceedings (the "CAT Guide")).[2] Common to both regimes are four fundamental pillars:

i. there is a duty on the Court/CAT to restrict expert evidence;[3]

ii. experts have an overriding duty to the Court/CAT and not to the party instructing them;[4]

iii. the Court/CAT has the power to appoint a single joint expert; and[5]

iv. the Court/CAT must ensure that cases are dealt with "justly and at proportionate cost".[6]  

In light of these Rules, the Court was of the view that a direction for a single joint expert to give evidence (in place of separate experts) "like any other direction giving permission for expert evidence is governed by two primary dimensions": the first is to ensure that the case is dealt with justly and at proportionate cost and the second is to restrict expert evidence to only that which is reasonably required to resolve the proceedings.[7]

Question of conflicts

ZF and Autoliv both submitted that a conflict of interest did arise between the Defendant groups and that this precluded appointment of a single expert. They cited in support of this the Court of Appeal's decision in UK Trucks Claim Limited v Stellantis NV and Others, where the Court held that a single economic expert could not be used for both classes of Claimants.[8] In UK Trucks, the Court noted that the regression analysis was "highly sensitive" to data and the associated assumptions, both of which contained an element of subjectivity.[9]

The Court of Appeal disagreed. It identified that the true conflict in UK Trucks was the fact that the single joint expert was being instructed by one organisation which represented the whole class in circumstances where there was "divided loyalty" within that organisation to two different sub-classes.[10] The Court was clear that UK Trucks is therefore not authority for the proposition that a conflict of interest in the context of econometric modelling meant that separate experts should be ordered.

Outcome

Strikingly, the Court concluded that the CAT had been incorrect in finding that where a conflict of interest does exist, "it will not ordinarily be appropriate to order joint experts".[11] This conclusion seems to have flowed in substantial part from (i) the ability of the Court under the CPR/CAT Rules to order a single joint expert for both claimant and defendant; and (ii) the fact the overriding duty of the expert is to the Court and not the party who instructs them.

While the Court acknowledged that the existence of a conflict of interest is a "material factor" in deciding whether to order the use of a single joint expert, it is not "a trump card" which requires the appointment of separate experts.[12]  The Court held that "It may well be appropriate in the interests of justice to order a single expert even if there is a conflict of interest between the instructing parties on the matter to which the expert evidence is directed."[13]

By allowing both Defendant groups to instruct separate experts, the economic issues in dispute (which the CAT would have to reconcile) would increase threefold. The fact that the existence of rivalling models would increase disproportionately as the number of experts increased was why the existence of conflicts of interest between Defendants in the present scenario could not operate in such a way as to "undermine the ability of the court or Tribunal to manage the just disposal of a case like this to address the problem".[14]

The Court also found that the CAT had not erred in finding that a conflict of interest did not arise on the facts of the case. The Court disagreed that there were conflicts between ZF and Autoliv in relation to apportionment of liability, noting that the pleaded case is an undifferentiated claim in which both parties are jointly and severally liable. Further, while "in theory" a conflict of interest could arise in relation to the issues of umbrella damages and causation, "in practice" there was no such conflict because the relevant expert evidence (as advanced by the Claimant's expert) would not be capable of making distinctions relevant to this question. 

Stepping back from the specific reasoning of the decision itself, this case needs to be seen in the broader context of the issues facing the Courts, and in particular the CAT, in managing an increasing volume of complex cases.  Actions before the CAT frequently pose very specific challenges, engaging large amounts of critical expert evidence, which is very difficult even for a specialist Tribunal to engage with. Mrs Justice Bacon (speaking extra-judicially) has noted that lawyers and experts are not yet providing the Tribunal with material which is most helpful from a judicial perspective – often, experts are so far apart and there are so many unresolved issues in the analysis that the Tribunal is not able to reach a decision by following one or other (or even a combination) of the proposed approaches.[15] The CAT itself has publicly recognised the difficulties it faces in this regard, noting on several occasions that there is a need in large complex cases to manage the expert process carefully in order to ensure cases are disposed of efficiently.[16] The Court and the CAT's  judgments in these proceedings can be seen as a further step in the CAT seeking to make cases (and its caseload) manageable. 

The judgments of the Court of Appeal and the CAT are also notable in referring to the tension between the expert's duties of independence, and the tendency for the Claimant's expert to find a significant positive overcharge and the defendant's expert to find no overcharge. While the Court of Appeal observed that an expert's overriding duty to the court may mean that an expert "expresses views on matters which the party calling them would rather were put in a different way or not at all",[17] such comments may not accord with what actually happens in practice i.e. that an expert's evidence typically does seek to advance the instructing party's case.  The issue was also addressed in the Royal Mail decision where the CAT went as far as to say that perhaps this "is an inevitable consequence of the adversarial process and one should expect a party to have an expert that supported their case".[18] Indeed, in cases which turn so greatly on the expert evidence, there is a question as to whether it is consistent with the English adversarial system to strip a party of the ability to choose its own expert.   

Both the Tribunal and the Court of Appeal acknowledged that given the size of the claim, instructing multiple experts would not have been disproportionate,[19] but concluded that proportionality, while an important factor, is not determinative. Parties will therefore no longer be able to contend that they have cleared the "proportionality hurdle" in order to justify the use of separate experts.

In this case, the Court found (similarly to the Tribunal) that the conflicts identified by ZF and Autoliv were "theoretical" and did not impact the expert evidence. It will be interesting to see whether the outcome of future decisions in this area will be different if the relevant conflicts are more tangible – for example, if the defendants in a multi-defendant claim have served contribution notices against one another (which were notably absent in this case). It is not clear what will happen in the event the single joint expert comes to the view that there should be separate analyses for each of ZF and Autoliv on the basis, for example, that the disclosure indicates differing levels of overcharge for each Defendant group.    

The Tribunal did acknowledge that one scenario in which separate experts may be permitted is if there are "multiple schools of thought" in relation to the relevant expert issue(s). While, superficially, this might encourage parties to portray differences between the experts as being rooted in different schools of thought, such a strategy would appear to be a non-starter in the context of competition claims where the central analysis is econometric evidence. As the Court noted, "the fact experts here might use different regression models or disagree about the input variables does not demonstrate the presence of multiple schools of thought".[20]