The recent decision of the High Court in Vattenfall AB v Prysmian SpA  EWHC 1694 (Ch) is another example of claimants being allowed to use non-addressee English subsidiaries as anchor defendants for their competition damages claims. It is also another example of the court considering but not actually having to decide the interesting legal points around attribution of liability which potentially arise in such cases.
There have now been several cases with the same basic structure: the European Commission decides that a company is liable for a competition law infringement and some claimants then start proceedings against that company’s English subsidiary in order to establish jurisdiction in England. The defendant objects to the claimants’ attempt to sue an entity which was not, after all, an addressee of the Commission’s decision simply to use it as a means of bringing proceedings in London.
The problem which defendants face in such a scenario, and the reason why they keep losing these cases, is that it is relatively easy for claimants to allege that the English subsidiary was knowingly involved in the infringement. If the Commission has found that there was an EU-wide infringement then it will often be entirely proper for the claimants to infer, at least sufficiently to meet the low threshold for establishing jurisdiction, that the infringement was implemented in England through the English subsidiary. In Vatenfall the claimants had the additional benefit of being able to point to concrete evidence in support of their knowing implementation plea. Provided that such a plea is properly made, a standalone claim can be run against the English subsidiary which can then be used as the anchor defendant.
The more interesting legal question is what the position would be if a claimant cannot properly plead a standalone case of knowing implementation against the English subsidiary. Could the parent’s liability be attributed to the subsidiary so that the subsidiary can be sued and used as an anchor defendant even though it was not involved in the cartel? The English judges who have considered this question have expressed different views about it.
One view is that the answer is that liability can be attributed from an infringing subsidiary to its parent company but not the other way round. Supporters of this view point to the fact that, when discussing attribution of liability in the line of cases starting with Akzo Nobel NV v European Commission  5 CMLR 23, the European Courts have been concerned only with imputing liability to a parent company, and that they only permit such attribution if the parent exercised a ‘decisive influence’ over the infringing subsidiary.
The problem, however, is that the reasoning in the Akzo Nobel line of cases is expressed in quite wide-reaching terms. The basic logic is that if the parent and subsidiary are part of the same single economic unit then they form a single undertaking, and that if an undertaking infringes competition law then all of its legal entities are liable for the breach. The ‘decisive influence’ test is really about determining whether the parent and subsidiary are part of the same economic unit, not an additional threshold for the attribution of liability between companies which are in the same economic unit.
Thus the alternative view is that the liability can be attributed between any and all companies in the same undertaking. This has caused some consternation among English judges because of its apparently wide-reaching consequences. Does it mean that liability could be attributed to a subsidiary with no knowledge of or involvement in a cartel? Or even from one subsidiary to another?
In the Sainsbury’s case ( CAT 11) at  the CAT suggested something of a compromise: liability could be attributed between companies in the same undertaking but only if they had “in some way” participated in the breach or otherwise exercised a decisive influence over a company which did. That is a sensible solution, but it might be said that it still sits somewhat uncomfortably with the reasoning in Akzo Nobel.
A different way of formulating the point might be as follows. In accordance with the reasoning in Akzo Nobel, liability can always be attributed between companies in the same undertaking. However, when asking whether companies are in the same “undertaking” one needs to keep in mind that the identification of an undertaking depends on the circumstances. In a competition infringement case, the question is whether the companies acted as a single economic unit for the purposes of the infringement. If they did then they are all part of the same undertaking and they are all liable. If they did not then they are not part of the relevant undertaking and liability cannot be attributed to them from any other company in the group.
If this approach is correct then a subsidiary which genuinely had nothing to do with an infringement will not be liable for it. But a subsidiary which, whether knowingly or not, acted in concert with other group companies such that they operated as a single economic unit to implement an infringement will be liable for it. That is essentially (with slightly different reasoning) the approach which appealed to the CAT in Sainsbury’s, and it is a solution which avoids some of the extremes of other proposed solutions.