Introduction

On 9 August 2017, the UK Competition and Markets Authority (CMA) issued a rare ‘no grounds for action’ decision following its investigation into whether certain rebate schemes offered by Unilever amounted to an abuse of dominance in breach of UK and EU competition rules (the CMA decision).

The correct framework for analysing rebates has long been the subject of debate under EU and UK competition law rules. The publication of the decision appears to be a deliberate step by the CMA to set a precedent for future enforcement in the UK which embraces an ‘effects-based’ approach.

Recent EU case law

The ‘three-category’ approach – Post Danmark II

Traditionally, the European courts have adopted a strict approach to rebates, often strongly influenced by the form of the rebate mechanism employed.

In its October 2015 judgment in Post Danmark II, the Court of Justice of the EU (CJEU) endorsed the approach of categorising rebates granted by dominant companies into three groupings, and set out the legal treatment applicable to them:

  • volume rebates, ie quantity discounts linked solely to the volume of purchases from the seller concerned, are presumed to be lawful. The CJEU suggested that to fall into this category, rebates must be granted for each individual order, rather than being aggregated for multiple orders placed over a given period;
  • loyalty rebates, requiring or incentivising exclusivity, are presumed to be unlawful. These entail obligations or incentives on the buyer to obtain ‘all or a given proportion’ – which appears to mean ‘all or most’ – of its supplies of a product from the seller; and
  • a third category, for non-volume and non-exclusivity rebates, which requires an assessment of ‘all the circumstances’ to determine legality.

The practical effect of this approach was to limit significantly the types of rebate scheme that would be presumed to be abusive without at least some level of investigation as to their potential effects on the market. 

Possible move to a ‘two-category’ approach – Intel

This three-category approach has been contradicted more recently by Advocate General (AG) Wahl’s advisory opinion to the CJEU in the Intel appeal, published in October 2016.

AG Wahl argued that there were in fact only two categories of rebates: volume-based ones, which are presumed to be lawful, and a second category for which there is no presumption of unlawfulness, but that require an assessment of ‘all the circumstances’, irrespective of whether or not the rebate involves exclusivity. AG Wahl also explicitly criticised several other aspects of the European Commission’s assessment of the rebates in question.

If the CJEU adopts AG Wahl’s approach, this would accentuate the move at EU level towards assessment of the likely effects of rebate conduct, and away from the relatively swift conclusion of unlawfulness based on the form of the rebate that has often been adopted by the EU courts.

UK enforcement

In the UK, a limited number of rebates cases have resulted in a decision in the last 10 years. Some earlier cases were discontinued on the grounds that they did not constitute an administrative priority, after an initial review which considered the likely effects of the rebates in question. Arguably, therefore, the CMA (and its predecessor, the Office of Fair Trading) has always been inclined to take an effects-based approach in this area.

Most recently, prior to the decision, the CMA looked into this type of conduct in the pharmaceutical sector. In May 2017, the CMA sent formal objections to Merck Sharp & Dohme for an alleged abuse of a dominant position through a discount scheme for its Remicade product. This step, combined with the decision, indicates that enforcement against rebates may now feature more prominently on the CMA’s radar.

The CMA adopts the ‘three-category’ approach

The CMA decision concerned promotional schemes offered by Unilever to retailers for single-wrapped impulse ice cream products: free or reduced-price additional products were available if the retailer purchased a minimum number of cases of product (eg ‘buy 8 get 4 free’ promotions).

The CMA followed the three categories proposed in Post Danmark II and found that the schemes fell into the third category (ie non-volume and non-loyalty rebates that required an assessment of ‘all the circumstances’).

Adopting some (but not all) of the language in Post Danmark II, the CMA set out the following framework for its assessment of ‘third-category’ rebates.

  • Effects analysis: it was necessary to assess whether such rebates could produce an exclusionary effect – ie whether, considering all the circumstances, the rebates were capable of making:
    • market entry very difficult or impossible for competitors;
    • and it more difficult or impossible for contractual counterparties to choose between different suppliers.
  • Likelihood: the anti-competitive effect of the rebate must be likely, ie not purely hypothetical (language in previous judgments such as ‘tends to’ and ‘capable of’ should be interpreted in this way).
  • No need to show actual effects: it was not necessary for the CMA to establish that there had been concrete exclusionary effects in order to find an abuse. Evidence of actual effects may be relevant, but were not required. Further, it did not need to show that a likely anti-competitive effect was serious or appreciable. In the event, the CMA assessed the effects of Unilever’s rebate scheme and noted that it did not in practice appear to have had an exclusionary effect on competition.
  • Intent: although not a prerequisite, an intent to exclude competitors is one of the factors that can be taken into account when determining the existence of an infringement. The CMA noted that Unilever’s internal documents contained no evidence of an intent to exclude competitors.

The CMA looked in detail at the rebates in question and ultimately concluded that the structure and availability of the promotional schemes, taken together with the purchasing patterns of the retailers, meant that Unilever’s promotional schemes were unlikely to have an exclusionary effect.

Implications of the CMA decision

The decision gives some welcome insight into why the CMA concluded the rebate scheme did not amount to an abuse of dominance. In addition to adopting the approach set out by the CJEU in Post Danmark II, the CMA gave more detail about its framework for assessment of effects in a manner that suggests it was intending to provide guidance for the future.

The following comments can therefore be made about the current approach to rebates in the UK:

  • The ‘three-category’ approach significantly limits the types of rebates that can be presumed to be unlawful simply due to their form – this now only applies to exclusivity rebates. If the CJEU were to follow AG Wahl’s opinion, exclusivity rebates would require an effects-based assessment as well.
  • Neither the CJEU nor the CMA require evidence that the rebate scheme produced exclusionary effects in practice. However, they have clarified that an authority must consider all the circumstances and conclude that exclusionary effects are likely
  • This shift towards overt effects assessment could introduce significant additional flexibility for dominant businesses. Although the CJEU in Post Danmark II concluded that there was no legal obligation on an authority to examine whether the dominant company’s rebate practices would have driven an equally efficient competitor from the market, it nevertheless held that this ‘as-efficient-competitor test’ was one of a number of tools that could be used to assess whether a rebate scheme amounted to an abuse of dominance. AG Wahl in Intel noted that the as-effecient-competitor test was 'particularly useful'. In the future, unless a materially distorted market structure precludes its application (as was held to be the case in Post Danmark II), this type of analysis by a dominant company may be hard to ignore if it suggests that a rebate scheme would not have led to exclusion.
  • Nonetheless, certain types of rebate scheme still carry significant risk. The rebate scheme in Post Danmark II was itself a volume-based rebate. However, it did not benefit from the presumption of lawfulness due to its retroactive nature, applying to aggregated purchases over the course of a year. The CJEU gave a strong indication that this rebate – in the relevant market context – was likely to have an exclusionary effect.
  • It is also noteworthy that the Unilever promotional scheme that featured in the CMA decision appears to have been a non-retroactive scheme of the ‘buy some get some free’ variety, which did not aggregate purchases over a given period. In order to conclude that this scheme was not likely to have exclusionary effects, the CMA conducted a wide-ranging investigation into the sales of manufacturers and retailers over several years, and reviewed large numbers of Unilever’s internal documents. It appears that the CMA’s initial concerns about this rebate scheme were prompted by the limited freezer space available in smaller retailers for the relevant ice cream products. It is to be hoped that this type of – relatively common – rebate mechanism does not elicit similar levels of attention outside of this context.

The CJEU’s judgment in Intel is due on 6 September 2017. AG Wahl’s opinion contained important criticisms of the European Commission’s analysis in that case, several of which would suggest that there should be greater scope in the future for dominant companies to defend rebate schemes based on an analysis of likely effects. If the CJEU were to follow AG Wahl’s recommendations, the Intel case may therefore signal a further advance in the law from the position set out in Post Danmark II and the CMA’s Unilever decision.