Key Points

  • The ECJ’s Tomra decision confirms that the use of individualised and retroactive rebates by a dominant firm is a per se abuse, and it is not necessary to prove anti-competitive effects or intent.
  • The Tomra decision is at odds with the European Commission’s 2009 Guidelines on the abuse of a dominant position, which advocate an effects-based approach to determine whether the proposed rebate had the actual effect of foreclosing competitors.
  • Until the legal position is clarified by an EU court decision reviewing a Commission action applying the 2009 Guidelines, dominant companies should avoid using individualised and retroactive rebates, and should also undertake a review of any rebate scheme that was adopted in reliance on the effects-based approach described in the 2009 Guidelines.


The European Court of Justice (ECJ) recently handed down a significant decision holding that the Tomra Group (Tomra) had abused its dominant position through the use of exclusionary strategies in the European Economic Area (EEA). The judgment follows the General Court and ECJ’s traditional formalistic, or “per se,” approach, which generally treats rebate schemes implemented by dominant entities as anti-competitive. The ECJ’s decision is in tension with the effects-based approach to abuses of dominance championed by the European Commission’s 2009 Guidelines.  

What Constitutes Dominance in the EU?

Dominance has been defined in the EU as a position of economic strength that enables a firm to prevent effective competition in a relevant market through unilateral behavior (i.e., behavior that is to an appreciable extent independent of the firm’s competitors, of its customers, and ultimately of consumers.)1 Market shares provide a useful first indication for the European Commission of the market structure and of the relative importance of the various competitors active in the market. The Commission’s experience suggests that a firm is not likely to have dominance if the firm’s market share is below 40% in the relevant market. However, there may be specific cases below that threshold where competitors are not in a position effectively to constrain the conduct of a dominant firm (e.g., because of serious capacity limitations).


The EU Per Se Position

The traditional approach represented by the ECJ’s Tomra decision is to condemn rebate schemes employed by a dominant company as an abuse of dominance under Article 102 of the Treaty for the Functioning of the European Union (TFEU). Under this per se approach, a rebate program implemented by a dominant firm will be condemned if the rebates are given in exchange for customer loyalty and not on the basis of genuine cost savings and efficiencies, particularly where the rebates are tailored to individual customers or constitute ‘all purchase’ rebates. It is not necessary to prove an anti-competitive effect under this approach; a rebate scheme with these characteristics is by its object deemed to be an abuse, because the following is presumed:  

  • Where a dominant entity uses information about its main customers to set rebate targets in such a way as to compel customers to make most of their purchases from the dominant entity, the firm thereby excludes other competitors from the relevant market (this situation is termed an “individualised” rebate); or
  • By applying the discount not only to purchases over a certain threshold, but to a customer’s entire order, the dominant firm thereby penalises customers who do not purchase all of their requirements from the dominant entity (such rebates are termed “retroactive”).  

Under the per se approach, such individualised and retroactive rebate schemes are viewed as preventing customers from making purchases from alternative suppliers, as doing so would in effect penalise the customers through the denial of the discount and the imposition of associated switching costs. In the per se view, these rebate schemes are therefore deemed to be a barrier to entry that prevents other competitors from entering into or expanding in the market.  

Tension Between the Per Se Position and the Effects-Based Approach

In 2009, the European Commission issued Guidelines detailing its approach to Article 102 investigations going forward. These Guidelines distanced the Commission from the traditional per se approach to rebates.

In the Guidelines, the Commission confirmed that rebate schemes are not an uncommon practice. A company may use rebates to incentivize customers to purchase not only those “must have” items that the customer will purchase from the company regardless of discounts, but also those “contestable” items that the customer might otherwise purchase from alternative suppliers. At the same time, however, the Guidelines recognize that where such rebates are instituted by a dominant firm, the result may be anti-competitive foreclosure of a necessary part of the market to competitors.

Applying its Guidelines, the Commission will look at the likelihood that the rebate will hinder entry or expansion by competitors that are equally or more efficient than the dominant entity.  

As well as looking at the terms of the rebate, the Commission will attempt to calculate the price a competitor would have to offer in order for a customer to switch. This switching price will be calculated as the dominant entity’s price less any rebates given over a relevant sales range and period. As long as this calculated price remains above the “long-run average incremental cost” of the dominant supplier (i.e., the dominant supplier is not operating at a loss), a rebate would generally not be seen as capable of foreclosing the market, as an equally efficient competitor should be able to compete on that basis. Where, however, the calculated price is below the “average avoidable cost” of the dominant supplier (i.e., the costs that could have been avoided if the company did not produce extra units), the rebate will generally be seen as foreclosing the market. Where the price lies between these two cost measures, other factors, such as alternative strategies available to competitors, will be considered.

The Tomra Judgment: Is the Tension Resolved?

Tomra produces reverse vending machines (RVMs), which collect empty beverage containers. A competitor, Prokent, complained to the Commission that Tomra had abused its dominant position, under Article 102 TFEU, to prevent new competitors from entering the market and to drive existing competitors out of the market.

In a decision issued three years before its 2009 Guidelines, the Commission held that Tomra had indeed abused its dominant position in the RVM market in five EEA countries, by implementing individualised loyalty rebates, exclusivity agreements, and individualised quantity commitments. The rebates were found to have the following characteristics:


  • Identified customers were entitled to rebates on their whole purchase order where they reached a given purchasing target during a period.


  • The thresholds related to the total requirements of the customer or a large proportion thereof. They were established on the basis of a particular customer’s estimated requirements or purchasing volumes achieved in the past.
  • Some rebates contained progressive discounts relating to two or more thresholds. The first threshold would correspond to a substantial proportion (over half) of the customer’s estimated or actual requirements, while the higher bonus thresholds often corresponded to the customer’s total demand.

These characteristics were found to be abusive, as customers who initially bought from Tomra had a strong incentive to continue to purchase from Tomra in order to reach the necessary threshold and reduce overall prices. This incentive artificially raised the cost of switching to a different supplier and thereby harmed competition.

As a result of these findings, the EC imposed a fine of €24million on Tomra.  

On appeal to the General Court, Tomra argued that the per se approach endorsed by the Commission with no analysis of actual effects of the discounts has no basis in business practice. Tomra pointed out that the Commission had failed to establish:

  • that Tomra’s conduct had in fact eliminated competition;
  • that Tomra had intended to eliminate competition; or
  • that the rebates had resulted in negative prices (i.e., below-cost pricing).

In a decision issued in September 2010, the General Court rejected these arguments and held (per Hoffmann-La Roche v Commission)2 that individualised retroactive rebates are intended to give the customer an incentive to obtain supplies exclusively from the dominant undertaking and are incompatible with the objective of undistorted competition, because they tend to remove or restrict the customer’s freedom to choose its sources of supply and thereby deny competitors access to the market. Further, the court held, there is no requirement under the law to analyse whether the rebates actually had this effect on the market or to establish that there had been below-cost pricing.

After the failed appeal to the General Court, Tomra took the case to the ECJ.3 Tomra contended that the General Court had failed to show that:

  • Tomra had intended to foreclose the market;
  • the portion of demand that was foreclosed was a significant portion of the market; or
  • there had been any analysis to show that there had been below-cost pricing in line with the 2009 Commission Guidelines.

The ECJ confirmed the General Court’s ruling and rejected the appeal.

Abuse of Dominance

The ECJ agreed with the General Court that “abuse of dominant position” is an objective concept and held that anti-competitive intent is only one factor relevant to whether there has been an abuse. Even if it were shown that the firm intended to compete on the merits, as Tomra asserted, there could still be an abuse of dominance, the ECJ held.  

Further, according to the ECJ, it is sufficient to show that the firm’s conduct has the possibility of producing an anti-competitive effect on the market. The ECJ held that it is not necessary to show that the conduct did in fact produce such an effect.  


The ECJ concluded that by foreclosing part of the market, Tomra had restricted competition, because competitors should be able to compete for the entire market. The ECJ further held that it is not necessary to determine what percentage of a market must be affected before conduct would be classified as abusive.

Retroactive rebates

The ECJ confirmed the General Court’s formalistic assessment of “retroactive” rebates (per the longstanding precedent of Hoffmann-La Roche v Commission). The ECJ held that it is not necessary to prove that a retroactive rebate actually resulted in below-cost prices. Instead, the specific characteristics of the rebate scheme will determine whether it is abusive.

The ECJ held that retroactive rebates have a “suction” effect that would inherently drive out Tomra’s competitors. Customers would choose Tomra to supply the contestable part of demand over other competitors, as this would mean that the effective price of the last units bought by the customer was very low. As this effect tends to be anti-competitive, according to the ECJ, it is unnecessary to carry out an analysis of the actual impact. The ECJ cited and by inference agreed with the Advocate General’s position that the 2009 Guidelines had no relevance to the 2006 decision under review.

Uncertainty in the Law

The approach taken in Tomra contrasts with the shift that has been seen in the ECJ’s approach in other cases, such as the French Broadband predatory pricing case,4 where a more effects-based enforcement approach was taken (see DechertOnPoint Predatory Pricing in the EU: The French Broadband Case” April 2009).

There is significant uncertainty surrounding rebate schemes following the latest Tomra judgment. The ECJ’s ruling that there is no obligation for the Commission or the General Court to examine separately whether the price charged by Tomra was lower than its “long-run average incremental costs,” in relation to the retroactive rebate scheme, is in direct contrast to the Commission’s Guidelines. As the Tomra case pre-dates the Commission Guidelines, it remains to be seen whether the ECJ will approach future cases differently. Further, the Guidelines are not binding, so the tension between the effects-based Guidelines and the formalistic approach of the Tomra decision may persist.

This tension is illustrated in the Intel case,5 where the Commission followed its 2009 Guidelines and applied the effects-based approach to find that the rebates in that case were abusive (see DechertOnPointUncertain Times for Dominant Firms: The EU Commission’s Intel Decision Finally Becomes Public” October 2009). However, the Commission in Intel also performed a separate analysis of Intel’s rebates under the traditional per se approach and found the rebates to be abusive under that alternative approach as well. The Intel decision has been appealed to the General Court, but no judgment has yet been handed down. No other judgments have been issued by the General Court or the ECJ on rebates following the introduction of the 2009 Guidelines.

In practical terms, if the General Court and the ECJ continue to use the formalistic per se approach when reviewing Commission decisions made subsequent to the publishing of the 2009 Guidelines, the effects-based approach may become irrelevant. In the meantime, the confusing tension between the two approaches persists.  

Rebate Schemes Can be Pro-Competitive

There are strong arguments for holding that rebates should in the first instance be seen as pro-competitive. Dominant firms may employ rebate schemes for justifiable commercial reasons, such as to achieve economies of scale or to assist in the introduction of new products to the market.

Where there have been no negative effects on the market following the introduction and during the span of a rebate scheme –– such as the dominant entity’s pushing out competitors and then raising prices, lowering quality, or slowing down innovation — it is difficult to justify the assertion that the practice is anti-competitive.

Indeed, preventing dominant entities from implementing rebates may harm consumers as they will end up paying higher prices for their goods. This potential may be a good reason to suppose that the Commission will continue to move towards a more effects-based approach in the future, despite the Tomra judgment.

Contrasts With the US Approach

While the US courts and antitrust enforcement authorities have not settled on a consistent approach to loyalty discounts, most US variations differ markedly from the approach taken by the EU courts and, to a lesser extent, by the European Commission’s Guidelines. The US approaches tend to break down into some form of either predation or a de facto exclusive dealing analysis. Either approach would impose on the governmental or private challenger the burden of proving that a loyalty discount would have an actual exclusionary impact across a significant portion of the relevant market.

Any predation analysis would also require proof that the dominant firm could recoup its losses from a bundled discount — a requirement that is also absent under the EU approach as applied by both the Commission and the courts.  


There is significant uncertainty created by the tension between the ECJ’s Tomra decision and the 2009 Guidelines. Until there has been an EU court review of a Commission decision applying the 2009 Guidelines, a dominant company should err on the side of caution and operate on the basis that the per se approach used by the ECJ remains good law. Accordingly, dominant companies should avoid using individualised and retroactive rebates, and should also undertake a review of any rebate scheme that was adopted in reliance on the effects-based approach described in the 2009 Guidelines.