Post Danmark II: A Clarification of the Law on Rebates under Article 102 TFEU

By Konstantinos Sidiropoulos

Post Danmark II constitutes the latest signal as to the view of the CJEU with regard to the assessment of rebates granted by dominant firms. As this was the first preliminary reference in a rebates case ever, there were high expectations with regard to the judgment (see e.g. here). It was seen as a golden opportunity for the Court to provide meaningful guidance, unconstrained by the limitations of judicial review in a truly fascinating and heavily disputed field of EU competition law. Indeed, this is the area where the European Commission made the most significant efforts to alter the current state of the law (see paras 37-45 of the Commission’s Enforcement Priorities Paper), albeit unsuccessfully (see judgments in Intel and Tomra). Hence, the key issue was whether the CJEU would ultimately yield to the increasing pressure to move to a more economically inspired approach to rebates under Article 102 TFEU. Overall, the ruling is valuable in that it clarifies the standard applicable to rebates granted by dominant undertakings.

Facts and Questions Referred to the CJEU

Post Danmark was controlled by the Danish State and held a statutory monopoly on all mail weighing up to 50 grams. This led to over 70% of all bulk mail in Denmark (the relevant market) being covered by that monopoly. Post Danmark had a market share of 95%, significant structural advantages and a nationwide distribution network, while the market was characterised by high barriers to entry and economies of scale. In 2007, Bring Citymail entered the bulk mail market, but withdrew in 2010 after suffering heavy losses. Following a complaint lodged by Bring Citymail, the NCA issued a decision founding that Post Danmark had abused its dominant position on the Danish bulk mail market, in 2007 and 2008, by operating a rebate scheme for direct advertising mail. The case was subsequently brought before the Maritime and Commercial Court which decided to send a preliminary reference to the Court.

The rebates implemented by Post Danmark had the following characteristics: i) the rebate scale was standardised ranging from 6% to 16%; ii) the rebates applied to periods of one year; iii) they were conditional; and iv) they were retroactive.

The questions referred to the Court were of general significance for the law on rebates under Article 102 TFEU (para 20). First, the Court was asked to illuminate the criteria to be applied for considering that a rebate scheme as the one at issue in this case is abusive. Second, it was requested to clarify the relevance that is to be attached to the ‘as-efficient-competitor’ test (‘AECT’) for the evaluation of rebates. Third, the referring court asked whether a rebate scheme must generate probable and/or appreciable exclusionary effects in order for it to fall within the scope of Article 102 TFEU.

Judgment of the Court

The Court considered that the rebate scheme at issue was neither a loyalty rebate nor a pure quantity discount (para 28). Rather, it treated it as belonging to the residual third category of rebates; namely, those that are neither formally conditional upon exclusively nor solely based on volume. Therefore, the Court held that for this particular rebate scheme to be found unlawful, it was ‘necessary to consider all the circumstances’ (para 29). These circumstances include the modus operandi of the scheme, the extent of Post Danmark’s dominant position and the particular conditions of competition prevailing on the relevant market.

As regards the criteria and rules governing the grant of the rebates, the Court held that the retroactive nature of the scheme in conjunction with the annual reference period, which was regarded as a relatively long reference period, could produce the so-called ‘suction effect’ (paras 32-35). The Court also stressed the very unique conditions prevailing on the relevant market which pointed clearly to the conclusion that competition was already very limited and that Post Danmark was an unavoidable business partner (paras 39-42). Moreover, the Court held that the widespread impact of the rebates may be regarded as a useful indication as to the likelihood of anticompetitive effects (para 46). Finally, the Court re-affirmed that a dominant firm may, as a matter of principle, provide a justification for its conduct (paras 47-49).

As regards the question concerning the relevance to be attached to the AECT, the Court held unequivocally that this economic test is not a legal prerequisite in order to find a rebate scheme abusive under Article 102 TFEU (para 56). It also pointed out that, considering the particularities of the case, the AECT is of no relevance at all. Two interrelated reasons were put forward by the Court to justify this statement. First, the emergence of an as-efficient competitor was by all practical means impossible due to the structure of the relevant market; and, second, in a market such as the one in this case, the presence of a less efficient competitor may contribute to intensifying the competitive pressure (paras 59-60; similarly AG Kokott’s Opinion at points 71-72). Interestingly, however, the Court made patently clear that the AECT should not be excluded altogether as a proxy for the assessment of rebates (para 58). Rather, it should be regarded as ‘one tool amongst others’ (para 61).

As to the issue of the probability and appreciability of the exclusionary effects, the Court held that, although it is a prerequisite that the likelihood of an anticompetitive effect is shown, there is no need to demonstrate the serious or appreciable nature of these effects for a practice to fall within the scope of Article 102 TFEU (paras 66-67 and 73-74). Put differently, the Court refused to fix an appreciability threshold for the purposes of determining whether there is an abuse of dominance.


Despite the weak reasoning concerning the necessity of setting a de minimis threshold for the application of Article 102 TFEU, the judgment should be regarded as a welcome contribution in the effort of clarifying the law on rebates under Article 102 TFEU.

To start with, the Court confirmed the three-fold categorisation of rebates which was expressly proposed for the first time by the GC in Intel (paras 74-78). Interestingly, this was one of the few points where the Court disregarded AG Kokott’s Opinion, where she argued that ‘it is ultimately immaterial whether the scheme can be assigned to a traditional category of rebate’ (point 29). Be this as it may, the case-law continues to classify business practices in categories depending on their form, and this in turn determines the corresponding legal tests for establishing whether the practices in question are abusive or legitimate. This should not be seen as problematic, insofar as legal rules are – and must be – formal in order to be administrable and predictable, and thus, in conformity with fundamental rule-of-law principles.

Furthermore, the ruling is important because it clarified the scope of the prima facie legality rule as regards quantity rebates. Although there has never been any controversy as to the presumptively legal nature of quantity rebates, the issue of whether this presumption applies to volume rebates that are retroactive as opposed to incremental was hitherto unsettled. The only notable example is Michelin II, a case concerning a retroactive volume rebate scheme, where the GC ignored the presumption of legality in its analysis. Following Post Danmark II, it seems that for a rebate to qualify as a quantity discount it must satisfy three cumulative conditions; namely, be at once volume-based, standardised and incremental (see Hans Zenger at p. 722, who was vindicated on this point). This inevitably leads to a narrow understanding of the ‘quantity rebates’ category.

The most marked aspect of the ruling is the evolution of the law on rebates that fall into the residual category, and which must be assessed under the ‘all the circumstances’ standard. Indeed, the EU Courts have never voiced in a systematic way the factors that must be taken into account in the context of the analysis of ‘all the circumstances’ for the assessment of the so-called ‘loyalty-inducing rebates’. As such, the criticism that it is sufficient to identify some factors hinting at the ‘loyalty-inducing’ nature of the scheme to establish a prima facie violation of Article 102 TFEU (see here and here at p. 15) was legitimate. In Post Danmark II, however, the Court departed from previous case-law both rhetorically and substantively. Rhetorically, it omitted the terms ‘loyalty-inducing’ and ‘fidelity-building’ which are slippery and potentially misleading. Most importantly, as a matter of substance, the Court clarified that under the ‘all the circumstances’ standard one must conduct a full effects-based analysis of the rebate. Therefore, contrary to the findings in Michelin II (especially, para 241) and British Airways (especially, para 288), when assessing this kind of rebates, the likelihood of any exclusionary effect cannot merely be assumed, but must be assessed on the basis, inter alia, of the factors outlined at paras 35, 39, 40 and 46 of the judgment in Post Danmark II. Consequently, the (un)lawfulness of these rebate schemes depends on an in concreto assessment of their effects in the specific market context. That is to say, where exclusivity is not a formal requirement for the grant of the rebate, is it necessary to conduct an economic analysis of the rebate in question. The factors that must be examined under this economic analysis include: i) whether the rebate is retroactive or incremental; ii) how large the barriers to entry are; iii) the significance of economies of scale and scope; and iv) what portion of the market is (non-)contestable.

In this connection, much turns on the facts of the case. In a highly regulated market, rebates granted by dominant firms should expect greater scrutiny. On the other hand, the approach would presumably be more lenient in the context of an open, less regulated and dynamic market. This not only is economically sensible, but it is also the very essence of a proper effects-based analysis.

The judgment is also innovative in that it is the first time that the CJEU recognised that the AECT may, ‘on principle’, be a relevant proxy in rebate cases under Article 102 TFEU (paras 58 and 61). Interestingly, the Court held so obiter dictum, since the special features of the case rendered the application of the test counterproductive and useless (see para 59). However, it would be naïve to claim that the Court abandoned its negative stance against the AECT as advocated in the Commission’s Enforcement Priorities Paper (see Tomra at paras 73, 74 and 80; Intel at paras 140-166). This is so, not only because of the flaws (see e.g. here) and practical challenges of the test (see e.g. here at p. 13), but also because of two other reasons. Firstly, the EU judiciary, as a general rule, refuses to deal with complex economic and technical issues. Therefore, the CJEU is not susceptible to this kind of analysis. Secondly, the EU Courts seem to be more concerned with the level of legal certainty across Europe, that is, at national level (see Regulation (EC) No 1/2003). As such, there is a sensitivity as to the adverse impact of legal uncertainty to the vertical relationship of the CJEU with the national courts.

Most disturbing is the finding that there is no need to demonstrate an appreciable effect on the market under Article 102 TFEU. This constitutes a recapitulation of the previous case-law (see Case T-155/06 Tomra at paras 231-242, confirmed on appeal Case C-549/10P at paras 42 and 46; Intel at paras 116-120). According to these dicta, an ‘abuse of minor importance’ is caught by the prohibition of Article 102. This, however, is problematic. If foreclosure covers too small a part of the market to actually marginalise the dominant undertaking’s rivals, rebates tend to lower prices, rather than generating anticompetitive effects. In any event, as pointedly stressed by Whish (see here), the demonstration of appreciable effects is a ubiquitous requirement in EU competition law; there is no convincing reason to move away from it in the context of Article 102 TFEU.

Be this as it may, a rebate scheme must at least generate likely exclusionary effects. This, combined with the statement of the Court that the widespread impact of a scheme is part of the assessment of ‘all the circumstances’, may lead to the argument that the Court introduced a de minimis rule through the backdoor. In other words, the appreciability issue is addressed perforce in the context of the examination of ‘all the circumstances’ (on this argument see here). This idea was neatly expressed by the AG who stressed that:

the use of a de minimis threshold […] seems to be unnecessary for two further reasons: first, […] the exclusionary effects […] are to be determined on the basis of a specific examination of all the relevant circumstances of the individual case and their presence must be more likely than their absence. Secondly, the prohibition of abuse contained in Article [102 TFEU] is in any event directed only at conduct capable of affecting trade between Member States (point 93).

According to the AG, these considerations rule out the possibility that Article 102 TFEU would apply to conduct which is ‘of wholly negligible significance’ (point 94). This is indeed a superior rationale for denying the de minimis rule compared to the one provided by the Court, which linked its reasoning almost exclusively to the special responsibility doctrine (paras 71-73). Even so, the AG’s reasoning is equally problematic because exclusivity rebates are not examined under the ‘all the circumstances’ standard. In any event, it would be appropriate to prescribe a safe-harbour for reasons of clarity and consistency.

The ruling is generally good law. It is a well-drafted and internally consistent judgment. It might be a disappointment for certain observers who argue that the case-law is excessively formalistic and not in line with the teachings of economics. Nevertheless, the Court was right in heeding the AG’s invitation to remain in the mainstream (see point 4).

Leave a Reply