The Role of the AEC Principle and Tests in a Dynamic and Workable Effects-Based Approach to Abuse of Dominance^†

I. Introduction

Since the 2008 guidance paper on exclusionary abuse of dominance,¹ European case law has evolved and moved away from its prior formalistic approach, and the associated per se prohibitions, to a more economically grounded effects-based approach. Building on the experience of the last one and a half decades, the European Commission announced on 27 March 2023 an initiative to draft new guidelines reflecting the latest case law on exclusionary abuse of dominance under Article 102 of the Treaty on the Functioning of the European Union (TFEU), as well as some immediate revisions to the 2008 guidance paper.² Concurrent with this announcement, the Commission published a policy brief explaining the background to both initiatives, with the title ‘a dynamic and workable effects-based approach to abuse of dominance’.³

Key Points

• The as-efficient-competitor (AEC) principle is a workable effects-based framework that recognises the importance of ‘appropriability’ in dynamic competition.

• AEC tests are nothing more than mathematical abstractions of the principle, but they can come with practical limitations depending on the conduct and circumstances.

• A strict application of the principle ignores the importance of not-(yet-)AECs and ‘contestability’ in dynamic competition.

• However, finding a workable alternative remains elusive, and the principle—with or without adjustments, and with all its caveats and nuances—remains a useful economic framework to conceptualise exclusionary abuse and the boundaries of enforcement.

In the policy brief, the Commission discusses the concern that an ‘overly rigid’ implementation of the effects-based approach could make enforcement ‘unduly burdensome or even impossible’.⁴ At the same time, the authors of the policy brief do recognise that the preceding form-based approach may carry the risk of capturing false positives.⁵ As a compromise, the Commission is now in pursuit of a ‘dynamic and workable effects-based approach’ to exclusionary abuse of dominance cases.⁶ However, this raises the following question: exactly when can an effects-based approach to exclusionary abuse of dominance cases be considered both dynamic and workable?

In this article, we focus on the role of the ‘as-efficient competitor’ (AEC) principle and tests and whether and how they can support a dynamic and workable effects-based approach to abuse of dominance cases. We start by outlining how the AEC principle—formally introduced in the EU the 2008 guidance—has provided authorities, courts, and companies with a comparatively predictable and administrable framework that recognises the importance of ‘appropriability’ in dynamic competition—i.e. the idea that incentives to innovate should not be dulled by the protection of less efficient competitors and that even dominant firms should be able to appropriate the gains of their efforts to innovate even if this goes at the expense of less efficient competitors (Section II).

We also discuss how AEC tests, as mathematical abstractions of the AEC principle tailored to the conduct under consideration, can play an important role in operationalising the AEC principle—albeit with varying degrees of success in practice, determined by the varying ease of implementation (Section III). We conclude with an economic perspective on the Commission’s criticism that an ‘unduly strict and dogmatic’ application of the AEC principle could lead to underenforcement by ignoring the potential positive impact of not-(yet-)AECs and ‘contestability’ in dynamic competition (Section IV). We conclude with some final reflections on a way forward for the principle (Section V).

II. AEC principle as a workable effects-based approach in support of dynamic competition?

The AEC principle presumes that conduct by a dominant firm is anticompetitive only if it leads to the exclusion of a hypothetical, equally efficient competitor. In this section, we discuss the origins and logic behind the AEC principle (and tests), as well as its gradual evolution within the EU since the Commission’s 2008 guidance paper on enforcement priorities.

A. Origins and economic logic

The first application of the principle can be found in the original price–cost tests for predation, which began to be applied in the USA in the late 1970s. The origins of this ‘equally efficient competitor’ concept in the context of predatory pricing can be traced back to at least Posner (1973) and Areeda and Turner (1975).⁷ As early as 1973, US judge Richard Posner argued that⁸

It would seem that only two practices should be forbidden as predatory or (a better term perhaps) exclusionary pricing. The first is selling below short-run marginal cost […] [and] the second practice […] is selling below long run marginal cost with the intent to exclude a competitor […]. If there is intent to exclude, however, pricing below long-run marginal cost will have the purpose and likely effect of excluding an equally efficient competitor—equally efficient because if the ‘predator’ is more efficient he can and will exclude his competitor by charging a price equal to or higher than his own long-run marginal costs.

Posner essentially proposed a method based on direct efficiency considerations to guide the identification of potential exclusionary effects and reduce the overreliance on intent.

Early application of a price–cost test in the EU can be traced back to the AKZO predatory pricing case in 1991, where, in relation to pricing below average total cost, the Court of Justice stated⁹

Such prices can drive from the market undertakings which are perhaps as efficient as the dominant undertaking but which, because of their smaller financial resources, are incapable of withstanding the competition waged against them.

The basic logic of these tests was that if a dominant company sets prices below its own costs, AECs (i.e. competitors with the same costs as the dominant undertaking) would not be able to compete on merit.

The key merit of the AEC principle is that it allows for the prosecution of exclusionary abuse of dominance whilst recognising the importance of ‘appropriability’ in dynamic competition—i.e. that even dominant firms should be able to appropriate the gains from their efforts to reduce cost and innovate, without being hamstrung by the protection of smaller competitors per se. Put differently, the AEC principle in competition law ensures that the law does not shield less efficient rivals from competition—which would otherwise decrease the incentives of these rivals, but also the incumbent, to compete. As such, the application of the AEC principle can be seen as an instrument to move away from the protection of competitors per se and towards the protection of a dynamic competitive process in which firms that manage to reduce cost and innovate benefit the most.¹⁰

In addition to avoiding the protection of competitors per se, another key advantage of the AEC principle is that it promotes a degree of predictability in competition law enforcement and facilitates compliance by providing a practical analytical framework for self-assessment. The reason for this is that the AEC principle primarily relies on information that firms themselves have access to. In particular, under the AEC principle, dominant firms would need to consider the consequences of their conduct primarily for a hypothetical, equally efficient competitor—rather than for their actual competitors, whose cost structures and demand they are unlikely to know. Importantly, this also means that the efficiency level of potentially excluded competitors are in principle irrelevant for the application of the AEC principle: it is about whether a hypothetical AEC could still profitably compete—not whether actual competitors were so much less efficient that they were therefore excluded from the market.

B. From the 2008 guidance paper to the 2023 policy brief

The AEC principle was formally introduced in the EU in the 2008 European Commission guidance on enforcement priorities in applying Article 102 TFEU to abusive exclusionary conduct by dominant firms.¹¹ In this section, we reflect on the past recognition of the importance of appropriability in dynamic competition—as well as other notable features of the 2008 guidance that are currently being discussed again, including the seemingly artificial distinction between price- and non-price conduct, the specification of the AEC test, the AEC principle as a necessary condition to finding an infringement, and the benefit from protecting not-(yet-)AECs.

1. Appropriability and dynamic competition as motivation behind the AEC principle

The adoption of the 2008 guidance contributed to moving away from a formalistic approach to enforcing Article 102 towards a so-called ‘effects-based approach’, guided by economic analysis, and in which enforcement priorities were less based on rigid criteria but considered the potential effects of conduct on the competitive process.¹²

The guidance designates ‘anti-competitive foreclosure’ as ‘foreclosure leading to consumer harm’—as opposed to elimination of any competitor. Thus, the guidance clarifies that¹³

With a view to preventing anti-competitive foreclosure, the Commission will normally only intervene where the conduct concerned has already been or is capable of hampering competition from competitors which are considered to be as efficient as the dominant undertaking.

As noted by the Commission itself at the time, the AEC principle can be seen as part of the broader objective to focus antitrust investigations on the preservation of ‘consumer welfare’.¹⁴

Since the publication of the guidance, the idea and motivation behind the AEC principle have been discussed in several landmark court cases. In these cases, the EU courts often referred to the principle that vigorous competition that benefits consumers may by definition entail the elimination of competition from less efficient competitors. Recently, in Intel (2023), for example, the Court of Justice of the European Union (CJEU) reiterates that¹⁵

[I]t must be borne in mind that it is in no way the purpose of Article 102 TFEU to prevent an undertaking from acquiring, on its own merits, the dominant position on a market. Nor does that provision seek to ensure that competitors less efficient than the undertaking with the dominant position should remain on the market […]. Thus, not every exclusionary effect is necessarily detrimental to competition. Competition on the merits may, by definition, lead to the departure from the market or the marginalisation of competitors that are less efficient and so less attractive to consumers from the point of view of, among other things, price, choice, quality or innovation […].

However, whilst the principle has featured in case law, there has been a lack of consistency in its implementation in practice. Some commentators have argued that some of these judgements have not been in line with the spirit of the 2008 guidance and that they cast doubt, at least to a degree, on the future of the AEC principle and effects-based approach more generally.¹⁶

Recently, the 2023 policy brief recognises that the 2008 guidance contributed to moving away from a formalistic approach to enforcing Article 102 TFEU and towards an effects-based approach. It notes that a form-based approach may carry the risk of capturing false positives; however, it does not elaborate further on the reasons why such ‘false positives’ are actually problematic from the perspective of dynamic competition. Instead, the policy brief emphasises that ‘an overly rigid implementation of the effects-based approach could set the bar for intervention at a level that would render enforcement against practices that restrict competition unduly burdensome or even impossible, with serious negative consequences for consumers, the EU economy and society at large’.¹⁷ The flip side—i.e. the risk of over-enforcement and chilling effect on incentives to invest and innovate, and hence on dynamic competition—gets little attention.

2. Price versus non-price conduct

In addition to the introduction of the AEC principle to EU competition policy, other notable features of the 2008 guidance inform much of the current debate on the AEC principle and tests. For example, the guidance clearly separates price-based exclusionary conduct from non-price abuses, with AEC tests applying exclusively to the former.

As discussed later in this article, whilst an AEC test more naturally lends itself to pricing conduct, there is no reason, in principle, why it cannot also be developed for non-price conduct, such as self-preferencing or bundling. More generally, the notion of ‘less efficient competitor’ has been clarified in case law to refer not only to prices and costs but also to the other economic dimensions of competition. The 2023 Intel judgement, for example, designates less efficient competitors as ‘less attractive to consumers from the point of view of, among other things, price, choice, quality or innovation’.¹⁸ The policy brief also recognises this point.¹⁹

3. Specifying the AEC test

Additionally, the 2008 guidance does not define in detail the specific conducts to which an ‘AEC test’ is thought to be more suitable or how the test should be performed, other than providing general references to ‘economic data relating to cost and sales prices, and in particular whether the dominant undertaking is engaging in below-cost pricing’.²⁰ This has sparked a debate in case law as to when and how it is appropriate to refer to the AEC principle or perform an AEC test. Again, this is a point that is also recognised by the policy brief (and one that we discuss in more detail in the next section).²¹

4. AEC principle as a necessary condition

The 2008 guidance outlines that if it can be clearly demonstrated that an AEC could compete, ‘the Commission, will, in principle, infer that the dominant undertaking’s pricing conduct is not likely to have an adverse impact on effective competition, and thus, on consumers, and will therefore be unlikely to intervene’.²² In other words, the guidance presents the AEC principle effectively as a necessary condition for opening an investigation—and hence as a kind of ‘soft safe harbour’ that dominant firms, to some extent, can rely on for the self-assessment of compliance.

The opposite, according to the guidance, does not hold: if it can be shown that an AEC cannot profitably compete, then ‘the Commission will integrate this in the general assessment of anti-competitive foreclosure […], taking into account other relevant quantitative and/or qualitative evidence’.²³ The general assessment of foreclosure takes into account factors such as the position of the dominant undertaking, the conditions on the relevant market (e.g. entry and expansion, economies of scale and scope, network effects), the position of the dominant undertaking’s competitors, the position of customers and input suppliers, the extent and scope of the allegedly abusive conduct, the possible evidence of actual foreclosure, and direct evidence of exclusionary strategy. ²⁴

The 2023 policy brief, on the other hand, downplays this characterisation of the AEC principle (and tests) as a necessary condition. More precisely (although only discussed in the context of pricing abuse), the policy brief notes that ‘while notion of foreclosure of as-efficient competitors may be conceptually justified as a general proxy for intervention in pricing abuses, it is important to avoid an unduly strict and dogmatic approach to such a standard’.²⁵ The Commission considers that ‘[t]he fact that a dominant undertaking’s pricing conduct “passes” an AEC test should not be considered as a conclusive indication that such pricing conduct is not capable of negatively affecting competition’.²⁶ Moreover, in the context of the AEC principle in general, the policy brief notes that ‘in certain instances, genuine competition may also come from undertakings that are less efficient than the dominant firm’.²⁷

5. Protecting not-(yet-)AECs

The 2008 guidance already recognises that the AEC principle may not always be appropriate in the presence of important competitive pressure from not-(yet-)AECs. More specifically, the guidance notes that it may be beneficial to consumer welfare to protect also less-efficient competitors:²⁸

[T]he Commission recognises that in certain circumstances a less efficient competitor may also exert a constraint which should be taken into account when considering whether particular price-based conduct leads to anti-competitive foreclosure. The Commission will take a dynamic view of that constraint, given that in the absence of an abusive practice such a competitor may benefit from demand-related advantages, such as network and learning effects, which will tend to enhance its efficiency.

The 2023 policy brief emphasises this point further, particularly in markets where the emergence of AECs may not be possible because of the structure of the market, or where the relevant market is protected by significant barriers to entry.²⁹

However, this does raise the question of how to assess whether particular exclusionary conduct by a dominant firm is permissible—and perhaps even desirable, from the perspective of ensuring sufficient appropriability under dynamic competition—despite the fact that it leads to the exclusion of a not-AEC. And how can one ensure that such an assessment is still reasonably predictable and administrable, and hence suitable for supporting ex ante compliance efforts by companies?

C. Alternative principles or tests

In implementing an effects-based approach to abuse of dominance cases, the academic literature has offered two alternatives to classify exclusionary conducts: the welfare-balancing test and the no-economic-sense test.³⁰

The welfare-balancing test is more accurate than the AEC test (or principle). Similar to the competitive assessment in mergers, it aims to directly identify whether the conduct reduces competition to the detriment of consumers, without creating efficiencies that are sufficient to offset this detriment.³¹ Although it provides an accurate description of the anticompetitive effects of exclusionary conduct on the market as a whole, it still rests on ‘comparing the magnitudes [and credibility] of various effects to predict the likely overall impact on consumers’.³² Indeed, even though an argument could be made for its predictability properties, the welfare-balancing test appears weak in terms administrability precisely because measuring consumer welfare is not empirically clear-cut but ‘require[s] strong and often untestable assumptions that substantially affect the estimates’, rendering its application—particularly in terms of ex ante self-assessment—as error-prone.³³

At the other extreme there is the no-economic-sense test, which establishes the existence of anti-competitive exclusionary conduct by detecting an ‘improper intent’ that would make no economic sense but for its predisposition to eliminate or lessen competition.³⁴ The main benefit of the no-economic-sense test is that it is much more predictable and administrable—with companies only needing to reflect on the rationale of their own conduct. However, in many scenarios, certain behaviour can be commercially rational irrespective of whether it leads to anticompetitive exclusion. Setting below-cost prices to gain scale economies may be one example. Moreover, not all types of exclusionary abuse conduct are loss-making. For example, exclusionary loyalty rebates may involve discounted prices that are still above long-run marginal cost (as discussed in more detail later).

Compared to these alternatives, the AEC principle—and its associated tests—can be seen as a sort of middle ground, offering higher levels of administration and predictability that the welfare-balancing test, and more accuracy that the no-economic-sense test.

III. AEC tests as mathematical abstractions of the AEC principle

The AEC principle is a conceptual framework. However, its practical application and usefulness often depends on how the principle is translated into a formal ‘test’ based on observable evidence. To aid authorities and courts with the application of the AEC principle in specific cases, economists have developed various tests that abstract the general AEC principle into testable mathematical propositions—tailored to the specific conduct under consideration.

After reflecting on the relationship between the AEC principle and tests, we outline the different ways in which the AEC principle has been or could potentially be formulated into mathematical propositions, depending on the type of conduct. In the process, we show that the ease with which this can be done depends on the complexity of the conduct under consideration and the availability of data and that this is reflected in the way which authorities and courts have applied or accepted AEC tests in specific cases.

A. Relationship between the AEC principle and AEC tests

There are three important things to note about the relationship between the AEC principle and AEC tests that may be overlooked. First, both the AEC principle and AEC tests aim to assess the exact same thing: could a hypothetical AEC profitably compete given the conduct of the dominant firm? As such, there would be no basis to state, as a matter of principle, that the AEC principle would be relevant in any particular case, whereas the AEC test would not.

Second, if in any one particular case an AEC test is not considered feasible, this does not mean that the AEC principle itself is no longer relevant. The reason for this is that in any one particular case, the ease of robust implementation of the AEC test can vary substantially, depending on (i) the ability to translate the principle into a mathematical abstraction that captures the AEC principle given the specific conduct under consideration and (ii) the availability of data. If it proves impossible to formulate a sufficiently clear abstraction of the AEC principle in the context of a particular case, or to test it to the required standards, this does not mean that the AEC principle itself is no longer relevant; it only means that there is no mathematical abstraction that can give it weight.

Finally, it would be incorrect and potentially confusing to refer to the AEC test as a singular noun. AEC tests aim to abstract the AEC principle into a testable mathematical proposition, but this needs to be bespoke to the conduct under consideration. As such, the AEC test looks different depending on the conduct.

B. AEC tests under specific types of conduct

Whilst not offering an exhaustive list of potentially exclusionary conduct, we discuss below how the AEC test has been or could be formulated in specific cases—and identify exactly where the practical challenges lie.

1. Predation

The mathematical application of the AEC principle is most straightforward in the case of predatory pricing. In predatory pricing cases—e.g. AKZO (1991)³⁵ and Post Danmark I (2012)³⁶—the AEC principle and test would simply ask whether the price set by the dominant firm (⁠|$p$|⁠) is indeed lower than its marginal costs per product (⁠|$c$|⁠) plus relevant fixed costs (⁠|$F$|⁠) divided by total quantity (⁠|$q$|⁠):

$$\begin{equation} p<c+F/q \end{equation}$$

(1)

Much of the discussion then focuses on exactly which prices, costs, and quantities to consider and over which timeframe.³⁷

The main benefit of this formulation, as in any AEC test, is that it relies only on data that are readily available to the dominant company. This means that it is able to self-assess its own conduct ex ante, which in turn gives more legitimacy to any ex post infringement decision by authorities. Moreover, note again that it also means that the actual efficiency levels of potentially excluded competitors are in principle irrelevant here—as with the AEC principle more generally. The principle and tests are about whether a hypothetical AEC could still profitably compete—not whether actual competitors were so much less efficient that they were therefore excluded from the market.

Note, finally, what the AEC test does not actually do here: it does not provide any theory of harm or benefit, or proof thereof. It does not explain why a dominant company prices below cost, or what the conditions would be for this to be anti- or procompetitive. As such, as with the AEC principle more generally, a concurrent review of associated theories of harm and benefit would need to be considered as well.

2. Margin squeeze

In margin squeeze cases—e.g. Deutsche Telekom (2010)³⁸, TeliaSonera (2011)³⁹, and Telefónica (2012)⁴⁰—the allegation is that a vertically integrated firm with a dominant position upstream is raising the upstream input price and/or lowering the downstream retail price to squeeze the margin available to downstream rivals to the point that they are forced to exit the market.

The AEC principle and test would then ask whether the difference (i.e. margin) between (i) the downstream retail price charged by the vertically integrated firm (⁠|${p}_D$|⁠) and (ii) the upstream input price charged by the vertically integrated firm to downstream rivals (⁠|${p}_U$|⁠) is sufficient to cover the marginal costs per product (⁠|${c}_D$|⁠) and relevant fixed costs divided by total quantity (⁠|${F}_D/{q}_D$|⁠) of its own downstream operations:

$$\begin{equation} {p}_D-{p}_U<{c}_D+{F}_D/{q}_D \end{equation}$$

(2)

Note, importantly, that the AEC test in margin squeeze cases only considers the prices charged by the vertically integrated firm and its downstream costs. As with predation, the prices and costs of the potentially excluded competitor do not play any role—as the principle is to look at whether a hypothetical AEC is able to compete profitably. This aids the dominant company in its ability to self-assess. Additionally, note that the upstream cost of the vertically integrated firm does not play any role. This is because the aim here is to evaluate exclusion on the downstream market.

As with predation, much of the discussion focuses on exactly which cost benchmark to consider. As we also discuss in more detail later, this can be particularly relevant in network industries such as telecommunications, where economies of scale and supply-side externalities mean that nascent competitors are unlikely to have the same low costs as the dominant company. Instead of comparing to an AEC, one alternative would be to compare to a ‘reasonably’ efficient competitor, where ‘reasonably’ would need to be clearly articulated ex ante to ensure legal certainty and the ability for firms to self-assess.

Interestingly, this version of the AEC test may also be used to identify exploitative abuse of dominance in the presence of a more efficient downstream competitor, as it is able to detect all types of margin squeezes.⁴¹ Under exploitative conduct more specifically, the integrated firm sets its prices at such levels that allow it to capture the surplus introduced by a more efficient entrant, which remains in the market.⁴²

3. Rebates

In rebate cases—e.g. Intel (2009)⁴³ and Tomra (2012)⁴⁴—the allegation is that firms set conditional discounts such that customers are disincentivised to switch all or part of their new or substitutable demand to a competitor, as they would lose out on these conditional discounts.

Rebates have a ‘suction effect’ on customers when the rebates apply to a part of the purchases that customers have done or will in any case do from the dominant firm. The reason for that is that any competitor to the dominant firm will then have to set a price sufficiently low to also compensate the customer for the loss in discount experienced on this part of its ‘non-contestable’ demand (such as past purchases, as in the case of retroactive rebates).

The AEC principle and test would then ask whether the discounted price charged by the dominant firm (⁠|${p}_d$|⁠) is less that the average price a customer would end up paying if it switched the contestable share of its demand (⁠|$s$|⁠) to a competitor charging at cost (i.e. the marginal and fixed costs per product, so |$c+F/q$|⁠) whilst having to pay the undiscounted price (⁠|${p}_u$|⁠) over its uncontestable share of its demand (⁠|$1-s$|⁠):

$$\begin{equation} {p}_d<s\cdotp \left(c+F/q\right)+\left(1-s\right)\cdotp{p}_u \end{equation}$$

(3)

In such cases, even if a competitor would set its price equal to cost, the customer would not switch its contestable share of demand to this competitor—as doing so would lead to a higher average price.

To ease the assessment of the AEC test, it is possible to replace the cost benchmark |$c+F/q$| with what is called the ‘effective price’—i.e. the highest price that a competitor can set to still be competitive. Replacing |$c+F/q$| with effective price |${p}^{\ast }$| in the above formula (set now as an equality) and solving for |${p}^{\ast }$| provides:

$$\begin{equation} {p}^{\ast }=\frac{p_d-\left(1-s\right)\cdotp{p}_u}{s} \end{equation}$$

(4)

This provides the maximum price that a competitor can set in order to compete with the dominant firm—which decreases when the contestable share of demand |$s$| decreases. Whenever this price is below the relevant cost benchmark, an AEC would not be able to profitably compete.

Note that—unlike in predatory pricing—the dominant firm does not actually have to incur a loss to exclude an AEC, as long as the contestable share of demand is sufficiently low. The reason that a dominant firm may still make a profit despite setting an effective price that is below its own cost is intuitive: if a customer switches to a cheaper competitor, this means not only a lower price over the contestable share but also a higher price over the non-contestable share. This, in turn means that a competitor to the dominant firm has to offer a price that is even lower the actual, discounted price set by the dominant firm, and potentially even lower than the relevant cost benchmark of the dominant firm.

Again, much of the discussion then focuses on exactly which prices and costs to consider. Moreover, in rebate cases, much of the discussion will be on exactly how rebates are structured (in many practical cases, it is less straightforward as the simple numerical example provided here) and what the non-contestable share of demand would be.⁴⁵

4. Bundling

In bundling cases—e.g. Google Android (2018)⁴⁶—the allegation is that a discount applied to the joint purchase of two products by a firm dominant in one of those products forecloses the market to a firm that is only active in the competitive product.

The AEC test in bundling cases—also referred to as the attribution test—involves the attribution of the entire discount received on the bundle to the competitive product.⁴⁷ Taking product |$A$| as the competitive product and |$B$| as the product in which the dominant firm is dominant, the attribution test asks whether the standalone price charged by the dominant firm for product |$A$| (⁠|${p}_A$|⁠) minus the total discount received when buying the bundled product (⁠|${p}_{AB}$|⁠) instead of both products separately (⁠|${p}_A+{p}_B$|⁠) is sufficient to cover the marginal costs per product (⁠|${c}_A$|⁠) and relevant fixed costs divided by total quantity (⁠|${F}_A/{q}_A$|⁠) of selling product |$A$|⁠:

$$\begin{equation} {p}_A-\left[\left({p}_A+{p}_B\right)-{p}_{AB}\right]<{c}_A+{F}_A/{q}_A \end{equation}$$

(5)

which can be straightforwardly rewritten as

$$\begin{equation} {p}_{AB}-{p}_B<{c}_A+{F}_A/{q}_A \end{equation}$$

(6)

Whenever this condition holds, it means that an AEC that only sells the competitive product |$A$| is unable to profitably compete by setting a price that incentivises the consumer to forego the bundle discount |$\left({p}_A+{p}_B\right)-{p}_{AB}$|⁠.Whilst in principle straightforward, one challenge in the AEC test in bundling cases (as indeed in other exclusionary cases) is accounting for the relevance of scale economies. In particular, the test does not take into account cases where ‘a bundled or loyalty discount exceeds cost but has worsened rival efficiency by denying it economies of scale’.⁴⁸

5. Self-preferencing

Finally, the AEC principle (and potentially test) may also be applied in non-pricing conduct such as self-preferencing. In self-preferencing cases—e.g. Google Shopping (2017)⁴⁹—the AEC principle would ask whether the reduction in demand of a hypothetical competitor with the same marginal cost still allows it to operate profitably.

Unlike in the pricing cases discussed above, the issue in non-pricing conduct is that there is no obvious price–cost comparison to make. Although the AEC principle is agnostic on the parameter of competition, the AEC tests as developed above do rely on price as the relevant parameter. However, this does not mean that a mathematical formalisation of the AEC principle into a test is in principle impossible. For example, one can also consider whether the conduct leads to a decrease in demand that would have made an AEC loss-making.⁵⁰

The mathematical formalisation of this actually takes a similar framework as margin squeeze. However, rather than looking (only) at the relevant upstream and downstream price |${p}_U$| and |${p}_D$| and at downstream marginal cost |${c}_D$|⁠, one would also need to consider how the conduct affects the downstream quantity |${q}_D$|⁠, and hence the base of the downstream per-product fixed cost allocation |${F}_D/{q}_D$|⁠.More specifically, the AEC test in this case would take the margin squeeze formulation as discussed above, but now replaces quantity |${q}_D$| with a hypothetical quantity |${q}_D^{-}$|⁠, which captures the quantity that the dominant firm would have sold if it would have been the one disadvantaged by the self-preferencing.

$$\begin{equation} {p}_D-{p}_U<{c}_D+{F}_D/{q}_D^{-} \end{equation}$$

(7)

Note that the hypothetical quantity |${q}_D^{-}$| is not actually observed. However, it can be inferred when you observe actual quantity and have reasonable estimates of (i) the benefit enjoyed by the dominant firm from the self-preferencing and (ii) the reduction in sales suffered by competitors.

As in the margin squeeze case, the AEC test in self-preferencing cases can come with challenges when economies of scale play an important role. That is especially relevant for platform economies characterised by significant direct and indirect network externalities—i.e. demand-side externalities. However, note that the above formulation of the AEC test actually adjusts the quantity to account for demand-side effects of the conduct as well, and as such at least in principle would account explicitly for the relevant economies of scale. Additionally, the AEC test may be corrected by applying a supplementary minimum efficiency scale test, and testing whether the self-preferencing conduct has prevented existing or nascent rivals from maintaining or reaching that scale.⁵¹

C. Conclusion on the application AEC test for different types of conduct

As was set out in this section, even where the AEC principle is seen as appropriate, operationalising can create challenges, and the question of whether numerical AEC tests are useful and in which type of abuses is in itself evolving with case law and new economic thinking.

Broadly speaking, there appears to be broad consensus that the AEC test can be usefully applied to certain price abuses, such as margin squeeze and predatory pricing cases, because the translation of the principle into a mathematical abstraction is less controversial and the necessary input variables are more readily available. In other pricing cases, however (and particularly in rebates), the application of the AEC test is more challenging—not for principled reasons, but for practical ones. For example, defining the non-contestable share is generally not obvious. And for non-pricing conduct, such as self-preferencing, economic thinking is still very much developing.

Moreover, a common challenge in any AEC test is how to determine, in a consistent and transparent way, an alternative reasonable cost or demand benchmark, when supply- or demand-side economies of scale imply that the as-efficient benchmark is not considered a reasonable comparison.

Finally, it bears remembering that AEC tests—as the principle in general—do not actually provide or prove any particular theory of harm or benefit. As such, a concurrent review of associated theories of harm and benefit (including motives and conditions) would need to be considered as well.

IV. An economic perspective on the future of the AEC principle

Having reflected on the origins and economic logic of the AEC principle and tests, and the relationship between the AEC principle and its tests, in this section, we reflect further on the key developments in the Commission’s approach to the principle—and in particular on the role of not-(yet-)AECs.

A. Key developments in the Commission’s approach

The 2023 policy paper explains that the Commission’s enforcement practice and priorities need to evolve to take into account developments such as ‘increasing evidence of market concentration at macro-economic level and growing importance of digital markets and services […]’.⁵² It also explains that ‘[i]n such fast-moving markets, often featuring strong network effects and “winner-takes-all” dynamics, it is paramount to ensure an effective and swift enforcement of Article 102 TFEU to intervene before tipping occurs and entrenched market positions are created’.⁵³

In relation to the notion of anticompetitive foreclosure, the Commission also amended the 2008 guidance to clarify that the notion of anticompetitive foreclosure extends beyond full market exclusion:⁵⁴

such notion should be understood as not only referring to conduct that can result in the full exclusion or marginalisation of actual or potential competition, but also to conduct that ‘has the effect of hindering the maintenance of the degree of competition still existing in the market or the growth of that competition’, in other words, conduct that weakens an effective competitive structure even without necessarily producing the full exclusion or marginalization of competitors.

The expression ‘has the effect of hindering the maintenance of the degree of competition still existing in the market or the growth of that competition’ is quoted from Servizio Elettrico Nazionale (2022).⁵⁵ The Commission also cites recent judgements, which, in its view, confirm that the relevant test to be applied to establish an abuse is whether the conduct adversely impacts an effective competitive structure.⁵⁶

As to the AEC principle, the Commission explains that it does not see it as useful in markets where barriers to entry and expansion (e.g. due to economies of scale or network effects) lead to the emergence of dominant firms, as challengers may not be expected to be ‘as efficient’ as the incumbent. Yet, their presence in the market may nonetheless be expected to benefit consumers, for example, if there is a prospect that they could become ‘serious challengers for all or a significant portion of the dominant firms’ customers in the future’.⁵⁷ Mandating evidence of foreclosure of AECs would then likely lead to underenforcement. In support, the Commission refers to case law that supports the notion that genuine competition may also come from undertakings that are less efficient than the dominant firm, and thus, that the competitive constraint thereby exerted may warrant protection under Article 102 TFEU. The Commission specifically cites Post Danmark II (2015)⁵⁸ and Unilever (2023),⁵⁹ where the emergence of AECs may not be possible because of the ‘structure of the market’, or because the relevant market is protected by significant barriers.⁶⁰

Finally, in relation to AEC tests, the Commission considers that ‘a generalised use of the AEC test to determine which cases of price-based exclusionary conduct to pursue as a matter of priority is not warranted and, if such test is carried out, its results should in any event be assessed together with all other relevant circumstances’. The Commission considers this approach to be consistent with case law highlighting that the application of an AEC test is not legally required to prove an abuse.⁶¹

B. Is the AEC principle still fit for purpose?

As was set out at the outset of this paper, the AEC principle and related tests are part of the broader effects-based approach to enforcement against exclusionary conduct. Although we acknowledge that there is an ongoing debate as to the merits of the effects-based approach itself, as hinted in the Commission’s own caveats to its continued application, we start from the premise that an effects-based approach to the assessment of exclusionary abuse remains appropriate. This is because no compelling case has been advanced in the economic literature that any alternative ‘form-based’ test can deliver the kind of precision that would be required to justify reliance by enforcers on such simplistic heuristics.

Thus, the key question is how to effectively operationalise the effects-based approach; in other words, identifying the tools and principles that can best help policymakers to identify correctly exclusionary conduct that is detrimental to consumer welfare, striking a balance between the risks of both over- and underenforcement in supporting dynamic competition. Specifically, we explore whether the AEC principle is still fit for purpose, and the extent to which it should be relied on for identifying anticompetitive conduct, in the context of recent policy debates. Two points are particularly relevant here.

First, we note that just as a formalistic use of ‘form-based’ criteria is often rejected as too blunt a tool, a formalistic approach to the application of the AEC principle can be inappropriate in many circumstances, even where the AEC principle remains a useful framework for thought. The AEC concept and related tests do not in themselves prove or disprove any theory of harm (or benefit), but merely illustrate whether the conduct is capable of excluding a hypothetical AEC. As such, the principle remains agnostic as to the possible anti- or procompetitive incentives behind the conduct. This in turn means that the AEC principle should not be considered in isolation, but within the broader fact and evidence base, and considering aspects such as the likelihood of successful monopolisation, the degree of market power, the degree of market coverage, and so on. At the same time, it is important that sufficient guidance is provided to achieve a degree of predictability of enforcement and to ensure that firms know how to comply with competition law. In the context of exclusionary abuse in particular, where the conducts and notion of abuse itself are not well defined and have evolved with case law, clear principles such as that of the AEC can provide a degree of legal certainty.

Second, in considering the weight that policymakers should give to the AEC principle and tests (or any other tests), it is important to acknowledge that no single answer or approach is inherently superior, and the selected approach needs to take into account a range of factors including developments in case law, new economic evidence, and technological change as well as the policymakers’ own objectives and priorities. The precise implementation and calibration of enforcement is an exercise in balancing the risk of false positives (known, in statistical terms, as ‘type I error’, that is, the risk of finding anticompetitive conduct where none exists) with the risk of false negatives (the ‘type II error’, i.e. failing to identify anticompetitive conduct where it does exist). Whilst the risk appetite over one or the other type of error can vary over time, it is important that the pendulum does not swing too far in any one direction.

Thus, there is no one-size-fits-all answer to the question of whether and to what extent the AEC principle and tests remain appropriate and useful in discerning anticompetitive conduct. The appropriateness and usefulness will vary depending on the circumstances of the case, the nature of the conduct, and the broader evidence. The usefulness of the AEC principle in assessing exclusionary abuse boils down to whether, in principle, one expects that the protection and presence in the market of less efficient competitors is detrimental to consumer welfare.

As was illustrated in Royal Mail (2021),⁶² entry can affect consumer welfare through its effect on

• Allocative efficiency—this occurs when consumers pay a market price that reflects the marginal cost of production. Normally, a greater number of competitors will improve consumer welfare by exerting downward pressure on prices.

• Dynamic efficiency—this occurs when the output of firms, from a given amount of input, is the greatest it can be. A greater number of competitors can strengthen incentives for all firms in the market to innovate and cut costs. However, this is not always the case, as will be explained further below.

• Productive efficiency—this occurs when a firm is combining resources in such a way as to produce a given output at the lowest possible average total cost.

If a less efficient competitor enters, productive efficiency will be expected to decrease, as the same output is produced in the industry but with higher total industry costs. However, entry could still have a positive impact on allocative and dynamic efficiency, e.g. by constraining an incumbent’s ability to raise prices and/or by spurring innovation. In such a scenario, although outcomes will be worse when compared to a scenario with ‘efficient entry’, it can be argued that it is not a priori clear whether they would be worse with respect to a ‘no entry’ scenario, and the answer would depend on which effect dominates. However, as will be elaborated further below, the effect of entry on dynamic competition can also be ambiguous, and depends on the circumstances of each market; this adds further complexity when trying to identify cases in which ‘inefficient’ entry can be expected to have a net positive impact on consumers.

C. Should competition law protect not-(yet-)AECs to ensure sufficient contestability?

The notion that ‘inefficient’ entry can be net-positive for consumers has gained traction with enforcers and in case law, particularly in industries with large scale or network effects, which tend to be dominated by large firms and entrants would not be expected to be as efficient as the incumbent. Much emphasis is thus placed on the notion, that whilst not-yet-as-efficient as the incumbent, potential entrants could become challengers in the future and constrain the dominant firm’s future conduct, and thus, competition law should afford them protection.

Whilst the Commission in particular has recently placed renewed emphasis on the concept of not-yet-AECs, especially in the context of digital markets, it is worth noting that this is not a novel concept, and it is routinely debated in cases where the AEC principle is translated into tests. For example, in the context of margin squeeze cases in telecoms, much debate has taken place on the extent to which, in formulating an AEC test, one should consider the costs of lower-scale entrants instead of the dominant firm’s cost.⁶³ Thus, the not-yet-as-efficient principle can also be thought of as an extension or policy-based adjustment to the strict adherence to an AEC principle.

However, whilst economic theory recognises that high-cost firms can impose competitive constraints, it is also well understood—in theory and in policy—that artificially keeping inefficient firms in the market and preventing dominant firms from competing strongly may not be in the interest of efficiency and consumers in the longer term.

This debate is not new, and it can be captured by the wider discussion in the economic literature of the effect on competition on innovation. This is encapsulated by the seemingly conflicting theories of Schumpeter and Arrow dating back as far back as the early 20th century. On the one hand, Arrow argues for the intuitive concept that more competition will lead to more innovation.⁶⁴ On the other hand, Schumpeter argues that, as long as the market remains ‘contestable’ (i.e. the incumbent can be displaced if it does not maintain high levels of innovation and investment), less competition can lead to more innovation. This is because the incumbent would be rewarded for its innovation through higher margin, and it would have an incentive to continue to invest to maintain its position.⁶⁵

This seeming contradiction only highlights that the relationship between innovation (an important aspect of consumer and societal welfare) and competition is more complex than it may appear at first glance, and that it depends on the characteristics of the market, and how these affect firms’ ability and incentives to innovate. Shapiro attempts a synthesis of the Schumpeter’s and Arrow’s perspectives by describing (albeit in the context of mergers) incentives to innovate as being influenced by three principles: (i) ‘contestability’, which focuses on the ‘extent to which a firm can gain profitable sales from its rivals by offering greater value to customers’; (ii) ‘appropriability’, which focuses on the ‘extent to which a successful innovator can capture the social benefits resulting from its innovation’; and (iii) ‘synergies’, which emphasises that ‘firms typically cannot innovate in isolation’.⁶⁶

Whilst competition authorities have expressed concerns that some markets, especially in the digital and technology spheres, tend to be dominated by large players, where contestability at platform level can be seen as lacking, at least in the short term, it is also important not to lose sight of the role of the appropriability principle in producing desirable outcomes for consumers and society, to the extent that innovation requires high levels of initial and ongoing investments, as well as the role of synergies, whereby value is created by ‘systems’ that are made up of many individual components.

Thus, the conundrum remains that the nature of competition on the merits in many industries is likely to entail the marginalisation of less efficient rivals. Altering the standard in a way that excessively favours inefficient entry may have unintended consequences that are detrimental to consumers, especially in the medium to long term, by chilling innovation. Also, whilst maintaining contestability in markets is a legitimate policy objective, the question arises as to whether full contestability in the short term is desirable or whether the focus should be on preserving medium- and longer-term contestability, to avoid the risk that valuable innovation is lost in the first place. Indeed, whilst markets with scale and network effects may be prone to giving rise to market power, these same effects can also work in the opposite direction, as a lack of continued investment on the part of the incumbent can open the way to challengers and lead to the unravelling of the pre-existing system.

D. Setting ex ante standards for effective ex post enforcement

There is also a question as to whether ex post enforcement is the right tool to achieve those objectives as opposed to other policy initiatives. Ex post enforcement is notably more difficult and unpredictable when the standards that companies should rely on for self-assessing their conduct are unclear. Thus, to the extent that more weight is given to the objective of preserving entry, there is much to be gained from the competition authorities tightly defining the conditions under which an entrant that is ‘not-as-efficient’ is nonetheless seen as worthy of protection under competition law. In other words, there is no sidestepping the issue that defining the boundary between anticompetitive foreclosure and vigorous competition is complex and case-specific; moving the bar away from stricter adherence to the AEC principle should be accompanied by a clear-enough re-drawing of such boundaries.

Moreover, it is hard to think of an alternative approach to exclusionary abuse of dominance that does not retain the key merit of the AEC framework—i.e. its focus on protecting competition rather than competitors per se and its positive effects on predictability and compliance. Indeed, other approaches such as the no-economic-sense test and the welfare-balancing test have been considered by the Commission and rejected in the past, either because they could not easily be generalised to all types of abuses or because they were too vague. There is still a need to conduct thorough effects-based analysis to determine whether the conduct is structurally affecting competition to the detriment of consumers, as opposed to harming individual competitors.

V. Conclusions

In the 2023 policy brief, the Commission appears to usher in a potentially more interventionist approach, with less reliance on the AEC principle and tests, and broadly cautions that an ‘overly rigid’ implementation of the effects-based approach could make enforcement ‘unduly burdensome or even impossible’. The Commission is said to be in pursuit of a ‘dynamic and workable effects-based approach’ to exclusionary abuse of dominance cases,⁶⁷ one which recognises ‘[…] increasing evidence of market concentration at macro-economic level and growing importance of digital markets and services […]’,⁶⁸ and the importance of preserving entry and contestability by not-yet-AECs.

Whilst the attraction and the potential merits of adjusting the AEC framework to accommodate potential challengers exist, and although economic theory also recognises that less efficient firms can impose competitive constraints, it is also well understood, both in theory and in policy, that artificially keeping inefficient firms in the market and preventing dominant firms from competing strongly could reduce incentives to innovate and may not be in the interest of efficiency or consumers longer term. This may also apply to digital and technology markets, which may be seen as being characterised by high levels of innovation and investment, and as giving rise to synergies.

Thus, whilst there is a place for adjustments and evolutions to the AEC principle, the conundrum remains that the nature of competition on the merits in many industries is likely to entail the marginalisation of ineffective rivals. Therefore, there is still a need to conduct thorough, effects-based analyses to limit the risk of type I errors (i.e. finding abuse where none exists). Rejecting the AEC standard and tests outright, or ignoring its merits, is unlikely to be helpful. The AEC principle, with or without adjustments, and with all the caveats and nuances, remains a useful economic framework to conceptualise exclusionary abuse and the boundaries of enforcement.

Footnotes

1

Guidance on the Commission’s enforcement priorities in applying Article 82 of the EC Treaty to abusive exclusionary conduct by dominant undertakings (2009) OJ C45/7.

2

Commission, ‘Amendments to the Communication from the Commission Guidance on the Commission’s enforcement priorities in applying Article 82 of the EC Treaty to abusive exclusionary conduct by dominant undertakings’ 1923 (Brussels, 27 March 2023)

3

Linsey McCallum, Inge Bernaerts, Massimiliano Kadar, Johannes Holzwarth, David Kovo, Marie Lagrue, Edouard Leduc, Luca Manigrassi, Jorge Marcos Ramos, Isabel Pereira Alves, Vera Pozzato and Pinelopi Stamou, ‘A dynamic and workable effects based approach to abuse of dominance’ (2023) 1 Competition Policy Brief. Although the policy brief is published under the Commission label, it notes that it does not reflect the official position of the European Commission, with the views expressed only attributable to its listed authors.

4

McCallum et al. (n 3), 4.

5

McCallum et al. (n 3), 4.

6

McCallum et al. (n 3), 4.

7

Richard A. Posner, ‘Exclusionary Practices and the Antitrust Laws’ (1973) 41 U. Chi. L. Rev, 506–535; Phillip Areeda and Donald F. Turner, ‘Predatory pricing and related practices under Section 2 of the Sherman Act’ (1975) 88.4 Harvard Law Review; Martin Mandorff and Johan Sahl, ‘The Role of the “Equally Efficient Competitor” in the Assessment of Abuse of Dominance’ (2013) 12 Competition LJ;

8

Posner (n 7), 518–519.

9

Case C-62/86, AKZO Chemie BV v Commission, EU:C:1991:286, para. 72. See also Case T-271/03, Deutsche Telekom v Commission, EU:T:2008:101, para 194.

10

Carl Shapiro, ‘Competition and Innovation: Did Arrow Hit the Bull’s Eye?’, in Josh Lerner and Scott Stern (eds), The Rate and Direction of Inventive Activity Revisited (University of Chicago Press, 2012)

11

Guidance on enforcement priorities in applying Article 82 to abusive exclusionary conduct by dominant undertakings (n 1).

12

Richard Whish and David Bailey, Competition Law (8th ed, Oxford University Press 2015), section 5.2.

13

Guidance on enforcement priorities in applying Article 82 to abusive exclusionary conduct by dominant undertakings (n 1), para. 23.

14

Commission, ‘Antitrust: consumer welfare at heart of Commission fight against abuses by dominant undertakings’ IP/08/1877 (Brussels, December 2008)

15

Case C-377/20, Servizio Elettrico Nazionale SpA and Others v Autorità Garante della Concorrenza e del Mercato and Others, EU:C:2022:379, para 45, 73; Case C-413/14, Intel Corp. Inc. v Commission, EU:C:2017:632, paras 133–134. See also, among others, Case C-209/10, Post Danmark A/S v Konkurrencerådet, EU:C:2012:172, paras 21–22; and Case C-680/20, Unilever Italia Mkt. Operations Srl v Autorità Garante della Concorrenza e del Mercato, EU:C:2023:33, para 37.

16

See, for example, Derek Ridyard, ‘Calibration and consistency in Article 102: Effects-based enforcement after the Intel and Post Danmark judgments’ (2016) 3 Art. N° 80,734 Concurrences, 29–38.

17

McCallum et al. (n 3), 4.

18

Intel (n 15), paras 133–134.

19

McCallum et al. (n 3), footnote 48.

20

Guidance on enforcement priorities in applying Article 82 to abusive exclusionary conduct by dominant undertakings (n 1), para 25.

21

McCallum et al. (n 3), section III.C

22

Guidance on enforcement priorities in applying Article 82 to abusive exclusionary conduct by dominant undertakings (n 1), para. 27.

23

Guidance on enforcement priorities in applying Article 82 to abusive exclusionary conduct by dominant undertakings (n 1), para. 27.

24

Guidance on enforcement priorities in applying Article 82 to abusive exclusionary conduct by dominant undertakings (n 1), para. 20.

25

McCallum et al. (n 3), 5.

26

McCallum et al. (n 3), 7.

27

McCallum et al. (n 3), 5.

28

Guidance on enforcement priorities in applying Article 82 to abusive exclusionary conduct by dominant undertakings (n 1), para. 24.

29

McCallum et al. (n 3), 5.

30

Xingyu Yan and Hans Vedder, ‘Minimum Efficient Scale, Competition on the Merits, and The Special Responsibility of a Dominant Undertaking’ (2023) 19.1 Journal of Competition Law & Economics.

31

Gunnar Niels, Helen Jenkins, and James Kavanagh. Economics for competition lawyers (3rd edn, Oxford University Press 2023) para. 5.27; Keith N. Hylton, ‘The Law and Economics of Monopolization Standards’, in Keith N. Hylton (ed), Antitrust Law and Economics (2nd ed, Edward Elgar Publishing 2010).

32

Steven C. Salop, ‘Exclusionary conduct, effect on consumers, and the flawed profit-sacrifice standard’ (2005) 73 Antitrust LJ, 331–332.

33

Gregory J. Werden, ‘Identifying exclusionary conduct under Section 2: The no economic sense test’ (2004) 73 Antitrust LJ, 20.

34

John Vickers, ‘Abuse of Market Power’ (2005) 115 The Economic Journal.

35

Akzo (n 9).

36

Post Danmark (n 15).

37

Niels, Jenkins, and Kavanagh (n 31), section 5.3.

38

Case C-280/08, Deutsche Telekom v Commission, EU:C:2010:603.

39

Case C-52/09, Konkurrensverket v TeliaSonera Sverige AB, EU:C:2011:83.

40

Case T-336/07, Telefónica and Telefónica de España v Commission, EU:T:2012:172.

41

Germain Gaudin and Despoina Mantzari, ‘Margin squeeze: an above-cost predatory pricing approach’ (2016) 12.1 Journal of Competition Law & Economics, 165.

42

Bruno Jullien, Patrick Rey, and Claudia Saavedra, ‘The Economics of Margin Squeeze’ (2014) CEPR Paper No. DP9905.

43

Intel (Case COMP/C-3/37.990) Commission Decision of 13 May 2009 relating to a proceeding under Article 82 of the EC Treaty and Article 54 of the EEA Agreement.

44

Case C-549/10 P, Tomra Systems ASA and Others v Commission, EU:C:2012:221.

45

Oxera, ‘Intel and the AEC test: “Do. Or do not. There is no try.”’ (Agenda, February 2022). A discussion on this is also provided in McCallum et al. (n 3), section III.C.

46

Google Android (Case AT.40099) Commission Decision of 18 July 2018 relating to a proceeding under Article 102 of the Treaty on the Functioning of the European Union (the Treaty) and Article 54 of the EEA Agreement.

47

Niels, Jenkins, and Kavanagh (n 45), section 5.8.8.

48

Einer Elhauge. ‘Tying, bundled discounts, and the death of the single monopoly profit theory’ (2009) 123.2 Harvard Law Review.

49

Google Search (Shopping) (Case AT.3970) Commission Decision of 27 June 2017 relating to a proceeding under Article 102 of the Treaty on the Functioning of the European Union (the Treaty) and Article 54 of the EEA Agreement.

50

Germain Gaudin and Despoina Mantzari, ‘Google Shopping and the As-Efficient-Competitor Test: Taking Stock and Looking Ahead’ (2022) 13.2 Journal of European Competition Law & Practice.

51

Yan and Vedder (n 30), 142–144.

52

McCallum (n 3), 2. Also, see Commission Communication ‘A competition policy fit for new challenges’, COM(2021)713, p.2, 11 and 12.

53

McCallum (n 3), 2. Also, see press release of 3 December 2008, Antitrust: consumer welfare at heart of Commission fight against abuses by dominant undertakings, IP/08/1877.

54

McCallum (n 3), 4.

55

Servizio Elettrico Nazionale (n 15), paras 44, 68.

56

Servizio Elettrico Nazionale (n 15), para 44; Unilever (n 15), para 36.

57

McCallum (n 3), 5.

58

Case C-23/14, Post Danmark A/S v Konkurrencerådet, EU:C:2015:651, paras 59–60.

59

Unilever (n 15), para 57.

60

McCallum (n 3), 5.

61

Specifically, the Commission refers to Unilever (n 15), para. 62; Post Danmark (n 58), para. 61; and Case T-604/18, Google and Alphabet v Commission, EU:T:2022:541, para 643.

62

Case C3/2020/0151, Royal Mail v Ofcom, EWCA Civ 669.

63

Niels, Jenkins, and Kavanagh (n 31), section 5.7.3.

64

Kenneth J. Arrow, ‘Economic welfare and the allocation of resources for invention’, in Universities-National Bureau Committee for Economic Research and Committee on Economic Growth of the Social Science Research Council, The Rate and Direction of Inventive Activity: Economic and Social Factors (Princeton University Press 1962).

65

Joseph A. Schumpeter, Capitalism, Socialism, and Democracy, (3rd edition, Harper Perennial 1942)

66

Shapiro (n 10).

67

McCallum et al. (n 3), 4

68

McCallum et al. (n 3), 2. Also see Commission, ‘Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions: A Competition Policy fit for new challenges’ COM(2021)713 (Brussels, 18 November 2021), 2 and 11–12.

Author notes

Oxera Consulting LLP, United Kingdom. e-mail: [email protected]

Oxera Consulting LLP, Germany. e-mail: [email protected]

Utrecht University, the Netherlands, and Oxera Consulting LLP, the Netherlands. e-mail: [email protected]

We are grateful to Gunnar Niels and Richard Whish for valuable comments and discussions. Any opinions or errors remain our own. This article represents our views only and not necessarily those of our affiliations. Oxera advices, amongst other, companies, lawyers, and policymakers on economic issue connected with competition—including in cases related to exclusionary abuse of dominance. We have not received any funding for writing this article and have no further interests to disclose