9 min read

Introduction

The last two years have seen a series of landmark rulings from the European Court of Justice (CJ) in seminal Article 102 cases, including Google Shopping, Android Auto and most recently Google Android. Ostensibly clarifying some of the most contentious aspects of Article 102 enforcement policy and practice, the rulings in fact fuel uncertainty around key concepts, significantly expand the scope of the prohibition and impose more exacting standards on dominant undertakings. This briefing examines the key aspects of these rulings and their implications.

Meaning of “abuse”

Following years of debate over the precise formulation of the concept of “abuse”, the recent CJ rulings have coalesced around a test with two cumulative conditions: a dominant undertaking’s conduct must (i) use methods which depart from competition on the merits, and (ii) have the actual or potential effect of restricting competition. Both concepts have been at the heart of the recent rulings, with the CJ taking an increasingly expansive approach to each.

Conduct departing from competition on the merits

The notion of “competition on the merits” remains far from settled. Recent CJ rulings – in particular SEN and Google Android – provide limited guidance at the extremes:

  •  Conduct which has the effect of broadening consumer choice – by placing new goods on the market or by increasing the quantity or quality of goods already on offer – should fall squarely within the scope of competition on the merits.
  • Conversely, conduct which holds no economic interest for the dominant firm other than eliminating its competitors – so as subsequently to raise prices – falls outside competition on the merits.

But for conduct not falling within either extreme, considerable debate and uncertainty remain.

For example, the CJ has held that practices that rely on resources or means inherent to the holding of a dominant position, such that they cannot be adopted or replicated by a hypothetical as-efficient competitor, also fall outside competition on the merits. While this approach makes sense on the facts of SEN – where it applied to very specific circumstances in which the “resources” were derived from a statutory monopoly – what it means in other contexts is less clear. It arguably opens the door to the European Commission taking an overly broad view of the relevant “resources” (for example, treating demand won through competition on the merits as inherent to the holding of the dominant position) in an attempt to penalise dominance itself. This is at odds with longstanding case law from the European courts, reiterated in SEN itself, that Article 102 does not sanction the existence of a dominant position in and of itself.

According to the CJ in Google Shopping, whether conduct departs from competition on the merits depends on the specific circumstances of the case. In Google Shopping, while the CJ found that self-preferencing by a dominant firm does not necessarily depart from competition on the merits, it did on the facts of the case, given the relevant market characteristics and specific circumstances. This affords authorities considerable latitude to find that conduct departs from competition on the merits, while providing very little practical guidance for dominant or potentially dominant undertakings seeking to ensure compliance.

And while there are categories of conduct for which analytical frameworks exist to determine whether they depart from competition on the merits – including exclusive dealing, tying and bundling, predatory pricing, margin squeeze and refusal to supply – this list has been expanded following Android Auto and Google Shopping to include access restrictions and self-preferencing.

Access restrictions and self-preferencing

Access restrictions and self-preferencing have come under increasing Article 102 scrutiny with the rise of digital platforms. Dominant firms have sought to rely on longstanding case law – known as the Bronner case law – holding that dominant firms are only obliged to grant access to their infrastructure in very limited circumstances, namely where such access is indispensable to allow effective competition in a neighbouring market.

But in Android Auto, the CJ held that dominant firms may also be required to grant access in certain situations where that access is not indispensable, provided that doing so would not dampen investment incentives or alter the fundamental economic model of the dominant firm’s infrastructure (and absent security, integrity or other technical reasons why doing so would not be feasible). On the facts of Android Auto, the CJ found that the infrastructure in question had been developed with a view to third-party use, such that requiring Google to grant access would not adversely affect its investment incentives or economic model.

Even in Android Auto, where the facts suggested access would not undermine the economic model, the outcome is somewhat perverse – requiring a dominant firm that developed an “open” platform to provide access to actual and potential rivals, including by building the technology to enable interoperability. Coupled with the CJ’s ruling in Google Shopping, that access must (depending on the “specific circumstances”) also be granted on equal terms with the dominant firm’s own downstream services to avoid allegations of self-preferencing. In contrast, firms that develop “closed” platforms for their own use may continue to deny rivals access, unless access is indispensable and refusing to do so would eliminate all effective competition.

Android Auto leaves a number of open questions. It has introduced an exception to the strict Bronner case law, imposing a lower standard for mandating access to a platform which is “open”. But it is not clear what will and will not constitute an “open” platform for these purposes. In particular, although Android Auto specifically recognises the importance of preserving investment incentives and the dominant undertaking’s economic model, these cannot be neatly aligned with a binary “open” vs “closed” distinction. And while according to the CJ access need only be granted where the access-seeker’s product actually or potentially competes with the dominant firm, this might not always be obvious. The CJ also seems to take a narrow view of when interoperability restrictions might be justified, referring only to situations where such interoperability would compromise the platform’s security or integrity or otherwise be technically impossible.

Ahead of judicial clarification on these issues, dominant or potentially dominant firms should proceed with caution, conducting careful analysis before opening any feature or platform that is currently closed, and clearly documenting the economics of any decision to open, so as to maximise the chances of resisting other types of access requests. Where a feature or platform is already open, there may be arguments that the requested access does not relate to the currently open feature, or that the access-seeker does not actually or potentially compete with the dominant firm. If these arguments are not available (such that Android Auto applies) a dominant firm should consider whether there are security/integrity or other technical justifications for refusing access.

Actual or potential effects on competition

The CJ’s expansive approach to Article 102 is also evident in its treatment of the requirement that conduct must have actual or potential effects on competition in order to be abusive. In particular, the long-running debate over whether, as a general principle, conduct must be capable of excluding a hypothetical as-efficient competitor (AEC) to be abusive has been addressed in the court’s most recent ruling in Google Android. The court has also confirmed that, while the Commission must establish a causal link between the abusive conduct and its potential effects, a counterfactual analysis is not mandatory.

Relevance of as-efficient competitors

The CJ confirmed in Google Android that, while it is not the objective of Article 102 to ensure that competitors less efficient than the dominant undertaking remain on the market, it is not necessary for conduct to be capable of excluding an AEC in order to be abusive. The CJ has gone further by suggesting that, in the digital economy – where variables such as innovation, access to data, multi-sidedness, user behaviour and network effects give rise to high barriers to entry and complex interactions – it may be neither possible nor appropriate to base the analysis of exclusionary effects on whether an AEC could replicate the impugned conduct. In the CJ’s formulation, “in such circumstances, conduct that does not fall within the scope of competition on the merits is capable of making the entry, maintenance or even emergence of an as-efficient competitor practically impossible.”

While the ruling has immediate implications for abuse of dominance cases in digital markets, the principle appears to extend to any sector characterised by innovation, reliance on data, multi-sided markets and network effects – and potentially to any market that exhibits high barriers to entry. In other words, dominant firms should be aware that they may be found to have abused their dominant position even where their conduct is not capable of excluding an AEC, if it is nevertheless capable of foreclosing potential entrants or competitors that are not yet as efficient.

Causal link between conduct and effects

The CJ confirmed in Google Shopping and more recently in Google Android that while the Commission must prove a causal link between the abusive conduct and its potential effects, there is no obligation to do so by way of a counterfactual assessment. Rather, the Commission can rely on a range of evidence, including for example the views of market participants, suppliers, customers, professional and consumer organisations. Nevertheless a dominant firm is entitled to put forward a counterfactual analysis to challenge the existence of a causal link.

Conclusion

The CJ’s recent rulings significantly expand the reach of Article 102, affording the Commission considerable discretion in finding both that conduct departs from competition on the merits and that it has an actual or potential effect on competition.

The practical implications for dominant and potentially dominant undertakings are considerable. They will need to bear in mind when self-assessing their conduct that the broader context is key – market characteristics and the specific circumstances in which conduct takes place can be determinative of whether that conduct could be regarded as falling outside competition on the merits. And where once they might have been able to rely on a favourable AEC analysis as sufficient comfort against enforcement risk, in digital markets at least this is no longer dispositive – consideration must be given to the impact of any behaviour on potential entrants and less-efficient competitors. 

The Commission has been waiting for the CJ’s ruling in Google Android before finalising its Article 102 guidelines, which are currently expected in the third quarter of this year. Whereas the Commission had received criticism for its attempt in the draft guidelines to downplay the relevance of the as-efficient competitor principle, the court’s ruling in Google Android will no doubt have emboldened it on this point.