Dominance abusers see EU regulator refresh its approach and bin burdensome guidance

Big antitrust cases against big companies have become a big headache for the European Commission in recent years.

The increased sophistication of companies, lawyers and economists has persuaded judges to tear holes in some of the EU regulator’s most eye-catching cases, meaning enforcers grew hesitant, cases took longer, and new laws such as the Digital Markets Act were drafted to solve the problems.

Today, the commission told the world it hasn’t forgotten about its once all-powerful law against dominance abuse — Article 102 of the Treaty on the Functioning of the EU. It announced it was conducting some temporary repairs ahead of deeper surgery expected for 2025.

In the diplomatic language of a commission document today, the aim is to avoid “undesirable outcomes” of the sort that had appeared increasingly frequent. Life had gotten tough for the commission, and it was restraining enforcement. In more forthright language: We need cases to be simpler and faster; we don’t want companies blocking our paths at every turn; and judges have been making our lives a nightmare.

Companies will be wary of an enforcer overhauling its policy and plotting a path to more vigorous enforcement. But the excuse for the change is that EU judges have issued 32 rulings in the last 15 years, meaning no need to issue “guidance” for a legal terra incognita. Rather, the landscape is clear now — or at least clearer.

In short, the regulator can plot a route — as it did in 2008 — but it’s the courts that set the highways and byways.

Guidance v Guidelines

At the source of today’s announced change is a 2008 policy approach that ushered in an era of increased economic sophistication. The commission adopted a “guidance paper” that set out what kinds of corporate behavior it would investigate as a priority.

It was the product of a debate between lawyers and economists that enforcement should move away from rigid rules and look more at economic impact. The idea it sought to capture was that something that might look harmful might not be so in reality, if you looked closely at the market effects.

And so the “effects-based approach” was born, which is now fully embedded in EU case law. But the guidance paper itself has been a continual bugbear for the commission, forcing it to justify all its cases in that light. Earlier this month, Qualcomm told judges at the EU’s General Court that enforcers were riding roughshod over that policy approach.

Internally, many in the commission linked recent court defeats — most notably in Intel and Qualcomm — to some of the ideas in the 2008 policy declaration, and so it should surprise no one that the commission wants to rid itself of the millstone around its neck. It always said the paper wasn’t the law — only judges set that — but it had become a continual tripwire in court proceedings.

So, the guidance is dead. Long live the guidelines.

Officials will draft full-blooded “guidelines” to appear in draft form in 2024 and come into force end of 2025. Note the change in nomenclature: "guidance" was a weasel word that had judges scratching their heads, working out what weight to give the document. Guidelines generally mean a formal EU policy that the commission should follow, or else explain why it won’t.

So, in 2025, the guidance paper will be ritually interred, never to be seen again. Until then, the commission has introduced changes that give a clear sign of where it sees enforcement going.

Complex, all too complex

EU officials have publicly and privately questioned developments in Article 102 enforcement in recent years, saying it has become too difficult to prosecute cases. With economics so central to constructing a case, every corporate action could be explained away with counter-evidence and economic modelling, they said.

Companies welcomed this, saying it was only fair given the complexity of markets — in particular for technology and digital services — and the billion-euro fines that awaited them for breaking the rules. EU officials should have to meet a high standard, they said.

Intel, Google and Qualcomm have all scored notable victories in court after judges ruled the commission didn’t reach that standard.

Now, after landmark rulings in several other Article 102 cases, the commission believes it can chart a course through those judicial edicts, setting out a “dynamic and workable effects-based approach,” to quote a policy paper published today by the watchdog.

If a regulator has to prove that the corporate conduct would have a harmful effect, what does that look like? Courts had variously toyed with concepts such as “likely effects” and “capability to foreclose.” The commission has now set its store out for “potential effects,” meaning effects that aren't merely hypothetical. Behavior can be abusive even if those potential effects don’t actually end up happening, the commission adds.

It also explains that to harm competition, a large company doesn’t have to “fully” exclude a rival. Simply harming the “effective competitive structure” of the market is enough.

This will make it easier to prosecute cases as well: Enforcers can protect potential rivals who are still not in a position to challenge the dominant company. There don’t have to be corporate corpses strewn across a market to justify launching an investigation.

'As-efficient' competitors

The EU courts over the last decade have witnessed a recurring nightmare for the commission: It’s called the notion of “as-efficient competitors,” and it has made antitrust officials' lives most uncomfortable.

Its apogee was in the 2008 guidance paper, cementing an approach whereby corporate behavior was bad only if it shafted rivals that were “as efficient” as the dominant company. That meant enforcers should pay less attention to the suffering of lesser companies simply on the grounds that they weren’t up to the fight.

And so companies such as Intel, Unilever and Qualcomm could argue that they didn’t harm rivals, rather those rivals were weaker and were leaving the market because of inefficiency.

Intel, in particular, managed to hold the commission to this policy approach, enshrining its importance in the investigative process even where officials said it wasn’t relevant. And a judgment in January involving Unilever spread the influence of the approach more broadly.

Now, the commission has sought to cauterize that wound, playing down the as-efficient competitor test's role.

“While using the 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 application of such a standard,” it said in the policy paper.

Officials note that the judgments in Intel and Unilever highlighted various ways in which rivals can be “as efficient” other than on price: For example, quality, choice or innovation.

“Mandating evidence of foreclosure of as-efficient competitors [in certain scenarios] would likely lead to under-enforcement. Enforcement action should instead take into account the dynamic nature of competition and be aimed at protecting the competitive process, with a view to allowing entry and expansion.”

So, the as-efficient competitor approach and its accompany economic test have been pushed into a corner, with the commission stressing that it is “not legally required.”  

And in a direct reference to cases involving the likes of Intel and Qualcomm, the commission said such a test was “generally not warranted” because exclusivity rebates — which were at stake in those cases — were “by their very nature capable of affecting competition.”

Such utterances have grown in frequency and volume in recent years, but now they are in black and white on EU-headed notepaper.

Indispensability

Officials have also seized on recent judgments on cases that dealt with dominant companies imposing conditions on rivals’ access to a key input. Think Slovak Telecom’s unfair terms for rivals wanting to access its local loop, or Google’s demotion of rivals on search results pages.

Lawyers have pointed to case law from 1998 that says regulators need to show that the input is “indispensable” for the competitor to do business. Otherwise there is no EU obligation to grant access.

But the courts have generally favored the commission on this point, distinguishing between situations where a dominant rival doesn’t want to grant access at all and where it does. In the latter scenario, the terms need to be fair, judges have said.

And so investigators have tweaked the guidance paper, underscoring that they can pursue a wider range of mischief.

What’s new?

Much of the above will be recognizable to students of the EU courts’ rulings in recent years. And, indeed, many of those ideas are enshrined in the 2008 guidance paper.

But the concepts have been refined and developed by judges, and the commission has found some of its desired tactics thwarted.

Now that body of legal interpretation is being codified in the new guidelines, and the commission will explain how it views the sometimes contradictory messages included in the courts’ output.

That won’t stop the fighting. The application of Article 102 remains one of the most protean — and contested —areas of EU law. But at least now everyone can look to the judgments instead of the defunct guidance paper.