Toe to toe with the tech giants

Online innovation moves at a rapid pace and legislators can struggle to keep up. Global Insight examines whether competition law provides the necessary tools to keep the major players in check.

In 2009, the European Commission (EC or the ‘Commission’) imposed a fine of €1.06bn on chip maker Intel. Brussels antitrust investigators concluded that, between late 2002 and the end of 2007, Intel abused its dominant position under Article 82 of the EC Treaty and used illegal sales practices in the market for x86 central processing unit (CPU) computer chips.

At the time, the fine was the largest ever imposed on a single company for breaching European Union antitrust regulations. It’s since been surpassed by the €2.42bn fine imposed on Google in June 2017 for abusing its dominance in online searches.

It’s probably no surprise that Intel’s appeal against the fine is still ongoing. In September last year, the company won what lawyers are calling a significant victory when the Court of Justice of the European Union (CJEU) said the General Court, a lower tribunal, would have to re-examine the appeal. The CJEU ruled that the Court should review all aspects of Intel’s arguments, including whether the sales practices it had used (allegedly encouraging computer makers to use its chips over rival Advanced Micro Devices (AMD) through so-called exclusivity incentives or rebates) had actually harmed competition.

‘The ruling was mainly about the extent to which the effects of the exclusivity incentives in question should be taken into account in the review by the Commission,’ says Thomas Janssens, Global Head of the competition team at Freshfields Bruckhaus Deringer and Senior Vice-Chair of the IBA Antitrust Committee. ‘Previously, the approach was more form-based than effects-based, in that the rebates would have been considered anticompetitive and that was very much the end of it.’

The case was the main topic of discussion at a panel session co-chaired by Janssens on ‘dominance in the high-tech sector and beyond’ at the IBA’s 21st Annual Competition Conference, held last September in Florence. One of those on the panel was Pablo Ibàñez Colomo, Associate Professor of Law at the London School of Economics and Political Science. The CJEU’s ‘clarification’ of the law is ‘a valuable step towards legal certainty,’ he said.

‘It’s quite a significant judgment in terms of the evidence that the Commission now has to gather,’ adds Mark Tricker, a London-based antitrust and competition partner with Norton Rose Fulbright. ‘In the past, the Commission thought it was allowed to make certain assumptions, without having to prove effects, and now, assuming the dominant company challenges this, it has been told it has to go through a full analysis to show the anticompetitive effect of a rebate scheme.’

In response, the Commission said its original decision on Intel had included a detailed analysis of the anticompetitive effects of the conduct investigated. ‘This has also been the case for Commission decisions sanctioning abuses of dominance taken since, and of course will continue to form part of the Commission’s analysis in ongoing cases,’ a spokesperson added.

Europe’s regulators vs Big Tech: a timeline

Too big to regulate?

Some have argued that Intel’s victory is good news for other big firms facing similar allegations in the EU. However, it could just as easily have little impact. ‘The case was quashed by the CJEU on what were essentially procedural grounds and referred back to the General Court,’ say Marc Wiggers and Robin Struijlaart of Dutch law firm Loyens & Loeff. ‘There appears to be a reasonable chance that the General Court [will] attempt to correct its errors and thus uphold the fine, or alternatively allow the Commission to substantiate its decision and thus impose a new fine on Intel for the same conduct. The Commission did so, for example, in the Air Cargo case, which was also lost on purely procedural grounds.’ In the Air Cargo case, the Commision fined 11 air cargo carriers almost €800m for operating a price-fixing cartel. The General Court annulled the decision on procedural grounds, only for the Commission to amend and re-adopt it, thereby re-establishing the fines.

Competition lawyers will be closely watching the Commission’s current investigation of Qualcomm – which is based on allegations that the company abused its dominant market position and paid Apple to use its chips exclusively – as well as its multiple investigations of Google for any indications of the effects of the Intel ruling on the standards of economic evidence Brussels will have to produce to support its decisions in abuse of dominance cases.

The Commission’s competition regulator (the Directorate-General for Competition) has confronted large United States tech firms several times over the years, with Google being the most obvious recent example (see box). ‘There are lots of questions about the extent to which the Commission has the tools and the ability to deal with competition issues in the tech sector,’ says Tricker. Some of these questions may be at least partly answered when the Commission publishes the full version of its Google decision. There is still a lot of uncertainty about exactly why Google has been found to have abused its position and the precise way that the Commission has framed its argument.

‘A number of recent cases have raised all sorts of tricky tech-specific difficulties for the Commission in terms of getting to grips with the sector, analysing it using their traditional tools and applying traditional concepts of how a company might abuse a dominant position, which don’t necessarily fit the facts very neatly,’ Tricker adds.

Concerns about market dominance in the tech sector are shared in the US. ‘The current framework in antitrust – specifically its pegging competition to “consumer welfare”, defined as short-term price effects – is unequipped to capture the architecture of market power in the modern economy,’ says Lina Khan, Director of Legal Policy at the Open Markets Institute and a visiting fellow with Yale Law School. ‘We cannot [understand] the potential harms to competition posed by Amazon’s dominance if we measure competition primarily through price and output.’

Similarly, European regulators can’t prevent deals like Microsoft’s acquisition of LinkedIn and Facebook’s acquisition of WhatsApp under current rules, says Alexandre Lacresse, an antitrust lawyer at French law firm Fidal. ‘Quite often, the acquired company has a very low turnover, so it does not reach the threshold at which the competition authorities have to get involved,’ he says. ‘Most of the time, this is because services are offered rather than sold.’

Competition authorities are currently rethinking the way they look at these kinds of deals. Some countries, led by Germany, are pressing for quite a hard line on the technology sector. In 2017, for example, the ninth amendment of the German Act against Restraints of Competition entered into force, introducing new notification thresholds in order to cover acquisitions of startup companies active in the digital economy or in innovative markets.

‘Turnover thresholds are not appropriate for businesses of the digital economy that do not necessarily have a high turnover, frequently offer their services for free, are able to quickly increase their customer base and so gain significant market shares,’ says Marco Plankensteiner, a Paris-based competition partner at Kramer Levin. ‘German regulators are also investigating Facebook and its use of data,’ adds Tricker. ‘Some may not see that as a competition issue, but they are looking at it under competition rules.’

The Commission is also considering a proposal to the EU Council to amend turnover thresholds to also include a transaction value threshold. This would bring acquisitions of companies with valuable intellectual property (IP) and technology but relatively low turnover under the scope of EU merger control. Such an additional threshold was recently introduced in Austria.

Other countries argue that, as technology companies move at such speed, competition law is not an ideal tool for regulating them. For them, the market may well be able to resolve competition issues more quickly than the law can. ‘We talk about Google, but remember that, in the early 2000s, we were all talking about Yahoo,’ says Lacresse. ‘Then Google arrived and, in just a couple of years, it was leading the way. That’s the main problem: the market moves so fast.’

Similarly, a few years ago, people worried about the iPhone becoming dominant in the mobile space. ‘That is clearly no longer a concern, because there is a such a good range of competition now for mobile devices,’ says Tricker.

What more can regulators do?

Lacresse also notes that competition authorities are often not equipped, in terms of people or financial resources, to deal with the tech sector. This is worrying at a time when the reach of companies like Facebook, Amazon, Google and Apple seems to be greater than anything seen before.

Some believe regulators could and should be more active. One suggestion is imposing unbundling measures on dominant companies, which could ultimately have the effect of breaking up a company. In autumn 2014, for example, the European Parliament adopted a resolution calling on the Commission to unbundle Google’s search engine from its commercial services. But, although the Commission did recently impose a fine on Google, there are no (public) signs that Brussels is actually considering such a drastic step at the moment. The European Parliament does not have the power to impose such measures unilaterally.

Alternatively, dominant tech firms could be regulated as public utilities. ‘Forced licensing or providing access to digital infrastructures is already a possibility under EU and Dutch national competition law,’ say Wiggers and Struijlaart. ‘The requirements, and therefore the burden of evidence, for a competition authority or a complainant, however, are strict. The entity that is subject to forced licensing needs not only to occupy a dominant position, but the IP/infrastructure to which forced access is provided needs to constitute a so-called “essential facility”. [This] means [that] it cannot possibly be duplicated by anyone, and the refusal of access [to it] prevents the introduction of a new product or service, for which there is a market demand.’ The terms under which a forced licence or access is provided need to be fair, reasonable and non-discriminatory.

Lacresse also mentions the ‘essential facilities’ doctrine, which was first developed and applied in the US to a monopoly over a railway bridge to require companies to share the use of a railway line (the United States v Terminal R.R. Ass’n case). The doctrine was later extended and applied to other infrastructure, including electricity transmission lines, stadiums, jointly developed downhill ski facilities and natural gas pipelines.

There is no reason why it cannot be applied in the EU. ‘The principles are stated in Articles 101 and 102 of the Treaty on the Functioning of the European Union and can be adapted to any situation because they are very basic,’ says Lacresse. However, he adds, competition law is mainly case law. ‘Judges make the law most of the time.’

Lacresse thinks the Swedish set-up is interesting. Sweden has rules based on turnover, just like in France or the EU, but the national competition authority also has the right to come back for a second look at a merger and, even where thresholds are not met, may require a party to notify a concentration ‘where particular grounds exist’. It thus has the power to annul a deal ‘if it leads to problems in the market that were not foreseen at the time of the deal,’ he says. ‘It doesn’t use this power very often – only five times over the last 20 years – but the threat of action obliges the parties to take a very close look at their transaction. To prevent [action], they can decide to notify the concentration voluntarily.’ Perhaps a similar approach could work with Facebook and others.

Moving with the times

Market dominance is just one of many challenges thrown down to regulators by technology companies. Another hot topic at the moment is artificial intelligence (AI) and the extent to which competition law can deal with it. When computers become involved in pricing for goods and services (as is the case at Amazon or Uber), for example, the potential for collusion is even greater than when humans are setting the prices.

‘Sophisticated computers are central to the competitiveness of present and future markets,’ say Ariel Ezrachi of the University of Oxford and Maurice Stucke of the University of Tennessee in a recent research paper. ‘With the accelerating development of AI, they are set to change the competitive landscape and the nature of competitive restraints… We are shifting from the world where executives expressly collude in smoke-filled hotel rooms to a world where pricing algorithms continually monitor and adjust to each other’s prices and market data.’ Will regulators end up imposing fines on machines?

There’s also discussion about how the use of algorithms and data used online is allowing and will allow companies to price discriminate between individual consumers based on their online footprint, and the extent to which that raises issues from a competition perspective. ‘The traditional view of competition law is that price discrimination is not a bad thing; it’s something you might expect to see in a well-functioning market,’ says Tricker. ‘It can lead to efficiencies if you are able to distinguish between different customer groups and if those customers are prepared to pay different prices.’

However, with the massive increase in the amount of data that companies have, it’s less clear that that’s the case. ‘You end up with [price discrimination] going too far [and] individual prices [being set] for individual consumers based on the data that the company holds about them,’ says Tricker. That’s another of many issues that competition law will be forced to address as the influence of tech firms continues to grow.

Jonathan Watson is a freelance journalist and can be contacted on jwatson1@gmail.com