< Back

Share |

Changes afoot as big data and competition collide

January 2017

There is no doubt about it; the questions regarding the potential for the use of data to have a harmful impact on the competitive landscape are not going away. The issue has once again made mainstream headlines following Microsoft's acquisition of LinkedIn for $26.2 billion. Many privacy campaigners thought this would be the case they had been hoping for, when a competition regulator finally took action to prohibit or sanitise a transaction due to concerns about the future use of 'big data'. However, the Commission followed the lead of competition law regulators around the world, clearing the deal notwithstanding the concerns over the 'data market power' Microsoft would acquire.


This issue has been simmering since 2008, when Google's acquisition of Doubleclick prompted concerns about the ability of the entity to leverage its complementary strengths in search and ad-serving technology to more comprehensively track customers' online behaviour. It was revivified in 2014 during Facebook's acquisition of WhatsApp, when the European Commission (Commission) was urged to prohibit the transaction on the basis that it would lead to the denigration of WhatsApp's privacy policies (concerns which may have been justified, based on recent events). Nevertheless, the Commission found that, regardless of any privacy issues, neither transaction was sufficiently likely to harm competition. As a result, both deals were cleared.

In addition to these transactions, which sparked but failed to ignite the debate, privacy advocates across Europe have been diligently searching for more kindling. In 2014, the former European Data Protection Supervisor, Peter Hustinx, wrote an opinion on "Privacy and competitiveness in the age of big data", arguing for closer dialogue between regulators and a "holistic approach" to the enforcement of competition law and data protection. Mr Hustinx' successor, Giovanni Buttarelli, has picked up the torch and is gathering increasing support from European regulators. The French, German and UK competition authorities have already carried out enquiries into the issue, and the German Federal Cartel Office has even taken the potentially incendiary step of launching an investigation into Facebook alleging that its privacy practices have managed to fall foul of competition laws (see our previous article for further analysis).

Data as an input

The Microsoft/LinkedIn deal was somewhat different, however, and the concerns being put forward certainly had more of an 'antitrust' flavour than the previous transactions. The primary argument that was aired was that Microsoft might deprive its rivals of the trove of highly valuable information which LinkedIn possesses, without which existing rivals would be unable to offer effective competition, and potential market entrants would face insurmountable barriers (input foreclosure).

It was certainly credible that these arguments would gain more traction with the Commission than those raised in the Google and Facebook acquisitions. The Commission states in its Non-Horizontal Merger Guidelines (Guidelines) that input foreclosure may be anti-competitive and provide grounds to prohibit a merger. As such, there were genuine competition arguments for the Commission to get its teeth into.

However, the arguments based on input foreclosure were always going to prove problematic for a number of reasons. What was required was for Microsoft/LinkedIn to hold a dominant position on the relevant market for the provision of the data in question. Establishing dominance requires a "market definition" exercise, a fundamental tool in competition analysis which relies, in large part, on assessing the potential for one "product" to be substitutable for another.

The idea that an entity could be "dominant" in relation to data is undermined by the fact that, while a data set may be unique, that does not mean there are not adequate substitutes for it in the same market. Competition case law has clearly established, for example, that while certain intellectual property may be unique, that does not mean there are no other intellectual properties which might prove an adequate substitute if one is unavailable for in-licensing or prohibitively expensive.

Furthermore, arguments around "data dominance" also tend to ignore the fact that, unlike many inputs, data is not a finite resource. More data can always be collected and the same data can be simultaneously possessed by multiple market players. The same data may also be collected again at source, or acquired from third parties. The fact that equivalent, if perhaps not identical, data can invariably be replicated by rivals in this way further obfuscates attempts to establish that a company could dominate a market for data, and that rivals would be unable to effectively compete without access to it.

Finally and perhaps most importantly, the Guidelines state that input foreclosure occurs where:

"post-merger, the new entity would be likely to restrict access to the products or services it would otherwise have supplied absent the merger, thereby raising its downstream rivals' costs by making it harder for them to obtain supplies of the input under similar prices or conditions as absent the merger".

It is understood that LinkedIn does not currently supply the user 'interactivity data' which were the source of most concern. It was always unlikely, therefore, that input foreclosure could have ultimately provided a basis on which the deal could have been blocked. As such, the Commission chose to clear the deal following its initial Phase I investigation. Although Microsoft was required to offer "commitments" to the Commission in order to secure clearance, these were concerned with preventing Microsoft from pre-installing LinkedIn on all Windows PCs and integrating LinkedIn into Microsoft Office (similar practices to those which landed Microsoft in such hot water in the past regarding Internet Explorer and Media Player), as opposed to any concerns about Microsoft's use of big data.

Data – an essential facility?

Now that the merger has been cleared, it is possible that third parties that are denied access to the data could switch their focus and seek to establish an infringement of competition law based on the argument that the data constitutes an "essential facility" – a recognised abuse of a dominant market position.

Of course, any such argument would face the same difficulties in establishing dominance as outlined above in the merger control context. Moreover, the competition case law on essential facilities (and the broader category of abuses of dominance based on "refusals to supply") reveals a highly restrictive approach, and there have been very few instances where this has been successfully established. In part, this is due to the fundamental principle that all organisations, regardless of their market power, should be free to choose their trading partners. Forcing any company to deal with other companies is a highly invasive remedy, particularly when those companies are its rivals and all the more so when a dominant company is being compelled to relinquish a competitive advantage.

As such, in the leading case on the essential facilities doctrine, which related to an attempt to access a dominant company's highly developed distribution network, the European Court of Justice stated:

"It should be emphasised in that respect that, in order to demonstrate that the creation of such a system is not a realistic potential alternative, and that access to the existing system is, therefore, indispensable, it is not enough to argue that it is not economically viable..."

As such, even if the potential alternatives to accessing the facility (such as gaining access to alternative data sets from third parties or seeking to generate the data independently over time) are less advantageous, this will not be sufficient to compel access to the dominant company's facility – in this case LinkedIn's data.


Even though the Commission cleared the transaction without requiring commitments from Microsoft regarding its use of data, the clearance decision is unlikely to deter privacy advocates in this debate and there may even be some encouragement to be found. The Commission stated in the Decision:

"Privacy related concerns as such do not fall within the scope of EU competition law but can be taken into account in the competition assessment to the extent that consumers see it as a significant factor of quality, and the merging parties compete with each other on this factor. In this instance, the Commission concluded that data privacy was an important parameter of competition between professional social networks on the market, which could have been negatively affected by the transaction."

Although these were ultimately not deemed sufficiently significant to derail the transaction, such comments will be seized upon by Mr Buttarelli and his followers and for anyone with a reason to be concerned by the next 'big data' merger that comes along. Furthermore, recent comments from Margrethe Vestager, the European Commissioner for Competition, confirm that this initiative is gathering momentum:

"I think Mr Buttarelli is right when he says in his opinion that we all have something in common. We all care about fairness. And the rules on data protection, on competition and on consumer protection all play a part in making that fairness a reality.

Companies need to think from the start about how to use data without hurting consumers. And if they do, there's no reason why big data and competition should conflict…if we want to be able to deal with big data issues throughout the EU, then every national authority has to have the tools it needs to enforce the rules.

In the last few months, we've been talking to those national authorities, to understand what powers they need…I think there's a strong case for new EU rules as part of the answer… We’ll know more when we’ve finished our impact assessment. And if we do find that new EU legislation is the best way forward, I hope to put a proposal on the table early next year."

Given the unprecedented changes that have already occurred in the world of EU data protection law over the last few years, the prospect of further reforms regarding the permitted use of data will hardly be music to the ears of privacy lawyers across Europe.

If you have any questions on this article or would like to propose a subject to be addressed by the Global Data Hub please contact us.

Microsoft/LinkedIn approved, but changes afoot as big data and competition collide
Richard Craig

Richard Craig      

Richard looks at the interplay of 'big data' and competition following Microsoft's acquisition of LinkedIn.

"Given the unprecedented changes that have already occurred in the world of EU data protection law over the last few years, the prospect of further reforms regarding the permitted use of data will hardly be music to the ears of privacy lawyers across Europe."