Are some tech companies using their cache of big data to crush competition?
U.S. regulators don’t think so. At the U.S. Department of Justice (DOJ) Technology Policy Institute Conference in August 2018, Deputy Assistant Attorneys General Barry Nigro and Roger Alford said they’d be unlikely to bring antitrust cases against big tech companies based on their collections of large amounts of data. But this view contrasts with the European Union (EU), where a few companies already face enormous fines for what regulators there consider competitive dominance in big data.
In this hoganlovells.com interview, Hogan Lovells partners Falk Schöning, based in Brussels, and Logan Breed, based in Washington, D.C., discuss the contrasting perspectives of the EU and United States regarding big data, competition, and antitrust law, and how regulators are attempting to level the playing fields between large tech companies, smaller firms, and new entrants into the open market.
How do regulators in the U.S. view the competitive significance of big data in their antitrust investigations?
Logan Breed: The question of how the antitrust laws should handle the issue of big data is somewhat uncertain in the United States. There have not been any cases brought by U.S. antitrust enforcement agencies based on the idea that a set of data is competitively significant enough to cause any competitive violation to occur or for a transaction to be competitive. However, it is a topic that is getting attention in antitrust circles.
The U.S. Federal Trade Commission (FTC) recently embarked on a series of hearings to assess the state of competition and consumer protection law in the 21st century, and it will hold two days of hearings in November on the topic of “Privacy, Big Data, and Competition.” It will be an opportunity for both sides of the debate to air their views publicly, and for the FTC to either come to some conclusions and write a report or leave those comments out there. It’s unclear exactly how this will end up. Moreover, the U.S. Attorney General and the head of the DOJ Antitrust Division recently met with multiple state Attorneys General reportedly to discuss potential actions against big tech companies.
There are examples of merger cases where data was significant, and there has been at least one case that the FTC brought, in which the remedy involved a licensing of data. But that’s not exactly what the concept of big data and its competitive significance really means when people talk about it in the abstract.
One potential issue would be if an acquisition or merger between two companies that have big datasets will create a unified dataset that is an impossible barrier to entry for other companies to compete with. Again, we’ve never seen a case like that. I suppose in the abstract, it’s possible. But most big datasets are not unique and are not uniquely competitively significant. As a practical matter, I think it would be very difficult for a U.S. antitrust agency to prove that a transaction violated Section 7 of the Clayton Antitrust Act, which prohibits mergers and acquisitions that may tend to substantially lessen competition in a relevant antitrust market, solely on the basis of the fact that the merging companies have a lot of data.
So when companies with a lot of big data merge datasets, you don’t anticipate it will become a legal issue in the United States in the near term?
Breed: At a recent conference, two DOJ Antitrust Division front-office leaders talked about big tech companies that have large amounts of data and the protests by smaller companies against the bigger companies about their data. But the upshot of their comments was that they think it would be very difficult in the United States to bring that kind of a case. Nevertheless, the economic power of big tech companies is a hot topic at a political level in the U.S. on both the left and right of the political spectrum, and big data will likely continue to be one of the issues that the detractors of big tech use.
So the takeaway for clients is, there’s a lot of noise right now about whether antitrust laws are adequate to rein in big companies, including big tech companies, and whether big data is an existential competitive threat that’s going to undermine competition going forward. I think the answers to those questions are first, the antitrust laws are sufficient to take into account the competitive significance of big tech companies and their datasets. And second, big data in the abstract is not an existential threat to competition.
There is some general concern that companies with big data may impact competition, but is there a specific issue that’s up for debate?
Falk Schöning: The starting point, and the big picture question, is whether or not there actually is a problem that any company has control of big data. To be provocative, you could say, these companies may have vast amounts of data, but so what?
There is some agreement in both the United States and the EU that there could be an antitrust problem with those companies that control big data, but it is unclear what the problem is and for whom? Is it a problem for the consumers; for competitors, including the smaller digital companies out there; or the old, offline economy which is disrupted by big data companies? Depending on your perspective, you will come to totally different answers.
How this is being handled in Europe? Are attitudes similar or different to the United States?
Schöning: Indeed, they are somewhat different. This makes the topic interesting, relevant, and also difficult to handle for clients.
In Europe, all the fuss about big data and antitrust that Logan mentioned has somewhat resonated with the authorities and politicians. It’s not just an academic debate. The European Commission, the German Federal Cartel Office, the French Competition Authority and others — they all are not only talking at conferences about this. They have set up teams internally and they have published papers where they laid the groundwork for future antitrust investigations. So while I think we still have yet to have the big data case, I’m quite sure the authorities would want to look at the cases where all these theories are dealt with — and you can see from some of the enforcement actions that they’re trying to.
The other aspect, which I also find interesting, is whether antitrust law is fit for its purpose here. Should it actually be revamped to be brought into the 21st century? In Europe, there is a debate as to whether antitrust is really the instrument you want to use, or whether you need regulation of big data companies similar to telecommunications regulation. Such a real data regulator would be set up very specifically to protect companies from the abuse of a strong position by having that much data.
That’s the debate that politicians raise, and for our clients, that’s a debate to be followed and engaged in because this would eventually decide about leveling the playing field where they are active.
Breed: I think clients should think about whether they want to engage with these FTC hearings that are coming up, too. There will be opportunities for public comment.
For companies in Europe, are there forums similar to the upcoming FTC hearings that Logan mentioned, where companies can engage in the debate?
Schöning: They are plenty of public consultations on specific aspects of this, so that’s an opportunity to engage and respond. For the larger digital players, you can assume that they all have their offices in Brussels and in the EU Member States and they do talk to the regulators, sit on panels, and express their views. Smaller or more specialized companies tend to so less, but they could think about using trade associations to get their voice heard.
That part of the work is more public affairs. I think what’s more interesting from a legal perspective, for us and our clients, is the compliance aspect. In Europe antitrust authorities may pick up this potential theory of harm about having a lot of data. So when you have a lot of data, you need to be very careful about what you’re doing with it because otherwise someone could allege you are abusing it. So that becomes a compliance question.
And for clients, the challenge is to include this new world in their existing compliance programs, which is typically done by training salespeople, or having audits where you look at salespeople's e-mails. But in a world of big data, you actually have to talk to the IT and software engineers and understand what they are working on.
So you suggest companies rethink their compliance programs and involve their software engineers?
Schöning: Yes, and there’s even a term for it developed by the authorities in Europe: the European Commissioner for Competition is asking companies to apply something she calls “compliance by design,” which means you can’t hide behind an algorithm if you infringe antitrust laws.
So we say, for example, my engineer developed the software that monitors the prices of all my competitors online. Whenever they change their price, my algorithm automatically changes our price, too. That may be nice if prices go down, but if everyone applies a similar algorithm, and if these algorithms are probably even self-learning, the computer programs may conclude that they are all better off if they increase prices.
And then the question is, who’s responsible? The machine can’t be responsible in legal terms, so is the guy who developed the software responsible? Or the procurement person who bought the software from a third-party vendor? If you consider the aforementioned European Commission's approach, the company that applies the software could be responsible. But the legal conclusion to come to that point is not that straightforward.
What do you expect to happen next in the discussion about big data and antitrust law?
Schöning: The next level is whether it’s a problem that we should solve at all by means of competition law.
Here are two adjectives that describe competition law: it’s flexible and it’s case specific, so it will only be applied, say, to the company that has big data but not to all companies that hold some form of data. The sandwich shop, which happens to have a website where you can order your food, will not be the focus of the antitrust authorities.
In contrast, regulation applies to everyone meeting certain criteria. Whereas the regulation could say, if you operate a website, you need to do this — competition law would not prescribe specific conduct but rather enforce the law only afterwards if problems occur. For large global companies, that’s not a problem, because they have enough resources to deal with this. But for the sandwich shop, it could be a problem. That is — of course generalizing — the difference between competition law, which is flexible, and regulation.
On the other hand, regulation also has advantages: it gives you predictability of rules. Competition law is what the lawyer calls ex post — only after something has happened. It’s applied only if a regulator thinks your behavior is problematic.
As you may have seen, some large digital players have been fined already in Europe. Even for a large company, these have been very significant fines. And because there is no book that talks about how you run a search engine or a social network by the law, the antitrust enforcement is not predictable. As Logan said, in the United States, to date there is much less risk of a fine. But in Europe, this lack of predictability can cost companies a lot of money.
What’s your perspective as to why Europe and the U.S. have such different approaches to big data and antitrust enforcement?
Schöning: I think some Europeans are more concerned with data in the hands of companies than in the hands of the State.
This is very different from the understanding of freedom in the United States, which is more about ensuring that your data is not in the hands of the state. That’s very different, and it triggers why the EU has things like GDPR, which basically regulates how private companies can deal with your data, but not what the state can do with it.