Competition and Big Data: Some Trends Never Go Out of Fashion
What do haute couture and competition policy have in common? Trends come and go but some things just never go out of fashion. That is most obviously true for haute couture or fashion in general (see, e.g., Oprah’s list of timeless, mistake-proof fashion items) but once in a while some competition policy debates share that very same characteristic. A good example is competition policy and ‘big data’ – a topic that has been à la mode for quite some time and that looks like a strong contender for this year’s main trends.
That at least is the conclusion you could draw when reading a recent Wall Street Journal (WSJ) article referring to Commissioner Vestager’s statements on her having “an open mind” about enforcing competition rules against companies that accumulate large amounts of data. There is of course nothing wrong with keeping an open mind and adapting enforcement to new technological developments, but it’s interesting to see this topic coming on the back of the Commissioner’s previous statements stressing that data is not to be automatically equated with market power. It would probably go too far to take her recent statements as proof for a change of heart but given the Commissioner’s crucial role in enforcement priority-setting, these statements can certainly provide insights into the inner workings of Europe’s highest competition enforcer.
Given the trend of debating “big data” and competition enforcement is likely to continue, it’s worth to recall which potential competition problem commentators have raised with respect to data collection. In the WSJ article, Commissioner Vestager talks about data accumulation that could make it harder or even impossible for competitors to effectively compete. The theory is that companies that collect large amounts of data could achieve an insurmountable competitive advantage in the marketplace. This argument often assumes an endless, positive feedback loop where data accumulation leads to better services that in turn lead to more users which again increases the amount of data collected. Hence, how could a smaller competitor ever break this miraculous cycle of endless innovation exclusively reserved to today’s biggest data hoarders?
The most obvious answer to this question is to come up with a better, more innovative, more creative, or just different mousetrap that would attract users. The relatively short history of the Internet does not show any evidence for large amounts of data being an effective wall for fending off competition. In fact, the exact opposite is true. Think of how Tinder disrupted the till then data-heavy dating market, or how Uber broke into markets often dominated by a couple of companies holding valuable customer information, or how Snap’s value proposition of being a ‘camera company’ enabled it to compete in social media. The success of these companies has also been facilitated by relatively low switching costs in online markets and even in cases of higher switching costs, users have shown a remarkably strong tendency to multi-home. The average number of apps people use on a monthly basis has reached an all time high of 35 in 2017 and is predicted to increase even further. Users do not use a service because of a superior database but because of a superior experience.
Apart from that, it is by now well documented that the economic and some technical aspects of data simply do not allow it to be a good shield against competition. Data is non-exclusive and easily replicable. One company holding certain data does not prevent another one from gathering the exact same data. This is directly related to the better mousetrap argument above – data accumulation is the result of companies’ popularity among consumers, it is not the cause of their success. Furthermore, placing much emphasis on ‘data accumulation’ is in itself wrong. In simple terms, it does not matter how much data a company collects when it does not know what to do with it. World-leading economists have always emphasized that the bottleneck in innovation markets is not the availability of data, but the supply of high-skilled talent knowing how to derive meaningful insights from data. If you couple that with data being subject to diminishing returns and, in many cases, to considerable value reduction over time, it is not hard to understand why a seemingly insurmountable data advantage has not prevented some of yesterday’s digital market leaders getting into oblivion or even non-existence.
One could argue that the argument would be a different one if a given dataset was found to be unique or so important that competitors would need access to it to effectively compete. Competition lawyers would refer to an ‘essential’ input to which access would have to be granted. But so far, such a dataset has not been found in online markets. In the context of the more recent Microsoft/LinkedIn merger some have brought up the argument that LinkedIn might have unique data that other companies were not able to replicate. The Commission dismissed this argument entirely and pointed to other data sources readily available to competing companies.
This approach is not only sensible but also reflects the EU’s very high legal threshold for a finding of an essential facility to which competitors would have to be granted access. In cases going back to Oscar Bronner and many others, European Courts made clear that access to a competitor’s resource can only be mandated if it is absolutely indispensable for that competitor’s business and if the denial of access is likely to eliminate all effective competition in a given market. In fast-moving online markets it is hard to see which dataset would fulfill this legal standard. After all, the Courts had very good reasons to set the bar this high — mandating access to a company’s resources will always reduce that company’s incentive to come up with that resource or product improvement in the first place. Why invest if others would be able to free-ride on your investments?
It is remarkable that Commissioner Vestager’s comments seem to suggest that data-sharing could be imposed when competitors have a hard time competing. One is left wondering whether companies would be forced to share datasets that to most other companies would be useless or whether insights from data analytics would have to be shared, too. In light of the above, both options would be difficult to justify and, even worse, could lead to unintended consequences like the reduction in innovation – the exact opposite of competition rules’ aim. Whether competition enforcement and big data will remain a “timeless, mistake-proof item” (citing Oprah on fashion, as mentioned above) of competition policy will hence have to be seen. Doubts surely exist.