Data and Privacy Are Not Antitrust Concerns
The calls for stronger antitrust enforcement continue to grow, especially for Internet companies such as Google, Amazon, Facebook, and Apple. Several members of Congress have called for breaking up some of these tech giants. Attorneys general for 50 U.S. states and territories have announced a wide-ranging antitrust investigation of Google, and the Justice Department and Federal Trade Commission are also preparing to investigate the companies. Against this backdrop, the House Subcommittee on Antitrust, Commercial, and Administrative Law this week will hold the third in its series of hearings on antitrust issues.
This hearing will be devoted to the role of data and privacy in competition. It will be interesting to see what evidence the subcommittee hears about this connection, because the actual evidence is very thin. Antitrust law is an important tool for protecting competition and preventing abuses of market power. But the mere possession of data seldom confers lasting market power. And privacy issues, while very important, should not be an antitrust issue.
It is true that, in some cases, data, like other inputs to production, can be used in anticompetitive ways. But as ITIF has written, many aspects about data reduce this threat. First, the mere possession of data is seldom the main source of competitive advantage. What distinguishes companies is often the development of algorithms that add market value to the products they sell. Google’s search, for example, benefits from lots of data. But its biggest asset is the software that uses this data to deliver the most relevant search results. While more data can lead to better results, after a certain point, there are diminishing returns. Doubling the amount of data may lead to only marginal improvements in the quality of an algorithm’s output.
A great deal of data is widely available at little cost. Data, especially personal data, which is the focus of this week’s hearing, is often not exclusive to a specific company. Indeed, nothing stops users from sharing their personal information with more than one company. Data is also nonrivalrous. Unlike most resources, its use by one party does not diminish its value to anyone else. Many parties can use the same information, and once data is generated the cost of duplicating it is almost zero.
Another constraint is that lots of data is useful for only a short time. Thus, any advantage it confers is temporary. Knowledge that a consumer has been searching for hiking boots is only valuable until she buys a pair. Economists Anja Lambrecht and Catherine E. Tucker looked at the implications of data for market power and concluded that: “The unstable history of digital business offers little evidence that the mere possession of big data is a sufficient protection for an incumbent against a superior product offering.”
Privacy concerns pose even less of an antitrust issue. Some argue that if there was more competition, consumers would have more choice regarding the privacy policies of companies. In other words, they allege that even though many online services are free, and therefore there is no price harm to consumers from some platforms having a large market share, consumers are still “paying” by having to share their data and not having the choice of more privacy-protective services.
There are two important points with regard to this. First, the correct response to privacy concerns is not antitrust enforcement; it is privacy regulation. If elected officials believe that large Internet companies are not doing enough to protect privacy, the proper response is to enact national privacy regulation. Second, consumers may be less concerned with privacy than regulators think. And for those that are concerned, they usually have a choice among business models. For example, consumers can use DuckDuckGo, a search engine that markets itself as more privacy protective, or they can avail themselves of the privacy settings on Google or Microsoft’s Bing. Other companies offer subscription services without data collection.
Antitrust policy is an important lever to protect competition. Part of its effectiveness to date is the fact that it has been firmly based on an empirical analysis of whether business actions are likely to create identifiable harm to consumers and innovation. But antitrust has never been based on the view that gaining market share per se is illegal, especially, as in the case of Internet platforms, if that share increases consumer welfare. Moreover, antitrust has long been focused on competition issues, not social policy issues, like privacy. Modifying that tradition now on the assumption that digital is different would be a mistake and would jeopardize antitrust enforcement’s effectiveness without adequately addressing the underlying issues.
