Report: Antitrust Concerns About Data Are Overblown
While some US commentators are pushing for regulators to consider companies’ data holdings in antitrust analysis, a new report from the Information Technology and Innovation Foundation (ITIF), a leading global tech-policy think tank, finds that merely having data does not raise anticompetitive concerns. And if data-rich companies do thwart competition, regulators already have tools to address potential harms.
The report’s author, Joe Kennedy, a senior fellow at ITIF and former chief economist for the US Commerce Department, argues that data-rich companies are not an economic threat, but in fact an important source of innovation. If the simple possession of data becomes a new factor in antitrust analysis, it would depress innovation when policymakers should instead be encouraging it.
“Just like any other important resource, companies may use data to gain advantage over their competitors, but existing antitrust law already allows regulators to take effective action,” said Kennedy. “But merely having more data than a rival does not itself threaten competition any more than having more machines does. Regulators can do a lot of damage by restricting the collection and use of data to mollify a small group of privacy advocates. We should be encouraging data innovation, not discouraging it.”
In the report, Kennedy debunks three major arguments for expanding the scope of antitrust review to incorporate how companies collect and use data:
First, he explains why the mere possession of large amounts of data does not give a company a significant competitive advantage that its rivals will be unable to challenge. Data is non-rivalrous: One company’s possession of data does not come at the expense of another’s.
Furthermore, if the possession of large amounts of data were necessary for an entrant to compete successfully, that would not necessarily constitute an unfair competitive advantage. As Kennedy points out, “Many industries have high start-up costs. We do not say that Ford has an unfair advantage just because companies must first build an expensive factory before they sell a single car.”
Second, Kennedy explains that current competition policy can already respond to competitive threats stemming from large-scale aggregation of data, even though many of these markets do not involve prices—usually a common component of antitrust analysis. While the extent and importance of non-price competition can be difficult to quantify, this should not change the proper application of government powers. Furthermore, in the case of internet platforms—which are often held up as prime examples of why data possession should be included in antitrust analysis—they make money not on search tools or social networking, but on advertising, where there is clearly a price. There is no evidence that companies processing large amounts of data while also providing free services have significant power in the advertising market, or that they currently abuse any market power they do have to get away with providing worse services.
Finally, despite concerns that the acquisition of large amounts of user data presents a serious threat to privacy, US consumer-protection authorities have sufficient powers to ensure that companies comply with existing privacy regulations and honor pledges they make about their data policies. Regulators also have broad authority to address problems arising from the actual misuse of data.
In the report, Kennedy reviews recent mergers in the United States and Europe and finds that, in most cases, agencies have allowed mergers to proceed when they did not affect the amount of competition in existing markets—even if the acquiring company ended up with significantly more data. When a merger did threaten competition, the agencies were able to protect consumers, often by requiring that data be shared with competitors, as in the real-estate information industry when Costar and LoopNet merged. Opponents’ warnings proved to be almost entirely unfounded. In other cases, the agencies were able to reexamine the facts in response to new developments, such as in the Facebook-WhatsApp merger. Kennedy concludes that the current antitrust analysis approach has been effective in preventing harms, so there is no need to introduce a new metric that would chill innovation.
“The debate over whether the mere fact that a company possesses data is germane for antitrust analysis is usually not about whether control over large amounts of data could in some instances be used to thwart competition, because it certainly can,” Kennedy said. “The dispute is really about the degree to which antitrust regulators should take preemptive action to limit the collection and use of data to guard against possible harms that are often only tangentially related to competition. As an example, privacy concerns are not competitive threats. Any preemptive action in that area would only hurt innovation and provide no further benefits to market competition.”