Dissecting Professor Wu and Yelp's Local Search Study: An Antitrust Law Analysis of The ‘Experimental Evidence’
27 Pages Posted: 3 Aug 2017 Last revised: 5 Aug 2017
Date Written: August 1, 2017
In a 2015 article co-authored with three Yelp employees, Professor Wu purports to examine evidence from one experiment allegedly relevant to the question whether Google’s display of local search results in a defined space on the search results page violates the antitrust laws. The authors rely entirely upon this single experiment to support two central claims: that Google’s local results harm consumers and that this innovation violates the antitrust laws. A closer look at the authors’ experiment, however, makes clear it is does not substantiate either of these claims. I describe the methodological flaws in the authors’ approach in some detail, including their mistaken claim that the single experiment amounts to an “RCT,” or “randomized controlled trial.” I also identify the mistaken premise underlying their provocative legal claim – that is, that a firm can face antitrust liability for innovation that improves consumer welfare on the grounds that one can imagine an innovation that would have generated greater welfare improvements. Antitrust law has repeatedly considered and rejected this notion. Professor Wu and his co-authors draw even more provocative conclusions from this experiment, but ignore the fact that in the 5 years following Yelp’s initial complaints regarding local search, Yelp’s annual revenues grew from $83 to $713 million. In response to the common refrain of antitrust complainants that they could have performed even better but for the allegedly unlawful conduct, I explain the relevance of a rival’s economic success in the context of Section 2 claims involving exclusionary conduct. Professor Wu and his Yelp co-authors’ theory thus ultimately fails as a matter of economics, law, and data.
Keywords: Google, Search, Antitrust, Yelp, Local, Monopolization, RCT, Product Design, Innovation, FTC, Methodology, Section 2
JEL Classification: L12, L40, L41, L44, K21
Suggested Citation: Suggested Citation