The Competitive Effects of Search Engine Defaults

Abstract

The landmark antitrust case U.S. v. Google centers on novel vertical contracts in which Google pays web browsers and device makers to make Google Search the default search engine on their products. But the competitive effects of such arrangements are not yet well understood. I study them in a novel model of competition between two-sided search platforms, which earn all revenues on the advertiser side. Search algorithms "learn" and improve with use, a process that resembles network effects. Defaults create switching costs that nudge" consumers toward the default platform. Due to algorithmic learning, defaults can have significant competitive effects even if switching costs are small. Broad defaults (those affecting a large share of users) by a dominant platform reduce consumer welfare under most plausible conditions. They also suppress entry and investment by laggards. However, narrow defaults (particularly by laggards) can create positive spillovers and stimulate entry and investment. This casts doubt on the wisdom of polices (such as mandatory "choice screens") that would eliminate defaults altogether.

Keywords: Antitrust, platforms, search engines, defaults, two-sided markets, network effects, data, monopolization, exclusion

JEL Classification: L40, L41, L12, K20, K21