Replacing the Structural Presumption

Click here to read the full article online

Horizontal mergers are a focus of competition regulation throughout the world. In the United States, the so-called structural presumption—under which anticompetitive effects are presumed if market concentration and its merger-induced increase are high—is regarded to be central in merger challenges. But is it? Should it be? And should it be strengthened as well as extended to additional domains, as a U.S. House Majority Staff Report on Big Tech and a Senate bill recently suggest? Upon examination, despite its superficial appeal as a proxy to aid in the identification of problematic mergers, the structural presumption turns out to neither simplify assessments nor provide a useful indicator of anticompetitive danger. A core defect is that the structural presumption requires a choice of market definition in order to determine concentration. The market definition process is incoherent and, ironically, requires proof of the very sorts of effects that are to be presumed rather than demonstrated under the structural presumption. Moreover, the concentration measures used in the structural presumption are largely the wrong market characteristics for predicting anticompetitive effects under the leading economic models used in each standard setting. Finally, in operation the presumption and its associated burden- shifting framework are strange: they are deployed only after completing a full trial, battle of experts and all, and these are bench trials, so judges are supposedly deciding whether to take the decision from factfinders who are themselves. A full reset is required—that is, if the structural presumption indeed describes current merger decision-making rather than constituting a superficial rationalization for what actually transpires. Those seeking to toughen antitrust enforcement in the tech domain and elsewhere should do so in ways that effectively address competitive harms rather than creating seemingly simple but deeply misguided legal tests that would inevitably but unnecessarily allow some anticompetitive actions and block other, beneficial ones.