How Much Economic Analysis in Competition Law and Policy Do We Really Want?
Concurrences N° 4-2023, Art. N° 115114, November 2023
University of Groningen Faculty of Law Research Paper No.3/2024
12 Pages Posted: 24 Apr 2024 Last revised: 22 May 2024
Date Written: November 1, 2023
Abstract
Thorough economic analysis of the relevant markets and competitive constraints is central to most antitrust cases, especially since the Commission’s turn towards an “effects-based” approach in the late 1990s and 2000s. Recent case law of the EU Courts, moreover, has required the Commission to respond to a defendant’s substantiated counter-arguments by a comprehensive economic analysis of the circumstances of each case. But how thoroughly should competition authorities analyze the individual facts of a case? How many counterfactuals should it consider? What degree of evidence and certainty can we expect competition authorities to produce? Can there be such a thing as “too much” economic analysis?
This contribution argues against the “naturalistic” belief that defining relevant markets and identifying competitive relationships is just a matter of collecting enough data to ascertain objective economic truths that are “out there”. Sophisticated empirical methods are unlikely to solve the institutional and epistemic limits confronting competition authorities. Moreover, the combination of extensive use of economic analysis by competition authorities on the one hand and strict standards of judicial review on the other hand could jeopardize effective competition policy. Accordingly, the contribution argues in favor of a certain degree of deference to competition authorities so as to avoid a battle between increasingly sophisticated economic and econometric analyses, which is unlikely to improve the quality of decision-making, and which plausibly reduces the overall effectiveness of competition policy.
Keywords: competition law, competition policy, antitrust, economic analysis, market definition, complexity science, judicial review
JEL Classification: K10, K20, K21
Suggested Citation: Suggested Citation