Why One Should Never Define Markets or Use Market-Oriented Approaches to Analyze the Legality of Business Conduct Under U.S. Antitrust Law
Richard S. Markovits
University of Texas Law School
April 12, 2012
U of Texas Law, Law and Econ Research Paper No. 228
Since 1978, I have been arguing that one should never use market-oriented approaches to determine the legality of business conduct under either the Sherman Act’s specific-anticompetitive-intent test of illegality or the Clayton Act’s lessening-competition test of illegality. Somewhat more concretely, for over 30 years, I have been arguing that one should never apply these tests of illegality by defining relevant markets and basing the predictions or post-dictions the applicable test of illegality makes legally salient even in part on estimates of market-aggregated parameters — e.g., on estimates of the defendant’s market share or defendants’ market shares and the total market share of the leading four or eight firms in the relevant market (parameters on which the U.S. antitrust-enforcement authorities traditionally focused) or on estimates of the post-conduct HHI(s) of the defined relevant market(s) and the conduct-generated change in the HHI(s) in question (parameters on which the Department of Justice and FTC have focused more recently).
My argument for this conclusion has two components. The first is a demonstration that — regardless of whether the markets that any recommended market-oriented approach is supposed to define are ideal-type (classical) markets (whose definition is supposed to satisfy widely-shared professional assumptions about the competitiveness of products placed within a given market and the difference between the competitiveness of products placed respectively in the same market and in different markets) or functional (antitrust) markets (which are supposed to be defined so as to render the antitrust-legality-analysis decision-protocol that makes use of them “optimal” or “maximally cost-effective”) — market definitions are inherently arbitrary, not just at their periphery but more comprehensively. The second is a demonstration that — even if, contrary to my view, one could define markets non-arbitrarily — antitrust-legality-analysis decision-protocols that use such market definitions cannot be cost-effective because data on the non-market-aggregated parameters that are used to define the markets in question have more legally-relevant predictive power than data on any market-aggregated parameters that one could define could have.
In 2010, Professor Louis Kaplow published an article Why Ever Define Markets? that argues for the proposition that one should never define markets for the purpose of measuring a firm’s economic power, which is a corollary of the conclusion that I established in 1978. Kaplow’s article includes a lengthy footnote that — after stating that my 1978 article constitutes a “particularly harsh attack on market definition” — denigrates it on a number of accounts. The article I am posting (1) delineates slightly-improved versions of my 1978 arguments against the use of market-oriented approaches to analyzing the legality of business practices under U.S. antitrust law, (2) explains why those arguments and the “idiosyncratic” (Kaplow’s accurate if pejorative characterization) conceptual systems and competition theories they employ imply that Kaplow’s more limited conclusion is correct, (3) delineates and criticizes Kaplow’s “arguments” for his conclusion (the most relevant of which is a correct assertion of a proposition that is an analog to the conclusion of my second argument for the claim my 1978 article establishes — an assertion he does not and cannot justify because he does not develop and use any counterpart to my idiosyncratic conceptual systems and theories, which play a critical role in the justificatory argument), (4) demonstrates that all of Kaplow’s criticisms of my 1978 article are either incorrect or unjustified, and (5) asserts that at least some of the errors Kaplow makes when criticizing my article are important because they are made by others as well and militate against the correct analysis of the legality of various types of business conduct under U.S. antitrust law.
Number of Pages in PDF File: 126
Date posted: April 12, 2012 ; Last revised: June 21, 2012