Thursday, October 2, 2008
Posted by D. Daniel Sokol
Malcolm Coate of the FTC's Bureau of Economics has an interesting new paper on Unilateral Effects Under the Guidelines: Models, Merits, and Merger Policy. One conclusion is noteworthy, "Finally, the Bush administration had no statistically significant effect on FTC enforcement policy."
ABSTRACT: This paper models FTC unilateral effects merger policy using a broad sample of 153 investigations undertaken between 1993 and 2005. Standard unilateral effects analysis proposes a range of models for competitive effects. A review of the files suggests that half of the sample is evaluated with a dominant firm/monopoly model, while the bulk of the other cases use some form of generic spatial analysis. Bertrand and Cournot structures only appear potentially relevant in about a dozen cases each. Statistical modeling shows it is relatively easy to predict the outcomes of the investigations, using models based on either the full sample or a limited sample that excludes the two-to-one mergers. In addition to the standard structural variables (significant rivals, fringe competition, market leadership, and change in the Herfindahl), entry, performance evidence, and vertical variables significantly affect the probability of a merger challenge. Finally, the Bush administration had no statistically significant effect on FTC enforcement policy.