Via Geoff Manne at Truth on the Market, I see that FTC Commissioner Josh Wright has an important discussion of the burden of proof in merger analysis in his dissent In the Matter of Ardagh Group. In his blog post, Geoff discusses the history of how potential efficiencies get credited in merger analysis. Under the 1984 Guidelines, efficiencies had to be proved by clear and convincing evidence, which (in the view of Judge Ginsburg, who used to be AAG for Antitrust) was “well-nigh impossible”. Under the 1997 Guidelines, efficiencies had to be substantiated so that they were reasonably verifiable by the agency and weren’t “vague or speculative”, and this remains true today. Nothing wrong with that, of course, but various commentators do believe that this standard differs from the more lenient standard for showing anticompetitive harm. That’s where Josh’s dissent comes in. Josh writes:

The first issue is whether the magnitude of the burden facing merging parties attempting to demonstrate cognizable efficiencies should differ from the burden the Commission must overcome in establishing the likelihood of anticompetitive effects arising from the transaction in theory. The second is whether the magnitudes of those burdens differ in practice. The Commission appears to answer the first question in the negative. With respect to the second question, the Commission points to some evidence that the Agency does in fact consider efficiencies claims when presented in many investigations. There is little dispute, however, that the Commission gives some form of consideration to efficiency claims; the relevant issue is over precisely how the Commission considers them. More specifically, must merging parties overcome a greater burden of proof on efficiencies in practice than does the FTC to satisfy its prima facie burden of establishing anticompetitive effects? This question, in my view, merits greater discussion.

. . . [A] critical issue highlighted by this case is whether, when, and to what extent the Commission will credit efficiencies generally, as well as whether the burden faced by the parties in establishing that proffered efficiencies are cognizable under the Merger Guidelines is higher than the burden of proof facing the agencies in establishing anticompetitive effects. After reviewing the record evidence on both anticompetitive effects and efficiencies in this case, my own view is that it would be impossible to come to the conclusions about each set forth in the Complaint and by the Commission – and particularly the conclusion that cognizable efficiencies are nearly zero – without applying asymmetric burdens.

Merger analysis is by its nature a predictive enterprise. Thinking rigorously about probabilistic assessment of competitive harms is an appropriate approach from an economic perspective. However, there is some reason for concern that the approach applied to efficiencies is deterministic in practice. In other words, there is a potentially dangerous asymmetry from a consumer welfare perspective of an approach that embraces probabilistic prediction, estimation, presumption, and simulation of anticompetitive effects on the one hand but requires efficiencies to be proven on the other.

. . .

The lack of guidance in analyzing and crediting efficiencies has led to significant uncertainty as to what standard the Agency applies in practice to efficiency claims and led to inconsistent applications of Section 10 of the Merger Guidelines, even among agency staff. In my view, standard microeconomic analysis should guide how we interpret Section 10 of the 2010 Merger Guidelines, as it does the rest of the antitrust law. To the extent the Merger Guidelines are interpreted or applied to impose asymmetric burdens upon the agencies and parties to establish anticompetitive effects and efficiencies, respectively, such interpretations do not make economic sense and are inconsistent with a merger policy designed to promote consumer welfare. Application of a more symmetric standard is unlikely to allow, as the Commission alludes to, the efficiencies defense to “swallow the whole of Section 7 of the Clayton Act.” A cursory read of the cases is sufficient to put to rest any concerns that the efficiencies defense is a mortal threat to agency activity under the Clayton Act. The much more pressing concern at present is whether application of asymmetric burdens of proof in merger review will swallow the efficiencies defense.

Speaking of mergers and Truth on the Market, also check out Geoff Manne’s post on the Comcast-Time Warner merger.