In a new paper, Jon Baker and Carl Shapiro argue that the pendulum of horizontal merger enforcement has swung too far in the direction of non-enforcement. Using historic merger challenge data and survey evidence, they conclude that:
The recent figures for both agencies are low relative to the average of those reported by Commissioner Leary, particularly at DOJ. Indeed, they are for both agencies identical to those observed during the second term of the Reagan administration. The DOJ merger enforcement number in particular matches the lowest in modern history. We interpret these figures as indicating that merger enforcement during the first term of the current administration was surprisingly low, particularly at the Antitrust Division, even after accounting for any expectations that a new Republican administration might resolve close cases more in favor of permitting mergers than would the Democratic administration that preceded it.
Baker and Shapiro call for re-calibrating the significance of the structural presumption, which “identifies the minimum quantum of evidence the government must present in order to satisfy its burden of production, and hence shift a burden of production to the merging firms.” In practice, this leads to different approaches for establishing a structural presumption, depending on whether the theory of harm is one of coordinated or unilateral effects.
Unilateral effects As predicted unilateral effects do ultimately not depend on market shares but upon the diversion ratios between products sold by the merging firms, “it is important to allow the government also to establish its prima facie case with evidence that the diversion ratios between the products offered by the merging parties are substantial.” Using market shares (and the implicit assumption that diversion ratios are proportional to those shares) remains an alternative, but is usually a second best solution.
Coordinated effects The crux of a coordinated effects case lies in demonstrating changed incentives of the market participants in the post-merger world. Thus, traditional market definition, which essentially equates a market with the minimum viable scale for a cartel, is the correct starting point. In a second step, the agencies would have to establish that combining the merging parties is a difference that makes a difference in the post-merger world, either by eliminating a maverick or by reducing the ability of a maverick to disturb coordination. Baker and Shapiro propose that a 4-3 merger or a showing of post-merger HHIs in excess of 2,800 is sufficient to make a prima facie case for coordinated effects.
As to rebuttal arguments, Baker and Shapiro are critical of the uncritical acceptance of the three “E”s, expansion, entry, and efficiencies by the agencies and the courts.
In summary:
We certainly do not propose a return to the horizontal merger control policies and precedents of the 1960s. The presumptions we have described here would not be irrebuttable, though they would be influential. They would be based on aspects of market structure, but not solely on market concentration, and in some cases, not on market concentration at all.
Any article by Baker and Shapiro is worth reading, and this one is no exception. Most of their proposals are sensible as a matter of economic theory and legal practice. Few, I suppose, would object to more predictable antitrust enforcement, and rebuttable presumptions have always been a powerful tool to increase predictability without the collateral damage from the heavy artillery of per se rules.
That said, Baker’s and Shapiro’s proposals should include additional procedural transparency requirements for the agencies in the merger review process, not just in the few cases that actually go to trial. The agencies should only get to enjoy a presumption after sharing the evidence on which the presumption rests with the parties to the merger (with some limitations, of course, if third parties are involved).
Technorati Tags: antitrust, horizontal mergers