DOJ/FTC Workshop on Updating the Horizontal Merger Guidelines

On December 3, 2009, the Department of Justice and Federal Trade Commission held a joint public workshop to explore the possibility of updating the Horizontal Merger Guidelines used by both agencies to evaluate potential competitive effects of mergers and acquisitions. Below is a summary of the major points from the workshop's four sessions (which are available for re-play here):

Introduction

In his introduction, Jon Leibowitz (Chairman of FTC) said "we are a long ways from the bad old days" of the Schering brief and Section 2 Guidelines (when the FTC and DOJ disagreed on antitrust policy). Nowadays, he said, FTC and DOJ work well together. Christine Varney (from DOJ) spoke after him, but the only thing she said was "I have no comment on the Comcast/NBC merger."

Carl Shapiro was the moderator of the first panel. He spoke about the need to update the Horizontal Merger Guidelines because: (1) there may be gaps between the Guidelines and actual agency practice; and (2) new economic analysis and evidence may change the way we think about antitrust policy.

There were four major issues discussed: (1) Innovation Competition; (2) Direct Evidence of Competitive Effects, (3) Market Definition, and (4) Unilateral Effects.

Innovation

As for the first issue, everyone agreed that the current Guidelines do a poor job of explaining how to measure harm/benefits to innovation (or how to define an innovation market). The agencies have some experience measuring innovation in the pharma industry, but not in any other context. The consensus was that perhaps the agencies should produce some additional commentary or extremely general guidelines on innovation, but that we simply were not ready for formal Guidelines because no one really knows how to measure innovation.

Direct Evidence / Market Definition / Unilateral Effects

The real debate came during the discussion of the next three issues, all of which “merged” together. Generally speaking, some academics and former government attorneys were in favor of using direct evidence of competitive effects, to the extent it exists (i.e., smoking gun documents from officers and directors of the companies who were to merge; customer surveys; natural experiment data as in the Staples case; or unilateral effects evidence). The academics and former government attorneys thought one could use direct evidence of competitive effects as a way to "back into" the market definition. In other words, if you could prove an effect, then you automatically proved a relevant market--you don’t need to do anything else. If that's true, then what happens to the traditional market definition and the presumptions and safeharbors that go with them?

The private practice attorneys (and a few of the academics) were more skeptical of such an approach. They thought the traditional market definition was essential before you could identify direct evidence of competitive effects. Robert Willig, Professor of Econ at Princeton, said: "Direct evidence without competitive analysis is full of pitfalls for the foolish looking for magic bullets." The folks on this side of the debate said that even if you had direct evidence, the judge/jury will want to know more.

Everyone loved the SSNIP test and had nothing bad to say about it. But some folks thought there were other ways (direct evidence) to prove relevant markets. And there was some discussion of a "variable" SSNIP test in some circumstances.

As for unilateral effects, the panelists seemed to agree that if you did a unilateral effects analysis of differentiated products, then all mergers would look like they would raise prices. But this is not true. Some folks said unilateral effects analysis would cause over-enforcement and prevent mergers that were good for the economy. Others said not to worry because you could rebut the unilateral effects analysis with other evidence.

Conclusion

Everyone seemed to agree that the Guidelines should be revised to address the use of direct evidence of competitive effects and unilateral effects. But there was debate about what the Guidelines should say. Everyone wanted to keep the SSNIP test. And everyone also agreed that any revised Guidelines should be general and flexible (like the current ones).