Mergers, Merger Control, and Remedies: A Response to the Vita-Osinski Critique

It has been nearly four years since publication of my research monograph, Mergers, Merger Control, and Remedies: A Retrospective Analysis of U.S. Policy (MMCR). During this time, the book has received attention from economists and lawyers, from policymakers in the United States and elsewhere, from think tanks and consultants, and from reporters and others. This has been due in part to the fact that the focus of the research—mergers, their effects, and their control—have become matters of wider public concern in the U.S. economy. MMCR also has received attention because its methodology—distilling the results of numerous careful studies of mergers—is relatively novel and also because its findings include a number of conclusions that run counter to some strongly held opinions. As a result, I expected criticisms from those who disagreed with the findings in MMCR. Such criticism has now come from Michael Vita and David Osinski, economists at the Federal Trade Commission.

In this review and update, I will first provide a brief overview of what my book in fact says, and then respond to Vita and Osinski’s various criticisms. I acknowledge an oversight that affects the strength of one conclusion, but otherwise, as I will show, there is nothing in the Vita-Osinski critique that undermines the methodology or alters the conclusions of the book. Indeed, I believe their critique ultimately makes clear the fundamental objectivity and strength of the underlying research.

I. MMCR, In Brief

MMCR consists of several parts. It begins by lamenting the dearth of ex post evaluations of antitrust actions but notes that a sufficient number of merger retrospectives now exists to permit at least some systematic analysis of mergers rather than relying on mere anecdotes or on broad statistical evidence not closely related to specific mergers. Merger retrospectives involve careful isolation of the effects of specific mergers on price and occasionally other outcome variables. An essential part of my analysis rests on an exhaustive compilation of all such studies in the literature, using an explicit protocol for including studies so as to avoid any appearance of subjectivity, and then converting all the qualifying studies into a common format and metric for purposes of synthesis and analysis. This approach—essentially a meta-analysis of existing high-quality studies—summarizes what is reliably known about mergers, merger control, and remedies. While not the first such attempt, MMCR is the most comprehensive effort to extract insights from carefully studied experience.

The book develops several different databases, all but one consisting of merger retrospective studies. The largest database of retrospectives consists of some 80 studies covering more than 3000 mergers, many on an aggregated or “grouped” basis. Most of the research focuses on roughly 50 studies and a similar number of mergers and transactions that are individually analyzed. Those same mergers also yield about 120 product-level observations on price outcomes, since all studies report the effects on multiple products. Progressively smaller data sets are also used for questions, such as merger remedies, posed at a more granular level.

MMCR analyzes four major topics: actual case-bringing, the price outcomes of mergers, the effectiveness of merger control policy, and the effects of remedies in cases where they were employed. The evidence yields numerous substantive insights, some rather striking, into mergers and associated policy. The one non-retrospective database shows that merger enforcement has progressively narrowed its focus over time. The remaining meta-analyses of retrospective studies establish that these carefully studied mergers on average have resulted in nontrivial price increases, that antitrust policy has permitted too many anticompetitive mergers, and that remedies, especially conduct remedies, have often been ineffective. The book concludes as it begins, noting that the findings underscore the value of retrospectives and that these need to be conducted more often and systematically in order to inform and improve merger control and antitrust policy more generally.

The reception that the book has received has been gratifying. Many have credited its methodology for its objectivity and precision and found its substantive conclusions illuminating. Indeed, the European Commission’s Directorate-General for Competition almost immediately commissioned its own meta-analysis of studies of European mergers and merger policy, crediting MMCR for the design, protocols, and tests. The ensuing report noted that “[t]he most relevant paper for our purposes is an analysis of U.S. merger retrospectives by Kwoka,” and identified MMCR as “an important precursor” of its research. The authors of the EC report, in a separate paper, declared “The intellectual motivator of our article is Professor John Kwoka’s meticulously compiled study of US merger retrospectives.” The EC study, as well as reports by the International Competition Network and various competition agencies, have recommended merger retrospectives as devices for harnessing past experience for its implications for improving merger control policy.

Substantively, the EC report, the separate article by its authors, and at least four surveys of U.S. mergers arrive at conclusions broadly similar to those in MMCR, namely, that merger and remedy policies have to varying degrees been deficient. An increasing number of other observers and policymakers agree with that assessment. Indeed, the current Assistant Attorney General for Antitrust, Makan Delrahim, has cited MMCR approvingly for its critical review of conduct-type remedies.

While acknowledging the value of retrospectives, the Vita-Osinski critique takes a different view of my work. It challenges the methodology that I use and the conclusions that I draw (or at least those that it asserts I draw), and raises a host of specific issues. Apart from a classification error affecting two data points and the strength of one conclusion, however, the Vita-Osinski critique is misguided and incorrect. It relies on incomplete and selective reading, exaggerates readily addressed concerns, dismisses key evidence, and insists on inappropriate methodology.

This response to the Vita-Osinski critique will first address its criticisms of the research regarding mergers and merger control—the bulk of the book— followed by its criticism of my findings with respect to merger remedies.

II. Mergers and Merger Control: The Evidence

The Vita-Osinski critique focuses most of its attention on the analysis of mergers and remedies in Chapter 7 in MMCR. They object to its overall thrust as well as specifics in that chapter. Here I show that their various points have no merit.

A. Misreading MMCR

Vita and Osinski begin by quoting passages from Chapter 7 of MMCR that summarize evidence from the compilation of merger retrospectives analyzed in that chapter. They then assert that this evidence does not justify my “central finding . . . that federal . . . merger enforcement is excessively lenient and ineffective” and that I “find[ ] that FTC/DOJ merger policy has failed systematically.” This an erroneous reading of MMCR, since neither in those passages nor elsewhere in MMCR do I conclude that merger policy is simply ineffective and a failure. Rather, there is evidence of both effective and ineffective policy actions, but, in the actual words of MMCR, on average the agencies “fail to challenge a considerable fraction of [mergers] that result in price increases.” Those more nuanced statements, not those of Vita and Osinski, characterize the actual findings of MMCR.

In this passage and elsewhere, Vita and Osinski juxtapose evidence and statements from MMCR in a fashion that creates the appearance of inconsistency or lack of support. The statement concerning policy leniency that they quote is not the conclusion of the chapter on merger retrospectives that they reference, but rather appears in a later chapter that explicitly relies on “additional data and information.” This additional evidence is support of that conclusion—evidence not noted, much less quoted, by Vita and Osinski—derives from two sources. One is another compilation of merger studies covering a total of more than 3000 mergers in “grouped” or aggregated form. These are methodologically just as sound as the retrospectives analyzed in Chapter 7 but are less detailed and so are analyzed separately. They arrive at very much the same conclusion as the individual merger studies, namely, that post-merger prices are on average higher.

Additional evidence cited by MMCR as supporting the conclusion regarding increasing policy leniency would be difficult for Vita and Osinski to dispute since it derives from the FTC itself. These data cover all of the agency’s merger investigations from 1996 up through 2011—more than 1200 in total— and indicate how many resulted in some kind of enforcement actions. Tabulations of these data in MMCR make crystal clear that between 1996 and 2011, the FTC progressively shifted away from challenges to any mergers that resulted in more than four remaining competitors. Indeed, by the end of this period the agency had abandoned such cases altogether—not bringing a single such case in the final four years. The Vita-Osinski critique never mentions this evidence of increasing leniency, despite the fact that it represents unambiguous, and explicitly cited, support for my conclusion.

Continue reading the full text of this article in PDF format.