We found the article by Erwin Blackstone, Larry Darby, and Joseph Fuhr, Jr. (“The Case of Duopoly,” Winter 2011–2011) an interesting and useful piece on the lack of empirical or theoretical support for the market structure/​performance relationship. Economists focused on antitrust have understood the difficulties in showing such a relationship for decades, but the piece rightly showed how this result extends even to duopolies.

Starting with the Reagan revolution, Chicago school insights have dramatically reformed merger policy. A sample of data from the Federal Trade Commission indicates that, between 1993 and 2010, the agency challenged three‐​to‐​two mergers only 78 percent of the time—a high fraction, of course, because often a third firm does provide a strong competitive constraint to the other two, but far from the 100 percent one would expect under the old Structure‐​Conduct‐​Performance (S‑C-P) school. (See Malcolm B. Coate, “Benchmarking the Upward Pricing Pressure Model with Federal Trade Commission Evidence,” Journal of Competition Law and Economics, Vol. 7, No. 4 (December 2011.))

In our view, several factors account for this shift. First, in the 1970s, the Chicago school revolution exposed the intellectual bankruptcy of the S‑C-P paradigm. Over time, courts came to accept the bulk of the Chicago analysis. Because merger review is effectively law enforcement, the federal antitrust agencies had to follow suit, with the consensus growing in the post‐​Reagan era.

Second, politicians understood, at least implicitly, that these changes were good for them. An active market for corporate control—easy entry and exit from asset ownership—encourages investment by reducing the risk of being locked into failing assets and encourages efficiencies in asset ownership that maximize the value of those assets. These factors are good for the stock market and, with the shift from defined benefit to defined contribution pension plans, good for the middle class. While Republicans have generally supported business, an entire class of “Dow Jones” Democrats, led by President Bill Clinton, arose to support the idea that economic growth is good for America—and, more importantly (to them), good for political careers. Recent natural experiments in which some professional politicians deviated from the consensus on non‐​antitrust issues have proved the risk of seriously advocating archaic populist rhetoric. Today, real populism involves more of an attack on “crony capitalism” than big business.

Third, a regulatory approach based more on in‐​depth, case‐​by‐​case analysis of competition within a market, rather than a cookie‐​cutter approach that summarily condemns high concentration, had great appeal to the antitrust bar. Lawyers representing merging firms could get more deals through while simultaneously increasing billable hours for the more extensive analysis that the regulators expected. The revolving door nature of bureaucracy left the antitrust agencies staffed with past or future members of the defense bar. While there may be a fringe of true believers left who may occasionally find themselves in positions of authority within the agencies, even attorneys with a pro‐​enforcement bias don’t want to kill the goose that lays the golden egg.

Lastly, because the Chicago challenge to S‑C-P was two‐​pronged, involving conceptual questions and empirical tests, an effective challenge to the Chicago school that reinvigorates the S‑C-P paradigm would likewise require theory and empirics to make the point that structure does matter, or matters a great deal in four‐​to‐​three or three‐​to‐​two mergers. However, industrial organization academics have largely abandoned careful empirical work regarding structure/​performance relationships (one exception involves a series of retrospectives done by FTC economists, focusing principally on the petroleum industry, but also including the case study of the Heinz/​Beech‐​Nut baby food merger that Blackstone, Darby, and Fuhr cite). Instead, these academics focus on increasingly complex game‐​theoretic models of competition. The problem with this research is that the resulting claims of welfare reductions from mergers are weakened by the number of assumptions, some of which are impossible to check, embedded in the models. (For more on this, see our 2009 paper “Daubert, Science, and Modern Game Theory: Implications for Merger Analysis,” SSRN # 1268386, Supreme Court Economic Review [forthcoming].) Some of these models could be applicable to specific cases when the market‐​specific facts line up with a particular model (and, we argue in our paper, the model is empirically validated with data from the market), but the models are far from general propositions about the relationship between structure and performance. Overall, Chicago school analysis remains the controlling intellectual authority in antitrust.

Blackstone, Darby, and Fuhr are right to highlight the lack of empirical and theoretical support for the old Philadelphia National Bank presumption that high concentration is a good predictor of poor market performance. Well‐​done antitrust review already recognizes that investigations are often fact‐​intensive. While it is an academic cliché to end by calling for more research, we think it particularly relevant here. More merger retrospectives and other empirical analyses of structure and performance would be helpful in understanding and perhaps generalizing the conditions under which high concentration is injurious to consumers and when, in contrast, mergers enhance competition along one or more important dimensions. An even more interesting line of research would focus on how to apply the lessons learned from the Reagan merger reforms to other areas of public policy. The fundamental problem involves designing structures to align the political and public choice interests with those of the reformers.

The opinions in this letter are those of the authors and do not necessarily represent the views of the Federal Trade Commission, any individual commissioner, or any commission bureau.