For nearly a century, courts and commentators have wrestled with the question of when U.S. antitrust laws should apply extraterritorially. Initially, U.S. courts adopted a territorial approach, applying the Sherman Act' only to anticompetitive conduct that occurred on U.S. soil.2 In time, the territorial approach gave way to an effects approach, which extended the Sherman Act's reach to foreign conduct causing anticompetitive effects in the United States.3 A third alternative-the balancing approach-developed to moderate jurisdictional conflicts created by the effects approach.4 The story of U.S. extraterritorial antitrust enforcement ever since has been a contest between balancing and effects,5 with effects winning a narrow victory in the Supreme Court's Hartford Fire decision. Yet it is IMAGE FORMULA13
only recently that scholars have begun to consider the extent to which these three approaches-territorial, effects, and balancing-are economically efficient.'
To assess the economic efficiency of these three approaches, it is useful to divide them into two basic groups: (1) those that would assign legislative jurisdiction over an activity exclusively to one nation and (2) those that would allow concurrent legislative jurisdiction by more than one nation over the same activity. The territorial and balancing approaches belong in the exclusive-jurisdiction category. Each attempts to assign regulatory authority over a particular activity exclusively to one nation-the territorial approach to the nation where the activity occurs; the balancing approach to the nation with the strongest interest in the activity. The effects approach, by contrast, favors concurrent jurisdiction, allowing any nation that feels a sufficient effect from an activity to regulate it. Professor Weintraub and I (among others) have argued in favor of concurrent antitrust jurisdiction.8
The strongest argument against concurrent jurisdiction is that it results in overregulation.9 Professor Andrew Guzman has recently argued that concurrent jurisdiction is inefficient because it allows any country that suffers a loss in national welfare from an activity to prohibit it, even if the activity causes a net gain in world welfare.' In this article, I offer an economic defense of concurrent antitrust jurisdiction. Whether such jurisdiction is efficient, I will argue, depends critically on how one defines efficiency.
In Part II, I explain the three basic approaches to extraterritoriality and classify them as promoting either exclusive or concurrent jurisdiction. Part III sketches the economic argument against concurrent jurisdiction-that it may result in overregulation by allowing countries to prohibit activities that cause a net gain in world welfare. Part IV presents an economic defense of concurrent antitrust jurisdiction in two parts. First, exclusive jurisdiction is just as likely to lead to inefficient underregulation as concurrent jurisdiction is to lead to inefficient overregulation. Second, the level of regulation produced by concurrent jurisdiction can only be considered inefficient if one adopts a Kaldor-Hicks, rather than a Pareto, definition of efficiency. I argue that, at least in this context, a Pareto definition is more appropriate. In Part V, I note that even under a system of concurrent antitrust jurisdiction there are some changes in procedure that could improve efficiency, particularly in the area of merger regulation. Part VI concludes. IMAGE FORMULA16
II. APPROACHES To EXTRATERRITORIALITY
U.S. antitrust laws are generally silent on whether their extraterritorial application should be limited in any way," and for the most part Congress has been inclined to leave this question to the courts. 12 The courts have applied three quite different approaches to determine the extraterritorial scope of U.S. …