Document Type


Publication Date



Antitrust is rightfully concerned about the structure of markets as well as the bargaining that occurs in them. As a result, the absolute cost of redeploying resources can be just as important as the transaction costs of arranging for their movement. This paper examines several broad themes in antitrust, considering the role of various assumptions about the costs of getting resources moved toward superior positions and the ability of the antitrust system to facilitate this movement. Part II very briefly examines structuralism as a theory underlying antitrust enforcement, particularly its assumptions about the difficulty and costs of moving resources. Harvard School structuralism assumed that the costs of moving resources from lower value to higher value uses was high and often precluded competition from emerging. At the other extreme, the Chicago School assumed that resource movement was nearly cost free and that long-term monopoly was rarely sustainable as a result. Transaction cost economics then emerged as a welcome and unifying compromise.

Part III turns to barriers to entry or rival expansion, looking particularly at the differing definitions provided by Harvard and Chicago School economists and showing why the Harvard definition is superior for antitrust purposes today. That may not have been true in the 1960s when proposals for various forms of “no fault” monopolization were under serious consideration. Part IV discusses antitrust’s two principal tests for welfare, total welfare and consumer welfare, and shows how they are related to our assumptions about the costs of movement. Consumer welfare tests generally reflect significant doubt that surpluses that accrue to producers from economies or otherwise will be passed on to consumers.

Part V turns to practices, arguing first that we need to rethink current antitrust doctrine about refusal to deal in dominated networks, which are networks that both dominate the markets in which they operate and are themselves dominated by a single firm. Although their advantages are many, one important effect of networks is to magnify the costs of resource movement. Next this paper examines some problems of vertical integration and product complementarity, as well as the contributions that transaction cost analysis can provide in cases involving asset specificity and the possibility of double marginalization. We also examine some specific problems of pricing and vertical control, looking in particular at the wide range of theoretical attacks on and defenses of so-called loyalty discounts and bundled discounts. In particular, it faults policy making based on models with restrictive and sometimes idiosyncratic assumptions and untested conclusions.


78 Antitrust Bull. 67 (2012)