JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. DESPITE the fact that collusion among American firms has been actively prosecuted under the antitrust laws for many years, little is known about the effects of this policy. Such ignorance is partly attributable to the problems that accompany empirical investigation. It is, quite simply, difficult to isolate the effects of any mode of business conduct on observable market phenomena. A second source of ignorance is the per se illegality of collusion under the Sherman Act. Since the existence of the offensive conduct is virtually the sole issue considered by United States courts, exploration of surrounding circumstances has been minimal. This paper attempts to shed some light by examining evidence on the characteristics of collusive firms during the period 1958-67.
BACKGROUND AND HYPOTHESESCollusion is widely regarded in the literature as a plausible, and even a commonplace, method of resolving oligopolistic indeterminancy [Bain, Chamberlin, Machlup, Scherer, Stigler]. A number of hypotheses, relating both to causes and effects, can be drawn from the traditional body of competition theory, and have been usefully supplemented by writers such as Stigler. For several reasons, however, these hypotheses are rather tentative, even as a broad guide to empirical investigation.Initially, expected relationships between a specific manifestation of business conduct and actual firm or market performance, tend at best to be imprecise. Knowledge of the stratagems employed by companies-e.g., 'price fixing'-simply fails to provide much basis for predicting the nature or quality of final market results.Secondly, hypotheses about collusion often involve not only assumptions about firms' objectives, but also such matters as the subjective probabilities that may be assigned to 'getting caught' and the degree of discomfort that antitrust prosecution may engender. This simply reflects the uncertainty of the collusive outcome; but whereas the resulting theoretical difficulty is hardly unique, it is not obvious how alternative contingencies are to be assessed objectively. Measurement of the conditions under which collusion is expected to occur is therefore complicated. All use subject to JSTOR Terms and Conditions 224 PETER ASCH AND JOSEPH J. SENECA Finally, and perhaps most seriously, empirical investigation is hindered by the fact that observable collusion in the United States is effectively a legal phenomenon. The study of large numbers of legal cases will necessarily reflect the criteria employed by public policy. This implies that it is impossible to examine the characteristics of collusion in some pure sense. Rather, what is examined is a mixture of collusive episodes and institutional responses to the...