A Simultaneous Equation Analysis of the Structure and Performance of the United States Petroleum Refining Industry

Abstract

formulation of the Structure-Conduct-Performance paradigm [I3], [14] has been predominantly concerned with the relationship between industrial concentration and profitability. In fact, it may be one of the most frequently tested relationships in economics. Professor Leonard Weiss compiled a table of 46 concentration-profitability studies for the I974 Columbia Law School Conference on Industrial Concentration and referred to another eight in a footnote [35]. Without exception, these studies specify and estimate single-equation models in which causality runs from structure to performance. The thesis of this paper is that these studies are misspecified-that structure may indeed influence profitability, but that profitability may, in turn, affect capacity expansion and net entry, two determinants of industrial structure. A correct specification of the economic phenomena would be with a set of equations that are necessarily both dynamic and simultaneously related.1 If this more detailed specification of the Mason paradigm is to have any value, however, it must have implications alternative to those accruing from single-equation studies. In this paper a dynamic simultaneous equation model will be specified, and two such implications will be examined. One relates to a possible source of bias in the single-equation results, and the second concerns the impact that changes in exogenous variables may have on endogenous variables. Although there has been debate as to the validity of the findings of the