Estimation of Structure-Profit Relationships: Comment

Recently, the Federal Trade Commission (FTC) has accused breakfast cereal manufacturers of illegal monopolization (1971).A major basis for the charge is the FTC belief that high advertising expenditures by the companies create effective entry barriers. One FTC study (1969) reached this conclusion after finding that the coefficient of the advertising/sales ratio in a multiple regression equation explaining firm profit rates was positive and significant. Several other of these so-called market structure-performance studies have recently been completed (Marshall Hall and Leonard Weiss, Vernon (1972)). However, in a recent issue of this Review, Blake Imel and Peter Helmberger (hereafter, I-H) presented some findings which cast serious doubt upon the validity of these studies. According to their carefully specified profit equation for a diversified firm, the significance of advertising as a barrier to entry depends critically on the size of an unknown parameter termed the "omega ratio." For example, for one specification of the profit rate equation, the estimated t-ratios varied from 3.31 to 1.86 to 1.29 as the omega ratio ranged from 0.0 to 0.5 to 0.8. Since I-H were unable to estimate this ratio and it is not known a priori, it is important to seek evidence as to its magnitude. In this paper, after a brief review of the I-H model, we suggest a method for estimating the omega ratio. We then present some estimates for a body of data similar to that used by I-H. Our results indicate that the estimated ratio (and hence the significance of the estimated coefficients) is approximately the same for two alternative specifications of the model. More importantly, the estimated advertising coefficients, conditional on our estimated omega ratios, are positive and statistically significant. Thus, this note tends to corroborate the hypothesis that advertising serves as an effective barrier to entry.