Academic journal article Quarterly Journal of Business and Economics

The Choice of Structure Measure in Industrial Economics

Academic journal article Quarterly Journal of Business and Economics

The Choice of Structure Measure in Industrial Economics

Article excerpt


Recent empirical research from the field of industrial economics fails to find the strong positive correlation between industry profits and market concentration that is hypothesized by the structure-performance model. Moreover, recent findings suggest that the correlation between firm market share and industry profit is stronger than the profits-concentration correlation and that when market share is included as a variable in regression models, concentration often is found to have a negative impact on profits (Ravenscraft, 1983; Branch, 1980; Gale and Branch, 1982). Kwoka (1979) offers an explanation for the observed negative coefficient for the four firm concentration ratio (CR4). His analysis uses nonpublic Federal Trade Commission data. His regressions of industry profit rates on the market share of the four largest firms reveal a negative coefficient for the share of the third firm and a statistically insignificant coefficient for the share of the fourth firm. Kwoka concludes, therefore, that the difficulty with CR4 arises from the inclusion of one irrelevant firm share (the fourth firm) and one firm share with an incorrect sign (the third firm). The recent literature discredits the tacit collusion or shared monopoly view that generally is associated with the structure-performance model and lends credibility to the revisionist or efficiency view which suggests that individual firm efficiency explains profit variation.

Proponents of the structure-performance model claim that the market concentration ratio is not the preferred structure measure for empirical models. Industrial economists generally regard the Hirschman-Herfindahl index as superior to the concentration ratio as a measure of market structure. A major advantage of the Hirschman-Herfindahl results from its sensitivity to both the number of firms and the relative firm size distribution.(1)

Jacquemin (1987) analyzes a number of concentration indexes and defines desirable properties of concentration measures in terms of their ability to capture changes in numbers of firms as well as changes in the firm size distribution. He places particular value on the capacity of a concentration measure to capture the influence of large firms on the market. Although Jacquemin cites the properties listed above as desirable, he argues that the ultimate test for the usefulness of a measure is determined by the statistical properties associated with the measure.

Using various statistical criteria such as adjusted [R.sup.2] and comparisons of statistical significance levels for the coefficients on structure variables, prior studies conclude that the Hirschman-Herfindahl provides no more than modest improvement over CR4 in the structure-performance model. Several studies use foreign data (Hause, 1977; Cowling and Waterson, 1976; McFetridge, 1973; and Jones, Laudadio, and Percy, 1973) to compare the explanatory power of alternative structure measures. In general, these studies conclude that the Hirschman-Herfindahl index provides little or no additional explanatory power as compared to the concentration ratio in the structure-performance model. Conversely, the work of Cowling and Waterson, who conclude that the Hirschman-Herfindahl is consistently superior to CR4 in explaining cross-sectional variation in the price-cost margin, provides the one exception to this conclusion.

Although the foreign studies provide some insight, differences in the social and legal frameworks restrict their direct application to the U.S. Studies of the U.S. economy generally cover a particular sector such as banking (Fraser and Rose, 1976) or the savings and loan industry (Marlow and Wright, 1987). These studies provide insight regarding specific industries, but they are of little value in assessing the relative usefulness of the Hirschman-Herfindahl index and CR4 in explaining variations in profit across all industries.

John Kwoka's research utilizes U. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.