The Structure Within Industries and Companies' Performance

THE theory of industrial organization has by and large viewed the industry as a homogeneous unit. Firms in an industry are assumed to be alike in all economically important dimensions except for their size. In this context, a considerable body of research posits that many industries are characterized by the existence of market power shared among their firms.1 This market power results, following Bain and others, from the presence of structural barriers to the entry of new competition and from industry characteristics (such as seller concentration) which lead to the recognition of mutual dependence among competitors and thereby stop interfirm rivalry short of the competitive ideal. Barriers to entry equally protect all firms in the industry from new entrants and the fruits of mutual dependence recognition accrue symmetrically to all firms, as well. Thus market power is an asset shared by all firms in an industry in proportion to their sales. Above-normal profits are the manifestation of this market power, and the profit rates of firms in an industry should be equal except for random (and hence uninteresting) disburbances. This theory of industrywide or "shared asset" profit determination, versatile as it has proven to be, is at odds with both commonplace observation and a small but growing body of systematic empirical studies. All firms in the typical industry are clearly not alike: they follow very different strategies along dimensions such as their degree of vertical integration, breadth of product line, distribution arrangements, and so on. An industry's member firms also frequently earn rates of return on invested capital that exhibit considerable variance. For example, General Motors has persistently outperformed Ford, Chrysler, and American Motors.2 IBM outperforms other computer manufacturers. Crown Cork and Seal (a smaller firm) persistently outperforms National Can, American Can and Continental Can. Finally, there are several statistical investigations of profitability that have produced results inconsistent witth the shared asset theory of market power. Demsetz (1973) has, for example, found that the profits of smaller firms are not higher in concentrated industries than they are in unconcentrated ones, though the profits of larger firms are.3 Shepherd (1972) argued that market power is firm-specific and dependent on the -firm's own market share, implying that profit rates increase systematically with size within an industry. Yet Marcus (1969) found that the relationship between firm size and profitability within an industry is erratic, with some industries exhibiting positive relations, some negative relations and others no apparent statistically significant relation at all. The purpose of this paper is to present a theory of the determinants of companies' profits which rests on the structure within industries as well as on industrywide traits of market structure. Built on the concepts of strategic groups and mobility barriers, this theory provides an explanation both for stable differences in competitive strategies among firms within an industry, and for persistent intraindustry profit differences among firms. I will show that the theory is consistent with the previously reported statistical results noted above. Next, I will present the supportive results of a new statistical test which examines the structural determinants of profitability for firms differently situated within their industries. Finally, I will show that the empirically supported theory refutes the Demsetz/Mancke view that large firms earn higher profits largely because they are more efficient or lucky, and not because they possess market power. Received for publication September 13, 1977. Revision accepted for publication March 20, 1978. * Harvard University. This study was supported by the Division of Research at the Harvard Graduate School of Business Administration and by the General Electric Foundation. It also benefited from comments by R. E. Caves and Michael Spence. 1 This is the familiar structure-conduct-performance paradigm of industrial organization. See Bain (1956). Scherer (1970) provides a comprehensive review. 2 For these and the other firm profitability data, see the helpful compilations in Forbes, January 1, 1977 and earlier years. 3 A consistent result is obtained by Osborn (1970), who finds that concentration has little (or a negative) effect on the profitability of small, fringe firms in an industry.