Do Markets Differ Much

This essay reports the results of a crosssection study of differences in accounting profitability that sheds light on some basic controversies in industrial economics. Most previous cross-section studies in this field have been concerned with testing hypotheses about structural coefficients in models meant to apply to essentially all markets. As we have learned more about the difficulties of constructing such general models and of performing tests on their structural parameters properly, structural cross-section analysis has fallen out of fashion. In contrast to most of the cross-section literature, the analysis reported here is fundamentally descriptive; it does not attempt directly to estimate or to test hypotheses about structural parameters. I hope to show by example that one can perform illuminating analysis of cross-section data without a host of controversial maintained hypotheses. Cross-section data can yield interesting stylized facts to guide both general theorizing and empirical analysis of specific industries, even if they cannot easily support full-blown structural estimation.' One can view the sort of search for stylized facts conducted here as either a replacement for or an input to interindustry structural estimation, depending on one's feeling about the long-run potential of that research approach. This study also departs from much of the cross-section literature by being fundamentally concerned with the importance of various effects, not just with coefficient signs and t-statistics. In particular, this essay provides estimates of the relative importance of firm, market, and market share differences in the determination of business unit (divisional) profitability in U.S. manufacturing. Using 1975 data from the Line of Business Program of the U.S. Federal Trade Commission (FTC), I find support neither for the existence of firm effects nor for the importance of market share effects. Moreover, while industry effects apparently exist and are important, they appear to be negatively correlated with seller concentration in these data. Section I relates firm, market, and share effects to current issues and controversies in industrial economics and thus supplies the motivation for our empirical analysis. The remainder of the essay treats the data and