| Abstract: | Strategy research explains why some firms outperform others, typically using profit rates, shareholder returns, and other continuous dependent variables. This paper investigates winning as the dependent variable, as measured by distributions of annual industry leadership in profits and returns to investors. This shift in dependent variable introduces alternative null models of competitive parity, including skew distributions derived from the natural sciences, and empirical distributions from nonbusiness domains such as chess, politics, sports, and beauty pageants. An empirical study of 20‐year leadership in U.S. industries shows that performance distributions in business follow statistical power laws resembling those in natural phenomena, and closely resemble distributions found in sports, politics, and other nonbusiness domains. The results support a presumption of persistent performance advantages in business, but show that business outcomes are indistinguishable from outcomes in the wider scientific and competitive landscape, and are amenable to explanation using relatively simple heuristics. The paper shows how the choice of null model shapes firm performance explanations, and explores the consequences of a more inclusive approach to null models in strategy research. Copyright © 2002 John Wiley & Sons, Ltd. |
