STRATEGIC LEADERS OR STRATEGIC GROUPS - A LONGITUDINAL DATA ENVELOPMENT ANALYSIS OF THE UNITED-STATES BREWING INDUSTRY

This paper builds on strategic group theory which was originally intro duced to help explain the observed variation in firm profitability acr oss an industry. This paper applies Data Envelopment Analysis (DEA) to identify both strategic leaders, the 'best practice' players in the i ndustry, and strategic groups to examine which are the source of the m ost sustained heterogeneity in the performance of U.S. brewers. Contra ry to the prevalent research approach in Strategy and Organization res earch which places a premium on empirical analyses that maximize expla nation of average behavior (Daft and Lewin, 1990), this paper examines extremal observations as a means of studying or inducing theories abo ut best practice, such as best strategies or most effective organizati on design (Lewin and Minton, 1986). The paper argues that significant insights, new knowledge, and unexpected theories can arise from studyi ng the best or worst of a population. The paper applies DEA to a longi tudinal reanalysis of U.S. brewing data to demonstrate its usefulness for obtaining new insights from identifying and studying industry outl iers. The paper raises important questions about the construct of stra tegic groups and illustrates how competing theories can be deduced and analyzed for explaining performance heterogeneity in the U.S. brewing industry.