Dynamic strategic groups: Deriving spatial evolutionary paths

Recent theoretical developments in the domain of strategic groups, specifically those related to cognitive groups and strategic group identity, seem to suggest that strategic group membership is likely to be relatively stable over time and that firms in a strategic group co-evolve. Yet appropriate data analytic approaches that use information about firms over time to identify stable strategic groups and their evolutionary paths have been lacking. To overcome such limitations, this research proposes a new clusterwise bilinear multidimensional scaling model that can simultaneously identify (1) the number of strategic groups, (2) the dimensions on which the strategic groups are based, and (3) the evolution of the strategy of these groups over time. Our discussion encompasses various alternative model specifications, together with model selection heuristics based on statistical information criteria. An illustration of the proposed methodology using data pertaining to strategic variables for a sample of public banks in the tristate area of New York, Ohio, and Pennsylvania across three time periods (1995, 1999, and 2003) identifies two underlying dimensions with five strategic groups that display very different evolutionary paths over time. Post hoc analysis shows pronounced differences in firm performance across the five derived strategic groups. This article concludes with a discussion of the implications of the findings, as well as potential future research directions.