See Change? Conceptualizing and Measuring the Temporal Dynamics of Industry Clusters

Abstract

This article outlines a temporal dynamics approach to the study of industry clusters. Despite extensive work on clusters, little attention has been paid to their temporal dynamics. We propose that understanding cluster dynamics is important, however, because clusters are seldom stable entities, and cluster dynamics may have strategic implications not accounted by existing approaches. Within this context, we develop a novel measure of cluster dynamics. Applying this measure to data on establishment population in the U.S. computer and semiconductor industries, we document wide variation in cluster size over time, both within and across regions. Furthermore, utilizing data on patents, we find that cluster dynamics correlate with localized knowledge spillovers in ways different from cluster size, suggesting that our approach may offer novel insight to strategy researchers.