Industrial district heterogeneity and competitive advantage: Evidence from Italy

This paper seeks to contribute to the ongoing debate concerning the role of heterogeneity for the GC competitive advantage. With this aim, we focus on different sources of heterogeneity and study how the different level of heterogeneity affects the performance of GCs. Specifically, we consider the GC heterogeneity due to: 1) the variety in the firms' technological specialization; 2) the diversity in the firms' organization; 3) the variety of the external knowledge brought into the GC; 4) the absorptive capacity of the GC. We conduct an empirical research on 32 Italian District Provinces (DPs), by measuring for each DP the level of heterogeneity and the level of performance. To do this, a set of variables both for the sources of heterogeneity and for the performances has been defined. In particular, the variety in the firms' technological specialization has been measured by using two different concentration indexes: the Gini coefficient and the Theil index, both aim at capturing the level of concentration of both firms and employees in a specific manufacturing sector. The diversity in the firms' organization has been computed by using the Gini coefficient and the Theil index. In this case the indexes measure the level of concentration of both firms and employees in a specific class of employees. As for the variety of the external knowledge brought into the DP, we have evaluated the breadth of the international trade linkages of each DP. Specifically, by using the Gini coefficient, we have calculated the level of concentration of both imports and exports in a specific country of destination/origin (Boschma and Iammarino, 2009). As for the absorptive capacity, we have calculated the number of graduates in technical-scientific fields in each DP. Finally, as measures of performance, we have considered three indicators: the GDP, the number of patents developed in each DP, and the firms' death rate. A cluster analysis has been applied to process the data collected.