Determinants of knowledge flows and firm performance in industrial clusters

Abstract

Industrial clusters have been in existence for more than a century. Numerous explanations have been provided for the existence and continued dominance of industrial clusters. Most of the explanations before 1990 focused on cost/resource based advantages. Some of these advantages include better access to raw materials / intermediate inputs, better infrastructure facilities, availability of skilled labour to firms in clusters. Recent explanations highlight the role of knowledge spillovers based learning for f m in clusters. These explanations suggest that clusters exist due to local knowledge spillovers i.e., knowledge flows that take place between entities (firms, institutions, etc.) in a cluster. The importance of non-local knowledge flows i.e., knowledge flows From outside firm's location has been largely ignored. Although, studies identify knowledge flows I knowledge spillovers as key for existence and continued dominance of industrial clusters, no single study has focused on exploring the determinants of these knowledge flows and knowledge spillovers and if these determinants help in better performance of firms in clusters.

The study hypothesizes that capabilities help in better performance of firms in clusters. Apart from capabilities, performance differentials are also possible due to differences in locational characteristics. As mentioned, these locational characteristics could include availability of raw materials I intermediate inputs, presence of certain kinds of infrastructure facilities and availability of skilled labour leading to better performance of firms in clusters. If capabilities affect performance, what determines the level of capabilities among cluster and non-cluster firms? Since knowledge flows are facilitated through networks, this study hypothesizes that networks/ network capital is a significant determinant of capabilities. We contend that knowledge flows and knowledge spillovers are important determinants of firm capabilities. Moreover, key locational differences such as availability of skilled labour, presence of R&D institutions, training facilities could lead to differences in sources of knowledge. These differences in sources of knowledge can lead to differences in capabilities across cluster and non-cluster locations too.

In order to study the above relationships, we undertook a large sample survey of firme in two industries - electronics and information technology (IT) to generalize results across industries. The survey was carried out in Bangalore, Pune and National Capital Region (NCR) clusters as well as multiple non-cluster locations to capture the differences between firms in clusters and outside clusters. To explore factors underlying performance differentials between firms in cluster and firms outside clusters, data was collected on performance, capabilities, linkages (network/ social capital), locational advantages, infrastructure, government policy and sources of capabilities at the firm level. Our study is the first of its kind to empirically operationalize the constructs of capabilities and social/network capital at the firm level. The capabilities were captured as a product, process and practice capabilities. However, due to problems of measurement, we could only focus on process and practice capabilities in our empirical analysis. Further, social/ network capital was measured on the basis of number, variety and importance of linkages (customer, supplier, consultants, universities, etc.) that the firms have. The linkages (networks) have been further distinguished on the basis of their geographic origin: (a) Linkages with entities located within the city/cluster where the firm is located(local networks); @) linkages with entities that are located outside the city/cluster but within the country (national networks); and (c) linkages with entities that are located outside the country (international networks).

Overall , our results show that capabilities are a significant determinant of performance for IT and electronics firms in clusters, whereas capabilities do not significantly affect performance of IT firms outside clusters. Thus, we can infer that capabilities of firms in cluster help in their better performance. Further, apart from capabilities, location is significant too. From this, we can infer that firms perform better in clusters due to reasons other than capabilities too. Apart from capabilities, we show that availability of raw material, support services, availability of information on fairs and exhibitions and presence of industry associations are significantly better in cluster. This might have contributed to better performance of cluster firms. Further, we show that network/network capital is a significant determinant of capabilities of IT and electronic firms across cluster and non-cluster locations. However, there are interesting differences between networks 1 network capital contributing to capabilities of IT and electronic firms across cluster and non-cluster locations. Our results show that international customer network capital and other local network capital are determinants of capabilities of IT firms in clusters, whereas national network capital (customers and others) significantly determine capabilities of IT and electronic firms outside clusters. Thus, non-local networks significantly determine capabilities of non-cluster firms, whereas local network contribute to capabilities of IT firms in clusters. Our results also show that economies of scale drive capability formation in clusters, whereas scale economies do not contribute to capability formation of firms outside clusters. Apart from the above, differences in important sources of knowledge, availability of skilled labour, presence of R&D and other training facilities could also have lead to differentials in capabilities of firms within and outside clusters. Bigger firms outside clusters might not have derived economics of scale for capability formation due to absence of these facilities. To summarize, this study has empirically shown using large sample data that linkages build capabilities and capabilities contribute to performance of firms. Being the first study of its kind, several gaps remain. For example, despite our best efforts many problems related to the measurement of capability and network capital remains unresolved. Similarly, we have not been able to explore the lags between knowledge flows, capability formation and performance. These can be pursued in future research.