Innovation, knowledge and relations – on the role of clusters for firms’ innovativeness

ABSTRACT

For more than two decades, theories on regional clusters have inspired economic and structural policies at the European, national and regional levels. Based on the assumption that clusters generate innovation, policy-makers at all levels of governance have adopted instruments and mechanisms to stimulate, resource and sustain clusters. Despite the considerable attention paid to the clustering phenomenon, empirical evidence on to what extent firms’ innovation activities benefit from operating in clusters is scarce and inconclusive. This paper contributes to the micro-foundation of clustering effects by examining the characteristics and activities of cluster firms in relation to their innovativeness. Bridging innovation, management and cluster theories, it is argued that structural and relational embeddedness, relational capital and absorptive capacity influence clustered firms' innovativeness. Partial least-squares structural equation modelling of data from 104 firms in two software and information technology service sector clusters reveals that firms’ structural embeddedness (i.e. frequency of interactions) in clusters and external networks facilitates innovation cooperation. Firms’ absorptive capacity reinforces this positive effect of cluster-internal interactions on innovation cooperation. Results also suggest a substitution effect of trust as relational control mechanisms for formal control mechanisms within the cluster. However, the study finds no significant impact of firms’ innovation cooperation within the cluster (i.e. relational embeddedness) on their innovation success.

KEYWORDS:

• Cluster
• relational embeddedness
• collaboration
• innovation
• relational view
• knowledge-based view
• structural equation modelling

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Agglomeration economies, as used herein, refer to Marshall-Arrow-Romer (MAR) externalities, i.e. knowledge spillovers, of firms in the same industry (Beaudry & Schiffauerova, 2009; Hervás-Oliver, Sempere-Ripoll, Rojas, & Estelles-Miguel, 2017).

2 Routines refer to firm-specific competences, repeated interaction patterns, operational processes etc. (Boschma & Frenken, 2006, 2011).

3 These numbers refer to core NACE sections B, C, D, E, H, J, K, and divisions 46, 71, 72, 73.

4 These results are consistent with the Community Innovation Survey (CIS 2010) for Germany, which showed a share of 38% for firms with continuous innovation activities (2009), and shares of 91% (NACE 62) or 93% (NACE63) for innovating firms (Rammer & Schubert, 2011; ZEW, 2012).

5 Variables compared are firm age, degree of specialisation, number of employees and turnover development.

6 SmartPLS setting included a path weighting scheme, a maximum number of 104 iterations, a stop criterion of 10–7, and initial outer weights of +1.

7 HTMT of correlations ‘is the average of the heterotrait-heteromethod correlations (i.e. the correlations of indicators across constructs measuring different phenomena), relative to the average of monotrait-heteromethod correlations (i.e. the correlations of indicators within the same construct’ (Henseler, Ringle, & Sarstedt, 2015, p. 121).

8 The model’s predictive relevance is calculated by applying blindfolding procedure with an omission distance of 7.