Search

The provision of public funds to private firms for the purchase of services, particularly knowledge-intensive ones, has received so far little attention from the evaluation literature (Bakhshi et al., 2015; Bruhn et al., 2018 are notable exceptions). These interventions often target small and medium-sized enterprises (SMEs), providing them with a small amount of public funds that reduce their cost of purchasing services (Storey, 2003). Public funding can take the form of a direct subsidy or a voucher, which firms must use to purchase services from accredited service providers, or sometimes from any provider freely chosen by the beneficiary firm (OECD, 2000; Storey, 2003; IEG, 2013).

Stochastic actor oriented models (SAOM) are of growing importance to study network dynamics focusing on the theoretical micro-mechanisms that induce the evolution of relations among a set of social actors. SAOM represent a suitable methodological framework to investigate the evolution of policy-driven innovation networks among heterogeneous actors. Drawing on a set of network policies implemented in the Italian region of Tuscany during the 2000s this paper investigates the time evolution of policy-driven innovation networks at regional level. Specifically we analyse how actors' relationships have evolved according to the following aspects: (i) propensity to collaborate with actors who are part of their existing network as opposed to experimentation of new relationships outside the group (trust/reputation effect); (ii) propensity to collaborate with actors sharing similar characteristics (homophily effect); (iii) propensity to collaborate with more popular actors, leading to the formation of a core of agents "controlling" the policy programme ("Matthew" effect).

The last twenty years have witnessed the diffusion of regional innovation policies that support networking among heterogeneous organisations (e.g.: firms and universities; small and large firms). The implicit assumption is that policies are needed to stimulate interactions that would not occur spontaneously, but whose presence would be desirable. Since it would not be possible, nor appropriate, to identify the specific partners with whom an organisation should collaborate, what policies can do is provide incentives in order to encourage collaboration among heterogeneous organisations, establish a general framework of rules and then leave the participants free to organise their innovative activity. However, the presence of general and flexible incentives does not ensure that undesirable behaviours (such as the formation of an oligarchic core of agents driving the funded projects, or the reiteration of established linkages) are not adopted anyway. Our study seeks to deepen the reflection on these points, introducing two main original elements with respect to the existing literature. First, we observe small scale regional policies targeting SMEs, instead that the large-scale EU programmes. Second, we adopt a dynamic approach, rather than a comparative static approach as usually adopted. Our empirical analysis explores how collaboration has evolved within a set of policies targeting innovation networks, which have been implemented by the regional government of Tuscany (Italy) in the programming period 2000-2006. By using a stochastic actor oriented approach, we investigate how their relationships have evolved according to several aspects: (i) reiteration of pre-existing linkages as opposed to experimentation of new relationships (stability or experience effect); (ii) collaboration with agents possessing complementary rather than similar knowledge and competencies or abilities (homophily effects); (iii) formation of a core of agents “controlling” the policy programmes (Matthew effect).