Empirical evidence from economic literature suggests that innovative activities based on extensive interactions between industry, university and local government have shown high developing performances. These studies placed emphasis on spatial proximity as the basis for dense and embedded networks of interaction and trust. Then, in many countries, institutional arrangements have been designed to start regional technological districts. A starting point for the evaluation of the policies to implement innovative networks (create and/or strengthen) could be the use of a network approach. It appears that networks are very different (size, density, reachability, diversity, openness, stability) and that the kinds of links that connect two organizations could have different key dimensions (formality, intensity, symmetry, reciprocity, multiplexity and trust) (Jackson, Wolinsky, 1996). In our opinion, evaluation criteria, beyond directly measurable input and output indicators of R&D policy, could be based on the effect of these policies on the network structures and network relationships. The strategy proposed in this paper aims to analyze the policy started from ten-years to support the start-up of the Italian Technological Districts (TDs) located in different regions. In particular, the main purpose is to explore how network structures influence network efficiency, in terms of knowledge exchange and innovation diffusion among DTs members. The main steps of the strategy are: i) to define networks through the identification of relationships between actors involved in geographical districts; ii) to describe and compare the observed networks in order to highlight their structural differences; iii) to find relationships that links network indicators to the characteristics of the environment in which such policies are implemented in order to analyze the effectiveness of different subsidies to networks.
