This paper investigates the hypothesis that social capital is a crucial factor in explaining technological innovation in industrial districts or, more generally, in firms' clusters. Nevertheless, the fundamental idea is that social capital is not, as generally suggested by the socio-economic literature, an individual attitude towards 'something good', within a public good or collective good conceptual framework: the effect of SC in networks depends on its complementarity to private benefits and on the opportunity costs of purely private capital factors.
A dynamic theoretical model that assumes social capital as the public component of the impure public good R&D is developed. It shows that the 'civic culture' of the district area in which the firm works is not sufficient as an incentive to increase her investment in social capital, because this investment strictly depends on the economic convenience in investing in the impure public good. Social capital /networking dynamics might positively and complementary evolve only if the opportunity cost of investing in innovation is sufficiently low.
We consequently focus our attention on a specialized industrial district located in the Emilia Romagna region - the biomedical district of Mirandola (Modena) - characterised by a strong pattern of innovative activity. Using technological innovation as dependant variable, we observe that R&D and networking/social capital arise as complementary driving forces for innovation outputs. When empirical evidence confirms that this complementarity plays key role, and consequently strong links exist between market and non-market dynamics relating to firms, the role for policy actions targeted to social capital is wider. The policy effort should be targeted toward both market and non-market characteristics taken together, rather than solely to the production of (local) public goods (social capital) or innovation inputs as independent elements of firm's processes. Social capital alone is not a sufficient input for innovations and growth: economic incentives matter. On the other hand, whenever SC dynamics are crucial for R&D private investments, the effect of economic incentives depends on the presence and degree of their complementarity.