ABSTRACT
The concept of innovation system conveys the idea that innovations do not originate as isolated, discrete phenomena, but are generated by means of the interaction of a number of entities or actors/agents. The set of actors and interactions has some specific features that are conserved over time, and it behaves as a whole in a large number of circumstances. These characteristics are shared by national (Lundvall, 1988, 1992; Freeman, 1987; Nelson, 1988, 1990, 1992; Niosi et al., 1993), regional, sectoral (Breschi and Malerba, this book; Guerrieri and Tylecoate, this book) or technological (Carlsson and Stankiewicz, 1991; Carlsson, 1994) innovation systems. They can all be represented as sets of institutional actors and interactions, having as their ultimate goal the generation and adoption of innovations at some level of aggregation (country, region, industrial sector, technology, etc.). The important theoretical and policy problem posed by these systems is that innovations are generated not only by individuals, organizations, and institutions, but by their, often complex, patterns of interactions. It follows that institutional and organizational configurations are important determinants of economic development and growth. This conclusion represents an important challenge at the theoretical and policy levels, because traditional theories of growth have been largely a-institutional. The historical specificity and the institutional nature of national systems of innovation (NSIs) cannot be predicted or explained by traditional economic theories. In this chapter an attempt will be made to justify the existence and to explain some of the main features of NSIs on the basis of evolutionary theories. However, a number of the conclusions are also applicable to other systems of innovation.
