Introduction

Today the term national innovation system appears in several different domains within social science and engineering, and it is widely used in policy circles all over the world. The concept reflects an assumption that the pattern of innovation differs across countries and that such differences can be explained by systemic features. The components of the innovation system are different, and they are linked differently to each other and such differences in economic structure and institutional set up are reflected in the rate and direction of innovation.

We will take as starting point ideas presented in the very first contributions that made use of the innovation system concept, Freeman (1982) and Lundvall (1985). There is some overlap between them but the perspectives are quite different. Freeman's analysis refers to macro-phenomena and to international trade and development, while Lundvall (1985) refers to the micro level where innovation is seen as shaped by user– producer relationships. We will argue that they are complementary and that they can be used to span and dissect important themes in the more recent literature on innovation systems and global value chains.

The concept national innovation system may be seen as a new combination of two different perspectives, one developed within the IKE group at Aalborg University and one developed at Science Policy Research Unit (SPRU) at Sussex University. The concept came out of bringing together an understanding of innovation as rooted in the production system (Aalborg) and an understanding of innovation as rooted in the science and technology system (Sussex).

The Aalborg approach was inspired by the concept ‘national production systems’ as it was used by French Marxist structuralists such as Palloix (1969) and de Bernis (1968). Esben Sloth Andersen (1992) criticized and developed these ideas by introducing an evolutionary perspective with focus on innovation with the aim to overcome the limitations of what he saw as a too static framework. Another important inspiration for the Aalborg group's work on innovation systems came from Björn Johnson (1992) who linked innovation and learning to the socioeconomic characteristics of national institutions. Lundvall (1985) took inspiration from early works by Andersen and Johnson while studying user– producer interfaces as reflecting economic structure as well as institutional characteristics.