Multisectoral Growth and National Innovation Systems

Terje Lensberg and other participants of the NOPEC conference on "Innovations, Policy and Society" (Oslo, December 1997) are gratefully acknowledged. 2 Because of their emphasis on bounded rationality and on technology as a very special kind of good, many economists dealing with the economics of technical change and the broader field of evolutionary economics tend to apply a form of theorising that de-emphasise formal abstraction and the related modelling criteria of rigour, consistency and robustness (see e.g. Dosi et al. 1988, Freeman and Soete 1997). Instead they emphasise relevance for description and policy-making in relation to the real world. The term used by Nelson and Winter (1982, 46) for this style of research is "appreciative theorising". attempt to avoid formal modelling per se. For them it is often the only way of theorising about important economic phenomena, given the gap between these phenomena and the scope of available formal modelling tools. There is, however, an inherent contradiction in ambitious appreciative theorising: it tries with rather simple and informal means to capture phenomena that are more complex that the ones treated by mainstream modelling tools. This is especially obvious in the recent studies on hyper-complex phenomena like national innovation systems and technology systems. Such studies started with e.g. papers Now we see many national studies, intensified descriptive work by international organisations as well as collective works edited by e. In relation to this new wave of innovation systems studies there is a need of a new type of model which is on the one hand open to the issues confronted by appreciative-style researchers and on the other hand able to capture systematically the core interdependencies of an economic system with heterogeneous firms engaged in exploring and exploiting an expanding set of economic activities. The paper outlines a type of model which is designed for this purpose. To begin with it is, however, useful to consider the debate on appreciative theorising. Appreciative theorising is good for exploring a new problem area and for establishing a preliminary agenda for empirical and conceptual research. But in a certain sense 3 appreciative theorising is too flexible when confronted with the huge number of ways of reducing the complex phenomenon of an innovation system into a manageable form: it does not constrain each researcher from choosing an idiosyncratic specification. This problem becomes obvious if a research area like that of innovation systems grows both in terms …