The applicability of the concept of national innovation systems to transition economies.

The institutional set-up regulates the relations between individuals or groups and thus influences the pattern and subject of communication and interaction in the economy. As innovations are mainly results of interactive learning processes, institutions affect innovation. Influencing innovation processes by the institutional set-up takes place through different channels emerging from three basic functions of institutions (Edquist & Johnson 2000: 176f). First, institutions reduce uncertainty about the behavior of others by providing information. As a result, relationships are stable enough to make communication possible. Second, they manage conflicts and cooperation between different parties. Third, institutions provide incentives by establishing salaries and wages, income taxes, intellectual property rights, etc. These functions do not only create stability to society, but also the needed stability for technical change. Because of institutions, innovative activities 'can to some extent be handled systematically and in organized forms, and this makes more concentrated efforts and more rapid progress possible' (Johnson 1992: 26). Even if the institutional set-up of an economy is characterized by a large degree of stability, completely rigid institutions would prevent economic development. Institutions do change continuously. In general, there are two different forces behind institutional change. First, different conflicts concerning political and social relations or income and power distributions repeatedly have been forces behind institutional and social change. Second, the classical force behind institutional change is technological change. Freeman and Perez (1988) illustrate this in their concept of techno-economic paradigm. They argue that a change of techno-economic paradigm has such extensive consequences for all sectors of the economy that its diffusion is accompanied by a major structural crisis of adjustment, in which social and institutional changes are necessary in order to achieve a better 'match' between the new technology and the old socio-institutional framework.

Lundvall (1992a) argues that innovation is rooted in the existing economic structure of a nation, as innovation reflects learning and learning is partially a result of routine activities. A second characteristic of innovation systems is the importance of the institutional set-up of an economy, because the relative stability of institutions over time enables organizations to act and survive in an uncertain world and provides the opportunity of innovative activities. The impact of the institutional set-up on the learning process appears in manifold ways. For instance, if a great part of the society resists change, as they tend to be critical towards innovations even if the advantages are obvious, then this informal institution is an innovation-blocking one. The stock of knowledge will not change, even if necessary and this will hamper economic development. Furthermore, institutions can directly affect the rate of innovations. Adequate incentive systems are necessary to support innovative activities of organizations. One possibility is to design educational policies in such a way that human resources of the economy are able and willing to gain and master new technology. This may help to increase the rate of realized innovative ideas. Simultaneously, the institutional set-up depends to a certain degree on the learning process of an economy, as new knowledge influences the norms and habits of society or gives reason to new policies and governmental rules. Figure 2 illustrates this inter-relationship.

[FIGURE 2 OMITTED]

'Structure-based' versus 'process-based' approaches of innovation systems

Comparing the definition of NIS with important aspects of technological development shows that the elements contained in a NIS are necessary for a successful process of development. The combination and effectiveness of different institutions, their interactivity and the emphasis of the relevance of knowledge and diffusion are elements constituting a NIS. Both concepts argue that factors like culture, factor endowment or institutional arrangements are important for the future rate and direction of economic development, focusing both on the path-dependence of technological change. Trying to analyze the prospects of a technologically backward country to develop raises the question about the actual state, effectiveness and development of the country's NIS.

The elements of NIS described in this article so far are components and relations among these in terms of main actors and institutions shaping the innovation process of a country. The main literature on innovation systems defines organizations and institutions as its main components, as they constitute the 'structure' of such a system. But this 'structure-based' approach includes a fundamental weak point. Different definitions of innovation systems point out different institutional and organizational set-ups and particular set-ups vary among systems; every definition in this respect picks out certain subjective determinants influencing an innovation system. But what are the determinants of innovation? Until now, this question can not be answered systematically and in detail. How can we decide about the importance of certain determinants? Edquist (1997) tries to solve this problem by defining NIS in a more general way as 'all important economic, social, political, organizational, institutional and other factors that influence the development, diffusion and use of innovations' (Edquist 1997: 14). Even if this definition includes all determinants influencing an innovation system, the overall problem remains: One has to pick out certain determinants when trying to analyze NIS. But, this will cause biased results, as institutions and organizations can have different duties and responsibilities in terms of activities in different countries. Therefore, we will follow an approach that has been developed at the 'Centre for Innovation, Reserach and Competence in the Learning Economy' (e.g. Edquist 2004), Edquist & Hommen (forthcoming), Hogselius (2005). The point of departure is the overall function of innovation systems that is considered to be creating, diffusing and using innovations (Edquist 2004: 190). Different kinds of activities are needed to fulfill this function,. Activities are defined as factors that influence the development and diffusion of innovations and can therefore be interpreted as the determinants of the overall function. In the following, thinking of activities as determinants of innovation systems will be labeled as a 'process-based' approach of NIS, because this approach is appropriate to describe the actual processes that take place in a system (Hogselius 2005). In the literature there is no agreement on activities that should be included in an innovation system, only that activities taking place in an innovation system do outreach R&D activities (Galli and Teubal 1997; Johnson & Jacobsson 2003; Liu & White 2001; Hogselius 2005). In the context of this article, we will not try to argue about the relevance and importance of certain activities, but we will survey activities on a more general level. Activities are defined as determinants serving the overall function of innovation systems in terms of creation, diffusion and utilization of innovations. Innovations are new knowledge or a combination of existing knowledge. As we will see, change of knowledge occurs on the basis of learning processes. On this account, even without specifying a certain list of activities, we can analyze the applicability of a 'process-based' approach by interpreting activities as determinants building new and combining existing knowledge with the help of learning and forgetting. This brings us back to our starting point, as our definition of a process-based approach corresponds to the Lundvallian approach, who recognizes that 'the most fundamental resource in the modern economy is knowledge and, accordingly, that the most important process is learning' (Lundvall 1992a: 1).

Applicability to transition economies