Over the past 20 years, the innovation system approach has significantly enhanced our understanding of the innovation process, stressing its non-linear, systemic, interactive and evolutionary character. The notion of regional innovation systems (RISs) highlights the regional dimension of new knowledge generation and exploitation and constitutes a powerful concept for explaining regional differences in innovation capacity. RISs can be conceptualised as the set of firms, organisations and institutions which influence the innovative behaviour and economic performance at the regional level. They are shaped by existing industry structures and technology paths, the set of knowledge organisations, and the prevailing institutions and networks. As a consequence, they exhibit a high degree of inertia. This may lead to phenomena of path dependency and "lock in" in particular regions and to a certain degree of stability in terms of regional disparities in innovation and economic development. Regions and their innovation systems, however, are not static entities. In fact, one can observe considerable changes of industry structures, innovation activities and patterns of networking in particular regions in the longer run, often reaching beyond the existing development paths. We find phenomena of innovation-driven catching up processes in lagging regions, restructuring processes in industrial regions leading to new industries and technology paths, as well as sometimes an erosion of innovation capacity and competitiveness in leading regions. Most research on RISs, however, has so far not dealt with such changes. The RIS literature suffers from a key weakness, that is, its static view brought about by a focus on existing structures and relations. As a consequence, the reconstruction of RISs and their evolution over time remains poorly understood. The aim of the paper is to enhance our understanding of how processes of RIS transformation take place. We will identify key actors and drivers of path renewal and new path creation and we seek to find out to which extent such changes are related to the existing economic and institutional structures. Based on a discussion of relevant theories and a critical literature review we will develop a conceptual frame for analysing RIS changes. Besides the RIS approach we will use ideas from evolutionary economic geography (EEG) which provides valuable insights into the long-run regional trajectories and sources of change in regional economies. We will also discuss empirical examples of such shifts based on evidence from Austria and other countries.
