Innovation as an essential ingredient in the modern conduct of business: a performance analysis.

The purpose of this project is to study, in depth, the link between innovation and firm performance by means of an econometric analysis. During recent years the innovation management of firms has undergone profoundchanges (Chesbrough, 2003; Chesbrough and Crowther, 2006). Firms used to invest a lot in specialized R&D departments in order to generate innovations, largely by themselves. This model has recently come under pressure by a more 'open' model, in which firms increasingly depend in their research process on external knowledge. This fosters research cooperation and it implies a reorganization of the research process in order to be able to internalize external ideas. Open innovation also implies a changing attitude towards the appropriability of innovative ideas, by selling them in disembodied form through licencing, etc., rather than merely through product innovation.

In a first phase of the project, the attention is directed towards firms that are active in Belgium, before extending it in a second phase to other European countries. Use will hereby, each time, be made of the third and the forthcoming fourth Community Innovation Survey (CIS) cross-sectional micro databases, which have been compiled in an uniform manner according the directives of the Oslo manual (OECD, 2005).
A structural model will be conceived that explains the innovation input and output stages, as well as their impact on sales and employment growth performance. Earlier models of this type have been estimated by Crépon et al. (1998) for France, Lööm and Heshmati (2002) for Sweeden, and by Klomp and Van Leeuwen (2001) for the Netherlands. But, their estimation procedure can only provide information on the mean behaviour of firms. Management science is, however, at least as interested in the behaviour of the extremes - this is in firms that perform exceptionally well, or poorly. Some recent studies have, therefore, presented estimates of innovation input or output equations, while making use of quantile regression analysis (Nahm, 2001, Ebersberger et al., 2006). This allows to identify systematic differences in behaviour between weak and strong innovative performers. But they have only concentrated on specific stages of the innovation process. An important variable such as the level of research cooperation is not considered. In view of the increasing tendency of firms in undertaking open innovation, it could be highly interesting to find out whether the relative impact of cooperation on research intensity and innovative sales differs between strong and weak innovators.

The objective of this project will be to present and apply a more systematic estimation procedure than has been used until now, and to obtain, thereby, a clear quantitative insight in the main determinants of innovative behaviour and consequent economic performance of firms in Belgium compared with other countries. Relevant economic policy conclusions will, hopefully, be derived from this.