Public Funded Research and Innovation in the Green Economy

The analysis investigates the relationship between universities, public research institutes and innovation in companies belonging to the green economy. By adopting a ‘general to specific’ specification strategy ‘a la Hendry’, the analysis sheds light on previously unexplored determinants of the value of information from public research. It discovers how motivations for innovation are a significant determinant of collaboration with the public research base. Motivations which are determined externally to companies (reducing environmental impact, health and safety) are much more important to explain the value that companies assign to information from public research than motivation coming from within companies, such as increasing market share, value added or flexibility. Moreover, motivations which are related to the introduction of new products or the replacement of outdated ones (hence linked to the introduction of radical innovations) are a powerful predictor of the usefulness of public research information to companies’ innovation strategies. Some constraints on innovation also appears to determinate what source of information is valued most. This is especially the case for regulation, both from national as from international sources. When companies are constrained in their innovation by regulation, they are more likely to turn to public research for information. The comparison of results from regulation seen as a motivation for innovation and regulation as a constraint on innovation and – especially ‐ the detailed analysis of different types of regulation shed light on the role of regulation in promoting innovation in the ‘Green economy’, casting serious doubts on the idea that a one size fits all deregulation approach could promote growth. The analysis also unveils the existence of a significant decrease in the value that companies assign to information from universities when the number of types of innovation undertaken by companies increases. A suite of competing hypothesis are proposed and discussed to explain this novel result. However, a definite conclusion on the sources of these decreasing returns to information requires the modification of some questions in the CIS questionnaire or additional data. Finally, the study unveils some potential problems with the econometric estimates used in previous models. On one hand it uncovers issues with the way the variable representing the breadth of searching pattern that companies use has been measured in previous studies and discusses the strategy to address this concern. Then, it also shows that many previous analyses have failed to take into account the potential presence of sample selection but finds results to be robust to this potential problem. More worryingly, previous analyses have failed to test the assumption of parallel lines which is made in the models which have been used. This assumption is violated in some cases and the consequences of this violation are discussed.