Articulating a compelling economic rationale to justify investments in research-by definition furthest removed from direct, immediate economic benefit-is perhaps one of chief challenges of R&D managers, policy makers and science analysts in the years ahead. Although several innovation studies and surveys have provided some convincing empirical evidence of impacts and benefits of research to technical progress, there is still an urgent need for comprehensive models, reliable data and analytical tools to describe and monitor links between R&D and industrial innovation in more detail. As for the role of scientific and engineering research in the innovation process, this paper reports on the findings of a novel methodology to increase our understanding of the contribution of research efforts to successful technical inventions. The approach is based on a nation-wide mail survey amongst inventors working in the corporate sector and the public research sector in The Netherlands. The inventors' inside information regarding their patented inventions-and related technological innovations on the market-provided a range of quantitative data on the importance of the underpinning research activities. Statistical models attempting to explain the degree of "science dependence" of the inventions identify a range of relevant variables, covering the inventor's own capabilities and previous R&D achievements, external information sources, as well as the inventor's R&D environment in general. Some 20% of the private sector innovations turned out to be (partially) based on public sector research. Furthermore, citations in patents referring to basic research literature were found to be invalid indictors of a technology's science dependence. (C) 2002 Elsevier Science B.V. All rights reserved.
