Science in Policy-Making: Neutral Experts or Strategic Policy-Makers?

Abstract

Although many policy process and diffusion theories follow the premise that scientific and technological knowledge plays a crucial role in a wide variety of policy fields, very few empirically assess the impact that institutional and process-relevant factors may have on the position of science within a process. The present study addresses the question of what role science plays in policy processes. To answer this, we apply the Advocacy Coalition Framework (ACF) and investigate three complementary assumptions using a qualitative comparison of four cases: the ACF claims that scientific experts can take very different positions in the policy process, depending on how conflictive or consensus-oriented the relations among actors and coalitions are within a so-called policy subsystem. Put differently, the type of subsystem impacts on the position of science within the process. The results show that subsystem-specific factors impact upon whether scientific representatives act at the periphery of a process or as policy brokers seeking feasible policy solutions.

Acknowledgements

Data for three (PM Visp, RDC and PJV) out of the four case studies was gathered at the Swiss Federal Institute for Technology (ETH Zurich) and within the research project “Sustainable Land- Use Practices in Mountain Regions: Integrative Analysis of Ecosystem Dynamics under Global Change, Socio-economic impacts and Policy implications” (MOUNTLAND), supported by the Competence Center Environment and Sustainability.

Notes

1. Tabu search clustering provides a relative goodness of fit (R²) that is 0.632 and 0.533 for Zurich Energy and RDC. In the case of PJV, the goodness of fit is rather low (0.255), but splitting the subsystem into three or more clusters worsened the goodness of fit. When using only one coalition (as in the case of PM Visp; and thus not relying on any cluster analysis), it became evident that the belief compatibility within this majority cluster became much weaker than when using two clusters.