In this thesis, I examine knowledge production and science-policy interaction associated with the management of chemical substances in the Baltic Sea under conditions of uncertainty and complexity. The thesis is primarily based on a qualitative analysis of policy documents and reports produced by the European Union (EU), the Helsinki Commission, the Swedish Chemicals Agency, and Gothenburg municipality, as well as 30 semi-structured interviews with scientific experts, policy makers, and government officials with knowledge and working experience in the relevant policy arenas.

I identify and examine key challenges of science-policy interaction associated with the management of chemical substances in the Baltic Sea, such as lack of data, uncertainty, and complexity (Article I). I further argue that the current model of science-policy interaction fails to adequately handle and account for these challenges (e.g. uncertainty in Article II).

Scientists and policy makers have made substantial (and sometimes successful) efforts to understand and counteract negative environmental trends in the Baltic Sea. However, on the basis of this empirical analysis, I conclude that the uncertainties tied to the complex chemical risks in the Baltic Sea region are too large and multifaceted to be adequately addressed by the “modern science-policy model” underpinning most contemporary risk assessments. Linked to this analysis, I identify several possible ways to improve the situation, for example new tools and methods for handling uncertainty as well as alternative models for science-policy interaction.

As a consequence, I explore the potential of alternative models of science-policy interaction, giving particular attention to the participatory model and the associated idea of post-normal science. The results highlight the substantial amount of rhetoric in EU sources about recommending wider public involvement in policies, but also reveal that there is a different situation in practice. The introduction of more radical approaches (such as post-normal science) to knowledge co-production and participation would require epistemological, institutional, and constitutional changes that are not feasible in the foreseeable future – at least for the case of chemical substances (Article III).

Improvements (methodological, institutional and so on) in the current modern model of science-policy interaction are just as important as the development of alternative modes of science-policy interaction (Article IV). However, the general conclusion of this thesis is that there is a need to rethink current science-policy interaction and in the process “break through” the widespread institutional denial of irreducible uncertainties.