The last decade is marked by tendencies of de-democratization and autocratization (Lührmann & Lindberg 2019), often characterized by polarization of politics and society, by undermining facts- and expertise-based policymaking, and by decreasing transparency and accountability of political and policy processes. As part of these democratic erosion processes, civil society has also been increasingly under attack, particularly if critical of the incumbent government.

Think tanks stand at the crossroads ofseveral ofthese tendencies, for several reasons. First, think tanks are members of civil society, and actors in this increasingly polarized political arena. Second, they are organizations that commonly contribute to policy processes with policy knowledge and expertise, which in de-democratizing and polarizing contexts appears to be increasingly troublesome. Third, think tanks are actors that traditionally claim independence beyond party and political lines in their activities. This position may be difficult to uphold in a context where governmental and specific party allegiance is of accelerating importance.

Our research aims to bring together scholarship on de-democratization and polarization seen from the vantage point of think tanks, in an attempt to improve our understanding of the specificities of political knowledge production in the context of democratic backsliding. The context for the paper is a research project funded by the Baltic Sea Foundation on policy advice in electoral regimes. In the project we are focusing on how think tanks as knowledge producers, providers of policy advice and advocacy are relating to the broader processes of de- democratization We propose to do this through a comparison of two extreme cases of de- democratizing countries in Europe: Poland (classified by https://www.v-dem.net/en/ as electoral democracy) and Hungary (classified as electoral autocracy), both of which were downgraded from relatively high and solid democracy prior to 2010 and 2015 respectively.