When evidence does not matter: The barriers to learning from science in two cases of environmental policy change in Brazil

Abstract It is generally accepted in public policy debate that expert knowledge tends to contribute to more effective formulation and implementation of policy. Most of the literature, however, has tended to be exclusively focused on the science–policy interface, ignoring the necessary pre-conditions of the broader national and institutional context for the effective use of scientific evidence. This shortcoming becomes particularly pronounced in analysis of developing in less pluralist countries. This article analyses two cases of Brazilian environmental policy-making and discusses the institutional pre-conditions for learning from science. By textually coding instances of direct and indirect participation of scientists in congressional debates and assessing the extent of their influence in final decisions, this article shows that despite being largely consensual to the scientific community, clearly communicated, and relevant, scientific information had no influence on the policy-decisions taken in two highly-technical areas of environmental policy in Brazil: forestry and pesticides. This article engages with the literature on the necessary institutional structures for learning from science and provides support for the hypothesis that countries with lower levels of political openness, and medium-to-low consensus requirements for decision-making, will tend to have considerable barriers to effectively producing evidence-based policy-making.