Browsing by Author "Frantsi, Valtteri"

The Nobel Prize in Economic Sciences in 2019 was awarded to Banerjee, Duflo, and Kremer for their fight against poverty, as well as for their methodological contributions to development economics. This thesis discusses their methodological approach, the use of randomized field experiments (RFE) in policymaking, which according to the advocates of the evidence-based policy (EBP), provide better and more objective evidence. This claim will be examined, and rejected, in the light of the methodological literature of field experiments in economics and a case study of the Finnish Basic Income Experiment (BIE Finland). It will be argued that EBP’s view on RFE’s objectivity is rooted on the narrow view of mechanical objectivity, which overemphasizes methodological norms, such as randomization. This hinders various value choices regarding the research process and ignores the fact that the quality and nature of the evidence can change in the process. Co-creation of the scientific methods and interaction of the science and policy, thus, challenge EBP to reconsider their normative guidelines. This thesis examines BIE Finland and demonstrates how ethical values can become as constitutive values of the research via decisions over (i) the experimental design and (ii) theoretical content, and via (iii) interpretation of the results. It will be argued that these three routes present epistemic risks, but also opportunities to increase the relevance and validity of the research. Ultimately these routes show how scientists are troubled by uncertainty and the risk of error, providing also an avenue for subjectivity. While these routes show complex trade-offs between epistemic and non-epistemic values, their implications for the objectivity of the research are also not clear. This is not only because, as will be illustrated with BIE Finland, RFE’s are compatible with various epistemic aims and inferences that are not always clear, but also because the consequences of inductive risk for the normative guidelines and evidential standards is neither obvious. It will be argued that EBP should clarify the constituents of the trained judgment and the role of epistemic and non-epistemic values throughout the research process, because it ultimately shows how researchers are troubled with uncertainty and the risk of error. This requires them to abandon the value-free ideal and move beyond narrow mechanical objectivity in order to address the epistemic risks and potential disappointment associated with the evidence-based policymaking.