Influences on norms of ‘good evidence’ among policymakers

Interest in evidence-based policymaking has grown steadily over the last few decades. Similarly, there has been a mushrooming of attention from researchers on the conditions for using evidence in policy, as well as the more adverse ways of using evidence. But how does policymakers’ assessment of evidence as ‘good’ or ‘useful’ develop? Conditions for using evidence may differ according to the attributes of policy fields, such as degree of conflict, media attention and tractability (Schrefler, 2010; Zarkin, 2020; Rimkute, 2015). It can also differ based on the supply of evidence within a policy field (Sarewitz and Pielke, 2007). Sources of evidence within a knowledge regime are manifold (cf. Campbell and Pedersen, 2015), and different knowledge producers –from universities to consultancies and international organizations – have different preconditions for and ways of producing evidence. Finally, evidence is not only used to make more effective policy decisions; evidence may also be used for strategic purposes and to bolster settled preferences (Weiss, 1979; Boswell, 2009). How policymakers within a given policy field assess evidence is accordingly likely to be conditioned by the interplay of several different factors, including a) characteristics of the policy fields; b) the purpose of and experience with using evidence, and c) the supply of different knowledge producers in the system of knowledge production. Using data from 1900 respondents to a survey about research use in the central administration in Norway, the purpose of this study is to examine how the interplay of these aspects conditions policymakers’ assessment of evidence. Preliminary findings suggest that policymakers’ assessments of what they consider to be salient characteristics of evidence can be clustered according to the following attributes: usefulness, closeness and academic orientation. These ‘norms for good evidence’ are significantly and differently influenced by where policymakers seek out knowledge, characteristics of the policy field, and for what they apply research-based knowledge. Valuing the usefulness of knowledge – evidence based on specific methods, showing ‘what works’ and pronounced policy implication – is associated with a preference for evidence that is produced by organizations with a formal relation to the central administration, policy fields with frequent changes in the political agenda, and instrumental use (and non-use) of evidence. Valuing knowledge that is based on Norwegian data developed by familiar researchers (closeness) is linked to sector-specific knowledge producers and instrumental uses of evidence. Finally, valuing evidence with an academic orientation – published in peer reviewed journals by recognized researchers – is linked to the use of evidence provided by consultancies and legitimizing uses of evidence. Surprisingly, use of knowledge produced in universities is not a significant influence on the ‘norms for good evidence’.