Conclusion and Perspectives: Risks and promises of the Evidence Movement

In the introduction chapter we identified a number of inter-related and over-lapping questions to be addressed in this book about the evidence movement. In answering these questions we hope better to understand the nature, origins, immediate impacts and long-term implications of the evidence movement. It should be clear that this is not only relevant to the narrow issue of the evidence movement itself (however defined) but also to much wider issues concerning evaluation, knowledge, and decision-making. These questions were prompted by our experiences as evaluators. These experiences were influenced by the evidence movement in at least two ways. First, there were changing expectations about how research and analysis was used in the conduct of evaluation. Impact Assessments in the European Commission, performance audits in supreme audit institutions, and invitations to tender for evaluations all reflected changing expectations of what constitutes compelling evidence and the acceptable principles of argumentation. Second, the programs and projects we were called upon to evaluate were changing. At best, programs had more clearly identified outcomes and were associated with more specific and measurable impacts. Even without this, program managers and practitioners were more used to collecting evidence of impact and were less surprised to be asked for data about costs, processes and impacts. At the same time, we were unconvinced by all of the claims made on behalf of the new evidence-based world. We were not only concerned that actual practice often failed to meet the high standards set by the evidence movement, but also we were not fully convinced that the evidence movement had fully grasped the density of relationships between evidence, theory, action and benefit. This led us to pose some questions at the start of this book. The questions included the following: We conclude this book by considering what the preceding chapters had to say in response to each question.