Abstract
This chapter elaborates the concept of credibility as a yardstick for the assessment of model-based policies. We further distinguish between epistemic and strategic credibility and show how this distinction helps us to understand the relation between models and the assessment of public policies. For policy assessment purposes, credibility depends mostly on extrapolation. Such extrapolations have a better chance of success if we draw on a causal mechanism, as structural models do. The sort of empiricism about causal interventions promoted, among others, by field trialists in economics has a lower chance of being epistemically credible. Furthermore, the reductive relationship between strategic credibility to epistemic credibility would lack of stability.