Mechanism-Based Policy Design: Expanding Designers’ Toolkit

The aim of the paper is to explore the potential of the use of causal mechanisms for guiding policy design. There is a vast literature on mechanisms, generally intended as mid-range theories explaining how inputs are related to outcomes. The concept has diffused from natural sciences to sociology (Elster 1983), behavioral economics (Bowles and Gintis 2011), political science (Tilly 2001; Falleti and Lynch, 2009), the organizational and managerial sciences (Bardach 2004; Barzelay 2007) and policy evaluation (Pawson and Tilley, 1997). Some applications are germane to policy analysis, but no explicit call has come forward for the use of the concept in the policy sciences. Bardach (2004) and Barzelay (2007) have suggested that – in order to reproduce policy success – analysts should go beyond actual design and understand causal mechanisms underlying the working of policy programmes. The analyst should understand how the programme produces change, isolate and reproduce mechanisms and go beyond surface and formal features. This permits adaptation and re-design of causally relevant policy features. The paper explores the prospects of re-focusing policy design on devising and reproducing causal mechanisms, with a specific interest in avoiding implementation gaps due to the progressive decrease of the needed actors’ engagement and co-operation. In order to do so, it will review and discuss different strands of literature on policy design, in order to show the place of mechanisms as a tool for designers. Anticipating part of the discussion, mechanisms have a middle level of abstraction, possibly permitting more specific guidance than policy tools and considering a wider range of causal factors than nudgers.