What works: When and why are nudges sticky, scaleable and transferable?

Evidence-based policy making assumes that the evidence in question has external validity i.e. that the conclusions drawn from research can be generalised to contexts outside of the research setting. In other words, that those conclusions are sticky, scalable and transferable. The use of so-called 'nudge'-style behavioural interventions has increased considerably over the last decade in both public policy and commercial design contexts. This has put pressure on a fledgling evidence-base to answer questions of external validity with premature confidence. In turn, this can lead to replicability failures and the risk losing public confidence and implicit consent for the use of these approaches. This paper argues that (at present) the only responsible answer to 'what works?' questions in relation nudges is that 'we don't know. . . yet'. It then goes on to argue that this is a perfectly natural and valid state for a young discipline to be in and points out the key areas that future research and practice needs to address to facilitate more confident and actionable claims of external validity.