Read the report
Randomised Controlled Trials (RCTs) are among the best and easiest ways of identifying whether an intervention has a causal impact on outcomes that we care about. RCTs are the accepted standard way of testing the efficacy of new drugs, and trials in medicine number in the tens of thousands each year. Social policy has been slower to embrace these methods, but has seen very substantial increases in the UK in the ten years since the foundation of the Education Endowment Foundation, and the formation of a further 11 What Works Centres in the years since.
As a result of medicine's leadership in this area, the majority of the methodological approaches used in social policy have borrowed from medicine. Although this makes sense up to a point, challenges do emerge because of the differences between medical trials and those in the various domains of social policy. Theoretical underpinnings of social interventions are necessarily less robust than those based on biological and chemical relationships that can be tested in the lab. Identifying the "ideal conditions" under which an intervention can be tested first, is also harder, as the relationship between people's characteristics and baseline outcomes and the effectiveness of treatments is not known.
In this paper, we aim to address this challenge. Reflecting on what we've learned from ten years of the What Works movement existing outside of medicine, as well as the changes to the standard medical model that allowed the unprecedentedly rapid development of the Oxford-Astrazeneca Covid-19 vaccine, we propose a new, leaner, and cheaper method for running social policy trials to help us find out What Works, faster.