Is no evidence better than any evidence when controlled studies are unethical?

Rigorous evidence on the health effects of social interventions is scarce1 2 despite calls for more evidence from randomised studies.3 One reason for the lack of such experimental research on social interventions may be the perception among researchers, policymakers, and others that randomised designs belong to the biomedical world and that their application to social interventions is both unethical and simplistic.4 Applying experimental designs to social interventions may be problematic but is not always impossible and is a desirable alternative to uncontrolled experimentation.3 However, even when randomised designs have been used to evaluate social interventions, opportunities to incorporate health measures have often been missed.5 For example, income supplementation is thought to be a key part of reducing health inequalities,6 but rigorous evidence to support this is lacking because most randomised controlled trials of income supplementation have not included health measures.5 Current moves to increase uptake of benefits offer new opportunities to establish the effects of income supplements on health. In attempting to design such a study, however, we found that randomised or other controlled trials were difficult to justify ethically, and our eventual design was rejected by funders.