experiment

What are randomised control trials, and what is the best way to use them to inform policy decisions?

Karthik Muralidharan and Paul Niehaus argue that when we test things at a small scale, they might not be predictive of how they perform at a larger scale

Providing information and assistance to informal firms increases formalisation, but the cost of the intervention is likely prohibitively high

An attendance-monitoring intervention in schools reduced absenteeism, but the impact on health care workers was limited