CDI Practice Paper
Loevinsohn
Institute of Development Studies, Brighton, March 2013

Natural experiments are observational studies of sharp, well-defined but unplanned changes. They hinge on identifying an uncontrolled but opportune 'intervention', typically of a kind or on a scale that could not – ethically or feasibly – be implemented deliberately, and communities, groups, or individuals that are affected and not affected, or that are differentially affected by that intervention. More than a method, natural experiments can also be understood as a resource – opportunities that must be recognised and wisely exploited.

This CDI Practice Paper explores some of the roles that natural experiments have played in evaluation and considers their potential and limitations and the prospects for their wider use. In the context of development, natural experiments' two most salient features are first, that they make it possible to assess scales and types of impact that are difficult or impossible to deal with through other approaches, and second that, being centred on large, notable events, they can lead to more open evaluations and wider public involvement in them.

Of particular interest as experience of the effects of climate change widens is the potential they hold to support evaluation and learning around resilience and adaptability to extreme events.

Related resources: View this publication
Partner(s): Institute of Development Studies