Abstract:

How can the quality of our decision-making with small sample sizes be enhanced? When can one generalize on the basis of a well-identified single case? In an array of prominent fields (e.g., law, medicine, business), case studies have a long and venerable history as tools for both research and teaching. They have recently risen to prominence in international development, as exemplified by the Global Delivery Initiative (a 30-organization consortium based at the World Bank), which produces and shares case studies on development interventions from around the world. But lingering concerns remain about the methodological veracity of cases and case studies as a basis for deriving defensible claims about causal inference, extrapolation, and their utility for problem diagnosis. Recent scholarship, however, seeks to identify more precisely the conditions under which these concerns can be assuaged. For 'complex' development interventions in particular, singular cases and case studies can offer distinctive insights into how such interventions work; and when coupled with theory and broader quantitative data, they can in fact be used to make causal inferences, and to guide decisions about the likely impact of replicating the intervention for different groups, in novel contexts, at larger scales of operation, or when implemented by different agencies. Engaging epistemological strategies from across the social sciences can greatly enhance development effectiveness.

Download the paper: 

Using Cases and Case Studies in Development: Causal Inference, Extrapolation, Diagnostics