This preview shows page 1. Sign up to view the full content.
Unformatted text preview: Problems with using expenditure Problems of intervention variables being too broad interesting but not quite sure what they are telling us Natural Policy Experiments The way forward can look at very speciFc types of interventions Problems of identiFcation Duos work in this area exemplary Randomised Trials Experiments: treatment/control only way of getting truly exogenous policy variation Big area of work now, e.g., textbooks, devolution of school funding; Typically researcher-NGO partnerships see MIT Poverty Action Lab Development Economics, LSE Summer School 2007 140...
View Full Document
- Spring '10