Methodological advances in policy impact assessment

Research in this area has examined the consequences of a misclassified treatment indicator, how to address the resulting bias and how to identify and estimate the misclassification patterns. We have also developed a thorough methodological framework for the analysis of non-participation in randomised experiments and assessed its implications for the experimental impact estimates by proposing a novel combination of both non-experimental methods and the experimental set-up.

Contacts

Lorraine Dearden

Lorraine Dearden

Research Fellow

Barbara Sianesi

Barbara Sianesi

Senior Research Economist

Jack Britton

Jack Britton

Senior Research Economist