|
Causal Inference and Program Evaluation
Instructor: Alberto Abadie Program evaluation comprises a set of statistical tools designed to assess the impact of public interventions, such as job training programs, on outcomes of interest, such as earnings. This is a methodological course, developing skills in quantitative program evaluation. The course will focus on recent advances in nonparametric and semiparametric techniques for program evaluation, as well as on the most relevant empirical applications. We will study a variety of evaluation designs, from random assignment to quasi-experimental evaluation methods. We will analyze the strengths and weaknesses of alternative evaluation methods. Students will be offered the opportunity to analyze data from the National Job Training Partnership Act Study, an actual evaluation of a training program for disadvantaged workers.
Outline
1. Introduction: counterfactuals and causal parameters
2. Randomized experiments
3. Matching, propensity score matching.
4. Difference-in Differences
5. Instrumental variables: local average treatment effects
6. Distributional effects
7. Regression discontinuity design
Alberto Abadie is Professor at the John F. Kennedy School of Government at Harvard University. His main research areas are econometrics, labor economics, and public finance. In addition, in his research he uses data and economic models to analyze the causes and consequences of political violence and terrorism. Since 2007, he serves as an editor of the Review of Economics and Statistics. He received his PhD in economics in 1999 from the Massachusetts Institute of Technology. He lives in Cambridge, Massachusetts. Personal Webpage
| 
RSS