Mini-Course: Causal Inference without Experiments

This mini course will discuss various strategies for obtaining causal estimates without an experiment, with empirical examples from labor economics and other applied micro fields.  Methods which allow for selection on observables as well as selection on unobservables will be covered.  The course will explore the pros and cons of using various methods to arrive at causal estimates.  It will not concentrate on the formal derivation of estimators, but rather on the assumptions and properties of various approaches and how to implement them.  An emphasis will be placed on current best practices, with empirical examples ranging from the evaluation of social safety net programs, to the effect of educational policy reforms, to the identification of peer effects in networks.
Preliminary outline of topics:
1.  Causality and Identification
2.  Panel Data, Fixed Effects, and Event Studies
3.  Matching and Propensity Scores
4.  Social Experiments
5.  Regression Discontinuity

About the Professor:

Gordon Dahl is Professor of Economics at University of California San Diego since 2012. Professor Dahl obtained a Ph.D in Economics at Princeton University in 1998. He has extensively published in Labor and Applied Economics in top journals including American Economic Review, Econometrica, Quarterly Journal of Economics, Review of Economic Studies, Journal of the American Statistical Association, Review of Economics and Statistics, Demography, among several others.

Dates:

Novemeber 17 & 18, 10:15 – 13:30