DGPE PhD course: Causal Machine Learning

Course topic

After reviewing basic identification strategies in econometrics as well as reviewing main ideas and methods of supervised machine learning, the main part of the lectures concerns causal machine learning, i.e. how to combine the prediction methods of the machine learning literature with the causal research designs to obtain reliable causal inference in empirical studies. We will discuss on how to improve the estimation of effects commonly targeted by empirical papers, like the average treatment effect on the treated, as well as discuss new possibilities to uncover finer grained causal heterogeneity. Having uncovered the latter, we discuss its use in some of the recent literature on optimal policy allocation. 

Contact

Joachim Kahr Rasmussen, CEBI, University of Copenhagen, jkr@econ.ku.dk