Causal inference with big data: an instrumental variable and model selection perspective


In this short talk, I briefly introduce some recent work on causal inference using the instrumental variable (IV) method in high-dimensional models. Some of the common issues of the instrumental variable method, namely, weak instruments, invalid IV, control variable, and hypothesis testing, are covered selectively. Some potential empirical applications will be discussed. An R package, ‘naivereg’, which implements some of the aforementioned methods, is available online.




Professor Fan Qingliang
Department of Economics
The Chinese University of Hong Kong

Related Posts