ECON-7300
This is an informal note to record what I thought about or what I thought was important in the Metrics I class.
GMM
If I can only choose one thing to remember in Metrics I, I would definitely choose GMM. This is mostly because many of the methods we learn in Metrics (at least in the first semester) are in the end some specific form of GMM. For example, IV is just a particular type of GMM. Also, I feel lot of the metrics portion of the applied economics papers are just about finding some moment conditions and estimating the model.
Data cannot speak for itself
As long as we are concerned about some endogeneity problem (which is a common problem in economics), we always needs some assumption to move forward. Just having more data will not solve your problem. In order to identify the parameter of interest, we need economic theory to guide us.