Course paper

There are three options here:

1. Apply a concept from this course to a research idea you have. Note this can be exploratory – don’t worry about having the perfect dataset. See how far you can get / what you can learn from a dataset that is readily available.

2. Use simulation to explore the implications of an existing paper. Take the reported estimates (or possibly even published simulation code) from an existing paper, generate new outcomes. Construct new counterfactuals designed to explore either the economics of the model or the assumptions underlying estimation.

3. Explore the robustness of an existing paper to the application some newer concept from this class. For a structural paper, can you relax some estimation assumption? For a reduced form paper, can you reanalyze the and reanalyze the problem using some of the ML methods discussed in the first part of the course. You can choose any paper you like, although we will need to verify it fits well in one of these two categories.

Important dates: Paper topics are due on 2/19. We will respond with feedback shortly after. There will be in class presentations on progress on 3/11, and final presentations on 5/6. Final projects are due on 5/10.

Additional requirements All projects must be hosted on a Github repository which you create.