Assignment

Applied Econometrics

You are allowed to work with any statistical software of your choice, but the preferred package for this course is R.

The goal of the project is to produce an essay that addresses a realistic empirical question by conducting a thorough econometric analysis. You may investigate any empirical question you choose: you will ?nd the data, decide on the analyses to perform, and draw all the conclusions. The project is meant to be an open-ended exercise.

The main goal is to demonstrate conceptual understanding of the course material and its implementation. Your essay should clearly state your empirical question and write down your regression models as we did in class. You should provide a motivation for which variables you include in your regressions and clear de?nitions of any constructed variables. You should carefully address diagnostic issues and model selection and testing.

There is no formal requirement on the length or format of the essay. The goal is to write something that is clear, readable, and thorough; however you feel you can best accomplish those goals is ?ne. Most write-ups are around 15 pages including tables and graphs.

A project proposal should be submitted NOW for feedback on adjustment. You can continue to work on it after you submitted the proposal. Proposal should include:

1. A description of your essay's empirical question.
2. Why it is important.
3. The data sources you will use and how you will get access to them.
4. The methods you will use.
5. Preliminary results you have obtained (if any).

I will provide feedback via email, discussing your preliminary results, details of your approach, additional strategies in addressing the empirical question, and any problems you bring up. But proposal does not count toward course grade.

Some useful data sources:

- Macroeconomic Data from Federal Reserve Economic Database (FRED)
- Wharton Research Data Services (WRDS)
- Center for Research on Securities Prices (detailed securities data)
- Global Insight (?nancial and economic data)
- Compustat (?rm level data for publicly traded ?rms)
- IPUMS (U.S. census data)
- Prediction and data mining competitions (all sorts of application areas)
- Harvard's open source research data repository
- Several data sets in R format.