Courses:

The project description is available as a file here: (PDF)

This project asks you to review, replicate, and extend an empirical paper from our reading list. The paper should be one that we have not covered in detail in class.

Papers must be chosen and the selection approved by the instructor by Lecture 13, the last class in the first half of the course. The paper is due in Lecture 21, the last class before Thanksgiving.

1. What was the purpose of the research? What questions were asked and what hypotheses were tested? Why are these questions and hypotheses of economic interest? What are the most important findings in the paper?
2. How does this paper fit into the relevant literature? What were the findings at the time the paper was written? What was the contribution of this particular paper? What has been done on this topic since this paper was published?
3. What data were used? Are they reliable or relevant? Are they rich enough to give meaningful answers to the key questions motivating the paper? What would constitute an ideal data set?
4. How was the research conducted? Do the techniques used make sense for this problem and do they appear to have been correctly implemented? What assumptions are needed to draw inferences about causation from the results presented in the paper?

1. Identify the main findings and use the authors' data to replicate these results if possible. If the author's data are not obtainable, construct the same sort of estimates using a data set of your choosing. Choose a data set that you would expect to generate similar results. Summarize and compare your results to the original results in a table. Discuss why you think your results differ from the author's if they do.
2. Extend the work in some way. Do this by either (a) estimating alternative interesting specifications that the author might have tried or that would shed further light on the issues raised in the paper (e.g., specification checks), or (b) collecting new data and producing results for this new sample. If you also collected new data to perform the replication, then either collect additional data as a robustness check on your findings, or extend as described under point (a).