Fall 2024 • Instructor: Kyle Butts
Monday, Wednesday 1 -- 2:15PM at RCED 103
Office Hours: Monday, Wednesday 11am -- 1pm at WCOB 408
This course covers the principles of causal inference. Methods include panel data models, instrumental variables, regression discontinuity designs, difference-in-differences, and matching. Emphasis on developing a solid understanding of the underlying econometric principles of the methods taught as well as on their empirical application.
All empirical analysis will be based in the R programming language
Our primary text for the course is Causal Inference: The Mixtape by Scott Cunningham. The textbook is available for free online. You may buy a print version, but it is not necessary for the course. The course will also use review articles for overviews of methodologies and academic articles for examples of empirical usage. Articles will appear in the Literature/ folder
You will need to download two programs:
- Install R from https://cloud.r-project.org/.
- Install Positron from https://github.com/posit-dev/positron/releases.
Positron is a new (and better) coding environment than RStudio, so I recommend it. However, you can use RStudio if Positron is giving you problems.
Mastering R will take time and dedication, but it is a powerful and adaptable tool that is highly valued by many employers. Invest the necessary effort and time, and you will see the benefits.
Problem sets will usually be due two weeks after assigned. The problem sets will require you to analyze datasets using techniques discussed in class. Students are encouraged to work in groups to discuss how the approach the problem sets, but each student must hand in his or her own set of answers. Missing or late problem sets will receive no credit.
Exams will be held during class time on October 7th and November 13th.
For the final project, you will write an empirical paper using microeconometric data. A complete project must include a research question in the form of "the impact of X on Y", descriptions of context, empirical strategy (including why the chosen strategy is valid for the question/data), results, and a brief discussion on limitations. It is okay if your project has issues in establishing causality; you will be graded on your ability to discuss these issues and not whether or not you have a flawless research project. You also will need to present a preliminary draft in the last 3 weeks of class.
Non-PhD economics students: (i) may work in a group of two; and (ii) are recommended but not required to include a brief literature review. PhD economics students should submit a single-authored paper and include a brief literature review.
This is a tentative schedule. This is the first time I've taught this course, so take this with a heavy dose of skepticism. In particular, do not set up holidays a class before or after the midterm.
| Week | Dates | Monday | Wednesday | Assignments |
|---|---|---|---|---|
| 1 | 08/19 - 08/21 | Syllabus + Potential Outcomes | Randomized Control Trials | |
| 2 | 08/26 - 08/28 | Randomized Control Trials | Randomized Control Trials | |
| 3 | 09/02 - 09/04 | No Class | Regression Mechanics | |
| 4 | 09/09 - 09/11 | Regression Mechanics | Selection on Observables | |
| 5 | 09/16 - 09/18 | Selection on Observables | Selection on Observables | |
| 6 | 09/23 - 09/25 | Selection on Observables | Selection on Observables | |
| 7 | 09/30 - 10/02 | Selection on Observables | Regression Discontinuity | |
| 8 | 10/07 - 10/09 | Regression Discontinuity | Regression Discontinuity | |
| 9 | 10/14 - 10/16 | No Class | Regression Discontinuity | |
| 10 | 10/21 - 10/23 | Midterm | Instrumental Variables | |
| 11 | 10/28 - 10/30 | Instrumental Variables | Instrumental Variables | |
| 12 | 11/04 - 11/06 | Instrumental Variables | Panel Data | |
| 13 | 11/11 - 11/13 | Panel Data | Panel Data | |
| 14 | 11/18 - 11/20 | Panel Data | Panel Data | |
| 15 | 11/25 - 11/27 | Panel Data | No Class | |
| 16 | 12/02 - 12/04 | Project Presentations | Project Presentations |
This course was inspired by a lot of material that I have blended together into the course. These include: