Syllabus for Gov 2003 at Harvard University

Head Instructor

Matthew Blackwell
Office Hours: Mondays, 1-3pm Sign up here

Teaching Fellow

Sooahn Shin
Office Hours: Mondays, 3-5pm @K105


This course will cover a growing field in political science and the social sciences more generally: causal inference. That is, we will primarily be concerned with how and when we can make causal claims from empirical research. This is a difficult task and, in general, there are no formulas or computer programs that can give your research a causal interpretation. We always require assumptions. This course will focus on providing you the analytic and quantitative framework to assess and implement these assumptions. Perhaps at the most basic level, this course will sharpen your senses toward the issues of causal reasoning in the social sciences.


In this course you will be expected to

Course objectives

After taking this course we hope you will:


Students in the class should plan to work hard at understanding some difficult material. You’ll need to have some proficiency with R. You should have taken GOV 2001 and GOV 2002 or the equivalent. Of course there are always exceptions to this, so feel free to chat with me about your background to see if the fit is right. Qualified undergraduates and graduate students from other departments are welcome to join the class.


Course structure


Lectures will cover the basic theoretical issues in lecture along with some applied examples. We’ll try to foster discussions when we can and perhaps even attempt some “breakout” time to work through issues (if possible given the health and safety situation). We hope to record the lectures via the camera in K354 and make the recordings available to those who cannot attend due to illness or class conflicts. Having said that, your attendance at lectures is very strongly recommended and your participation counts toward your grade.


At section, your TF will review the material for the week and demonstrate how to implement the methods we describe in lecture.


There are readings for each topic and they mostly cover the theory of the method along with some applications. Obviously, read the required readings and any others that pique your curiosity. In addition, though, engage with the readings: take notes, write down your impressions or confusions, talk with your classmates. All of your classes should be pushing your research forward and you will be more creative the more you actively read.

Course assignments


Participation is crucial component to the class and we will grade your participation based on your engagement during class meetings along with your contributions to discussions and feedback via Slack and the message boards.

Problem Sets

Methods are tools and it isn’t very instructive to read a lot about hammers or watch someone else wield a hammer. You need to get your hands on a hammer or two. Thus, in this course, you will have approximately 8 problem sets. They will be a mix of conceptual questions, analytic problems, computer simulations, and data analysis. These problem sets should be typed and well-formatted, preferably using RMarkdown. Each problem set will be equally weighted. I encourage you to work in groups on the homework, but you always need to write your own solutions including your computer code. Also, it is hugely beneficial to attempt the problems sets on your own before working in groups.

The schedule for the problem sets is below:

Problem Set Release Date Due Date
Problem Set 1 Thu, Sep 9th 12:00pm ET Wed, Sep 15th 11:59pm ET
Problem Set 2 Thu, Sep 16th 12:00pm ET Wed, Sep 22nd 11:59pm ET
Problem Set 3 Thu, Sep 23rd 12:00pm ET Wed, Sep 29th 11:59pm ET
Problem Set 4 Thu, Sep 30th 12:00pm ET Wed, Oct 6th 11:59pm ET
Problem Set 5 Thu, Oct 14th 12:00pm ET Wed, Oct 20th 11:59pm ET
Problem Set 6 Thu, Oct 21st 12:00pm ET Wed, Oct 27th 11:59pm ET
Problem Set 7 Thu, Oct 28th 12:00pm ET Wed, Nov 3rd 11:59pm ET
Problem Set 8 Thu, Nov 4th 12:00pm ET Wed, Nov 10th 11:59pm ET
Problem Set 9 Thu, Nov 11th 12:00pm ET Wed, Nov 17th 11:59pm ET

Student Project

In lieu of a final exam, this course requires students to write a short paper applying or extending the causal inference methods we learn in this class. These papers can be substantive (applying the methods to a problem of interest to you) or methodological (deriving a new method that innovates over existing ones). It should be no longer than 20 double-spaced pages and focus on the research design, data, methodology, results, and analysis. Literature reviews or background material should be omitted or included in an appendix. Co-authored projects are strongly encouraged and working as an individual requires approval. Working with collaborators will be the cornerstone of your career from now on, so it’s crucial to get to know this process sooner rather than later.

Here is a brief timetable for the projects:


We’ll draw heavily on the following texts for this class:

The following books are optional but may prove useful for additional coverage of some of the course topics. Always check to see if the books are available to download from Harvard servers (true for many CUP and Springer books).