Week 12: Quasi-experiments

Overview

Experiments are helpful tools in trying to determine whether a policy works. But in many cases experimentation is unethical or unfeasible. In these contexts, researchers can try to leverage features of naturally occurring data that provide comparable levels of control. These are called quasi-experiments because we are assuming we can analyze them as if they were experiments.

The lab introduces regression discontinuity designs and difference-in-differences as popular examples of quasi-experimental research designs.

Reading

• RD section 16

• Hoekstra, Mark. 2009. “The Effect of Attending the Flagship State University on Earnings: A Discontinuity-Based Approach.” The Review of Economics and Statistics 91 (4): 717-724

• Leninger, Arndt, Marie-Lou Sohnius, Thorsten Faas, Sigrid Roßteutscher, and Armin Schäfer. 2022. “Temporary Disenfranchisement: Negative Side Effects of Lowering the Voting Age.” American Political Science Review

Slides

Navigation

Click on the frame below and use the arrows to navigate. You can type ? (or shift + /) to see useful commands.

Open in a new window Download PDF

Lab

Quasi-experimental designs

Due

Monday, April 1, 11:59 PM