Published online by Cambridge University Press: 08 April 2015
Political scientists often attempt to exploit natural experiments to estimate causal effects. We explore how variation in geography can be exploited as a natural experiment and review several assumptions under which geographic natural experiments yield valid causal estimates. In particular, we focus on cases where a geographic or administrative boundary splits units into treated and control areas. The different identification assumptions we consider suggest testable implications, which we use to establish their plausibility. Our methods are illustrated with an original study of whether ballot initiatives increase turnout in Wisconsin and Ohio, which illustrates the strengths and weaknesses of causal inferences based on geographic natural experiments.
Luke Keele is Associate Professor in the Department of Political Science, 211 Pond Lab, Penn State University, University Park, PA 16802 ([email protected]). Rocío Titiunik is Assistant Professor in the Department of Political Science, P.O. Box 1248, University of Michigan, Ann Arbor, MI 48106 ([email protected]). This paper was prepared for the conference Spatial Models of Politics in Europe and Beyond, Texas A&M University, 2012. The authors thank the editor Cameron Thies, two anonymous reviewers, Lisa Blaydes, Matias Cattaneo, Don Green, Justin Grimmer, Danny Hidalgo, Simon Jackman, Marc Meredith, Clayton Nall, Ellie Powell, Wendy Tam Cho, Jonathan Wand, Teppei Yamamoto, and seminar participants at the University of Michigan, Stanford University, Yale University, Duke University, and Penn State University for comments and discussion. The authors also thank Mark Grebner for assistance with acquiring the Wisconsin Voter File. Titiunik gratefully acknowledges financial support from the National Science Foundation (SES 1357561). Parts of this manuscript were previously circulated in a working paper entitled “Geography as a Causal Variable.”