EITM-OSLO Summer Institute 2022

27 June - 29 June

Theoretical Foundations of EITM

One of the center goals of the EITM program is to promote theoretically informed empirical research. Game theory is a primary tool for building rigorous theories because politics most often involve strategic interactions which lend themselves to game-theoretic analysis. For example, challengers in a parliamentary or congressional race will consider the incumbents' response to such a challenge when deciding whether to run for candidacy or not. Likewise, rebel groups will consider the ability and willingness of an autocrat to fight a rebellion when challenging him. At the same time, standard introductions to game theory often imply that such theories generate only point predictions, unsuitable for testing. In the theoretical foundations seminar we first review basic ingredients of game-theoretic models. We then look at important varieties of rational choice models, specifically non-cooperative game theory and spatial voting models, in a form that emphasizes the techniques by which these models can be used to generate testable implications through comparative statics analysis and the analysis of parameter variations across a population. See syllabus.

Statistical Foundations of EITM

The basic motivation of EITM is that it pays to be more conscious and purposeful about how game theoretic models and statistical tests are connected. Unfortunately, game theory is often taught in a manner that submerges its relations­hip to testing, and statistical training typically begins with the hypothesis for testing having already been developed. This seminar presents statistical methods that lend themselves well to testing predictions from formal models. Our course begins with an introduction to probability models, covering random variables, important probability distributions, and Bayes' theorem. We then discuss the use of maximum likelihood as a method to conduct statistical inference, as well as its relations­hip to the Bayesian framework of inference. Our discussion of Bayesian inference includes coverage of conjugate bayesian models, as well as common techniques for computation such as the Gibbs sampler and Metropolis-Hastings algorithm. We conclude with applications of these ideas to the study of bargaining models, and to models of latent measures. See syllabus.

James Lo is Assistant Professor at the University of Southern California

Experiments for Evidence-based Political Science

Experiments are a key methodology in evidence-based political science. In contrast to observational research, where nature or individuals or groups determine who is treated, experiments randomly assign units of analysis to treatment and control conditions. The course will first introduce students to the potential outcome framework of causal inference and the fundamental logic of randomized experiments. It will then discuss the actual design and statistical analysis of modern experiments as well as their methodological strengths and weaknesses in establishing valid causal inferences. Throughout the course, we will look at several classic and more recently published examples of field, survey, and online experiments in the study of political processes (mainly but not exclusively related to issues of globalization). The course will be a mix of lectures and hands-on lab sessions and exercises. See syllabus.

Richard Traunmüller is Professor at the University of Mannheim

Quasi-experimental approaches

Causal inference with observational data is hard. We carefully examine compelling examples from Political Science to see why they work. Building on the materials from the previous modules, this module introduces three major quasi-experimental designs to generate causal evidence from observational data: the regression discontinuity design, the difference-in-differences design, and the instrumental-variable design. We discuss the identifying assumptions for each design, classical and novel approaches to estimation and inference, as well as computational tools to implement the designs. The course combines a theoretical introduction to each design with in-depth discussion and lab sessions in which students perform replications of published work. The course will be particularly helpful for students writing their prospectus or designing a major research project. See syllabus.