EITM-OSLO Summer Institute 2024

First week

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 involves 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.

Statistical Foundations of EITM

The basic motivation of EITM is it to closely connect formal theoretical models and empirical statistical tests. This course presents flexible Bayesian methods that lend themselves to testing predictions from formal models and to producing meaningful quantities of interest along with their uncertainties. Bayesian analysis involves two key aspects – inference based on probability theory and estimation using stochastic simulation. The course will spend some time on the basic principles of both aspects and then apply them to some of the workhorse models of the social sciences: choice models and IRT models for ideal point estimation. We will also discuss practical issues of applied Bayesian analysis, such as MCMC convergence diagnostics as well as Bayesian model checking and parameter summary.

Richard Traunmüller

Theoretical Implications of Empirical Models

In this module, we will turn the main focus of this program on its head. Just as formal models have testable empirical implications, many of the research designs commonly used in political science are premised on strategic interactions, the structure and outcomes of which are relevant for the interpretation and validity of empirical results. A nascent scholarly literature has recently begun to unpack these theoretical implications of empirical models (TIEM). We will learn a general framework for connecting empirical research designs to their theoretical underpinnings. Then, we will think through the connections between the TIEM approach and causal inference. Finally, we will work through several applications to research designs such as randomized control trials, instrumental variables, regression discontinuity designs, and conjoint experiments, and to research topics central to the study of representative democracy, including how governments respond to citizen protest, the role of the media, and voter rationality.

Link

Strengthening the External Validity of Causal Designs

At first glance, formal models and causal research designs seem like natural partners: while the former generates ceteris paribus predictions, the latter allow us to estimate causal quantities of interest holding all-else-constant. At the same time, modellers and empiricists may find themselves at loggerheads. Specifically, good theory should strive to make general predictions, but treatment effects are (almost) always locally-estimated. As a consequence, there often exists a mismatch between the target population of the formal model and the effective samples used to evaluate these predictions. Strengthening the external validity of empirical approaches is therefore crucial for constructing stronger empirical tests of theory.

Link