Lab Experiments in Political Science through the Lens of Experimental Economics

Abstract

Abstract Lab experiments have become an increasingly popular tool in political science to test theory, generate empirical regularities, and serve as a test bed for new mechanisms and institutions. Despite their close connection and overlapping lineage, lab experiments in political science are still at a relatively early stage of development, and only now starting to attain general acceptance in the discipline, in comparison to lab experiments in economics. Consequently, we evaluate the current state of lab experiments in political science by drawing on the philosophy and methodology of experimental economics. We begin by identifying a set of principles of experimental research in economics and outline a series of potential confounds confronting even the best experiments applying those principles. We then conduct a systematic review of all the lab experiments in the economic tradition published in the Journal of Experimental Political Science since its inaugural issue. We find that many studies are subject to the payoff-dominance critique, susceptible to incentive-compatibility issues, and do not make sufficient use of what we call the theory-experimental design-econometric trinity. We suggest, therefore, that lab experiments in political science (and economics) could often benefit from a more careful appreciation and application of the philosophy and methodology of experimental economics.