Good Luck, Tough Luck: Filtering, Attribution and Rationality in Economic Voting

Abstract

Do voters rationally make use of all the information available to them when assessing government performance? This thesis explores the information extraction problem faced by voters in open economies to distinguish between economic growth that is due to governments’ competence and growth that is due to exogenous luck. I analyze electoral and economic data for a panel of 93 countries covering a total of 565 democratic and competitive national elections for chief executive leaders. Using a two-stage estimation procedure to (1) identify the influence of clearly exogenous and observable factors on growth, and then (2) estimate the effect of previously-instrumented “luck” on incumbent governments’ electoral prospects, I find that voters systematically make mistakes when attributing responsibility for economic growth due to factors exogenous to their national governments. In other words, I find that voters reward incumbent governments for good luck and punish them for bad luck. In addition, my results suggest that increasing globalization attenuates the strength of the economic vote and, as a consequence, mitigates the deleterious effects of inefficient filtering. Combined, these findings lead to the conclusion that voters do not display fully-rational behavior. More broadly, my results cast some doubt on the effectiveness of the democratic vote as a selection and sanctioning mechanism, and raise serious implications for political accountability in today’s globalized world.