An event study analysis is performed on consumer confidence around election times using an effective sample of 84 national elections from 1985 through 2005 in the EU-15 countries. Consumer confidence is measured monthly using identical surveys, which began in 1985 and were gradually adopted by all EU-15 countries. Consumer confidence increases a number of months before the date of national elections and falls subsequently, a pattern that differs depending on whether the incumbent party wins or loses. Consumer confidence is able to predict positively the probability of re-election of the incumbent party and has extra predictive power over and above several macroeconomic variables that are used in the political business cycle literature. A similar behavior characterizes consumer confidence in the United States.