Working Paper, Department of Economics, Uppsala University 2009:14
This paper investigates how information affect voting behaviour. There exist a large literature suggesting that uninformed voters can use informational shortcuts or cues to vote as if they were informed. This paper tests this hypothesis using unique Swedish individual survey data on the preferences of both politicians and voters. I find that uninformed voters are significantly worse than informed voters at voting for their most preferred politicians. This suggests that uninformed voters can not make up for their lack of information using shortcuts. Furthermore, the errors uninformed voters make do not cancel out in large elections. Estimates suggest that the ruling majorities would have switched in almost 5% of Swedish municipalities had all voters been fully informed. The effects are estimated with both parametric and nonparametric estimation techniques.
Voting behavior Citizen candidates Information Cues