Working Paper: NBER ID: w12709
Authors: Kelly Shue; Erzo F. P. Luttmer
Abstract: If voters are fully rational and have negligible cognition costs, ballot layout should not affect election outcomes. In this paper, we explore deviations from rational voting using quasi-random variation in candidate name placement on ballots from the 2003 California Recall Election. We find that the voteshares of minor candidates almost double when their names are adjacent to the names of major candidates on a ballot. Voteshare gains are largest in precincts with high percentages of Democratic, Hispanic, low-income, non-English speaking, poorly educated, or young voters. A major candidate that attracts a disproportionate share of voters from these types of precincts faces a systematic electoral disadvantage. If the Republican frontrunner Arnold Schwarzenegger and Democratic frontrunner Cruz Bustamante had been in a tie, adjacency misvoting would have given Schwarzenegger an edge of 0.06% of the voteshare. This gain in voteshare exceeds the margins of victory in the 2000 U.S. Presidential Election and the 2004 Washington Gubernatorial Election. We explore which voting technology platforms and brands mitigate misvoting.
Keywords: Voting Behavior; Cognition Costs; Election Outcomes; Ballot Design; Misvotes
JEL Codes: D01; D72; D83; J10
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| ballot layout (K16) | voting behavior (D72) |
| candidate adjacency (Y80) | voteshares of minor candidates (D79) |
| misvotes (K16) | total votes cast (K16) |
| voting technology (K16) | level of misvotes (D79) |
| cognitive costs (D91) | likelihood of misvotes (K16) |
| misvotes (K16) | election outcomes (K16) |