Does Choice Architecture Affect Voters’ Decision to Cast Null Ballots?

• In Event: How Ballot Structure Contributes to Voter Errors and Omissions

Abstract

Latin America registers some of the highest rates of invalid voting in the world. Individual survey data suggests that at least some invalid votes are cast as a form of protest, including from high-knowledge voters dissatisfied with government performance across the region. Nevertheless, invalid voting is a multicausal phenomenon: it can equally be the product of intention or error, by either the voter or an electoral authority. Parsing out the relative weight and influence of these causes has thus far proven empirically impracticable, as data is either generated without distinction as to the causal mechanisms or is restricted to a subset of the universe of causation (e.g., self-reported voting behavior in individual surveys, leaving out unintentional spoiled ballots as well as ballots nullified by the electoral authority). I address this gap in the literature by providing an estimation of deliberate invalid voting. Exploiting a natural experiment in the state of Coahuila, Mexico, where a pilot of electronic voting machines included an explicit option to annul in several elections, I calculate that deliberate invalid voting accounts for less than a third of the share of null votes. Because it is not possible to rule out that including the option to annul enhances the likelihood of casting a null vote by providing a “protest cue,” I argue the estimation offers a ceiling for the relative proportion of voluntary spoilage within invalid voting. The findings pose normative and practical implications for the design of electoral rules and electoral material.

Author

• Raquel Chanto, INCAE Business School