R. Michael Alvarez, D. Roderick Kiewiet, and Lucas Núñez, California Institute of Technology

Protest voting is a phenomenon of growing interest to both academic and other observers of

elections throughout the world. In this paper we create a taxonomy of the wide range of behaviors

that have been labeled as protest voting, and review the literature studying them. Our review

shows that the analysis of protest voting faces significant challenges, first and foremost because

researchers lack a clear conceptual definition of the term. Lack of conceptual clarity is particularly

patent in the analysis of support for fringe or insurgent parties, as what is often labeled

as protest voting is consistent with patterns of ideological attraction and retrospective voting

that characterize support for mainstream parties. Protest voting can also take the form of voters

casting blank, null, or spoiled (BNS) ballots, but a major problem for analyses here is that it is

difficult to distinguish between BNS ballots that are cast intentionally from ballots that are voided

due to unintentional mistakes. Protest votes can sometimes take the form of tactical voting, and

in a few cases protest voting is organized and directed by political elites. In such cases, however,

protest voting has been either ineffectual or counterproductive.