Information, Exposure, and Support for Changing Election Rules

• In Event: Politics of Electoral System Reform

Abstract

Plurality elections produce results where winners can be elected without majorities and losing candidates can play a 'spoiler' role by fragmenting votes. Ranked Choice Voting (RCV), or the Alternative Vote, may mitigate these issues and has recently been adopted in several US jurisdictions by popular vote. Surveys conducted around the time of such popular votes allow for tests of demographic, attitudinal, and self-interest hypotheses about support for electoral system change (e.g. Banducci, Donovan & Karp 1999; McCarthy & Santucci 2021), but may provide less leverage for assessing how attributes of election systems affect support for change.

We seek to expand on such work by using survey experiments to test how differently framed depictions of plurality voting, and 'interacting' with a hypothetical RCV choice, affected support for RCV. Broadly, we seek to understand how people reason about proposals for electoral system changes. Specifically, we seek to understand how people might reason about changing from plurality to an alternative such as RCV.

We placed two experiments on the 2020 CCES specific to RCV. The first asked respondents if they favored RCV after being randomly assigned to one of three treatments describing plurality voting: A prompt about plurality voting producing winners most people did not support, a prompt that notes RCV can produce winning candidates despite votes being divided across three or more candidates; and a prompt that explicitly describes a 'spoiler' effect. We expect that respondents assess proposed election rules based on attributes (or 'failings' if prompted) of rules, and that each condition is associated with greater support for RCV.

Two versions of a second experiment employed a different item asking if RCV should be used in the US. This was designed to test if making a ranked choice similar to an actual ballot affected support for RCV. Half our sample was randomly assigned to participate in a mock-RCV choice where they ranked (on screen) four former US presidents, then asked their opinion about RCV after making their ranking. Our expectation is that interacting with a mock RCV choice increases familiarity with the system, and corresponds with greater support for RCV.

Our CCES module also included several additional questions on perceptions of how well democracy is functioning in the US that will also be used for non-experimental tests of how perceptions of democracy affect support for a proposed electoral system change, independent of being presented information about the proposed electoral system change. Our hypothesis is that discontent with democracy generally will lead to support for electoral rule change, independent of information of about the consequences of the change.

Data will be provided to us soon from CCES, these tests have not yet been performed. This proposal can be considered a pre-registration of the experiments.

Authors

• Todd Donovan, Western Washington University

• Caroline J. Tolbert, University of Iowa