Learning by Cue-ing? How Parties and Voter Guides Affect Candidate Perceptions

• In Event: Primaries and Candidates

Abstract

Efforts to educate citizens about the candidates and issues at stake in elections are widespread. These include efforts by political parties to circulate endorsements conveying their recommendations about which candidates or positions to support. They also include nonpartisan groups’ efforts to compile information about candidates’ policy views and distribute it via voter guides. Much previous research demonstrates that political party endorsements increase support for candidates among co-partisans in the electorate. More recently, scholars have found that both party endorsements and voter guides can improve voters’ ability to identify candidates who share their policy views in low-information local elections (Boudreau, Elmendorf and MacKenzie 2018).

But while previous research shows that these types of information influence voters’ candidate choices, several unanswered questions remain. One question is what exactly do voters learn from party endorsements and voter guides? Formally, an endorsement conveys nothing more than that a party’s membership supports the candidate. In practice, however, party endorsements can communicate much more than this. For example, voters might infer that an endorsed candidate is more likely to win the election than non-endorsed candidates. Because the political parties have well-established ideological reputations, a party endorsement could also help voters learn about candidates’ policy views, enabling them to identify the candidate whose views best resemble their own. Indeed, voter guides are specifically designed to confer this information. Further, given substantial evidence that party identification is an affective attachment that transcends both electoral prospects and issue positions, it is possible that party endorsements lead voters to infer additional information about candidates’ personal attributes. Indeed, partisans could view candidates endorsed by their party as more intelligent, empathetic, or as better leaders even when they lack information to corroborate such judgments.

Another open question is what types of voters benefit from party endorsements and voter guides when they choose to receive them? Previous experimental studies of these types of information force all respondents within a given treatment group to receive party endorsements and/or voter guides. In these studies, respondents do not choose what information they receive about candidates, as they do in real-world election settings. There is reason to believe that voters who choose to receive information are different from those who do not. Indeed, Mummolo and Peterson (2015) find that few voters took advantage of an opportunity to view a voter guide compiled by their local newspaper, and that those who did so exhibited high levels of political interest and knowledge. These findings suggest that information about candidates, whether circulated by political parties or civic groups, is likely to have minimal effects in practical settings.

We address these questions by conducting survey experiments that randomly assign a nonpartisan voter guide, political party endorsements, or no information about candidates running for mayor in San Francisco in 2018. We also provide a novel treatment condition where voters choose whether to receive the voter guide, party endorsements, or no information about the candidates. We assess the effects of party endorsements and voter guides on voters’ perceptions of the candidates’ policy positions, likelihood of winning, and personal attributes (intelligence, empathy, and leadership). Importantly, we compare the effects of party endorsements and voter guides among those who did and did not choose to receive information about the candidates. In randomly assigning party endorsements, voter guides, and a “choice” condition where voters decide what information to consume, we shed light on why different types of information work to improve citizen competence and what types of voters (low versus high in political knowledge / interest) are most likely to benefit.

We find that both party endorsements and especially voter guides improve voters’ ability to accurately describe the candidates’ policy views. These gains are concentrated among voters with low levels of political knowledge. Party endorsements also increase perceptions of candidates’ intelligence and leadership qualities among co-partisans whereas voter guides do not. In contrast to previous scholarship, we also find that a large majority of voters choose to receive information about candidates when given an opportunity and that most prefer a nonpartisan voter guide summarizing the candidates’ positions to receiving party endorsements. This is true among both low- and high-knowledge voters. The gains in these voters’ ability to correctly identify the candidates’ policy views are comparable to those “forced” to receive information in the experiment.

Authors

• Cheryl Boudreau, University of California, Davis

• Christopher S. Elmendorf, UC Davis

• Scott A. MacKenzie, University of California, Davis