Without Party Labels, Voters Lean Toward Experienced Candidates — But Parties React Differently

Why Voter Information in Nonpartisan Races Matters

Patricia A. Kirkland and Alexander Coppock investigate how voters choose candidates when party labels are unavailable. Nonpartisan elections remove a common cue that voters use to infer ideology and competence, so voters must rely on other candidate attributes. Understanding which attributes gain importance without party labels matters for how nonpartisan institutions shape accountability, candidate selection, and electoral competition.

How the Authors Tested It

Kirkland and Coppock use conjoint survey experiments to measure how much weight voters place on different candidate attributes. The experiments manipulate whether partisan information is shown and vary candidate features such as career (job) experience, political experience, and signals of partisan or ideological alignment. Experiments were run on both a nationally representative sample and convenience samples to assess robustness across respondent pools.

Key Findings

• When party labels are removed, voters place greater weight on candidate experience overall.
• The shift toward experience differs by respondents' own partisan affiliation: Republicans increase the weight they assign to job/career experience, while Democrats give more weight to political experience.
• Voters also attend to signals that imply partisan or ideological alignment when labels are absent, but these signals do not fully substitute for explicit party cues.
• The results support the idea that visible partisan information can crowd out attention to other candidate characteristics in voters' decision processes.

What This Means for Elections and Research

The study provides microfoundational evidence that institutional features—specifically whether elections are labeled partisan or nonpartisan—reshape the information voters use. For scholars and practitioners, the findings suggest that nonpartisan ballots may advantage candidates who can credibly demonstrate experience or who tailor nonlabel signals to voters’ expectations, and that these dynamics play out differently across the electorate.

Methods Note

The conclusions rest on experimental manipulation of candidate attributes in conjoint designs carried out by Kirkland and Coppock across multiple samples; the paper emphasizes attribute-level causal inference rather than aggregate turnout or institutional change effects.