Election trust campaign shifted one belief, but left most others unchanged
A preregistered field experiment during the 2024 US presidential election found that a nonpartisan election information campaign produced mostly null effects among skeptical conservative voters. The campaign, built around messages from election officials and designed to explain safeguards around election integrity, did not meaningfully change turnout, overall confidence in elections, or perceptions of how often fraud occurs. One exception emerged: trust in the vote-by-mail process increased.
The gap between survey findings and field behavior was large. In prior survey experiments, informational messages about elections have often looked powerful; in this real-world setting, the effects were much smaller. That divergence matters because election trust is not an abstract attitude measured in a quiet questionnaire. It is filtered through partisan identity, repeated elite cues, and the pressure of an actual campaign season.
A real-world test during a polarized presidential election
The experiment randomized conservative voters to either a control condition or a multimodal messaging campaign delivered during the 2024 contest. The campaign was crafted by political professionals and centered on explanations of election integrity protections. Because the intervention reached voters in a live political environment rather than a survey panel, it offered a harder test of whether corrective information can move entrenched views.
The study examined several outcomes across large samples. Turnout analysis included 147,488 voters. Measures of general confidence in elections ranged from 4,740 to 4,868 participants, while perceptions of the frequency of fraud were assessed in a sample of 4,925. The vote-by-mail outcome included 6,438 participants. Across those endpoints, the pattern was consistent: most measures did not shift in a detectable way.
The one positive result was narrow and specific. Trust in the vote-by-mail process improved, suggesting that voters may be more open to targeted reassurance about a single election procedure than to broad persuasion about the legitimacy of the system as a whole.
Why broad reassurance failed where a targeted message moved one process measure
Election trust is not a single belief. It is a bundle of judgments about counting accuracy, fraud prevention, administration, and the legitimacy of outcomes. A broad campaign may be too blunt to alter that entire structure, especially among voters already skeptical of election administration. Partisan identity can act like a filter, allowing some procedural information through while rejecting messages that sound too general or too corrective.
The vote-by-mail result fits that logic. Mail voting is concrete, familiar, and easy to describe in operational terms. Voters can picture ballots being sent, returned, and counted. When a message addresses a discrete process with visible safeguards, it may feel more credible than a sweeping claim that elections are secure overall. The broader judgments—whether elections are trustworthy in general or whether fraud is common—sit closer to identity and may resist simple informational correction.
A single sentence captures the core result: the campaign did not convert skepticism into trust at scale.
Field evidence came in far below survey expectations
The contrast with survey experiments is the most consequential feature of the study. In a survey setting, respondents are exposed to a message in a controlled environment, detached from campaign noise, social cues, and the emotional intensity of an election. Real-world campaigns unfold amid competing advertisements, partisan media, interpersonal discussion, and preexisting mistrust. The live political environment can dilute or overwhelm even carefully designed messages.
That difference points to a basic problem in political persuasion research. Responses measured in a survey do not always survive contact with the field. When people are asked abstract questions about electoral integrity, they may appear receptive to factual explanation. When the same people encounter messages during a heated presidential race, those effects can shrink quickly. The study offers direct evidence that field effects are materially smaller than the ones often seen in experimental survey work.
Turnout and fraud perceptions remained resistant to change
The intervention also did not move turnout in a meaningful way. This result suggests that an information campaign about election integrity is not, by itself, a mobilization tool strong enough to change voting behavior in a large conservative electorate during a presidential cycle. Voting decisions are shaped by partisanship, candidate preference, habit, and social pressure, none of which are likely to change simply because election officials are presented as trustworthy.
Perceptions of fraud also held steady. That outcome is clinically relevant in the civic sense: false beliefs about fraud are often durable, and they can anchor broader distrust in institutions. Corrective messages may need to do more than present facts; they may need repeated exposure, source credibility, and alignment with existing values. Even then, some beliefs may remain stubbornly intact.
The field trial therefore does not support the idea that a single nonpartisan campaign can quickly reverse skepticism in a polarized electorate. It does support a narrower claim: procedural messaging can move confidence in a specific voting method, but not necessarily confidence in elections overall.
What this means for election officials and public communication
Election officials and communicators looking for a practical use of the findings should be cautious about expecting broad trust restoration from a one-off campaign. Messages may work best when they are concrete, limited in scope, and tied to an election process voters can visualize. Vote-by-mail is one example. General assurances about election integrity are much harder to move, especially when skepticism is already partisan.
The study also raises a strategic question for future campaigns: should public messaging focus on broad legitimacy or on narrow procedural steps? The data favor the latter. In a contentious political environment, voters may accept explanations about how ballots are handled, tracked, and counted more readily than they accept a wholesale argument that elections are trustworthy.
Rigorous field evidence like this is rare, and the practical lesson is unambiguous. Campaigns built to change election trust will need more than polished messaging and official voices if the goal is to shift deep-seated skepticism during a live presidential contest. The current trial leaves open how durable the vote-by-mail effect is, whether repeated exposure would produce larger changes, and whether different voter groups would respond differently under the same conditions.