When Donald Trump pushed to stop the count of votes in the 2020 U.S. election, a wave of election fraud beliefs followed. Some made allegations online. Others stormed the country's Capitol on January 6, 2021. Today, more than a third of Americans believe that the election had marks of foul play (Weiner et al., 2024).
A number of studies have examined why such extreme beliefs arose from that election, linking the strong sentiments to loyalty to one's preferred party, the influence of a political party leader, or the susceptibility of certain personalities to conspiracy . But a new study, recently published in Psychological Science , shows that alliance to a political party isn't the only thing that contributes to wrongly questioning the legitimacy of an election. Instead, there's a deeper, more fundamental bias in our brains that can lead us astray.
Tear gas outside the U.S. Capitol on January 6, 2021. Photo credit: Tyler Merbler from USA, CC BY 2.0 https://creativecommons.org/licenses/by/2.0 , via Wikimedia Commons
"That's the other side of the research—cold cognition, which basically says some things actually don't need motivation to happen," said André Vaz, a postdoctoral researcher at Ruhr-Universität Bochum and a coauthor of the new research. "Just the way we process the information in a dispassionate way can still give rise to these wrong beliefs."
The new study zooms in on a mechanism called cumulative redundancy bias, where people reach wrong conclusions about cumulative results because of the mind's inability to filter out redundant information. A common example of this is sports teams. Often, the progress of a sports team is reported cumulatively, explained Vaz, where the points add up as time goes on. "The last observation, the total number of points at the end of the season, or the votes, is the most informative," Vaz said. However, because of cumulative redundancy bias, "we can't help but be influenced by all that we observed in the past," he added.
In other words, if a sports team is winning at the beginning of the season but loses its lead by the end, it feels like it was rigged. After all, the losing team performed better in the initial parts of the season, so it must be the better team overall. Normally, this bias doesn't have repercussions. But elections, Vaz noted, present information in a way similar to sports games. This means that cumulative redundancy bias could have contributed to the fraud beliefs in the 2020 U.S. election.
To test this hypothesis, Vaz and colleagues conducted seven different studies, recruiting seven sets of participants through Prolific, an online platform that connects researchers with participants. For the first three studies, the researchers tested various scenarios by exposing study participants to the results of a mock school election. In this mock election, two candidates were pitted against each other. One candidate took the lead initially but then lost their lead at the very end. The team then used real-life data from the 2020 U.S. election from Georgia for the following four studies, as the state had results that fit a similar pattern of progression. These numbers were initially presented as an undisclosed Eastern European election to the study's participants.
For the mock election, the researchers found that cumulative redundancy bias influenced perceptions of a candidate's strength, where the one who took the lead first was perceived as the stronger candidate. The bias persisted even after introducing the idea of election fraud and reporting the votes as percentages.
With the data from Georgia, "again, we find the effect," Vaz said. Just by the way the vote counts progressed, people believed that voter fraud was more likely. The study also found that reversing the pattern, where a candidate takes the lead and keeps the lead, lessens beliefs in fraud. Interestingly, the effect of cumulative redundancy bias remained even after the researchers explained to the study participants why the results were legitimate, with participants resistant to discarding the idea of fraud. The bias also acted on Republicans and Democrats alike, where fraud beliefs continued beyond party lines even after it was revealed that the numbers were from the 2020 Biden–Trump election.
These findings show that both cumulative redundancy bias and how election results are reported are powerful influences in fostering beliefs of fraud. Although there isn't much we can do to change these inherent biases, it is possible to change how votes are counted to prevent this late-lead pattern from appearing. Vaz noted that one of the factors that contributed to the 2020 U.S. election results was that mail-in ballots were counted late. Counting these mail-in votes earlier could prevent such a drastic change in results near the end of an election.
"It wouldn't kill the bias, but it would change the pattern of vote counting," Vaz said. "At least in a way where it wouldn't lead people into thinking one candidate was clearly ahead."
References
Vaz, A., Ingendahl, M., Mata, A., & Alves, H. (2025). "Stop the count!"—How reporting partial election results fuels beliefs in election fraud . Psychological Science, 36(8), 676–688.
Weiner, R., Clement, S., Guskin, E. (2024, January 2). Republican loyalty to Trump, rioters climbs in 3 years after Jan. 6 attack. The Washington Post.
