In the wake of the contentious 2020 U.S. presidential election, a significant segment of the American population embraced the notion that electoral fraud had tainted the democratic process. This surge in doubt and conspiracy theories culminated not only in vibrant online debates but also in one of the most violent assaults on the U.S. Capitol in recent history. Almost three years after these events, polling data reveal that more than one-third of Americans continue to believe that the election was marked by fraudulent activity. This pervasive skepticism has prompted researchers to investigate the cognitive and social underpinnings of such beliefs, seeking explanations beyond traditional partisan loyalty or political influence.
A recently published study in the esteemed journal Psychological Science sheds light on a less obvious yet profound cognitive bias that underlies the tendency to question election legitimacy. The research emphasizes that while political allegiance and leadership rhetoric play notable roles, a fundamental bias embedded in human cognition can distort perceptions of cumulative information, fueling erroneous conclusions. This revelation offers a novel perspective on why misinformation and fraud allegations stubbornly persist, even when confronted with clear evidence.
At the heart of this investigative effort is the concept known as cumulative redundancy bias. This cognitive phenomenon describes how individuals struggle to disentangle overlapping pieces of information presented cumulatively over time, hindering their ability to accurately interpret final outcomes. An illustrative domain for cumulative redundancy bias lies in sports reporting where cumulative scoring adds points sequentially throughout a season or game. Observers often disproportionately weigh earlier leads, finding it counterintuitive when a team that initially dominated falls behind by the conclusion. This dissonance leads to mistrust or assumptions of foul play, despite normal fluctuations inherent in competition.
Drawing an analogy to elections, the researchers suggest that sequential reporting of vote counts—where early tallies favor one candidate but are later overtaken by another—may simulate the same cognitive challenge. Such reporting patterns can inadvertently trigger the bias, prompting observers to suspect irregularities despite legitimate voting dynamics. The issue is compounded by real-world electoral practices, such as delayed counting of mail-in ballots, which can create these politically charged shifts in lead status.
To empirically probe this hypothesis, André Vaz and his colleagues conducted a robust series of seven studies involving diverse participant pools recruited via the Prolific platform, an online research recruitment tool. The initial trio of studies employed simulated school elections featuring two candidates where the initial front-runner ultimately lost. These carefully controlled experiments allowed the exploration of participants’ perceptions under varied conditions, including modifications to the presentation of vote tallies and introduction of fraud narratives.
Participants consistently rated the candidate who led initially as the stronger competitor, revealing cumulative redundancy bias at work. Remarkably, this bias endured despite transformations of vote counts into percentages and direct exposure to fraud-related information intended to clarify legitimacy. The persistence of this effect indicated a cognitive rigidity hard to dismantle through mere presentation adjustments or appeals to rationality.
Extending the investigation to real electoral data, the next series of four studies employed authentic vote count sequences from Georgia during the 2020 presidential election, masked as results from an ambiguous Eastern European election to avoid partisan bias. Here too, the researchers replicated their findings: the pattern of early leads followed by reversals elevated fraud suspicion, whereas consistent leads alleviated such mistrust. Even after disclosing the true identity of the election and explicating the legitimate procedural reasons for vote tallies—such as delayed mail-in ballots—participants’ beliefs in fraud remained remarkably resistant.
Strikingly, cumulative redundancy bias exerted its influence across ideological divides, affecting both Republican- and Democrat-leaning participants. This bipartisan susceptibility underscores that the bias is rooted in fundamental cognitive processing mechanisms rather than partisan identity alone. The study conclusively illustrates that political allegiance, while significant, is not the sole or even dominant driver behind election fraud convictions.
Implications from this research highlight a critical intersection between cognitive psychology and electoral reporting methodologies. The entrenched bias elucidates why simply presenting transparent vote tallies or publicly explaining counting procedures may fall short in mitigating fraud concerns. Instead, altering the sequence and timing of vote count announcements could better align with human cognitive processing and reduce misleading perceptions.
One tangible recommendation advanced by Vaz involves adjusting ballot counting protocols to integrate mail-in votes earlier in the reporting process. Such an approach could minimize the occurrence of dramatic lead reversals that heighten suspicion. Although this would not erase cumulative redundancy bias itself, it would modify environmental cues triggering the bias, thereby promoting more accurate interpretations among the electorate and media consumers.
Overall, this study represents an important leap in understanding the cognitive dimensions that shape political misperception. It shifts the explanatory lens from solely partisan or conspiratorial frameworks toward inherent limitations in human information processing. As democratic societies grapple with the challenges posed by misinformation, these insights offer practical avenues for reforming electoral communication strategies to foster trust and resilience against unfounded fraud allegations.
In conclusion, the persistence of election fraud beliefs transcends partisan lines and reflects fundamental cognitive biases affecting how we process partial and cumulative information. Addressing these biases through modifications in the timing and manner of vote tallies could attenuate unwarranted skepticism, strengthening democratic legitimacy. As the world continues to navigate contested elections and political polarization, integrating cognitive science findings into electoral policy and public communication emerges as an essential priority for safeguarding democratic institutions.
Subject of Research: People
Article Title: “Stop the Count!”—How Reporting Partial Election Results Fuels Beliefs in Election Fraud
News Publication Date: 24-Jul-2025
Web References:
- https://doi.org/10.1177/09567976251355594
- https://www.psychologicalscience.org/publications/journals/psychological_science
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.
Image Credits:
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
Keywords: Cognitive bias, Political science, Psychological science
