How to fight election misinformation
Science advances
How to fight election misinformation
by Natalia S. Bueno
How to fight election misinformation
What distinguishes democracies from autocracies is not the existence of elections but whether the public and political actors accept the results of fair elections as legitimate. When political leaders cast doubt on fair elections, they do more than contest outcomes—they corrode public trust in democratic institutions. These attacks make it harder for winners to govern and easier for citizens to justify antidemocratic actions, including violence and the refusal to concede power.
What can be done when leaders falsely claim that elections are rigged? This question is no longer theoretical. In recent years, prominent figures like President Donald Trump in the United States, former President Jair Bolsonaro in Brazil, and presidential candidates Keiko Fujimori in Peru and Luisa Gonzalez in Ecuador have promoted claims of voter fraud when results did not go their way (Fig. 1). These narratives fuel unrest and strain democratic norms. Finding effective, efficient, and scalable ways to contain the damage of election misinformation is critical to safeguard democracies.
In a study published in this issue of Science Advances, Carey et al. (1) provide evidence that simple, fact-based messages can help. Drawing on research conducted around the 2022 elections in Brazil and the United States, they show that factual information can increase confidence in elections and reduce beliefs about election fraud.
The study draws from two distinct strategies. The first, based on inoculation theory, aims to preempt misinformation. The key components of inoculation theory are the presence of a threat and prebunking (2). Threat is the psychological mechanism by which individuals perceive their own position as vulnerable, often achieved through a forewarning. Prebunking is a forward-looking message with information. Prebunking messages expose individuals to weakened versions of common falsehoods—such as claims about rigged voting machines—paired with accurate explanations and evidence. The idea is to prepare audiences before the misinformation spreads.
The second strategy uses quotes from credible sources, that is, political elites who defend the integrity of elections, even when doing so goes against their own political interests. In both countries studied, these sources were typically allies of the political leader casting doubt on the elections. Their unexpected stance made their statements highly persuasive.
At first glance, it may seem obvious that facts help, but several well-known psychological mechanisms and context impede accurate information from changing beliefs about election fraud. Misinformation is sticky because it aligns with people’s existing views, especially in polarized environments (3). When it comes to beliefs about election integrity, individuals also typically hold sincere beliefs about who they believe was the winner of elections—they are not simply endorsing beliefs about election fraud to express other sentiments such as support for a political leader (4). As such, those beliefs are likely to be resistant to change, posing additional challenges for factual information to create change. People are more likely to accept claims that reinforce their identity and are more skeptical of information that challenges it. Emotion, partisan loyalty, and even simple habits of information processing all pose barriers to correction (5).
In this context, Carey et al.’s findings are striking. Prebunking worked in both countries and was particularly effective among people who were most skeptical of election security, those who are usually hardest to reach. Messages from credible sources also worked, although their impact was more uneven. While the evidence in the paper cannot offer definite evidence on the durability of these strategies, the effects of both strategies, but especially prebunking, did not dissipate for several weeks. These findings suggest that exposure to factual messages can have a lasting impact.
Equally promising is the fact that these strategies are relatively low-cost. Prebunking messages can be delivered in brief posts using publicly available information. For example, Carey et al.’s study shared prebunking messages delivered in short texts, in which a rumor (a “weakened dose” of false information) was paired with a correct statement, followed by a brief explanation of the safety measures taken by the agency responsible for running the election. Credible sources, when they exist, can also increase trust in elections. Together, these approaches are practical, efficient, and scalable—key traits for real-world implementation by civil society groups, journalists, or election agencies.
These messages are simple, but there are aspects of these messages that can influence their effectiveness. In “inoculation-style” messages, a forewarning is often included as part of the prebunking message. Yet, Carey et al.’s study finds that prebunking messages work better without the forewarning message. Specifically, among Republicans in the United States, prebunking with forewarning was less effective than without it, possibly because the warning triggered skepticism toward perceived “fact-checking” or elite media narratives. These findings raise the concern that aspects of messages may alienate parts of the intended audience.
Another unexpected implication of the effects of prebunking is that information about future elections, given the forward-looking nature of prebunking messages, also influences beliefs about past elections. For example, prebunking about the 2022 US midterms increased belief that Biden legitimately won the 2020 election. This retroactive effect suggests that factual information about election security measures may have potentially broad implications.
At the same time, Carey et al.’s study shows that factual information about election integrity may be effective in improving the accuracy of factual beliefs and enhancing discernment between true and false statements regarding election fraud. Still, there are several other related attitudes regarding perceptions of threats to democracy, as well as trust in election systems, for which factual information may have a diminished impact. While no action against misinformation is a silver bullet against all sorts of misleading and false claims, using messages with limited impact in a contextiple sources of misinformation can be deflating.
Furthermore, one of the most important questions regarding the effectiveness of information-based tools in combating electoral misinformation is the issue of exposure. Carey et al. show that factual information can work—but only if people see it and, presumably, pay close attention to it. Yet, while claims of misinformation endorsed by prominent political leaders receive considerable airtime and individuals are potentially exposed to these claims repeatedly, less than 3% of individuals who read an article deemed false by a fact-checker viewed its correction (6). In addition, while factual information can persuade and change beliefs, it does not necessarily alter the sources of information that individuals seek or what informs their baseline attitudes.
For policymakers and the public concerned with misinformation, studies like Carey et al.’s are an important first step: They show whether specific messages and types of messages can be effective in correcting misperceptions. However, if civil society organizations and other groups adopt the types of messages used by Carey et al., what would the consequences be? Can factual information, either in the form of debunking, credible sources, or a combination, travel as fast as misinformation? Can these forms of information be incorporated into media diets on a routine basis? Will they be consumed by individuals more predisposed to believe misinformation? Next, we need equally rigorous, creative, and informative findings on how to make information reach voters who need this type of information and how to make it sticky.
REFERENCES
1
J. Carey, B. Fogarty, M. Gehrke, B. Nyhan, J. Reifler, Prebunking and credible sources corrections increase election credibility: Evidence from the U.S. and Brazil. Sci. Adv. 11, eadv3758 (2025).
2
C. S. Traberg, J. Roozenbeek, S. Van Der Linden, Psychological inoculation against misinformation: Current evidence and future directions. Ann. Am. Acad. Pol. Soc. Sci. 700, 136–151 (2022).
3
A. J. Berinsky, Rumors and health care reform: Experiments in political misinformation. Br. J. Polit. Sci. 47, 241–262 (2017).
4
M. H. Graham, O. Yair, Expressive responding and belief in 2020 election fraud. Political Behav. 46, 1349–1374 (2024).
5
G. Pennycook, D. G. Rand, Who falls for fake news? The roles of bullshit receptivity, overclaiming, familiarity, and analytic thinking. J. Pers. 88, 185–200 (2020).
6
M. H. Graham, E. V. Porter, Increasing demand for fact-checking. Polit. Commun. 42, 325–348 (2025).
Comments
Post a Comment