A new cross‑disciplinary study led by Kevin T. Morris of the Brennan Center for Justice and Ian Shapiro of Tennessee State University finds that racial resentment among White Americans helps shape beliefs in election fraud, with real‑world social media chatter aligning more closely with a city’s Black population than with voting patterns. The researchers say the findings illuminate how fraud claims can be used to reject election results while maintaining faith in democracy, and the work appears in Political Behavior.
Racial resentment, in the authors’ framework, is not simply overt hostility. It describes a subtler worldview that downplays the barriers faced by Black Americans and views minority political or economic gains as unearned. When members of an advantaged group perceive their social standing as threatened, the theory suggests, they may seek to defend their position.
The study builds on a thread of research linking racial attitudes to trust in elections. Earlier work in 2016 linked racial resentment to suspicions of widespread misconduct after Barack Obama’s election, and a 2021 study found that resentful White Americans tended to become less satisfied with democracy when their preferred candidate lost.
A further line of inquiry, a 2021 survey experiment, showed that exposure to unsubstantiated fraud claims lowered people’s confidence in election integrity without reducing overall support for democracy. The new study asks whether those fraud accusations carry an added racial charge that makes them more persuasive to resentful White voters.
Morris and Shapiro conducted three investigations to explore how racial attitudes interact with claims of election malfeasance. The authors say their motivation was to understand whether and how racial considerations shaped people’s trust in the 2020 general election. They posited that fraud narratives, with their crime focus, could let some White Americans see specific elections as corrupt while still trusting a system that had traditionally benefited them.
In one analysis, the researchers examined roughly 2.7 million public posts on Twitter in 2020 containing the phrase “voter fraud.” From that pool, about 95,000 posts mentioning American cities with at least 100,000 residents were isolated and measured against local demographics and election tallies. The social media data showed that cities with larger Black populations were cited more frequently in fraud conversations, a pattern that emerged after the election; before Election Day, the examined city characteristics did not predict how often a city would be mentioned.
Among the ten cities mentioned most often relative to their size were Detroit, Philadelphia and Atlanta. These places had an average Black population share of about 36 percent, versus around 16 percent for the large cities in the broader sample and under 7 percent for U.S. municipalities overall. In the statistical models, a city’s Black population share was the only reliable predictor of these mentions; how strongly a city voted for Joe Biden did not relate to the frequency of fraud discussions, all else equal.
“I think what surprised me most was how clearly racial composition mattered in predicting which places became targets of social media fraud narratives,” Shapiro said. “Partisanship and racial politics are closely linked, but the only predictor of how frequently a city was mentioned on social media was the Black share of the population. Not even Biden’s vote share was associated with fraud narratives, other things equal.”
“Prior to election night, narratives about fraud were all over the place. They were prevalent but lacked shape,” Shapiro explained. “On election night in 2020 and in the weeks after, attention narrowed to specific places that became focal points for fraud accusations. Those places were not random.”
The team then examined how public narratives aligned with shifts in individual attitudes. Using data from about 4,300 White participants in the 2020 Cooperative Election Study, they tracked confidence in the election before and after Election Day while controlling for conservative ideology, support for Donald Trump, age, gender, education, income and Fox News viewership. Confidence dropped most steeply among Trump supporters after the election, but after accounting for these factors, higher racial resentment correlated with an additional decline in confidence. The effect was evident among Trump supporters but not among White Biden supporters.
“Trump voters generally trusted the election less after it looked like he had lost. That’s normal,” Shapiro observed. “But confidence fell off a cliff among Trump supporters who scored higher in racial resentment, following the election and the accompanying focus on Black cities.”
To test whether racial cues directly drive fraud beliefs, the researchers ran an online survey experiment in July 2024 with 1,500 White American adults. Participants read one of three mock news reports: a neutral local restaurant, allegations of voter fraud in a fictional suburban city portrayed as largely White, or identical allegations in an “inner-city” described as predominantly Black. The stories included photographs of either White or Black election workers, voters and local election administrators. After the exercise, participants were debriefed and told the city was fictional and that election fraud is rare.
Across the entire sample, reading about fraud increased suspicion compared with the restaurant story, though participants as a whole did not view the Black city as more fraudulent than the White city. Yet important subgroup differences emerged. White Republicans who read about the predominantly Black city averaged 0.65 on a scale measuring belief in fraud, compared with 0.58 for those who read about the White city—a gap of 0.07 points, or about 12 percent. This partisan comparison was not part of the authors’ preregistered analysis plan.
Racial resentment also shaped responses. Participants with higher resentment showed elevated suspicion of fraud even when reading about the restaurant, and the fraud stories moved their beliefs considerably more than those of less resentful participants. Resentful participants tended to find accusations more believable when they targeted the Black city, though this pattern was statistically reliable for only one of the resentment measures, and the authors describe the evidence as somewhat mixed. Fraud beliefs aligned more closely with resentment toward Black Americans than with White consciousness, a measure of attachment to one’s own racial group.
“The survey experiment confirms prior findings that just talking about fraud can substantially affect how much of it people think is happening,” Shapiro noted. “Despite fraud being so rare, words have power. And when those words are tied to tangible places for a political purpose, the effects can be substantial.”
The findings echo earlier research suggesting White Americans who perceived discriminatory treatment of their racial group tended to be more sceptical about election fairness, while strong White identity without grievance correlated with greater faith in elections. The study also aligns with more recent work indicating that higher levels of racial resentment link to broader anti‑democratic views among White Americans and to anxieties about multiracial governance.
As with all research, there are caveats. The social media analysis depended on public Twitter posts and cannot capture offline discourse. The researchers could not distinguish posts that promoted fraud claims from those seeking to debunk them, and some messages may have been amplified by bots or foreign actors rather than ordinary voters. They also emphasised that the study focused on how narratives are received rather than how political leaders frame them.
Shapiro stressed that the investigation addresses reception rather than elite strategy: “The data we use can’t really tell us about the motivations behind elite-driven narratives. The paper didn’t directly show people Trump’s social media posts or anything like that.” He added that it is reasonable to suspect that elites provided some structure and encouragement for the narratives spreading on social media: “A lot of agenda setting and attitude formation happens from the top down. This study is more about where and among whom these racial narratives find, excuse me for this, fertile ground.”
Looking ahead, Shapiro said he is exploring how rhetoric‑driven policies might affect turnout, and he stressed the broader democratic stakes. “This study is important because we’re in an era when people are cynical about all sorts of institutions, including elections,” he said. “But things like losers’ consent—accepting election results even when your side loses—and the belief that elections are a viable way to hold leaders accountable are at the heart of democratic functioning. Politically convenient narratives about crime and race threaten these pillars of democratic life.”
The study, titled On Fertile Ground: How Racial Resentment Primes White Americans To Believe Fraud Accusations, was authored by Kevin T. Morris and Ian Shapiro.
