Our research shows that this attitude is more common when people lose elections than many might think. What is different this year is not that the losing side thinks it was cheated but that the leaders of their party (President Trump chief among them) are stoking and amplifying that view. The president’s refusal to concede and to begin the transition process — backed by prominent party leaders — has thrown the nation into tumult.
It’s important to keep the focus on Republican elites, because they are taking advantage of a phenomenon that has existed for years. For the past eight years, we have been asking nationally representative samples of Americans, before the presidential election, “If your preferred candidate does not win the presidential election, how likely do you think election fraud will have been involved?” In 2012, 2016 and 2020, the answers were consistent: About 40 percent of Americans responded that fraudulent behavior would “very likely” or “somewhat likely” explain the negative result. This stability of opinion is remarkable given the variability in candidates, levels of polarization, and campaign issues across these three election cycles.
These patterns harden after a winner has been declared. When we asked Americans, after each of these elections, if the outcome was the result of fraud, the losing side became more likely to suspect that suspicious activity was involved, the winning side less so. For example, in 2012 — once President Barack Obama’s victory became clear — the percentage of Democrats believing fraud was “very likely” or “somewhat likely” involved in determining the outcome decreased from 49 percent to 12 percent. Meanwhile, the percentage of Republicans expressing that view increased from 54 to 62 percent. Our findings complement a number of studies showing that losers tend to see election results as less legitimate than do winners.
Historically, these beliefs are roughly symmetrical between Republicans and Democrats. In our October 2020 poll of 2,014 U.S. adults, for example, 45 percent of Republicans and 44 percent of Democrats believed that if the other side won, it would at least “somewhat likely to be due to fraud. Other research finds similar patterns: When asked, in 2012, whether “President Bush’s supporters committed significant voter fraud in order to win Ohio in 2004,” 37 percent of Democrats said this was “probably true”; when asked if ” President Obama’s supporters committed significant voter fraud in the 2012 presidential election,” 36 percent of Republicans said this was “probably true.”
In general, everyone anticipates fraud, but only losers meaningfully retain such beliefs.
This research demonstrates that the major difference between 2020 and past elections has little to do with the innate, individual-level factors that motivate perceptions of voter fraud and everything to do with the behavior of political elites. In no other year that we studied did political leaders aggressively cast doubt on the outcome of the election (or implicitly cast aspersions on results by remaining noticeably silent).
Decades of research demonstrate that elite messaging sets the tenor of political discourse — that people follow the cues of their leaders. The longer Trump and his allies traffic in evidence-free claims about voter fraud, the longer associated perceptions of fraud among the mass public are likely to remain powerful, and perhaps even grow in intensity. The impact of sustained charges of fraud are speculative, to be sure — we are now firmly in uncharted territory. As courts reject Republican claims, will public opinion — or the signals sent by elites — shift, for example? But given that the public appears primed to believe charges of fraud in the best of times, Trump and many Republican elites are playing a dangerous game. If they persist, we should not be surprised if the anxieties they are exploiting lead to mass demonstrations and, perhaps even violence.