Fewer Americans in new survey expect Trump to concede if he loses in November

Nearly three-quarters of Americans, 74 percent, believe former President Trump will refuse to concede if he loses the 2024 election, a significant increase from similar surveys in 2020.

The CNN poll released Monday signals that many voters expect election denial rhetoric to continue if Trump loses again in November.

Just over half of respondents correctly expected Trump not to concede if he lost the 2020 election, about 55 percent in August 2020 and 58 percent in October of that year.

He did indeed refuse to accept the results of the election, with his supporters later assaulting the Capitol on Jan. 6, 2021, following months of rhetoric claiming the result was stolen or otherwise fraudulent. 

Only about a quarter of respondents said they believe President Biden would not concede if he lost in November, 23 percent. That figure is slightly lower than 2020 surveys about Biden, and in line with a 2016 survey regarding then-Democratic nominee Hillary Clinton.

Overall, an overwhelming majority of survey respondents, about 86 percent, said the losing candidate in a presidential election has an obligation to concede. That is identical to figures from 2020 surveys, and slightly higher than a 2016 survey.

The former president has already sown the seeds of 2020 election denial rhetoric for the 2024 race. Trump has repeatedly warned that Democrats could rig the general election against him, and he has repeated the 2020 election fraud claims that opponents have dubbed “The Big Lie.”

Trump faces a pair of criminal cases related to election denial and attempts to overturn the result of the 2020 election, a state criminal case in Georgia and a federal one based in Washington.

The Supreme Court is also expected to rule this month on whether Trump can be on the general election ballot after a court in Colorado ruled that his involvement in the Jan. 6 Capitol riots violated the Constitution and disqualified him from office.

The CNN poll surveyed about 1,200 people over the course of late January, with a margin of error of plus or minus 3.4 percent.