44% of Democrats Support Violence to Prevent Trump Presidency

University of Chicago political scientist Robert Pape says a new nationally representative survey found 25% of U.S. adults — including 44% of self-identified Democrats — support using force to prevent President Donald Trump from being president, as his research team warns political violence could intensify around the November midterm results.

By yourNEWS Media Newsroom.

CHICAGO — Support for political violence has climbed to levels that University of Chicago researcher Robert Pape says could make the period surrounding the 2026 midterm election unusually dangerous, with new polling finding nearly half of self-identified Democrats surveyed supported using force to prevent President Donald Trump from being president.

Pape, a University of Chicago political science professor and director of the Chicago Project on Security and Threats, disclosed the latest findings during a Sept. 24 interview with podcaster Chris Williamson.

According to Pape, 25% of U.S. adults in CPOST’s newest nationally representative survey said they supported “the use of force to prevent President Trump from being president.” He estimated that percentage represents approximately 65 million American adults.

Among respondents who identified themselves as Democrats, Pape said 44% agreed with the statement.

“This is a stunning number,” Pape said. “This is not just another turn of the data.”

The result represents support for a survey proposition, not evidence that 44% of Democrats are planning violent acts. Pape argues, however, that measuring willingness to justify political force is important because it identifies what he calls the social “kindling” from which actual violence can emerge when an election, assassination, protest or other event provides a focal point.

CPOST has conducted nationally representative surveys examining political violence since 2021. The surveys are fielded by NORC at the University of Chicago, which uses probability-based samples designed to represent the U.S. adult population. CPOST describes its research as examining political violence associated with both the political right and left.

Pape said the latest poll was CPOST’s 21st national survey and had only recently left the field when he discussed its findings with Williamson. As of Sunday, CPOST’s public survey page had not posted a full technical report for the newest poll detailing its sample size, margin of error and complete questionnaire. Previous CPOST/NORC national surveys have published those methodological details.

The latest figure continues a trend Pape has tracked over several years.

He said support for force against Trump had previously been around 6% to 9% in earlier CPOST measurements. By September 2025, shortly before the assassination of conservative activist Charlie Kirk, Pape said it had reached 14%.

The new measurement is 25%.

CPOST’s May 2025 survey had found 39% of Democrats agreeing that “the use of force is justified to remove Donald Trump from the presidency,” Pape said publicly last year. The latest 44% Democratic figure is five percentage points above that earlier measurement, although the wording Pape used to describe the newest question — preventing Trump from being president — is not identical to the earlier formulation about removing him from office.

Pape says his research team has also tried to determine what respondents understand the phrase “use of force” to mean rather than assuming everyone interprets it the same way.

Through focus groups and separate survey experiments, he said researchers asked people supporting force against Trump to explain what forms of action they had in mind.

Pape said 55% of those interpretations involved lethal violence such as assassination, while others involved collective violence resembling a riot.

“What we found is 55% of the time, they mean assassinating President Trump with rifles as the attempt that failed at Butler,” Pape said while describing the follow-up research.

That 55% figure should not be read as meaning 55% of all Americans or all Democrats support assassinating Trump. Pape was describing follow-up research into the meaning assigned to “use of force” among people who had already endorsed the broader proposition.

His polling also identifies violent or coercive attitudes on the pro-Trump side.

The latest survey asked whether respondents supported Trump using the U.S. military to suppress Democratic protesters. Pape said 16% of American adults agreed, representing an estimated 40 million people.

“16% of Americans, almost all of them of course Republican, 16% of Americans support President Trump’s use of the military to suppress Democratic protesters,” Pape said.

Pape argues that the simultaneous presence of substantial anti-Trump and pro-Trump support for coercive force makes the current environment especially unstable.

“We haven’t seen this level of violence, political violence, since the 1960s,” Pape said in a recent interview.

His concern is focused less on voting itself than on what comes afterward.

Pape has repeatedly identified the period beginning when polls close Nov. 3 and extending through the certification and seating of congressional winners in early January as the period carrying the greatest potential for escalation.

“That’s the real danger zone,” Pape said.

He said the central question is whether candidates and supporters who lose close elections accept the result or conclude that ballots, counting or certification were illegitimate. In another recent interview, Pape said roughly a quarter of respondents in a survey conducted this month were already worried about the counting of votes.

Pape has described the 2026 midterms as potentially the most dangerous election of Americans’ lifetimes and has discussed a worst-case possibility in which the contest becomes “the most violent election in American history.”

That is a risk assessment by Pape rather than a prediction that violence will necessarily occur.

Historical comparisons also require qualification. Elections during Reconstruction produced severe political and racial violence, including the disputed 1872 Louisiana gubernatorial contest, whose aftermath included the Colfax Massacre and other killings. Historians have described that episode as among the most violent elections in U.S. history.

Pape has labeled the country’s current political environment an era of “violent populism,” arguing that growing numbers of ordinary Americans — rather than only organized extremist groups — increasingly view political force as legitimate.

“Once you come to see that there are tens of millions of Americans that support political violence, that changes everything,” Pape said.

His University of Chicago research traces the phenomenon to overlapping political, demographic and economic pressures, while treating social media as an accelerant rather than the sole cause. Pape argues that rapid social change, disputes over national identity, economic inequality and weakened institutional legitimacy contribute to public acceptance of political violence.

CPOST has also documented a sustained increase in threats against elected officials.

An analysis of nearly 400 charged threats against members of Congress from 2001 through 2024 found that such threats increased approximately 600% from 2016 to 2017 and remained elevated afterward. Both Democratic and Republican members of Congress were targeted, according to the research.

Long-term research into lethal extremist violence nevertheless shows that ideological violence has not been evenly distributed historically.

A study using the U.S. Extremist Crime Database examined ideologically motivated homicides between 1990 and 2020 and identified 227 far-right homicide incidents and 42 far-left incidents. The far-right attacks caused more than 520 deaths, compared with 78 deaths attributed to far-left attacks.

Those statistics describe a particular category of extremist homicide over a specific historical period. They do not include every riot, threat, assault or episode of collective political violence, and Pape argues that the balance of political violence has changed considerably over the past decade.

“If you start in, let me just pick 1990, if you go for 35 years from 1990 to 2025,” the historical data show a majority of lone-actor violence coming from the right, Pape said. But he argued that beginning the analysis around 2015 produces a much more mixed pattern.

Trump has personally been the target of two major assassination attempts since 2024.

On July 13, 2024, Thomas Matthew Crooks fired eight rounds toward Trump during a campaign rally in Butler, Pennsylvania. Trump was struck in the ear by a bullet fired from Crooks’ weapon. Rally attendee Corey Comperatore was killed, while David Dutch and James Copenhaver were seriously wounded. A Secret Service counter-sniper killed Crooks.

The FBI investigated the shooting as an assassination attempt and potential domestic terrorism and said its investigation indicated Crooks acted alone.

Two months later, Ryan Wesley Routh positioned himself with a semiautomatic rifle near Trump International Golf Club in West Palm Beach, Florida, while Trump was golfing. A Secret Service agent spotted him before he fired.

Routh was convicted in September 2025 of attempted assassination and four other federal offenses. A federal judge sentenced him Feb. 4, 2026, to life in prison plus 84 months.

Federal authorities have also prosecuted alleged threats against Trump and other officials.

In April 2025, prosecutors charged Butler, Pennsylvania, resident Shawn Monper after accusing him of posting threats to kill Trump, administration officials and Immigration and Customs Enforcement agents. The criminal complaint alleged he had also discussed acquiring guns, ammunition and body armor.

In May 2026, federal prosecutors charged Miami Beach resident Nathaniel Sanders II with threatening Trump and transmitting threats through interstate commerce. Prosecutors alleged that his social media posts also threatened Secretary of State Marco Rubio and former Attorney General Pamela Bondi. Sanders is presumed innocent unless proven guilty.

Political violence and its public acceptance have not been limited to attacks involving elected officials.

Pape has repeatedly cited the December 2024 killing of UnitedHealthcare CEO Brian Thompson as an example of how an individual violent act can acquire substantial support among portions of the public.

Luigi Mangione pleaded guilty in federal court in August 2026 to stalking charges connected to Thompson’s killing and admitted that he followed and fatally shot the insurance executive outside a Manhattan hotel. His federal sentencing is scheduled for Dec. 18.

An Emerson College poll conducted shortly after Thompson’s killing found 17% of registered voters considered the killer’s actions somewhat or completely acceptable, while 68% considered them unacceptable. Younger voters were substantially more likely to express acceptance.

A separate Generation Lab survey conducted Dec. 19-23, 2024, sampled 1,026 college students nationwide from two- and four-year schools. The survey had a reported margin of error of plus or minus 3.4 percentage points.

Pape’s broader argument is that public approval of violence matters even when the overwhelming majority of people who express such views never become violent themselves.

He describes popular approval as creating a social environment in which a smaller number of unstable or highly motivated individuals may believe violence will bring recognition or support.

That distinction is central to interpreting the newest CPOST numbers. The survey measures attitudes toward force, not plans, weapons, organizational membership or an intention to carry out an attack.

At the same time, CPOST says more than five years of polling show persistent minorities on both sides of the political divide willing to endorse force for particular political objectives.

Pape argues that this changes the danger posed by disputed elections because a contested result could provide the focal event around which otherwise passive support becomes mobilized.

His concern for November therefore extends beyond whether polling places themselves remain peaceful.

The larger test, he says, will begin when the votes are counted and one side starts losing — and whether political leaders and voters respond through established legal processes or conclude that violence is an acceptable alternative.