���㽶��Ƶ

Ratings of American democracy by both the public and political scientists remained stable but still substantially lower than they were when President Donald Trump returned to office, according to the latest report by, which surveys the public and political scientists on the state of American democracy. 

The group is co-directed by , the James O. Freedman Presidential Professor of Government, and , the John Wentworth Professor in Social Sciences and interim dean of the Faculty for Arts and Sciences at ���㽶��Ƶ.

The Bright Line Watch report, from April through mid-September, focuses on political violence, redistricting, and support for democratic norms, as part of the group’s regular surveys monitoring democratic practices, their resilience, and potential threats in the United States. 

“While overall assessments of American democracy were unchanged as of mid-September, they could still decline further,” says Nyhan. “We certainly don’t see any evidence of those ratings rebounding; projections for the future are really pessimistic.”

The survey of 703 political science faculty from American colleges and universities and 2,750 members of the public was taken before the Department of Justice indicted former FBI Director James Comey and New York Attorney General Letitia James, both of whom President Trump has targeted.

Expert ratings of U.S. democracy in the survey placed the country closer to what is called an “illiberal or mixed democracy,” rather than a full democracy. 

An illiberal democracy has both democratic and nondemocratic attributes. For example, opposition parties might face systematic disadvantages, and prosecutorial decisions might be influenced by political considerations.

The faculty who responded to the survey rated U.S. democracy more similar to countries like Israel and Mexico than traditional democratic peers such as Great Britain and Canada.

Experts anticipate further declines in U.S. democracy when asked about their expected rating for 2027.

Among the public, the gap in ratings of U.S. democracy between Democrats and Republicans is the largest that it has been since Bright Line Watch began collecting data in 2017. The previous maximum difference of 12 points on a 0-100 scale was observed back in July 2018 (an average rating of 51 among Democrats and 63 among Republicans). The most recent survey revealed a gap of 15 points (43.6 among Democrats and 58.5 among Republicans) that grew to 18 points when respondents were asked for their predictions for rating U.S. democracy in 2027.

Bright Line Watch’s September survey period includes nearly a week after the killing of Charlie Kirk, the conservative American political activist. While most experts rated the shooting as a threat to democracy, ratings of democracy did not change measurably afterward among experts, the public, or Democrats or Republicans in the public. However, ratings of government performance at preventing political violence declined substantially among experts.

Bright Line Watch also examined public opinion toward Kirk’s slaying. Their prior work in February found that only 3% of Democrats and 2% of Republicans supported hypothetical political violence. This is consistent with collected after Kirk’s assassination by the, led by, an associate professor of government at ���㽶��Ƶ, which found that only 1% of Americans support partisan murder in a hypothetical scenario. 

But as Bright Line Watch reports, while 91% of the public disapprove of the Kirk assassination, including 88% of Democrats and 95% of Republicans, 12% of Democrats did indicate that they approved of it. According to the researchers, people may be more apt to approve of political violence against someone whom they oppose after an event has occurred, rather than in the abstract. 

In addition, the survey examines public and expert views on efforts by Texas and California to redraw voting districts to make them more favorable for Republicans and Democrats, respectively, in the 2026 U.S. House of Representatives elections.

Ninety-two percent of the political scientists surveyed rated Texas redistricting as a threat to democracy, but experts were split over California’s plan for redistricting.

Among the public, Republicans were more supportive of partisan redistricting with 64% approving of their party initiating it while only 34% of Democrats approved of initiating such an action. However, both parties tended to be more supportive of redistricting efforts when doing so aligned with their partisanship, especially in response to the other party moving first in what the co-authors refer to as the “fighting fire with fire” principle. To that end, 62% of Democrats approved of redistricting in response to another state doing so.

The survey also asked the public how often it is acceptable to violate a democratic norm in response to the other party doing so first. There was a good deal of hypocrisy in the results. While both Democrats and Republicans indicated that it’s never acceptable to retaliate against a democratic norm violation by the other party, 73% of Democrats and 52% of Republicans said it is acceptable for their party to retaliate.

Experts and the public were also asked to rate how recent and potential future events affect American democracy. For instance, 77% of experts rated Trump deploying the National Guard to Los Angeles as an extraordinary or serious threat to democracy.

To track events affecting democracy in real time, Bright Line Watch and the prediction platform Metaculus recently co-launched the, which uses human forecasters to evaluate the country’s democratic health and resilience after news breaks on an event. The researchers say that the index is intended to systematically track areas of concern during Trump’s presidency and could be used in future presidencies as well.

Bright Line Watch plans to frequently monitor the state of American democracy in the coming months. “There’s every reason for concern” about further decline, Nyhan says, "given that the indictment of former FBI Director James Comey and President Trump’s speech to the generals, among other events, took place after we finished data collection.’’