Americans Back Press Freedom but Accept Limits, Pew Surveys Find

Most Americans say a free press matters for society, but many would accept government limits on reporting in certain situations, according to two Pew Research Center surveys conducted this year.

Roughly three-quarters of U.S. adults described press freedom as "extremely" or "very" important to society's well-being. Yet at least half also backed government restrictions on reporting in specific circumstances, such as protecting national security or preventing inaccurate information.

The findings suggest the public does not treat the issue as all or nothing. "People are not all or nothing," said Katerina Eva Matsa, director of news and information at Pew, describing the data as showing Americans "negotiating" when presented with scenarios and options.

Where Americans draw the line

Most respondents said the government should not be able to require a news organization to remove a story because it is critical of the government or clearly favors one political side. But about half said the government should be able to force a retraction if a story contains false information, and about six in ten said so for a story posing a national security risk.

Asked to choose between a press free to report even if some reporting is inaccurate and prioritizing the prevention of inaccuracy even at the cost of press freedom, 53% chose preventing inaccuracy and 45% chose press freedom.

Americans also saw limits on government action. Just over half said it is "rarely" or "never" acceptable for the government to force journalists to reveal the identities of unnamed government sources or to take away a news organization's broadcast license. About two in ten called such actions "sometimes" acceptable, and only about one in ten said they are "usually" or "always" acceptable. Many also said it is never acceptable for the government to give journalists access to public figures in exchange for positive coverage.

A polarized picture

Views split sharply along party lines. About eight in ten Democrats called a free press extremely or very important, compared with about two-thirds of Republicans. Republicans were more likely than Democrats to say the government should be able to require a story's removal in national security cases — about seven in ten versus about half — with a similar gap on false information.

The share of Americans who described the media as "completely" free to report the news slipped from 33% last year to 28%. About 48% called it "somewhat" free, while roughly two in ten said it was "not very" or "not at all" free. Only about two in ten Democrats now see the media as completely free.

About 64% of Democrats said they were extremely or very concerned about potential restrictions on press freedom, compared with about three in ten Republicans — a reversal from the Biden era.

About seven in ten Americans said the Trump administration and the news media have a very or somewhat bad relationship, up slightly from 64% in 2025. On whether the press faces additional restrictions under Trump's second term, 44% said it has become less free, 26% saw no change, 12% said it is freer and 17% were unsure.

Matsa cited the fraught relationship between the media and the administration, the changing media environment, the evolution of what constitutes "the press," and tech companies' role as news gatekeepers as reasons for the report. She also pointed to the country's 250th anniversary as a fitting moment to examine public attitudes.

The study also found gaps in public knowledge. While 58% of adults know the Constitution protects the press from interference by the U.S. government, only 33% know it does not shield the press from interference by private companies such as social media platforms or web-hosting services. An accompanying eight-question quiz lets users test their knowledge of the First Amendment.

The analysis drew on two probability-based surveys of the Pew Research Center's American Trends Panel: one of 3,507 U.S. adults from March 23-29, 2026, with a margin of sampling error of plus or minus 1.9 percentage points, and another of 3,488 U.S. adults from June 22-28, 2026, with a margin of error of plus or minus 1.8 percentage points.