Subscribe now

This was an ordinary act of community speech, rooted in American soil for 250 years.

But federal officials took aim at Kae’s group. They didn’t attempt to silence her directly. Instead, they demanded that Facebook do what the government cannot do itself: silence people discussing and criticizing its activity. Facebook immediately complied, cutting off over 100,000 members from a space they relied on for community information. The attorney general and secretary of homeland security then publicly took credit for the takedown.

But this was unconstitutional. It was censorship by proxy. Jawboning. And the First Amendment forbids the government from laundering censorship through private companies.

Kae’s experience is a more blatant and unsubtle evolution of a pattern previously seen in cases like Murthy v. Missouri: Government officials coerce platforms to suppress certain viewpoints without employing formal legal process. The mechanism is indirect but powerful. Speech is not banned outright, but it disappears because government officials apply pressure and platforms understand the consequences of refusing.

That leaves users with a basic question: Who is really setting the rules of the online public square?

Americans do not seem eager to hand that power to either side. In FIRE’s 2024 Social Media Report, 64% of Americans said they do not trust social media companies to make decisions about what speech is allowed online. That same portion also said they don’t trust the government to do it either.

Americans care not only about what platforms decide, but how they decide. In the September 2023 survey, 78% said unbiased moderation is important, 77% prioritized transparency about government involvement, and 71% supported appeal processes.

In other words, people may accept some level of online content moderation, but they are far less willing to trust systems that operate in the shadows with the government.

That concern becomes even sharper when government involvement is explicit. In FIRE’s latest edition of the National Speech Index, 75% of Americans said they were at least somewhat concerned about government pressure on tech companies to suppress viewpoints. And 77% said the same about government access to user data for surveillance.

Those concerns cut across ideology, but they shift with power. In the fall of 2023, Republicans were especially likely to distrust government involvement in speech decisions. By April 2026, liberals were especially likely to express concern about government pressure on tech companies.

That shift shows how quickly fears about censorship can change depending on who holds power and whose speech appears to be at risk. The First Amendment cannot depend on who’s in the White House. Government power may seem useful when aimed at your opponents, but becomes dangerous the moment it is aimed at you.

Jawboning is particularly corrosive because it is difficult to hold the government accountable for an action they pressured a third party into taking.

The same problem is likely to intensify as new technologies reshape public debate. FIRE’s June 2025 survey on AI in political campaigns found majorities worried that regulation could be used to suppress criticism or chill lawful expression. Whether the issue is social media moderation, government access to user data, or AI regulation, the principle is the same: The government cannot pressure private companies into suppressing lawful speech.

The future of free speech requires clear lines now.

Kae could have let the group vanish. She could have accepted defeat against Leviathan and moved on. Instead, she fought back. Together with FIRE, Kae is suing the Department of Justice and the Department of Homeland Security to vindicate her First Amendment rights.

A federal court has already recognized the danger, finding Kae and her co-plaintiff likely to succeed on their First Amendment claims and blocking the government from continuing its coercion campaign.

Her story is not just about one Facebook group. It’s about whether Americans can document and discuss government activity in public without federal officials leaning on social media companies to make that speech disappear. It’s about whether the government can sidestep the Constitution by jawboning people into silence. Simply put, the government doesn’t get to decide what speech survives online. And it doesn’t get to conscript private platforms to do its censorship bidding for it.

Subscribe now

Data Byte

• Students who choose to identify as Protestant are the most tolerant religious group on campus.

• When given a choice between “Protestant” and “Christian,” most Protestant students seem to identify as Christian. Those who do are not nearly as tolerant as those who identify as Protestant.

• Mormon students are also very tolerant.

• Atheists and agnostics are especially tolerant of left-wing, but not right-wing, speakers.

• Buddhist and Hindu students are especially intolerant of right-wing speakers.

• Catholic students have the lowest average tolerance.

Deep Dive

We’ve seen before that students who frequently attend religious services as well as students who are studying religion are unusually tolerant of controversial speakers1, meaning they are willing to let them speak on campus. (We measure left- and right-wing tolerance by whether students say they would allow those controversial speakers on those sides.) This raises a few questions. Are religious people more tolerant? More specifically, which religions’ members tend to be more tolerant? (When I refer to people of a particular religion in this article, unless otherwise indicated, I generally mean those who identify as that religion.)2

Look at the following map. Being higher up means that a religion has greater average tolerance, while being more to the left or right means that a group shows bias in those directions.

Looks like God’s been busy.

Okay, there’s a lot going on here. Protestants are the most tolerant, so we’ll look at them in more detail. (Mormons are nearly as tolerant, though more right-biased, but we have more data on Protestants, so they’ll be easier to examine.) The first thing one might ask is whether Protestants are more tolerant because of their religion. We already know that, at least among college students, there’s a large gender tolerance gap (men are more tolerant than women) and a small ideology tolerance gap (conservative students are slightly more tolerant than liberal students). So then are Protestants more tolerant because they’re more likely to be male (or conservative)?

We can check by making a scatterplot binned by ideology, gender, and whether the person is Protestant or not, then seeing whether each Protestant point is more tolerant than the corresponding non-Protestant point:

So that’s hard to read. To make it clearer, we’ll make vector difference plots showing the vectors from non-Protestant to Protestant within each ideology and gender. Each line shows the path from non-Protestants of that ideology/gender to their Protestant counterparts. In other words, we can show how much “the Protestant effect” impacts tolerance for a certain group, regardless of how well that group does overall. We are not showing causation. That is, factoring in the Protestant effect doesn’t mean being Protestant is the cause of higher tolerance, just that the two are correlated. So, here’s a look at the Protestant effect for men and women:

When it comes to nearly every ideology or gender, Protestants are significantly more tolerant than their non-Protestant counterparts, and not only that, they’re usually more tolerant of both sides. So the Protestant effect isn’t (primarily) a result of gender or ideology. For some reason, Protestants are just unusually tolerant of controversial speakers. We can also show this (less intuitively, but more concisely) by just making a version of the religion free speech map that controls for gender and ideology:

Protestants remain unusually tolerant.

We can more closely examine what these tolerance differences look like by making heatmaps. Below, each small diamond is a particular set of left- and right-wing tolerances, and the brighter squares are where more of that group is:

Ah. So the big effect here is that Protestant men and women have roughly double the chance of being perfectly tolerant of their non-protestant counterparts. Whatever’s going on, it’s primarily about Protestants having a higher chance of being perfectly tolerant, rather than a general increased tolerance.

Okay, wait a minute. That’s about four thousand Protestants in a sample of over 68,000 students. Only four thousand? That doesn’t seem right. Let’s check how many students are in each religion:

There are a lot more students who identify as Catholic than Protestant, but even more who identify as “Christians.” Given that Protestants outnumber Catholics in the general population, while we can’t be certain, this sure looks like most Protestants tend to identify as “Christian,” whereas most Catholics tend to identify as “Catholic.”3

The Protestants who identify as “Protestant” are unusually tolerant, but the Protestants who identify as “Christian” aren’t. So what’s different about those two groups that could cause this? Well, maybe the Protestants who identify as “Protestant” tend to be more devout. Do the students who identify as “Protestant” attend services more than those who identify as “Christian”?

Well, there we go. Clearly, the answer is “yes.” On a separate note, it’s notable that Mormons, the other group with similar tolerance to Protestants, top the religious attendance chart.

On a separate note, remember that we’re looking at college students here. Colleges are often hostile environments towards conservatives or dissenting voices. What if there’s a selection pressure that means only the most tolerant Protestants decide to go to college? Ideally, we’d hope to ask the same questions to a large nationally representative sample. While we do have national data on free speech and religion (Protestants don’t do particularly well on allowing people to speak), the sample sizes are much smaller, it doesn’t ask the tolerance questions, and — very importantly — the religion data doesn’t include a separate Christian category.

So how else can we check for selection effects? Well, I’d imagine that pressures against Protestants or conservatives (which Protestants are more likely to be) are much smaller or nonexistent at conservative schools.4 Let’s make our map again, but this time we’ll exclusively use data from schools with conservative student bodies:

Unsurprisingly, there’s general rightward movement, but Protestants remain unusually tolerant. So it’s unlikely that Protestant tolerance is a selection effect.

We see a similar thing happen with atheists and agnostics compared to students who cite “Nothing in particular” as their religious belief. In both cases, the group willing to use more precise labels to describe themselves is significantly more tolerant.5

There’s an entertaining final piece of evidence that something like religious devotion is tied to greater tolerance. If someone selects “Other” for religion, it allows an open-ended response. Out of 68,510 students who responded to the survey, two used the open-ended response to argue with our classification scheme:

Anglican. Common mistake that there is an assumed binary between Catholic and Protestant (Anglican is neither).

Lutheran (LCMS) (the original Protestants, who are very different from Baptists, Methodists, Pentecostals, and especially Jehovah’s Witnesses, who are not Christian. I will not select an option that includes Jehovah’s Witnesses in the list of examples.) (By the way, JWs don’t even consider themselves to be Protestants either: https://www.jw.org/en/jehovahs-witnesses/faq/are-jehovahs-witnesses-protestants/)

Both students were perfectly tolerant of both sides, i.e. would definitely allow all speakers. This is only two students, so it shouldn’t be taken as strong evidence,6 but it’s still striking.

So okay, is Protestantism making people more tolerant? We can make a reasonable guess about which way the causality goes here. While we don’t have anything here that rules it out with complete confidence, it seems unlikely that being politically tolerant causes students en masse to convert to Protestantism. Much more likely is that either Protestantism teaches tolerance, that environments that raise people to be Protestants also tend to raise them to be more tolerant, or that some other factor causes Protestants to be more tolerant.

I don’t have immediate answers about what exactly seems to produce the Protestant effect. But if we can identify why this is happening, it might point us towards powerful tools to build a free speech culture. This one might be really useful.

The code, data, and codebook used to generate these plots are available here.

The Catholic church is a pedophilic institution.
The police are just as racist as the Ku Klux Klan.
Children should be able to transition without parental consent.

Transgender people have a mental disorder.
Abortion should be completely illegal.
Black Lives Matter is a hate group.

Subscribe now

Notably, the average American opposes censorship far more than college students in this country. Most Americans reject overtly violent censorship tactics. In fact, only 18% say it’s at least rarely acceptable to use violence to stop a speaker, compared to 33% of college undergraduates — and 27% last fall despite the murder of Charlie Kirk weeks before.

The new NSI data also shows liberals are more accepting than conservatives of shouting down speakers or blocking them. Very liberal Americans are also more accepting of shouting down speakers than moderate Americans.

There’s no significant difference when it comes to using violence to stop a speech, but a lot of Americans support such tactics and that support has remained relatively unchanged over the past two years, despite multiple assassination attempts against President Trump, an assassination attempt against Pennsylvania Governor Josh Shapiro, and the assassinations of United Healthcare CEO Brian Thompson and Charlie Kirk.

The NSI also asked Americans how acceptable it is to use legal protest tactics that are fully protected by the First Amendment, such as heckling the speaker before they deliver their remarks, silently holding up protest signs during the speech, or peacefully protesting outside the speech.

First, some good news. Almost no American surveyed, regardless of their ideological views, said that it is ever unacceptable to peacefully protest outside the speech. That said, significantly more liberals were accepting of peaceful protest than their moderate and conservative counterparts — but not their very conservative counterparts. If nothing else, these findings show that Lizardman’s Constant, which says about 4% of any survey will give bizarre or insincere results, is alive and well.

Yet, when it comes to other protected forms of protest, very liberal and liberal Americans were significantly more supportive of their use than moderate, conservative, or very conservative Americans. Just 4% of very liberal Americans surveyed, and just 6% of liberals, said that it was “never” acceptable to attend a speech and silently hold up a protest sign, compared to 9% of moderates, 12% of conservatives, and 18% of very conservative Americans.

Heckling a speaker before their remarks was less popular. Most moderates (52%), conservatives (58%), and very conservative respondents (62%) said this was “never” acceptable. But only 35% of liberals and 22% of very liberal Americans agree. And liberals are much more likely to say heckling before remarks is “never” acceptable compared to very liberal Americans. Simply put, the more you stand to the left, the less likely you are to oppose this form of protest.

Taken together, these results point to a clear pattern: liberals are generally more accepting of the protest tactics we asked about, including both protected forms of protest and tactics that can interfere with other people’s expressive rights.

They are more supportive of peaceful protest, silent sign-holding, and heckling a speaker before their remarks. But they are also more accepting of shouting down speakers and blocking others from entering speaking events. And these two types of tactics should not be treated as interchangeable.

In other words, liberals seem to support all forms of protest, even illegal ones, whereas conservatives are more likely to oppose protest tactics, even some protected ones. But a healthy culture of free speech requires defending the right to protest while drawing a firm distinction between dissent that adds to public debate and dissent that shuts it down. Thankfully, that line is clearly drawn by the legal contours of the First Amendment.

Subscribe now

But as a May Day rally on campus ended about an hour before Dean’s scheduled appearance, one student organizer told the remaining crowd that a speaker from Google’s Gemini AI was about to appear nearby and encouraged them to help him disrupt the event.

At 1:37 p.m., protesters entered the Grimes Engineering Center. Most of them chanted, “UC, UC, you can’t hide, you are funding genocide.” At least one walked onto the stage with a megaphone and asked Dean, “What do you say for your AI being used to kill Palestinians?” Dean replied that he was there to deliver a “scientific lecture.”

At 1:47 p.m., event organizers announced that the event would be shut down.

This episode is not an outlier, but part of a pattern that is accelerating. Since noting a little over a month ago that we are on pace for a record number of deplatforming attempts this year, we’ve recorded another 38. That increases this year’s tally to 108, over a span of basically four months. That’s already the fifth-highest total for a year in the entire database, which begins in 1998. And even though the success rate of these attempts declined from the staggering 93% we previously reported, it remains incredibly high at 72% (78 successful attempts).

By “deplatforming attempt,” we mean an effort to prevent a speaker, performer, artwork, or event from being heard or seen, whether through disruption, disinvitation demands, administrative cancellation, or pressure that leads a speaker to withdraw.

To be clear, protest is not the problem. Students have every right to picket, criticize, ask hard questions, and condemn speakers they find objectionable. The line is crossed when protest becomes a veto, when the goal is not to answer speech with more speech, but to ensure no one else can hear it, and incidents of that nature are becoming rampant. Indeed, the following ones all occurred over the past month:

• Administrators at Cape Fear Community College required the removal of a “No Kings” protest sign from the set of an opening-night student performance of the play The Bacchae due to political concerns.

• Noa Cochva, a former IDF combat medic and Miss Israel 2021, withdrew from speaking at the University of Oregon due to safety concerns after a large protest gathered at the reserved outdoor space. and organizers said protesters were blocking access and disrupting the event area. Cochva also had an invitation to speak at the University of Washington revoked due to safety concerns.

• Conservative political commentator and media host Stephen Crowder withdrew from a debate at the University of Pennsylvania with professor Jonathan Zimmerman, because the university refused to allow the debate to be livestreamed due to “safety concerns.”

• Students at the University of California in Los Angeles repeatedly attempted to disrupt a discussion between James Percival, the Department of Homeland Security’s general counsel, and Gregory McNeal, a Pepperdine law professor, by shouting questions, booing, holding signs with anti-ICE messages, playing sound effects, and making additional noise after the panelists were introduced.

• Pamela Evette, South Carolina’s lieutenant governor, had her invitation to deliver the commencement address at South Carolina State University revoked after a student-led petition demanding Evette’s disinvitation because of her political views garnered almost 9,000 signatures.

• Rutgers University revoked an invitation to alumnus Rami Elghandour, CEO of Arcellx, to deliver a commencement address after students objected to his social-media posts accusing Israel of war crimes and apartheid.

• Multiple attempts to disrupt the “Triggered: From Combat to Campus” tour — featuring Israeli Defense Force soldiers and/or reservists — on campuses across the country.

So, for those keeping score, that’s 108 deplatforming attempts in 17 weeks so far this year. Or basically one every day. These attempts are happening everywhere — at prestigious institutions like UPenn, UCLA, and UC Berkeley, and at lesser known schools like South Carolina State University, Western Washington University, and Cape Fear Community College.

The pace and volume of these attempts is breathtaking. If current trends hold, we can expect roughly 330 deplatforming attempts this year, which would shatter 2024’s record of 180. And, while that 330 figure probably won’t be reached, this year is still already the fifth-highest ever for these attempts and it’s only May 7.

This is alarming. Colleges do not have to invite every speaker. But once they do, they have an obligation to make sure the event can proceed. Students may protest, condemn, and challenge. They should not be allowed to veto. If campuses cannot maintain that distinction, the record being set this year will not just be a statistical milestone. It will be a warning about what higher education is teaching students to do with disagreement.

Subscribe now

We began by asking students whether they would allow six hypothetical controversial speakers on campus, three from the right and three from the left. Some students, let’s call them “Saints,” said they would definitely allow all six speakers. But this was only 4% of students. Then you’ve got “Demons,” who would definitely not allow any of the speakers. Finally, there are “Witches” who would only allow speakers whose views they support.

For each of those groups, we then did sentiment analysis of their own descriptions of moments when they censored themselves on campus. The results were fascinating.

One thing that immediately jumped out was that across all three groups, emotional reasoning is more common than rational thinking. On today’s campuses, the divide over free speech isn’t just about who gets to speak but about whether students are thinking or feeling their way to that answer, and the data isn’t good.

The selectively censorial Witches show the most emotional reasoning. If a particular viewpoint makes them feel bad, they don’t want that speaker platformed. By contrast, rational arguments are most common among Saints. As one student said, “I did not agree with the speaker, but everyone should be allowed to speak to create a diverse environment.” But Saints also lean more on emotional appeals than rational ones. In other words, even the most principled group operates less on principle than on how situations feel.

Things get bleaker still when you ask why students censor themselves.

Across all three groups, a common reason students self-censor is to avoid personal consequences, such as judgment from peers, especially the Saints. Another sentiment expressed in the self-censorship responses is a belief that certain viewpoints should not be openly expressed. Interestingly, more Witches mentioned this than avoiding personal risk. As one student said, “When they were talking about transgenders . . . I do not think that should be allowed.”

We then drilled down on more specific reasons for self-censorship. Fear of professors stands out as the top reason given by all groups, especially Saints. Saints also generally report higher sensitivity overall. This makes sense since they oppose censorship and are therefore more sensitive to factors that might cause it. Meanwhile, Witches are more worried about administrative pressure and protests compared to other groups. It’s also worth noting that grade fear is relatively low across the board, suggesting academic penalties usually matter less than social dynamics or pressure from authority figures.

What emerges from all this is a picture of a campus climate in which student self-censorship is less about formal punishment and more about perceived pressure, where even the students who are most committed to open expression are navigating that pressure with their instincts more than their principles. This has two implications. First, policy alone won’t fix the problem. Universities can adopt strong free speech statements, but if students expect social or institutional backlash, they’ll stay quiet. Second, the case for free expression itself is eroding where it matters most: how students actually think.

This is deeply disturbing because if support for free speech rests more on how situations feel, that makes it incredibly fragile. And if students’ willingness to speak up is being throttled by fear of their own professors, the impulse is being killed before it’s even fully formed.

Two of the most visible student groups in the country — Students for Justice in Palestine and Turning Point USA — have next to nothing in common except: both are frequently targeted for their protected speech. Together, they account for 24% of all group entries in FIRE’s Students Under Fire database going back to 2020.

Subscribe now

In case you don’t know these two groups, one advocates for Palestinians while the other advocates for conservatives, and they are no strangers to controversy. For example, in February 2026, Israel’s consul general in New York sent a letter to leaders at CUNY School of Law demanding the cancellation of an event organized by the school’s SJP chapter, which described Hamas’ tunnel network as “decolonial land use.” And in February 2021, student groups at the University of Nevada, Las Vegas demanded the derecognition of TPUSA over allegations that the national Turning Point Action “bused domestic terrorists to the Capitol and attacked our democracy.”

But this is not to suggest that the two groups are themselves perfect on free speech. SJP is known to disrupt campus events, such as the 2024 discussion between Bari Weiss and Frank Bruni at UNC Chapel Hill. TPUSA doesn’t really do that, though their national organization does maintain a Professor Watchlist to “expose and document college professors who discriminate against conservative students, promote anti-American values and advance leftist propaganda in the classroom.” Some entries in the watchlist include the contact information for administrators, ostensibly for visitors to file complaints.

Nevertheless, both groups are provocative. And both claim their speech is uniquely under assault. SJP says there is a “Palestine exception,” a pattern of discrimination against those advocating for Palestinians or against Israel. TPUSA says campuses are overrun with intolerant “woke leftists” who want to silence conservative voices. But the data tells a more complex narrative, one in which censorship on each side is real, yet also asymmetric.

Who started it?

Even a cursory glance at the data immediately reveals that censorship efforts against these two groups come from completely different places.

Roughly 75% of censorship efforts against TPUSA come from their fellow students, compared to only 20% against SJP. Students are by far the main censors targeting TPUSA. Administrators do too, to a lesser degree. And they are not really targeted by politicians. By contrast, SJP’s censors strike a closer balance among all three groups, but SJP experiences about seven times as many censorship efforts from politicians as TPUSA.

How are they punished?

Given the sources of these censorship efforts, it’s little surprise that these two groups would end up facing different sets of consequences for their speech. After all, administrators respond with a different range of disciplinary options than student governments.

As it turns out, 51% of all campaigns against SJP result in some sort of administrative punishment, compared to only 29% for TPUSA. But, interestingly, both groups are being punished at lower rates than all other student groups combined.

While SJP faces administrative punishment more often than TPUSA, these dealings are often less consequential. About 32% of administrative punishments against TPUSA result in the denial or loss of recognition. That’s above the average for all other student groups, which is only 10%, and far more than the 7% for SJP.

This pattern is even more pronounced when looking at student governments. Not only do student governments discipline TPUSA far more often than SJP — 26% vs. 2% — but they also deny or revoke TPUSA’s recognition practically every time (92%).

What about admins?

How institutions respond to speech controversies matters, which is why FIRE endorses institutional neutrality, because even well-intentioned statements of such a nature can have a chilling effect. And when it comes to SJP and TPUSA, the data is clear: administrators are far more likely to speak out against SJP than other groups.

When SJP is the target of a speech controversy, administrators issue public statements condemning its expression about a third of the time (30%), or double the rate of all other groups and nearly four times as often as they do in response to speech controversies involving TPUSA (8%). When it comes to actual punishments of these groups, the numbers get even worse: administrators condemn SJP’s expression up to 45% of the time, other groups 21% of the time, and TPUSA only 14% of the time.

On top of that, when TPUSA is the target, roughly 24% of the time administrators put out what we call an “Honor Roll” statement explicitly supporting the students’ expressive rights. In other words, the administrators speak up in the students’ defense. That’s twice as often as they do when other groups face down a censorship campaign, and thrice as often as they do when it’s SJP (8%).

What does it all mean?

As is so often the case when it comes to campus censorship, all sides have good reason to complain. The winds of campus censorship highlighted in last year’s Students Under Fire report have only grown stronger, with 2025 setting records for all the wrong reasons. So how do we slow them down and create a healthier climate for all students?

For starters, administrators can adopt an official position of institutional neutrality and better educate student governments, which must often act as agents of the institution, on the importance of freedom of speech. Student governments can likewise adopt neutral positions, recognizing that when they deny certain groups status or access to resources, they only deepen conflict and fuel resentment.

We live in divided times. If students are discouraged or prohibited from even talking across differences, that divide will not just persist, it will grow, and for the worst possible reason: not because they simply come to different conclusions on important issues, but because they don’t even know or understand what their opponents think.

So that’s not great.

In our National Speech Index, FIRE asks the general public a variety of questions related to free speech, including: How acceptable is it to use physical violence to stop someone giving a speech in their community? Gen Z are 9.6 times more accepting of violence against speakers than Baby Boomers1, and over 25 times more accepting of violence against speakers than the Silent Generation2. Each successive generation is more supportive of violence against speakers than the last, in most cases more than twice as supportive. About 43% of Gen Z say violence against speakers is at least rarely acceptable, and over a quarter say it’s sometimes or always acceptable3.

Subscribe now

I consider this an illegitimate (and ineffective) response to legitimate grievances. Whether one believes that there’s a lot wrong with our country or leadership; that Gen Z has gotten a raw deal economically; or even that our politics are completely broken — that break is in part because we have lost the ability to sit down and talk our problems out. Hurting or killing speakers doesn’t fix political issues. It doesn’t do anything about the problems driving the grievances people have. It just foments chaos and creates a new problem to fix.

But is this support for violence more a liberal problem? Young people tend to be more liberal, so maybe that’s the real source of the effect. To check, let’s separate the age groups by ideology:

Nope4.

Regardless of ideology, Gen Z is vastly more likely to endorse violence against speakers than Boomers, with that effect being strongest among the “very conservative.” This is primarily a generational divide, and the impact of ideology is a distant second. But the intersection of age and ideology does produce some very interesting effects. When we ask, for instance, about whether offensive speakers should face consequences, we see the opposite effect for old and young liberals:

Heck of a fall.

Compared to any other ideology, young and middle-aged liberals show more support for punishing offensive speakers. When it comes to older people, the opposite is true. In our data, young liberals are the most censorious group while old liberals are the least censorious5. One explanation is simply that we are seeing a semantic shift whereby for many measures, “liberal” has come to represent the opposite of what it once did. Think “woke” vs. old-school Progressives and Classical Liberals.

For those who care about free speech and a culture of open discourse, this should warrant serious attention. I don’t have a silver bullet here. I don’t think there is one. But when this many people start to say they are willing to resort to such a desperate solution, we have to confront the problem before the problem confronts us.

The code, data, and codebook used to generate these plots can be found here.

Subscribe now

Our data on the general population is less comprehensive than our student data. We don’t ask the same speaker-tolerance questions. But we do ask people whether they agree with the claim, “Words can be violence.” It seems likely that people who think words are violence are more likely to support censorship. The phrase is also a slogan among left-wing activists. But how widely held is this idea, and who holds it?

To figure that out, consider the scatterplot below, showing men on the left and women on the right. Each marker indicates a different political ideology, such as “conservative” or “very liberal.” The lower down each marker sits in a column, the more that group agrees with the claim, “Words can be violence.” The first thing to notice is that liberals, men and women alike, are more likely than conservatives to think words can be violence. That’s not too surprising. But take another look.

It turns out, this belief is not just a disproportionately liberal opinion — it’s also a disproportionately female one. In fact, even very conservative women are more likely to think “words can be violence” than very liberal men. Also, it seems the gender gap here is growing compared to last year, though there’s enough noise that it’s very uncertain.

Notably, the size of the gender gap varies significantly depending on what sort of question you ask, including in some ways that I wouldn’t have guessed. When we ask people whether offensive speakers should face consequences, we see a much smaller gender gap.

If I had to hazard a guess at why men and women are more in agreement here, I’d assume it’s because unlike with most of our questions, or public framings for that matter, there’s more emphasis here on how censorship can harm people. All you champions of free speech, take note. It might be that emphasizing the harm of censorship is a more effective way to get women engaged in the issue.

As one might expect, the gap reemerges when we look at aggression and ask whether Americans should have the right to curse at politicians.

Men are more okay with cursing at politicians than women, by and large. Very liberal women are also okay with it. Unsurprisingly, conservative women are not fans of profanity. But very conservative men are.

It’s striking to see how often speech issues don’t divide along simple political lines. Support for free speech isn’t a simple story about whether people agree with opinions or not. Oftentimes, it’s more influenced by social factors including gender, or driven by personality traits such as disagreeableness, or even more bizarre factors.

Subscribe now

An analysis of 72 campus deplatforming attempts — efforts to stop public expression on college campuses — from 1947 to 1999 shows that these efforts have their roots in the immediate post-World War II period.

In the 1940s, 50s, and 60s, almost every incident in the sample comes from someone to the political right of the expression they are targeting, and almost all of them were successful. Campus speaker-bans, ruled unconstitutional in 1968, were used to bar communists, communist sympathizers, and socialists from speaking on campuses across the country — including at Columbia University, UC Berkeley and UCLA, University of Michigan, and the University of North Carolina. But by the 70s, this pattern had reversed and deplatforming attempts now came overwhelmingly from the other direction.

The 1970s were the high-water mark of 20th century campus deplatforming with a total of 19 attempts. While earlier controversies centered around political views and the Cold War, deplatforming efforts in the 70s pivoted sharply to race. That shift was driven largely by repeated controversies involving William Shockley, a Nobel Prize-winning physicist turned eugenicist and, often, Roy Innis, the national director of the Congress on Racial Equality. Together, they account for nearly three-quarters (14 in all) of the deplatforming attempts that decade.

Shockley’s role is especially important. He had become a national symbol of controversy because of his views on race and intelligence. Invitations to speak on campus no longer functioned like ordinary academic events, but instead became flashpoints. When he was invited to speak at a symposium at the Brooklyn Polytechnic Institute in May 1968, a small group of faculty complained and the convocation was ultimately canceled, despite 493 scientists and engineers having already agreed to come, for fear of violence. Harold Taylor, the former president of Sarah Lawrence College, said of the scandal, “If we are never to discuss any controversial issues for fear there might be demonstrations, then the whole purpose of the university is destroyed.”

Innis matters here too because of his repeated connection to Shockley, and his willingness to debate him. In 1973, Innis was scheduled to debate Shockley on television on the topic of black intelligence, but Innis pulled out at the last moment because the student society at Princeton, which was organizing the event, would not grant access to the press or public. In 1974, Yale canceled a debate between Innis and Shockley on eugenics, an infamous decision that led to the publication of the Woodward Report, in which the university made the protection of freedom of expression an official school policy.

The 1980s brought geopolitics back. The decade’s deplatforming attempts were about the Cold War conflict, with Central America looming large. That makes the 80s distinct from both the anti-communist postwar years and the Shockley-driven scandals of the 70s. Now the targets were often appointees in the Reagan administration — figures like UN envoy Jeane J. Kirkpatrick, CIA Director William Casey, or Defense Secretary Casper Weinberger.

The 90s are the most politically mixed decade of all. There is a roughly even number of censors on the left and the right of the targeted expressions, the topics are all over the map, and the targets range from pro-choice commencement speakers to racial firebrands, conservative political figures, and culture-war provocateurs. Rather than one coherent wave, this decade looks like a transitional moment. This is the period when campus deplatforming began to take on the fragmented and multi-issue character it would have in later years.

But regardless of the decade, the attempts often succeeded. Of the 72 incidents in the sample, 53 ended by successfully deplatforming the target, including essentially all those prior to 1970. That’s a 74% success rate. But if that era was bad, the current one is even worse.

The 20th-century shows that campus deplatforming is not new. It has roots in the immediate post-World War II period, shifts ideological direction over time, and repeatedly clusters around the dominant controversies of a given era: anti-communism, race and intelligence, Central America, abortion, immigration, and the culture wars. But deplatforming attempts have become far more numerous since 2015. In the last few years, they’ve intensified even more: 171 in 2023, 179 in 2024, and 172 in 2025. We’ve already recorded 77 in just the first three months of this year. That’s more than the total number attempts in the entire dataset spanning from 1947 through 1999.

The post-World War II era saw widespread efforts to keep disfavored speakers off campus, often with remarkable success. But the modern era appears worse not simply because deplatforming still exists, but because it has become more frequent, more sustained, and in the most recent data, more successful. A 74% success rate in this 20th-century sample is troubling enough. But early 2026 data reveals a 93% success rate. The lesson is not merely that campus deplatforming has a long history, but that we are far from done with this fight. Indeed, by one crucial measure, this is our darkest hour.

“If it is protected, we will defend it … If you are a censor, we will fight you.”

– Bob Corn-Revere on maintaining principle in a time of polarization, Delaware Inns of Court keynote, March 2026.

Supreme Court revives protester’s lawsuit

The U.S. Supreme Court unanimously allowed a Mississippi man to proceed with a First Amendment challenge against a law restricting demonstrations to designated zones.

Subscribe now

The ruling clarifies that individuals can challenge speech restrictions prospectively — even if they were previously convicted under them — potentially expanding avenues for constitutional litigation.

Big Tech verdicts threaten speech

Recent verdicts in California and New Mexico, targeting social media platforms’ algorithms and design features, signal a push to treat online speech as a defective product despite longstanding First Amendment protections for editorial judgment.

Here’s to explain:

No matter how you feel about social media, the minute we start treating speech as if it were just another physical product is the minute we hand the government the power to decide what we can read, watch, and say.

Below the fold

• Two protesters in Australia were arrested — and could face up to two years in prison — on the first day of Queensland’s ban on the slogan “From the river to the sea,” as part of new rules regulating protest speech related to the Israeli-Palestinian conflict.

Terms of service

In his essay “The great chatbot panic,” Tyler Tone argues that the current wave of lawsuits blaming AI chatbots for suicides and violent behavior echoes past moral panics over media like Dungeons & Dragons, music, and video games, where courts ultimately rejected attempts to hold creators liable for harmful ideas rather than tangible defects. He warns that treating chatbot outputs as a “product” subject to liability would undermine First Amendment protections and push AI companies to sterilize speech to avoid risk.

This week in history

On March 23, 1775, Patrick Henry delivered a speech at the Second Virginia Convention urging the American colonies to take up arms against Great Britain. The speech is best known for the line, “Give me liberty or give me death.” He began by emphasizing the value of open debate and the right to express differing opinions. His remarks helped persuade the convention to adopt a resolution calling for the mobilization of troops, a pivotal step toward the American Revolution.

By the numbers

Only three months into the year, campus deplatforming is already on pace to set a disturbing new record, and if current trends hold, 2026 won’t just be a bad year for campus free speech. It’ll be the worst year on record for campus deplatformings.

In just the first three months of this year, there have been 70 such attempts. Even worse, 65 of those attempts succeeded — the highest success rate we’ve ever recorded in any year with 10 or more attempts.

Subscribe now

FIRE’s Campus Deplatforming Database tracks efforts to stop public expression on college campuses — disinviting speakers, canceling performances or film screenings, removing art, or disrupting events while they are happening. In just the first three months of this year, there have been 70 such attempts. Even worse, 65 of those attempts succeeded — the highest success rate we’ve ever recorded in any year with 10 or more attempts.

Universities’ responses to deplatforming attempts reveal whether their stated commitments to free inquiry and open debate mean anything in practice. Each successful cancellation teaches students, faculty, and organizers the same lesson: controversial expression is a liability, and institutions tend to care more about ducking trouble than standing on principle.

This chills future expression, and recent controversies show how broad that impulse has become.

This week, the University of Southern California scrapped a gubernatorial debate after excluded candidates complained about the race of those invited — they were all white. This shut down what should’ve been one of the clearest examples of a university serving as a forum for democratic exchange. Universities often claim to prepare students for civic participation. Canceling a debate involving major political figures because the controversy “created a significant distraction from the issues that matter to voters,” sends the opposite message. Namely, that even core political discourse can be treated as too difficult or too risky to host.

At New York University, the administration reportedly told student organizers that they could not invite certain music performers to a concert because the performers were affiliated with the No Music for Genocide boycott of Israel. That decision illustrates an especially troubling dynamic: universities are not only reacting to speech after the fact, but increasingly preempting it.

When administrators reserve the authority to decide which artists are too controversial to be invited in the first place, student autonomy shrinks and institutional gatekeeping expands. The result is a campus culture in which permission for expression becomes contingent on administrative comfort.

Other incidents this year further demonstrate this pattern.

• The University of North Texas removed an art exhibit after an anonymous tip alleged the show included artwork denouncing U.S. Immigration and Customs Enforcement. UNT also just revoked approval for a drag show after the university system lifted its systemwide pause on drag performances last August.

• The Catholic University of America rejected requests from the campus chapter of Students Supporting Israel to host Randy Fine and Dany Turza at two separate events because the discussions would not feature a “balanced presentation” of views about the Israeli-Palestinian conflict.

• And, the University of Southern Maine terminated its agreement to host a one-day conference, “The Consequence of Palestine,” because one of the scheduled speakers (Francesca Albanese, the U.N. Human Rights Council’s special rapporteur on human rights in the Palestinian territories occupied since 1967) has been sanctioned by the U.S. government and appears on the U.S. Department of the Treasury’s office of foreign assets control’s specially designated nationals and blocked persons list. Administrators said that hosting the event with a “sanctioned scholar” would violate federal law. Albanese was scheduled to appear remotely and organizers and free-speech attorneys argue that including Albanese as a remote, uncompensated speaker in a virtual event was not prohibited and required no authorization.

These cases differ in subject matter, politics, and format. But that is exactly the point. Deplatforming is not confined to one ideology or one type of event. It can target political debate, artistic expression, student entertainment, invited lectures, or academic programming. The common thread is a willingness to suppress expression rather than answer it, criticize it, or contextualize it.

Defenders of these cancellations often frame them as prudent, values-based, or necessary for campus harmony. Sometimes they invoke safety, reputational concerns, institutional mission, or community standards. But too often those rationales are vague, inconsistently applied, or selectively enforced. Universities rarely cancel events because they are truly incapable of supporting peaceful disagreement. More often, they cancel because controversy itself is treated as unacceptable.

That is a profound mistake. The university is one of the few institutions explicitly designed to accommodate disagreement. Students should encounter ideas, performances, and speakers they dislike, distrust, or find offensive. They should also learn how to respond: through protest, questioning, criticism, counterprogramming, and debate. Those are the habits of a free society. Deplatforming replaces those habits with passivity and permission-seeking.

That is what makes the 93% success rate so alarming. Failed deplatforming attempts can still cause harm, but they at least show that institutional safeguards are holding. A successful attempt signals that those safeguards are eroding. If nearly all deplatforming efforts are now succeeding, then the problem is not simply that controversial events are being challenged. The problem is that universities appear increasingly willing to fold under pressure.

These actions can have long-term consequences. Student organizers may stop proposing ambitious or provocative events because they fear resistance. Faculty may avoid controversial programming because administrative support feels uncertain. Artists and speakers may decline invitations if they believe universities cannot or will not stand by them. Over time, the chilling effect becomes more powerful than any one deplatforming. The lesson spreads quietly: do less, invite less, say less.

Seventy attempts in three months is troubling on its own. Sixty-five successful cancellations is worse. A 93% success rate suggests a campus speech environment in which those seeking to silence expression are winning.

This should concern anyone who believes higher education exists to test ideas rather than shelter institutions from them. If universities want to remain credible as places of inquiry, they must do more than praise free expression in mission statements. They must defend it even when it is inconvenient, contested, and costly.

A new survey of University of Wisconsin-Madison faculty, released this month by the school’s Tommy G. Thompson Center on Public Leadership, offers a clear look at how ideological imbalance shapes the campus climate at a flagship public university. Read alongside FIRE’s 2024 faculty survey, it tells a familiar story — just with sharper numbers.

Subscribe now

Among tenured and tenure-track faculty at UW-Madison, 70% identify as liberal, 21% as moderate, and just 9% as conservative — a ratio of almost 8-to-1. That’s not an outlier, but is part of a broader trend. Nationally, another recent analysis shows 74% of faculty identified as liberal and only 11% as conservative. By these numbers, compared to the general public, liberals are overrepresented in academia by 186% while conservatives are underrepresented by 70%.

This gap also may be widening as junior faculty at UW Madison skew more liberal than senior faculty, suggesting that as older professors retire, ideological homogeneity will deepen.

Ideology in hiring

The Thompson Center survey also tested for possible downstream effects of this ideological imbalance. Faculty respondents were randomly assigned to evaluate hypothetical job candidates expressing liberal or conservative views. The result: candidates with conservative views were consistently less likely to be supported for hiring, with gaps ranging from 11 to 38 percentage points depending on the topic. The topic of immigration produced the largest disparity.

Because the statements were randomly assigned, these results don’t reflect response bias or a general aversion to candidates expressing political opinions. Rather, the results reflect bias against academics with conservative viewpoints, plain and simple.

These findings mirror a growing body of independent studies and national surveys. FIRE’s 2024 survey also shows the same pattern. Among UW-Madison faculty in FIRE’s survey, only 44% said knowing an applicant was conservative would never count against them, compared to 69% who said the same about a liberal applicant. And while 76% said a liberal individual would be a good departmental “fit,” only 17% said the same about a conservative. These numbers were similar to faculty overall in FIRE’s survey. Importantly, in academic hiring, “poor fit” often translates into “not hired.”

Chilling effects

The imbalance doesn’t just shape hiring either. It also shapes who feels comfortable speaking.

At UW-Madison, only 23% of conservative faculty say they’re highly comfortable discussing controversial topics with colleagues, compared to 55% of liberals. The classroom also shows a similar divide. Similarly, FIRE’s 2024 national survey shows that nearly half of conservative faculty report self-censoring due to fear of backlash, compared to one in five liberals.

And when conservative faculty do speak up, they’re more likely to pay for it. Among UW-Madison faculty who expressed views on controversial topics in class, 20% of conservatives reported institutional consequences, compared to just 1% of liberals. On social media, the gap is 30% to 8%.

This is despite the fact that liberal faculty are more likely to say they express controversial views in the first place. In other words, one group speaks more freely and faces fewer consequences while the other speaks less and gets punished more when it does.

The bigger picture

None of this is happening in a vacuum, and this pattern extends beyond UW-Madison. FIRE’s 2024 survey of faculty finds that 55% of conservative faculty reported at least occasionally hiding their political beliefs from colleagues to protect their jobs, compared to 17% of liberals. And among the broader faculty population, the share of faculty self-censoring in their writing is nearly four times the rate recorded among social scientists at the end of the McCarthy era.

Taken together, the data point to a system where ideological imbalance doesn’t just describe who’s on campus but who gets hired, who speaks up, and who gets punished when they do.

The Israeli-Palestinian conflict.

Subscribe now

At these hyper-liberal schools, the topic is more than twice as difficult for liberals to discuss as any other issue.

Remember, the hyper-liberal schools discussed here are only a small group of schools, and don’t include any Ivies or other large schools, such as Columbia University. At those other schools, there are of course many examples of administrations going after students for protected pro-Palestinian speech.

While this phenomenon got worse after Oct. 7, the issue has been uniquely thorny for these students as far back as our data goes. In addition, disagreements among liberals or fear of social stigma may make the issue harder to discuss in general, but they do not explain why it is especially difficult at very liberal schools.

Students who don’t trust their administrations are significantly more likely to find it difficult to discuss the conflict. Liberal students at hyper-liberal schools are less trusting of their administration and more worried about discussing the Israeli-Palestinian conflict, even though they’re seemingly less likely to face punishment for pro-Palestinian protests there.

One generally assumes people are more comfortable sharing their views when surrounded by others who think the same way. So then why are liberal students at very liberal schools scared to talk about the Israeli-Palestinian conflict?

In FIRE’s College Free Speech Rankings survey, we ask students whether they’re comfortable discussing a variety of political issues.1 Which issues do they find difficult to discuss? Let’s begin by looking at students at moderate schools.2 In the scatterplot below, issues that conservatives find hard to discuss are closer to the top (measured on the vertical axis) while issues that liberals find hard to discuss are more to the right (measured on the horizontal axis):

This is about what I’d expect. The points fall along a diagonal line with a slope of roughly one, meaning conservatives and liberals roughly agree on which topics are difficult to discuss. Abortion and transgender rights are the thorniest. For women, the presidential election is also very difficult. China, crime, and SCOTUS are the easiest.

The offset from the center diagonal is interesting. At schools with close to perfectly moderate student bodies, liberal students still report greater difficulty discussing most topics. Given that most schools don’t exactly have administrations or environments that are disproportionately hostile to liberal students, it seems most plausible that this reflects a psychological difference between liberals and conservatives, perhaps greater neuroticism3 or anxiousness, a greater tendency to view one’s environment as oppressive, or something else. The fact that this gap is larger in women than men supports the notion that we’re seeing a psychological effect. That said, it’s not a massive effect.

But what about students at extremely liberal schools?

That’s a result.

For the most part, liberals at liberal schools are comfortable talking about every issue. We might’ve expected a diagonal with a larger offset or steeper slope, but instead we see a vertical blob. It’s not just that liberals have an easier time talking about most issues: the difficulty for the two sides becomes decoupled. Conservatives still have a difficult time talking about the normally difficult issues — abortion, transgender rights, and the 2024 Presidential Election — and an easy time with the normally easy ones — the Supreme Court, China, climate change — but liberals have an easy time with every issue. But these dynamics disappear for liberals at very liberal schools — echo chambers can seemingly make people more comfortable talking. To the small extent that some issues are tougher for liberals, it’s things like economic inequality, which aren’t normally thorny.

Until we get to the Israeli-Palestinian conflict. At moderate schools, it’s a normal political issue. For liberal students at liberal schools, it’s the difficult issue, with more than twice as many students finding it difficult to talk about than any other issue, even though they are surrounded by more like-minded peers than any other students.

So what’s going on here? Why is this issue so thorny? To understand why this happens, we look at four possibilities: the impact of Oct. 7, ideological divisions among liberals, fear of social stigma, and fear of administrative punishment.

October 7

First, the obvious: Presumably, this is a response to Oct. 7, the subsequent war, and the encampments. We can check that by making the scatterplot with (early) 2023 data:

Erm, what?

Oct. 7 and its aftermath coincided with the Israeli-Palestinian conflict becoming more difficult to discuss, but it’s been a uniquely difficult issue for liberal students at liberal schools for longer than that — in fact, since at least 2020, the first year we have data on this.

What’s so special about this particular issue?

Divisions among liberals

Well, there’s an unusual amount of disagreement amongst liberals of various shades. And the scatterplots binned together students who were anywhere from slightly to very liberal. At moderate schools, liberals tend to be more moderate. But at hyper-liberal schools, you tend to have more left-wing extremists. Maybe very pro-Palestinian students are uncomfortable discussing the conflict more often, and the difference between liberal students at various schools is just a question of how far to the left they are. To check this, let’s bin individuals by ideology (in addition to school leaning):

That’s part of the explanation. Very liberal students do have a tougher time discussing the conflict than moderately liberal students. But students at more liberal schools also have a harder time discussing it, regardless of ideology. A very liberal student at a far-left school is nearly 20 percentage points more likely to have difficulty than a very liberal student at a moderate school. For moderately liberal students, the gap is even larger. More liberal schools genuinely seem to be tougher places to discuss this issue.

But while I do suspect that liberal infighting contributes to the issue being difficult to discuss in general, I don’t think it’s the primary reason that it’s especially difficult at very liberal schools. At moderate schools, there are roughly even numbers of very and slightly liberal students. At hyper-liberal schools, very liberal students outnumber slightly liberal ones by roughly 6 to 1. While our ideology measurements are an imperfect proxy for opinion on this issue, I’d bet against numbers like that causing more liberal infighting.

Fear of social stigma

Okay, there’s research suggesting that college students often fear social sanction for speaking their minds. Even if hyper-liberal schools, which are all small liberal arts schools and half of which are all-women, have less disagreement of opinion, maybe what disagreement exists is met with more social stigma there. I don’t have any data that proves this isn’t a factor, and I suspect this has some effect, particularly on creating a baseline difficulty across a range of schools and contentious political issues.

But it’s probably not the whole story, or anywhere close to it. It can’t explain why conservative women report having an easier time discussing abortion or the 2024 presidential election at hyper-liberal schools than they do at moderate schools. It can’t explain why the presumably-apathetic students who haven’t thought much about their ideology have an easier time discussing the Israeli-Palestinian conflict. And it can’t explain a relationship between administrative distrust and difficulty, which we’ll discuss momentarily. There seems to be a larger factor here.

Fear of administrative punishment

Another obvious possibility is that fear of administrative ire is silencing students. After all, there were more than a couple heavy-handed administrative crackdowns in response to the encampments. We have data on a question similar to this: In our survey, we ask how clear it is that an administration protects free speech on campus. Are students who distrust their administrations more likely to be uncomfortable discussing such a charged issue?

That’s very information-rich, but also busy. Here’s a simpler plot showing the same core trend:

Yes, emphatically. At the most liberal schools, students who think it’s not at all clear that their administration protects free speech on campus are 30 percentage points more likely to find the conflict difficult to discuss than those who think it’s extremely clear. (Looking just at liberal students at hyper-liberal schools, we see a similar result.) On the plot above, notice how the higher rows (where students trust the administration less) are darker (more difficulty): there’s a very strong relationship between distrust of one’s administration and difficulty discussing the conflict, especially at liberal schools. While we can’t prove that it’s causal, that result suggests that fear of the administration is a significant force in deterring discussion on this issue.

So what do students at hyper-liberal schools say about the topic? We ask students to share a moment where they felt they could not express their opinions on campus. When they feel they’re unable to talk about the Israeli-Palestinian conflict, do they talk about the administration?

“Whenever we discuss the ongoing conflict in Gaza, it’s under the assumption that if we go against administration, we are kicked out of the college.” (Barnard College)

“Administration just expelled two students over speaking out about the Israel-Palestine conflict.” (Seemingly for disrupting an Israeli history class, which, regardless of whether it justifies expulsion, isn’t protected speech.) (Barnard College)

“Not being able to say from the river to the sea without risk of being honor coded” (Wellesley College)

“The college has been very pro-Israel, this has led to a lot of censorship away from pro-Palestinian conversations” (Mount Holyoke College)

“The admin here is pretty pro-Israel so sometimes I’m worried about expressing my pro-Palestine opinions” (Smith College)

These should be treated as an illustration, not a representative sample or strong evidence: there are students who give other reasons for silence on the conflict, including fear of social repercussions, or no reason at all. That said, sentiments like the ones above are quite common.

So is this fear actually justified? Are administrations actually cracking down on liberal students or pro-Palestinian protestors especially aggressively at hyper-liberal schools? We don’t have data on that exact question in its entirety, but we do have data on the responses to the encampments, which seems like a reasonable proxy. Were they shut down more often at very liberal schools?

No. Encampments were more common at more liberal schools, but, when they happened, they were less likely to result in arrests or other police action. (This is the size of the red/yellow bar as a fraction of the whole bar.) Administrations at more liberal schools were less likely to punish students for encampments.

But what about punishment for speech generally? We also ask students whether they’ve been punished or threatened with punishment for expression on campus:

Generally, the more liberal the school, the less likely the students were to be disciplined or threatened for expression. Yes, college administrators are famously unkind towards student speech, but the administrations at hyper-liberal schools seem more lenient towards pro-Palestinian protestors than their counterparts at other schools. Liberal students at hyper-liberal schools are less trusting of their administration and more worried about discussing the Israeli-Palestinian conflict, even though they’re seemingly less likely to face punishment for pro-Palestinian protests there. While I certainly agree with their general distrust of college administrations, in this specific regard, it looks like this might be a case of many pro-Palestinian students at hyper-liberal schools being scared of their own shadows or having deeply distorted views of what free speech means. We should of course be appropriately cautious when reading into the data something very related to what it says, but that seems the most reasonable conclusion from the available evidence.

Conclusion

I see two useful warnings here. The first is the danger of scaring yourself into silence with an invented or exaggerated fear. Yes, the administrative onslaught against student expression is often real, fierce, and brutal. Yes, even when not officially punished, speaking your mind can have social repercussions. But, if you’re lucky enough to be in neither of those situations (or neither particularly severely), it’s easy to overestimate the danger you actually face. If you’re an ardently pro-Palestinian student at an overwhelmingly liberal school, you probably shouldn’t be more afraid to talk about the conflict than someone who holds the same opinion at a school with a slightly liberal student body and an administration possibly deeply hostile to your cause.

The second is more obvious: the general state of discourse on campuses is bad. At moderate schools, over half of students find it difficult to discuss many important issues. At very liberal schools, those with minority viewpoints face a similar situation. Part of the problem here is censorship, but not all of it. If we want to make progress on issues vital to our nation, we need to be willing to discuss them honestly even when doing so is difficult, uncomfortable, or worse.

The code, data, and codebook used to generate these plots can be found here.

Since 2018, More in Common, a nonprofit focused on reducing political polarization, has tried to explain American politics less by demographics and more by the core values that shape how people see the country. That approach matters because it shifts the conversation from who people are to why they think as they do, and it often reveals divisions hiding in plain sight.

Subscribe now

Last month, MiC released Beyond MAGA: A Profile of the Trump Coalition, a study based on more than 18,000 interviews with Americans who voted for Donald Trump in 2024. The headline finding is simple but consequential: Trump voters are not one bloc. They fall into four distinct types, and their views diverge sharply on two issues now at the center of American cultural conflict: freedom of speech and whether presidents should use federal funding to pressure colleges and universities over DEI and CRT-related content.

Here are the four types of Trump voters:

• MAGA Hardliners are mostly evangelical Trump supporters who believe a “deep state” runs politics and that “the Left” hates America. About 25% hold college degrees.

• Anti-woke Conservatives say American identity is fading and “woke” ideology has ruined American education, news, and entertainment. They are the least religious group (31% atheist, agnostic, or unaffiliated) and about 40% hold college degrees.

• Mainline Republicans are the most optimistic about the American dream and least likely to say the country is in decline. About 25% have some college experience but no degree, and 38% hold only a high school diploma.

• Reluctant Right voted for Trump because he seemed “less bad” than Kamala Harris. They are the least likely to identify as Republican, the least hopeful about the next four years, and the most likely to say they vote across party lines.

Those distinctions aren’t just academic. They predict meaningful differences in how Trump voters weigh rights, institutions, and presidential power.

Consider freedom of speech. When asked whether it is an “unconditional basic right that the government should never be able to take away,” support among MAGA Hardliners and Anti-woke Conservatives is notably mixed: roughly half in each group say speech is unconditional, while the other half view it as conditional when expression poses a “threat of violence.” Mainline Republicans look similar, though slightly more supportive of an unconditional right (56%). The outlier is the Reluctant Right: roughly three-quarters say freedom of speech is unconditional and should not be taken away by the government.

If you’ve grown accustomed to assuming that the political right is uniformly absolutist on speech, this complicates that picture. The strongest free-speech instincts in the Trump coalition may be concentrated among voters who are also least committed to the Republican identity and least enthusiastic about the coming years.

The split is even clearer on federal funding for higher education and DEI/CRT-related issues. Nearly all MAGA Hardliners and Anti-woke Conservatives (96% in each group) agree that “younger Americans are worse off because of ideas pushed by the Left.” And at least nine in 10 (93% and 90%) say President Trump should punish public universities that promote ideas like DEI or CRT by withholding federal funding.

By contrast, Mainline Republicans are closer to a 50–50 coalition on these questions: 54% say younger Americans are worse off because of ideas pushed by the Left, and 53% support withholding funds to punish universities. The Reluctant Right matches Mainline Republicans on the first claim (55%), but breaks sharply on presidential intervention: 66% say it’s not the president’s role to determine what is taught in universities and that federal funding should continue.

Taken together, these findings point to a political reality that institutions, especially higher education, often miss: the internal debate among Trump voters is not only about ideology, but about authority. Even many voters who feel alienated by progressive (woke) cultural influence are uneasy about using the presidency to police what universities teach.

That nuance matters for anyone trying to lower the temperature. If universities respond as if all Trump voters want punitive crackdowns, they risk hardening opposition and confirming the most cynical narratives. And if policymakers assume a single “MAGA mandate” for sweeping funding threats, they may overread the coalition’s unity and underestimate the backlash from within it.

MiC’s report doesn’t argue that campus conflicts are imaginary; it shows they are politically complicated. The Trump coalition contains voters who want aggressive retaliation, voters who want cultural change without federal coercion, and voters who still hold a surprisingly robust commitment to unconditional free speech. Treating them as one bloc guarantees miscommunication and miscalculation. If we want fewer zero-sum showdowns, the first step is accuracy: recognizing that the next fight over speech and universities won’t only run left versus right. It will run through the right as well, and the outcome will depend on which faction leaders choose to represent.

Americans’ confidence in U.S. colleges and universities remains near historic lows. Although some have suggested that public opinion about higher education may be stabilizing or rising, our latest survey finds little evidence of meaningful recovery.

Subscribe now

In our latest poll, conducted in partnership with NORC at the University of Chicago, less than a third of Americans say they have “a great deal” or “quite a lot” of confidence in U.S. colleges and universities. That’s statistically indistinguishable from 2024 and 2025 levels. At the same time, almost the same proportion report having “very little” or no confidence at all.

These results stand in stark contrast to public opinion a decade ago. In 2015, 57% of Americans told Gallup they had a great deal, or quite a lot, of confidence in higher education, while just 10% expressed very little confidence. By 2018, these figures were starting to slip, with high confidence decreasing and low confidence increasing. Since then, the decline has accelerated. Compared to 2015, the share of Americans expressing high confidence has fallen by 26 percentage points, while the share expressing very little confidence has nearly tripled.

Confidence in higher education also differs sharply according to one’s politics, with long-term drops driven primarily by Republicans and, to a lesser extent, Independents. In 2015, over half of Republicans expressed high confidence in higher education. Today, only 15% say the same. That’s a decline of 41 percentage points. Independents have also lost confidence. While nearly half expressed high confidence in 2015, only 16% did in 2025, before rebounding.

Democrats remain more confident than other groups, but even Democratic confidence has eroded. In 2015, more than two-thirds expressed high confidence. In 2026, only 45% do, the lowest level recorded for Democrats across these data. The overall result is not merely polarization, though that does explain part of what’s happening here. But also, what we’re seeing is a collapse of confidence in higher education across all partisan political groups.

On their own, these confidence trends reflect an untenable dynamic for higher education leaders. But viewed alongside other indicators, they suggest deeper structural problems. Some are starting to voice concerns about an impending enrollment cliff facing higher education, but data compiled by researchers such as Kevin Wallsten suggest that demographic change is only part of the story. Just as important are signs that many high schoolers — most notably males and Republicans — are less interested in graduating from college, that ideological diversity among professors and administrators is low and still declining, and that highly visible speech controversies continue to pile up.

If colleges and universities want to remain socially relevant and financially stable, they’re going to have to confront the reality that large segments of the public increasingly see them as exclusionary rather than open.

This has consequences beyond institutional reputation. For young adults, college degrees remain one of the most reliable pathways to future success, even as growing numbers question whether it is worth pursuing. Allowing confidence to stagnate or erode further risks not only institutional stability, but also the broad-based access to opportunity that higher education has long promised. We need our colleges and universities. But confidence cannot be restored unless colleges and universities are willing to earn it.

It’s no secret that most colleges lean significantly liberal, but it’s easy to forget just how liberal. Describing many colleges as “liberal” doesn’t really do it justice when over 80% of colleges are more liberal than California. It’s therefore become a common refrain that such hyper-liberal colleges make students intolerant of conservatives and unwilling to engage them in productive conversation. But is that true?

We already know students on the extremes of both ends of the political spectrum tend to be less tolerant than moderates, suggesting that politically extreme schools would be less tolerant than moderate ones. We can check this by plotting minimum tolerance, or the student body’s overall willingness to let disfavored speakers speak, against the school’s ideology:

Being a conservative at Smith College must be fun.

As expected, more extreme schools (measured by the average student ideology) tend to be less tolerant. This suggests hyper-liberal schools (since hyper-conservative schools don’t really exist) are rough environments for dissenters. That’s hardly shocking. But the more interesting question is, do hyper-liberal schools make students less tolerant? Are liberal students at liberal schools less tolerant than liberals at moderate schools? The first question is tough to confidently answer from our data, but the answer to the second question is basically no. We can see this when separating students by ideology, which lets us ignore the of-course-it-happens-at-liberal-schools effect seen above. Let’s look at tolerance for right- and left-wing speakers separately:

For the most part, within each ideological camp, right-wing tolerance remains similar regardless of the leaning of the school. But left-wing tolerance is a different story:

Wait, what?

Both liberal and conservative students at liberal schools are more tolerant of left-wing speakers (making those schools rank as more tolerant overall). The effect is so massive that conservative students at the most liberal schools are even more tolerant of left-wing speakers than liberal students at the most conservative schools. Why on earth is this happening?

One explanation might be the contact hypothesis, famously articulated by the psychologist Gordon Allport, which is the idea that bringing people into contact reduces prejudice. According to this idea, conservatives, surrounded by liberals, would become more tolerant of liberals. But the effect doesn’t seem to work in reverse because tolerance for conservative speakers isn’t higher at the few conservative schools that exist. It’s hard to be confident without data from deep-red schools, but liberal schools do seem to have a unique effect.

What else could cause this? Well, more liberal colleges are more likely to have liberal arts majors, which tend to have higher (especially left-wing) tolerance. But that’s only part of what’s happening here. And beyond that, I don’t know yet.

This doesn’t mean that hyper-liberal colleges are tolerant or pleasant places for dissenters whose opinions cut against the grain, or that they’re bastions of good political discourse. Indeed, there’s evidence to the contrary. But something seems to be going bizarrely right there.

The code, data, and codebook used to generate these plots are available here.

Subscribe now

First, compared to last year, Americans are about five percentage points more likely to say the country is headed in the wrong direction when it comes to freedom of expression. Second, the small increase in trust the public had in President Trump to protect their First Amendment rights when he took office in January 2025 has now largely evaporated.

All together, the percentage of people who say things are headed in the wrong direction is consistently high. Recently, it’s even higher. In January 2024, 69% said things were headed in the wrong direction. The number dipped to 63% in April, then fell to its lowest level right after Trump took office in January 2025, reaching 59%. But that reprieve didn’t last.

By April, around the time Mahmoud Khalil and Rumeysa Ozturk were arrested and the Trump administration launched its campaign to “reform” higher education, two-third of Americans felt free expression was headed in the wrong direction. That spiked to about three-quarters in October, during the brief removal of Jimmy Kimmel, where it remained as of last month.

Ideology clearly impacts how people answer these questions, but it doesn’t explain everything. Conservatives appear especially sensitive to changes in party control. Throughout 2024, three-quarters or more said free speech was headed in the wrong direction. This dropped to one-third in January 2025 after Trump took office, but has been rising over the past year. Very conservative Americans show an even sharper relief pattern from roughly 80% in 2024 down to roughly a quarter in early 2025.

This is what you’d expect if “wrong direction” is partly a referendum on who’s running the executive branch. But the story can’t be reduced to just that. The liberal side’s volatility over the past few years indicates something beyond simple party-in-power logic and suggests that recent events have driven their perception of the country’s expressive environment. So while the color of the president’s necktie sets a baseline, it doesn’t lock perceptions in place. Events and institutions beyond the presidency also plausibly shape how safe people feel speaking freely.

The second big trend is even cleaner: confidence in Trump to defend First Amendment rights has declined compared to January 2025, regardless of one’s political beliefs.

In January 2025, shortly after Trump returned to office, 29% of Americans said they had “no confidence at all” that he would defend their expressive rights. One year later, 47% say the same, an increase of almost 20 percentage points in just a year. In April, a whopping 84% of very liberal Americans said they had no confidence in Trump to defend their expressive rights. Now 94% of them say this. The percentage of very conservative respondents who agree rose from 4% in July 2025 to 11% now, and among conservatives, from 3% in July to 13% now.

If we put the two trendlines together, the message is hard to miss. Americans increasingly feel less free to speak, and many of them increasingly doubt that the president will protect the right that makes that possible. The brief post-inauguration thaw in January 2025 now looks less like a durable reset than a momentary hope that the temperature would drop. Instead, the year that followed supplied vivid reminders that expressive freedom isn’t just debated online or litigated in courtrooms — it can be tested by arrests, pressure campaigns, and street conflict.

The most important takeaway from the NSI data may be that free speech is becoming a significant political issue again. When Americans see government power aimed at protected expression, they update their beliefs quickly and not just along partisan lines. If this trajectory continues, the country risks settling into a new normal where large majorities assume expressive freedom is deteriorating and where trust in leaders to defend it becomes the exception rather than the rule. Reversing that won’t come from rhetoric alone. It will require visible restraint from officials, institutional guardrails that hold in moments of anger, and a public commitment to the principle that the First Amendment protects unpopular speakers precisely when it’s hardest to do so.

In our surveys assessing university student tolerance for hypothetical controversial speakers, most students have not wanted to allow most speakers on campus. You can almost hear someone saying, “These kids today are too soft. Send ‘em to boot camp!” The funny thing is, that might actually help.

Subscribe now

One would hope that military duty would make people more principled defenders of free speech. But it also seems plausible that strict military hierarchies could make people more willing to use authority against dissent. So which is actually true? As usual, that’s tough to confidently answer with our data, but a related question is much easier: Are military students more tolerant?

Okay, that’s hardly shocking. Military students are, on the whole, more right-biased than others, and a bit more tolerant. But what if we break down the data according to ideology?

This looks very similar to the free speech map for gender and tolerance. Given that most students aren’t military (note the plot title), this is expected. So where do military students place?

Not a silver bullet, but that’s quite a result.

Military students are only a bit more tolerant overall, but they are significantly more fair to their political opponents. Male military students have 21% less tolerance bias than their civilian counterparts. Female military students have 29% less tolerance bias. (This is controlling for ideology. If you don’t control for ideology, then military men are 16% less biased than civilians while military women are 32% less biased.)

Two things to note here. First, there are 676 men and 526 women in this dataset. That’s a small enough sample that you shouldn’t put too much trust on any particular point (particularly when split by ideology). But it is high enough that the general pattern is very probably an accurate description of reality.

And, as usual, this doesn’t prove the direction of causation. Maybe being in the military decreases bias and increases tolerance. Or maybe students who are less biased and more tolerant are disproportionately likely to join up for the military. Or both.

It’s a common refrain that everyone is biased, and people are doomed by their own perspectives to never be fair judges towards ideas they oppose. But the data here reminds us that some people are significantly fairer than others. And if military experience indeed causes a decrease in bias, perhaps a principled impartiality really is a teachable skill.

The code, data, and codebook used to generate these plots are available here.

The assassination of Charlie Kirk at Utah Valley University last year sparked a wave of anger directed at scholars for their public comments following Kirk’s death. This reaction was faster, more targeted, and more severe than the response following the murder of George Floyd in 2020.

Subscribe now

Relatively few professors faced significant consequences for comments about Floyd himself. On the other hand, Floyd’s death appears to have spawned a broader, more sustained sensitivity to race-related discourse. In the months and years that followed, scholars who addressed racial issues — even when not speaking directly about Floyd — faced growing public backlash, administrative investigations, and in some cases, dismissal, as seen in the case of Charles Negy, who was fired, and Gordon Klein, who was placed on administrative leave.

It’s important to note that the data on race-related sanctions in 2020 covers the entire year, whereas the data on Kirk only spans about four months. In this compressed timeline there were 92 scholar sanction attempts, 44 sanctions, and 21 terminations. Notably, all 21 scholars who were terminated for their comments about Kirk had expressed views critical of him and were targeted by right-leaning activists, media outlets, politicians, administrators, and students. For example:

• Tamar Shirinian, a professor of anthropology at the University of Tennessee, Knoxville, posted on Facebook that “the world is better off without [Kirk] in it,” prompting outcry from lawmakers and media figures, and resulting in her termination.

• Darren Michael, a theatre professor at Austin Peay State University, re-shared a 2023 Newsweek headline quoting Kirk as saying that gun deaths were “unfortunately” necessary to preserve the Second Amendment, after which critics claimed the post signaled tacit approval of Kirk’s killing, Tennessee Senator Marsha Blackburn labeled it “reprehensible,” and the university fired Michael, condemning his post as “insensitive, disrespectful, and interpreted by many as propagating justification for unlawful death.”

• Lisa Greenlee, an instructor at Guilford Technical Community College, appeared to joke about Kirk’s death in a video that circulated online, prompting prominent political figures, including U.S. Representative Mark Walker and Representative Virginia Foxx, to call for her removal, after which Guilford Tech fired her.

Scholar sanctions hit a record high in 2025, with the aftermath of Kirk’s assassination being one of the primary contributors.

The long-term impact of Kirk’s assassination on campus speech — particularly for faculty — remains uncertain. A FIRE survey of 2,028 undergraduates, conducted Oct. 3–31, 2025, revealed a striking finding: while 81% of Utah Valley University (UVU) students opposed firing a hypothetical professor who said, “We are going to make America great again,” 57% of UVU opposed firing one who said, “These fascist Bible thumpers want to drag us back to the Dark Ages.”

When the survey results were broken down by political ideology, clear partisan divides emerged. Most liberal students opposed firing hypothetical professors who said, “It’s O.K. to punch a Nazi” or “These fascist Bible-thumpers want to drag us to the Dark Ages.” Similarly, most conservative students opposed terminating professors who said things like, “We are going to make America great again” or “Our colleges and universities are progressive indoctrination centers.” These differences suggest that what people consider acceptable speech often aligns with their politics rather than any consistent commitment to academic freedom.

Free speech after Floyd and Kirk: A call for consistent principles

The shift from left-leaning scrutiny over George Floyd to right-leaning backlash over Charlie Kirk reveals more than just ideological whiplash. It underscores the volatile, politicized landscape of academic speech in America. As faculty face growing pressure from all sides, the principle of intellectual freedom hangs in the balance.

If we want universities to remain spaces where difficult conversations can unfold, we must protect the right to speak — even, or indeed especially, when that speech is offensive. True academic freedom cannot depend on which side holds the cultural or political majority. In this polarized moment, recommitting to dialogue over censorial punishment may be our only path forward.

For the past two years, FIRE’s National Speech Index has tracked Americans’ views on free expression — whether the country is moving in the right or wrong direction, how secure people believe free speech to be, and how often they self-censor.

Subscribe now

That period has hardly been quiet. Americans have lived through a presidential transition, major political upheaval, and widespread campus protests. Our data offers a window into how public attitudes toward free speech have responded to those events.

The headline finding is consistent — and sobering. Over the past two years, Americans have generally said the country is headed in the wrong direction when it comes to people’s ability to freely express their views. But those judgments are highly sensitive to political loyalties.

Prior to the 2024 presidential election, most Democrats thought things were headed in the right direction, compared to only a minority of Republicans and Independents. After Donald Trump’s election, those views flipped. Clear majorities of Republicans now say the country is moving in the right direction, while Democrats have grown markedly pessimistic.

Perceptions of the security of free speech follow the same partisan arc. In October 2025, more Americans said the right to freedom of speech is “not at all secure” (22%) than said it is “very” or “completely” secure (15%). That represents a stark reversal from January 2025, when 23% viewed free speech as very secure and fewer than one in 10 believed it was not secure at all.

Among Democrats, nearly one in three now say freedom of speech is “not at all secure,” a sharp increase from roughly one in 12 who held that view prior to Trump’s election and inauguration. Independents show a similar, though less pronounced, shift. Republicans, by contrast, are less likely to view free speech as insecure than they did during the final months of Joe Biden’s presidency.

Yet while Americans’ perceptions of free speech rise and fall with political fortunes, other core indicators tell a different story. Self-censorship has remained remarkably stable. Across the entire two-year period, roughly one in four Americans report self-censoring at least a couple times per week.

That stability holds even across party lines. Independents consistently report higher levels of self-censorship than Democrats or Republicans, but rates for all three groups have changed little since January 2024. This suggests that self-censorship may reflect a relatively stable personal disposition rather than a reaction to political events.

Taken together, these findings paint a revealing portrait of Americans’ relationship with free expression. Beliefs about free speech appear to shift quickly with changes in political power. Actual behavior — whether people feel free enough to speak their minds — changes far less. In other words, political tides turn fast. But as for self-censorship, the chill lingers.