The Paradox of Political Freedom: When the Most “Liberal” Become the Least Tolerant

There is a strange paradox spreading through Western politics.

Almost everyone claims to defend freedom.

The progressive defends freedom from discrimination.
The conservative defends freedom from government intrusion.
The libertarian defends individual autonomy.
The classical liberal defends freedom of expression and pluralism.
The populist claims to defend the freedom of ordinary citizens against elites.

The vocabulary changes.

The word freedom does not.

And yet our political societies increasingly seem unable to tolerate the most elementary consequence of freedom:

someone else using it to reach a conclusion we despise.

That may be one of the defining political contradictions of our time.

Because perhaps the most revealing measure of political extremism is no longer simply how far left or right someone sits on an ideological spectrum.

Perhaps another question matters just as much:

Do you consider your opponent wrong — or do you consider your opponent’s right to hold that position illegitimate?

Those are radically different things.

The first is politics.

The second is potentially the beginning of something else.

And this distinction matters precisely because intolerance today rarely introduces itself as intolerance.

It usually arrives dressed as virtue.

The word “liberal” has become dangerously confusing

Before going further, one distinction is essential.

“Liberal” does not mean the same thing everywhere.

Classical liberalism emerged around individual liberty, property rights, freedom of conscience, limits on state power and pluralism.

Libertarianism pushes several of those principles further, particularly individual autonomy and restrictions on government coercion.

Modern American “liberalism,” meanwhile, generally refers to the centre-left or progressive political tradition.

European liberal parties constitute yet another family, often combining market economics, civil liberties, European integration and varying degrees of social progressivism.

And contemporary progressivism is not synonymous with classical liberalism at all.

Mixing these traditions together would produce a convenient polemic but poor analysis.

My argument is therefore not that liberals are extremists.

Nor is it that conservatives are inherently more tolerant.

There is abundant evidence of political intolerance, exclusion and authoritarian temptation on the right as well.

The more interesting proposition is this:

A political movement can call itself liberal while gradually abandoning one of liberalism’s most difficult principles: tolerating the political rights of people whose ideas it considers profoundly wrong.

And exactly the same contradiction can occur among libertarians, conservatives, nationalists or progressives.

That is why the paradox deserves attention.

Freedom is easy when everyone agrees

Most people support freedom of speech in the abstract.

The real test comes afterwards.

Should communists be allowed to speak?

Nationalists?

Religious fundamentalists?

Atheists?

Radical feminists?

Anti-abortion activists?

Pro-Palestinian activists?

Pro-Israel activists?

Climate radicals?

Climate sceptics?

People who challenge prevailing ideas about gender?

People who advocate dramatically higher taxation?

People who advocate eliminating much of the welfare state?

People who believe immigration should be radically restricted?

People who believe borders should be substantially more open?

Now freedom becomes uncomfortable.

And that discomfort is precisely the point.

Freedom of expression has almost no political meaning if it protects only expressions society already finds acceptable.

A society does not demonstrate tolerance by allowing people to say popular things.

It demonstrates tolerance by protecting the political space in which unpopular things can be said — subject, of course, to legitimate legal limits concerning matters such as threats, incitement and other unlawful conduct.

That distinction has become increasingly blurred.

The new definition of extremism

We normally define an extremist by asking:

What does this person believe?

Perhaps we should sometimes ask another question:

What does this person believe should happen to people who disagree?

That produces a very different political map.

Imagine two citizens.

Citizen A holds views that are far from the political centre but accepts elections, peaceful opposition, free expression, independent courts and the legitimacy of opponents participating in public life.

Citizen B holds relatively mainstream policy positions but believes people advocating certain lawful opposing positions should be removed from public platforms, excluded professionally or socially punished simply for expressing them.

Which citizen represents the greater threat to pluralism?

The answer is not automatically the person whose policy preferences are further from the median.

This is where conventional political labels become inadequate.

Ideological radicalism and political intolerance are related concepts.

They are not identical.

Pew Research Center itself made this distinction explicit in its major work on American political polarization: ideological consistency should not automatically be equated with political extremism. Someone can hold consistently liberal or consistently conservative positions without necessarily embracing extreme policy solutions. (Pew Research Center⁠)

That distinction is fundamental.

Extremism can concern what you want government to do.

But intolerance concerns whether you accept the political legitimacy of people who want something different.

From disagreement to moral contamination

Something important happens when politics becomes moral identity.

Consider the progression:

I disagree with you.

Then:

I think your idea is harmful.

Then:

I think supporting that idea says something bad about your character.

Then:

I don’t trust people like you.

Then:

I don’t want people like you in my social or professional environment.

And eventually:

Why should people like you be given a platform at all?

At that point politics has changed.

The opponent is no longer merely mistaken.

The opponent has become morally contaminated.

Researchers Eli Finkel and colleagues described a related phenomenon as “political sectarianism.”

Their framework contains three elements: othering, aversion and moralization. The opponent becomes fundamentally different from us, disliked and eventually regarded as morally suspect. The authors argue that the combination can make ordinary political defeat feel existential. (PubMed⁠)

That is an extraordinarily important insight.

Because once your opponent becomes evil rather than wrong, compromise changes meaning.

Compromise with someone who is mistaken can be pragmatic.

Compromise with evil feels immoral.

And suppressing evil can begin to feel virtuous.

That is how intolerance can be psychologically converted into morality.

The evidence is uncomfortable for everyone

Anyone looking for evidence that only one political camp has this problem will find examples.

But the broader evidence is considerably more interesting.

Pew’s landmark research into American polarization found increasingly distinct ideological camps and much less overlap between Republicans and Democrats than existed two decades earlier. It also found that ideologically consistent people were disproportionately politically engaged. (Pew Research Center⁠)

But consider what happened socially.

Among people with consistently conservative views, 66% said most of their close friends shared their political views.

Among consistent liberals, the figure was 52%.

Among people with mixed ideological positions?

Only 25%.

So conservatives in this dataset displayed the stronger tendency toward politically homogeneous friendship networks.

But another result reversed the pattern.

About 24% of consistent liberals said they had stopped talking to someone because of politics, compared with 16% of consistent conservatives and around 10% among people with mixed ideological views. (Pew Research Center⁠)

This does not prove that liberals are inherently less tolerant.

It proves something much more useful:

intolerance does not belong neatly to one ideology, and different ideological groups can express political closure differently.

One group may construct a homogeneous social environment.

Another may be more willing to terminate relationships over political disagreement.

Another may support restrictions on particular forms of expression.

Another may regard institutional constraints as obstacles when its own political objectives are frustrated.

The manifestations differ.

The underlying temptation is remarkably universal.

The further we move from moderation, the problem can intensify

Research into ideological intolerance adds another layer.

A peer-reviewed study indexed by the U.S. National Library of Medicine examined ideological attitudes and intolerance. Its results do not support the simplistic claim that liberals are generally more intolerant than conservatives: in the study, conservatives were on average more ideologically intolerant.

But the researchers also found something important for our argument:

differences in intolerance were associated with ideological extremity, with the patterns becoming stronger among more extreme ideologues. (PubMed⁠)

This should prevent both sides from declaring victory.

The intellectually serious conclusion is not:

“The left is intolerant.”

Nor:

“The right is intolerant.”

It is:

Ideological certainty can interact with intolerance in ways that make political opponents increasingly difficult to regard as legitimate.

The problem is not merely what people believe.

It is what absolute certainty encourages them to believe about everyone else.

The university is an extraordinarily revealing laboratory

Universities provide perhaps the clearest contemporary experiment because they combine young adults, politics, identity, intellectual debate and institutional authority.

The results deserve attention.

A 2024 Knight Foundation–Ipsos study found that approximately nine in ten American college students continued to regard free speech rights as very important.

Excellent.

But the same research found that 70% believed some forms of speech could be as damaging as physical violence.

Broken down politically, the figure was:

82% among Democratic students.

66% among independents.

59% among Republican students. (Knight Foundation⁠)

That does not mean 82% of Democratic students support censorship.

It would be misleading to claim that.

But it reveals a profound philosophical shift.

If speech itself becomes conceptually comparable with physical violence, the traditional liberal presumption in favour of allowing offensive speech becomes considerably harder to sustain.

Because societies legitimately restrict violence.

And once speech is morally categorized as a form of violence, restricting speech can consequently be reframed not as censorship but as protection.

That semantic transition matters enormously.

“I support free speech, but…”

The Knight research exposes another contradiction.

Two-thirds of students said self-censorship limits educationally valuable conversations.

Yet approximately two-thirds also reported self-censoring on at least one topic in classroom discussions.

Only around one quarter believed social-media dialogue was usually civil, down from 40% in 2016. (Knight Foundation⁠)

Think about the paradox.

We increasingly recognise that self-censorship damages intellectual exchange.

Yet we are simultaneously constructing social environments that encourage self-censorship.

That is not healthy pluralism.

It is pluralism on paper accompanied by conformity incentives in practice.

And the pressure does not come exclusively from the left

Here again, intellectual consistency matters.

FIRE’s 2025 College Free Speech Rankings surveyed 58,807 students at 257 American universities.

Very conservative students reported self-censoring particularly frequently: 34% said they did so very or fairly often, versus 15% of very liberal students. (thefire.org⁠)

But predominantly conservative institutions were hardly universally exemplary defenders of expression either.

FIRE placed institutions including Brigham Young University, Liberty University and Hillsdale College among schools receiving its “Warning” designation in that survey framework. (thefire.org⁠)

Again, the pattern is inconvenient.

Which is precisely why it is valuable.

Political tribes love evidence of their opponent’s intolerance.

They are much less enthusiastic about examining their own.

The conservative version of the same contradiction

There is a mirror image on the right.

A conservative may complain that progressive institutions suppress dissenting opinions.

That complaint may sometimes be justified.

But if the same conservative then wants government institutions to punish lawful political expression he finds offensive, interfere with universities because they teach ideas he dislikes, suppress peaceful opposition or treat journalists as illegitimate merely because their reporting is hostile, the principle has disappeared.

The argument has become:

Don’t censor me. Censor them.

That isn’t a defence of free speech.

It is a dispute about who gets to control the censor.

The same applies to progressive politics.

If someone defends minority expression, protest and dissent but wants lawful conservative opinions excluded from public discussion merely because they are offensive, the underlying principle is equally unstable.

Free speech for my tribe is not free speech.

It is tribal privilege.

The libertarian paradox may be even more fascinating

Libertarianism should theoretically be particularly resistant to this temptation.

Its intellectual architecture places extraordinary weight on individual autonomy.

You live your life.

I live mine.

Government interferes as little as reasonably possible.

But there is a potential psychological contradiction here too.

An uncompromising libertarian can become so convinced that almost every collective intervention constitutes illegitimate coercion that democratic disagreement itself starts looking morally suspect.

Taxation becomes theft.

Regulation becomes tyranny.

Public-health restrictions become dictatorship.

Redistribution becomes confiscation.

Central banking becomes manipulation.

Government itself becomes the enemy.

Some libertarians make none of these absolutist leaps.

Others do.

And once every policy disagreement is translated into the language of oppression, compromise becomes almost impossible.

Again, the paradox appears:

absolute commitment to liberty can become intolerance toward democratic decisions made by citizens who define liberty differently.

That does not invalidate libertarianism.

It demonstrates the danger of turning any political philosophy into an absolute moral system.

Social media industrialised moral certainty

Then came the machine perfectly designed to exploit this psychology.

Social media.

Its political economy rarely rewards:

“Interesting point. I disagree for the following three reasons.”

It rewards certainty.

Conflict.

Identity.

Humiliation.

Moral outrage.

The clever insult.

The devastating screenshot.

The enemy.

A 2026 Scientific Reports study analysed a very large sample consisting of roughly 67 million tweets and 30 million Reddit comments from 2010 through 2023.

The researchers found increasing estimated out-group animosity among the partisan users, politicians and journalists within their sampling framework.

But perhaps the most interesting finding was concentration:

a relatively small fraction of users generated a disproportionate share of negative political content. (Nature⁠)

This matters.

The internet may make societies appear even more intolerant than the median citizen actually is.

The most aggressive voices produce extraordinary quantities of content.

Moderates consume.

Extremists publish.

Algorithms distribute.

Journalists notice.

Politicians react.

And eventually everyone begins believing everyone else has gone mad.

Europe is not immune

It would be comforting for Europeans to dismiss all of this as another American pathology.

The evidence does not allow us that luxury.

Political scientists call the emotional dimension of partisan hostility affective polarization: the widening emotional distance between one’s own political group and competing groups. (Cambridge University Press⁠)

Comparative research across 22 European democracies found substantial affective polarization in European party systems, with particularly high levels in parts of Central, Eastern and Southern Europe. Importantly, ideological distance and emotional hostility are related but not identical: political systems can display considerable hostility even without extraordinary ideological distance. (Cambridge University Press⁠)

Research covering twenty Western democracies has likewise found that dislike between supporters of different parties is related to policy disagreement, with cultural disagreements becoming increasingly relevant over time. (Cambridge University Press⁠)

And European radical-right parties provide another instructive case.

Research covering 103 elections in 28 European countries found that populist radical-right parties occupy a particularly polarized position: their supporters express high levels of dislike toward mainstream parties while receiving high levels of dislike in return. (Cambridge University Press⁠)

The direction of hostility therefore matters.

So does reciprocity.

A society can enter a feedback loop in which:

“They despise us”

becomes the justification for

“therefore we are entitled to despise them.”

And both groups end up proving the other’s accusation.

But we should not exaggerate the apocalypse

There is another important warning in the evidence.

Polarization is real.

But claims that entire Western populations now hate each other can themselves become misleading.

A 2026 comparative study examining Denmark, Germany, Italy, Poland and the United Kingdom found a more nuanced emotional landscape than the popular “everyone hates the other side” narrative suggests. Strong negative emotions exist, but researchers also warned that citizens and commentators can overestimate their prevalence. (Cambridge University Press⁠)

Similarly, comparative work examining multiple dimensions of polarization across more than 100 countries over 25 years concluded that polarization has not simply increased everywhere and in every dimension. The trends depend heavily on whether we measure ideological distance, emotional polarization, elites or ordinary citizens. (Cambridge University Press⁠)

This distinction is essential.

We should not manufacture a crisis in order to write about one.

The danger is not that every Western citizen has become an extremist.

The danger is that politically intense minorities can increasingly define the boundaries of acceptable discussion for everyone else.

The silent centre

This may explain one of the strangest features of contemporary politics.

Millions of people hold combinations of views that political tribes consider ideologically impure.

Someone can favour immigration but want stronger border enforcement.

Support capitalism but favour universal healthcare.

Support abortion rights while disliking aspects of contemporary progressive culture.

Support LGBT rights while having reservations about specific gender policies.

Support environmental action but oppose particular climate regulations.

Support Israel’s existence while condemning Israeli government policies.

Support Palestinian rights while condemning Hamas.

Support Ukraine while questioning particular aspects of Western strategy.

Support markets while believing some industries require strong regulation.

Oppose Trump while distrusting progressive identity politics.

Support Trump policies while disliking Trump’s conduct.

These positions are not logically impossible.

Millions of citizens think this way.

Political tribalism increasingly tells them they shouldn’t.

Because ideological tribes demand packages.

Accept position A and you are expected to accept B, C, D and E.

Question D and suddenly your commitment to A becomes suspicious.

Politics becomes less a collection of judgments and more a membership contract.

Heresy is becoming more important than disagreement

This produces an extraordinary phenomenon.

People are often angrier at dissenters within their own political tribe than at predictable opponents.

Why?

Because an opponent is an enemy.

But a dissenter is a heretic.

The conservative who criticises his own conservative leader.

The progressive who questions progressive orthodoxy.

The environmentalist who supports nuclear power.

The feminist who disagrees with another feminist doctrine.

The pro-European who criticises EU institutions.

The patriot who criticises his own country’s foreign policy.

The socialist who defends free markets in one sector.

The capitalist who advocates aggressive redistribution in another.

These people disrupt tribal classification.

And tribal politics hates ambiguity.

Cancel culture is only one part of a much bigger phenomenon

The phrase “cancel culture” has become so politicised that it often obscures more than it explains.

The deeper phenomenon is social enforcement of ideological conformity.

It can operate through governments.

Universities.

Companies.

Media organizations.

Professional associations.

Political parties.

Online communities.

Religious communities.

Friendship groups.

Employers.

Consumers.

Sometimes social sanctions are entirely legitimate.

Freedom of speech does not mean freedom from criticism.

A company is not necessarily censoring someone merely by disagreeing with them.

Citizens have the right to boycott.

Editors have the right to choose what they publish.

Employers have legitimate interests.

Platforms require rules.

And other citizens possess their own freedom of association.

That distinction must be preserved.

But something changes when the objective is no longer:

“I want to challenge your argument.”

It becomes:

“I want to make expressing your argument socially impossible.”

That is where the liberal dilemma begins.

Tolerance does not mean moral relativism

At this point someone will inevitably object:

“So we should tolerate racism? Fascism? Communism? Religious extremism? Hatred?”

This question confuses tolerance with agreement.

To tolerate an opinion does not mean approving it.

It means recognising that a pluralistic society necessarily contains people whose beliefs we regard as wrong, foolish, offensive or occasionally repugnant.

Democratic societies can and do establish legal boundaries.

Threats are not simply opinions.

Incitement can cross legal boundaries.

Harassment is not identical to political argument.

Violence is obviously different from advocacy conducted within the law.

Different democratic legal systems draw these boundaries differently.

But if the boundary expands until offence itself becomes sufficient justification for exclusion, pluralism becomes fragile.

Because everyone is offended by something.

And today’s protected majority can become tomorrow’s offensive minority.

Karl Popper’s famous paradox — and its abuse

Discussions of tolerance often invoke Karl Popper’s “paradox of tolerance.”

The simplified internet version goes something like this:

A tolerant society must be intolerant of intolerance.

That sentence is frequently deployed as an intellectual permission slip for excluding opponents.

Popper’s actual argument was more conditional.

His concern was movements unwilling to engage in rational argument and prepared instead to answer arguments through coercion or violence.

That distinction matters.

Otherwise the paradox becomes circular.

Who is intolerant?

The people we have decided are intolerant.

Who decides?

We do.

What rights should intolerant people have?

Fewer.

At that point a principle intended to protect an open society can become a mechanism for closing it.

The most dangerous sentence in politics

Perhaps the most dangerous political sentence is not:

“You are wrong.”

It is:

“People like you should not be allowed to say that.”

Because hidden inside those words is a transfer of power.

Someone must determine which opinions qualify.

Someone must enforce the boundary.

Someone must decide tomorrow’s boundary.

And someone must decide who decides.

The authoritarian temptation rarely begins by announcing:

“I oppose freedom.”

It usually says:

“Freedom is extremely important, but this particular group is too dangerous to receive it.”

Every generation can invent its own exception.

The distinction we have forgotten

Liberal democracy depends on a distinction that increasingly feels almost archaic:

legitimacy is not agreement.

I can believe your policy would be disastrous.

I can campaign against you.

I can expose errors in your argument.

I can ridicule your contradictions.

I can vote against you.

I can publish twenty articles explaining why you are wrong.

And after doing all of that, I can still say:

you have the right to participate.

That final sentence is democracy.

Without it, politics becomes merely a struggle over which tribe controls the institutions capable of silencing the other.

Perhaps moderation is the wrong word

We often assume the opposite of extremism is moderation.

I’m no longer sure that is sufficient.

A moderate can be intolerant.

A radical can be pluralistic.

Someone can advocate enormous economic transformation while defending the civil liberties of opponents.

Someone else can hold perfectly conventional economic positions while wanting dissidents excluded from public life.

So perhaps the more meaningful opposite of political extremism is not centrism.

Perhaps it is pluralism.

Or, more simply:

tolerance.

Not tolerance as passivity.

Not tolerance as indifference.

Not tolerance as agreement.

Tolerance as the acceptance of an uncomfortable democratic fact:

Other people possess exactly the same political freedom that I claim for myself.

The ultimate liberal test

Here, then, is my test for anyone who calls himself liberal, libertarian, conservative, progressive or democratic.

Do not tell me how strongly you defend your own freedom.

That proves very little.

Tell me which freedoms you are prepared to defend for the person whose worldview you most dislike.

Would you defend his right to speak?

His right to vote?

His right to organise politically?

His right to publish?

His right to peacefully protest?

His right to challenge your ideas?

His right to criticise your identity, your religion, your political party, your country, your economic philosophy?

Subject to the same laws that apply to everyone else, would you defend his political citizenship even when doing so makes you profoundly uncomfortable?

That tells me far more.

The real paradox

The great political irony of our age may therefore be this:

We have never talked so much about freedom while becoming so suspicious of what happens when other people exercise it.

We celebrate diversity but often mean diversity of identity rather than diversity of thought.

We celebrate debate but increasingly prefer debate within boundaries established by our own moral community.

We celebrate authenticity until someone’s authentic opinion offends us.

We condemn censorship until the speech being censored is speech we hate.

We oppose ideological conformity — primarily when the ideology demanding conformity belongs to the other side.

And we describe ourselves as tolerant while steadily narrowing the category of people we believe deserve tolerance.

Perhaps this is why the conventional left-right spectrum increasingly fails to capture the most important divide.

The question may not simply be:

How progressive are you?

Or:

How conservative are you?

Or:

How libertarian are you?

The deeper democratic question is:

How much political freedom are you prepared to grant someone who would use that freedom to oppose everything you believe?

That is the difficult test.

Because freedom for ourselves is self-interest.

Freedom for our opponents is a principle.

And perhaps the political extremist of the twenty-first century should not be identified only by the radicalism of his own ideas.

Perhaps we should pay much closer attention to something else:

the moment he begins to consider illegitimate the right of another human being to hold and peacefully defend an opposing view.

That moment matters.

Because disagreement is the normal condition of democracy.

The attempt to delegitimise disagreement is not.

And when a society forgets the difference, the danger does not come only from the traditional extremes.

It can come from people standing beneath a banner marked:

LIBERTY.

Sources and further reading

Pew Research Center — Political Polarization in the American Public and Political Polarization & Media Habits. (Pew Research Center⁠)

Knight Foundation / Ipsos — College Student Views on Free Expression and Campus Speech 2024. (Knight Foundation⁠)

Foundation for Individual Rights and Expression / College Pulse — 2025 College Free Speech Rankings, based on 58,807 students at 257 institutions. (thefire.org⁠)

Finkel et al., Science — Political Sectarianism in America. (PubMed⁠)

Reiljan, European Journal of Political Research — comparative research on affective polarization in European party systems. (Cambridge University Press⁠)

Wagner, European Political Science Review — Affective Polarization in Europe. (Cambridge University Press⁠)

Gidron, Adams & Horne, British Journal of Political Science — affective polarization across twenty Western democracies. (Cambridge University Press⁠)

Harteveld, Mendoza & Rooduijn — research on affective polarization and Europe’s populist radical right across 103 elections in 28 countries. (Cambridge University Press⁠)

Fulay & Roy, Scientific Reports (2026) — analysis of approximately 67 million tweets and 30 million Reddit comments and changes in estimated out-group animosity. (Nature⁠)

Lawall, Versteegen & Tsakiris (2026) — comparative research on the emotions underlying affective polarization in Denmark, Germany, Italy, Poland and the United Kingdom. (Cambridge University Press⁠)