Multiculturalism cannot work. It’s not an opinion, it’s a law.

We were told that diversity was a strength. We were told it at school, on television, in advertising, in official speeches. We were told it with public funding, urban policy, high councils for integration. It wasn’t an idea in the air of the time, it was a state doctrine, built, embraced, institutionalized for forty years.

And like all doctrines that deny a natural law, it shatters against reality.

This op-ed isn’t an op-ed about immigration. It isn’t an op-ed about racism. It’s an op-ed about a mechanism. An anthropological mechanism, documented, reproducible, that history has validated on three continents and that political science is validating in real time before our eyes. Multiculturalism cannot work within a liberal democracy. It isn’t a question of dosage, of goodwill or of resources.

Let’s call it a law, if the word bothers the purists: a heavy trend, an empirical constant that three continents and a century of experiments reproduce without exception. In the social sciences, no one carves anything in stone. But when the same curve reappears in Beirut, in Sarajevo, in Baghdad, in the Val-de-Marne and in the figures of an OECD survey, you stop talking about coincidence. You can look for counterexamples: Singapore is often cited, Canada too. Singapore is a city-state with no liberal-democratic heritage, governed for sixty years by a single party that controls public space with surgical precision. Canada is a continental country with no land borders under pressure, whose cohesion rests on a civic identity built from a void, with no pre-existing nation to fragment. Neither one resembles France. The exceptions prove the rule: they work because they avoided precisely the conditions in which this mechanism runs wild. And laws don’t care about our opinions.

Democracy requires a we

Let’s begin at the beginning. What does a liberal democracy require, in its most elementary foundations?

It requires that a people agree to live under the same rules. Above all it requires that they agree to submit to electoral results even when they lose. That’s no small thing. It’s even extraordinary, when you think about it: millions of individuals with divergent interests accept that a majority decides for everyone, including for those who voted against. This democratic miracle rests on an invisible but indispensable substrate: social trust.

Not the trust you have toward your friends or your family. The other kind. The one you grant to the stranger on the subway, to your neighbor across the landing, to the official who checks your papers. That diffuse, generalized trust that makes you pay your taxes because you assume the others do the same. That makes you stop at the red light because you trust the other drivers to do likewise when you cross. That makes you leave your car outside your home without barricading it.

It’s what the American political scientist Robert Putnam calls social capital: the set of norms and networks that make cooperation possible between individuals who don’t know one another. When this capital collapses, individuals stop cooperating. They go home, close the door, and live for their narrow group. Society, in the broad sense, stops working.

The question is therefore simple: what produces this trust? And what destroys it?

Putnam against himself

Robert Putnam is an American left-wing academic, a convinced Democrat, pro-diversity. In the 2000s, he sets out to demonstrate scientifically that cultural diversity is good for societies. He spends years studying hundreds of American neighborhoods, measuring elementary indicators: do you trust your neighbor? Do you vote? Do you volunteer? Do you take part in collective life?

What he discovers destroys him.

The more culturally diverse a neighborhood is, the less individuals trust one another. And not only between different groups, within each group. A white person in a very mixed neighborhood trusts other white people less than a white person in a homogeneous neighborhood. Same for Black people, for Latin Americans, for everyone. Diversity doesn’t only create intercommunal tensions: it dissolves trust in general. It produces what Putnam calls “hunkering down”: individuals shrink like turtles, reduce their circle of trust, vote less, cooperate less, distrust everything and everyone.

Putnam would take years before publishing these results, in 2007, in an article titled E Pluribus Unum. He retests them, rechecks them, looks for the error in his method. He knows what his conclusions imply politically. He chooses to publish anyway, concluding that these results made him deeply uncomfortable, but that they’re very solid.

Since then, the meta-analysis by Dinesen, Schaeffer and Sønderskov covering 1,001 estimates from 87 studies conducted in dozens of countries has confirmed it: the more cultural and ethnic diversity there is, the less social trust there is. Independent of the wealth of the neighborhood. Independent of the level of education. It isn’t poverty that creates distrust: it’s cultural heterogeneity itself.

Putnam himself qualified his conclusions by suggesting that in the long term, societies could forge new common identities capable of restoring cohesion. That’s theoretically possible. But what he leaves out is that this reconstruction presupposes a cultural force of attraction that our societies, gnawed by identity doubt and self-shame, are no longer capable of exercising. You don’t remake a we with a State that has stopped believing in itself. And it presupposes that the State works at it deliberately and with constancy, exactly what France stopped doing forty years ago by substituting the celebration of diversity as an end in itself for the construction of a common identity.

Lebanon, Bosnia, Iraq: the law at the scale of a country

If you want to see what Putnam’s mechanism produces applied no longer to a neighborhood but to an entire State, you only have to look at the history of the twentieth century.

Lebanon was for decades the model of living together in the Middle East. Eighteen officially recognized communities, represented in the institutions according to a confessional system carved into the constitution: the president is a Maronite Christian, the prime minister a Sunni, the speaker of parliament a Shiite. Everyone is represented, everyone is protected. Beirut was called the Paris of the Middle East. In 1975, civil war breaks out. It will last fifteen years. A hundred and fifty thousand dead. The country has been trying to rebuild for fifty years, without succeeding.

Tito’s Yugoslavia: Serbs, Croats and Bosniaks live together as long as the State imposes a common identity by force. Tito dies in 1980. In 1992, twelve years after his death, comes the war, with ethnic-cleansing camps in the very heart of Europe, in 1995. A revealing detail: Serbian propaganda and Croatian propaganda during the conflict both draw on the memory of the Second World War. After fifty years of dictatorship, in less than ten years, the buried communal identities resurface as if nothing had happened.

Iraq: Sunnis, Shiites and Kurds coexist under the terror of Saddam Hussein. The terror is removed in 2003. In less than two years, sectarian civil war. The Islamic State is born directly out of this void.

Three countries, three continents, three different eras. The same scenario. When the imposed common we collapses (because the dictator dies, because the terror loosens, because the institutions crack), individuals return to their primary belonging. Without exception. It isn’t bad luck. It isn’t the doing of bad leaders. It’s an anthropological law.

Worth noting is an important difference with France: in Lebanon, multiculturalism was constitutionalized, confessionalism was the system. France, for its part, remains officially universalist, which makes the gap between the rhetoric and the reality even more dizzying.

Voting as an ethnic census

There’s a reality no one in France wants to name. Ethnic voting.

In a working democracy, individuals vote according to their convictions, their interests, their worldview. That’s the basis: a political offering, a deliberation, a choice. This choice can be influenced by social class, age, territory, but it remains permeable to ideas. You can convince someone to change their mind.

In fragmented multicultural societies, something tips over. Lars-Erik Cederman, a researcher specializing in ethnic conflicts, has studied this mechanism across dozens of countries. His conclusion is beyond appeal: in culturally fragmented societies, politics becomes ethnicized in an almost systematic way. Parties stop defending ideas: they defend communities. Voters stop choosing a project: they choose a side.

France is already there.

In 2022, 70% of France’s Muslims vote for Jean-Luc Mélenchon in the first round of the presidential election. Not because Islam is left-wing, all the opinion surveys show that French Muslims are mostly conservative on social issues. But because La France Insoumise spent ten years deliberately building a communal political offering, and that community returned the favor. Individuals who, according to their own convictions, should vote on the right, vote on the left because their group votes on the left. The platform doesn’t count. Belonging counts. It’s precisely what I analyzed in my article on the sociology of voting: identity determinants progressively crush programmatic ones.

As a mirror image, what Christophe Guilluy documents in La France périphérique and Jérôme Fourquet in L’Archipel français: the RN (Rassemblement National) vote in white rural areas has long since stopped being a working-class vote about purchasing power. It’s an identity vote, that of people who feel foreign in their own country. A vote of withdrawal, exactly in the sense Putnam means.

Two Frances that no longer vote together. Two Frances that clash symbolically at every election. When this stage is reached, the election stops being a deliberation. It becomes an ethnic census. And the loser of a census doesn’t recognize the legitimacy of the winner: it contests the legitimacy of the system. It’s moreover one of the deep flaws I had pointed out in my analysis of the French political legitimacy crisis: when institutions are no longer perceived as neutral, consent collapses.

Cederman goes so far as to say that when political communalization reaches this stage, the direction the logic takes is that of armed conflict. Not automatically. Not immediately. But it’s the slope.

The dead end no one dares to formulate

Faced with this observation, there aren’t thirty-six options. There are exactly two.

The first: to admit that multiculturalism cannot work within a liberal democracy respectful of individual rights. It’s the conclusion of Putnam, of Dinesen, of Cederman. A common culture isn’t decreed with slogans. It’s built with centuries of shared life, or with an assimilation no one was willing to carry out.

The second option is precisely that assimilation. We have to be precise about this word, because it has become a dirty word that its opponents have emptied of its meaning. Assimilation isn’t only a violence. It’s also a promise. A promise of belonging, of formal equality, of republican dignity. The Third Republic practiced it with the Bretons, the Basques, the Alsatians, the Occitans. Jules Ferry’s school wasn’t only a steamroller of regional identities, it was also an elevator. A Breton child who went through this process came out with the same rights, the same codes, the same access as any son of a Parisian notable. He lost Breton, he gained the Republic. You can debate the price: the ruler across the knuckles was real. But the promise was kept: in two generations, peoples who thought of themselves as Breton became French because the State didn’t only impose an identity on them, it offered them another one, universal, that opened doors.

Tito did the same in Yugoslavia. As long as the State imposed the Yugoslav identity (but without the promise of an authentic common belonging, without the rights and the dignity that make the contract acceptable), the communities coexisted under constraint. The State let go. We know what happened.

The question that then imposes itself is this one: if the Third Republic had to exercise this assimilation (at once coercive and full of promise) on Breton or Occitan populations whose cultural difference from the rest of France was infinitely smaller than that which separates an Afghan or a Malian from French culture, what brutality would have been necessary to carry out the same integration with populations from radically different civilizational spheres? And above all: who was ready to make the promise in return? Who was ready to say sincerely to these populations: you become French, truly, with no asterisk, no hyphen?

No one wanted to answer this question. It’s understandable. In a France of human rights where the protection of the individual has become the alpha and omega of political discourse, that demand (at once firm on the rules and generous on belonging) had become unthinkable to formulate. But then, you would have had to have the honesty to say out loud: without assimilation, the result won’t be a harmonious multicultural society. It will be a society of multi-conflictuality. What I developed in my reflection on the shadow of the Muslim Brotherhood over the Republic illustrates, moreover, how this assimilative void was systematically filled by other identity offerings, infinitely less compatible with the republican project.

We chose not to say it. We chose to call it “diversity is a strength”.

What we reap

Since 1980, every Western European country that has experienced significant waves of immigration records the same thing: a fall in declared social trust. The 2024 OECD survey documents it. The CEVIPOF has been measuring it for decades with its two ritual questions to the French (can most people be trusted and you can never be too careful when dealing with others), and the curves have gone in the same direction for forty years.

France is living in real time what Putnam had described at the scale of a neighborhood, applied to an entire country. And the collapse of European social democracy isn’t unrelated: the redistributive model presupposes a solidarity between strangers, precisely what cultural fragmentation erodes first.

The signs are in the street. Supporters who pull out the Moroccan flag in French stands during a France national team match. Memories that clash: here communities that proudly claim the heritage of the Barbary raids, there communities that carry the Algerian War like a wound for which France would be guilty for eternity. These narratives don’t seek to reconcile. They’re at symbolic war: on social media, in universities, in the press. Two relationships to the world, two incompatible memories, the same passport.

It isn’t hatred. It’s logic. When the French we disappears, individuals return to their primary belonging. It’s what their brain commands them to do. It’s what humanity has always done. It isn’t malice, it’s an identity protection as natural as the survival reflex. And meanwhile, as I analyzed in my article on the cartels and institutional collapse, non-state actors exploit with formidable efficiency the gray zones this disintegration creates.

France isn’t yet in civil war. But it’s living what Lebanon was living in the 1960s: communities that coexist, a democratic game that still works, but a vote that’s becoming ethnicized, memories that diverge, and no one in the institutions to name out loud what everyone sees in the street.

We made the choice of ideology against reality. We preferred the myth of living together to the brutality of what it would have taken to truly live together.

The price is being paid now. The question that remains (and it’s political, not sociological), is whether anyone, somewhere in the French institutions, is still capable of formulating the promise the Third Republic had managed to keep: join us, become us, and we’ll be serious about both words. Not diversity celebrated as an end. Not assimilation-as-punishment. Demanding belonging. It’s a narrow door. But it’s the only one that has ever worked.