In a famous essay from 1969, the Anglo-Jewish liberal philosopher Isaiah Berlin distinguished between “two concepts of liberty.” The first concept, negative liberty, implies freedom from external coercion. “By being free in this sense, I mean not being interfered with by others,” Berlin wrote. The second concept, positive liberty, is more prescriptive. It asserts that freedom lies in self-direction and self-mastery. 

These two concepts of liberty correspond to two types of liberalism: negative and positive. Negative liberalism, championed by Berlin, is value-neutral, maintaining that all forms of human endeavor are equally worthy and that people should be free to pursue them so long as they don’t interfere with others’ rights. Positive liberalism, on the other hand, asserts there are nobler and baser human values. For advocates of positive liberalism, such as John Stuart Mill, individual autonomy and freedom from tradition are themselves noble values, and people should use their liberty to pursue them. The implication here is that what is new and different is in some sense morally superior to what is familiar and comforting.

Many of the political fault lines in the West today—including Brexit, the election of U.S. President Donald Trump, the rise of Europe’s far right, and ongoing controversies over campus political correctness—are bringing the contradictions between these two types of liberalism into the open. Broadly speaking, populist right voters are negative liberals, whereas libertarians and those on the intellectual left are positive liberals. Right-wing populists are motivated by a desire for white ethnic community and want to freely express that sentiment, which they feel is taboo—even if they themselves often fail to respect the identity claims of minorities. But for positive liberals, diversity is a good in itself and populist opposition to it is immoral. Each side approaches politics with a log in its eye, rendering dialogue impossible. 

Why are the contradictions of liberalism only bubbling up now? Secularization, the decline of warfare, and the collapse of communism are all part of the story. Yet the most important factor by far is immigration. Large-scale immigration (combined with low native fertility) is transforming the ethnic composition of the West. The change is dramatic: Canada was over 86 percent white in 2001, but if current trends hold is projected to be around 20 percent white in 2106. The United States and western Europe are expected to see similar, if less dramatic transformations. The demographer David Coleman has termed this ethnic population replacement the “third demographic transition.” (The first was the transition from high birth and death rates to low birth and death rates; the second was to below-replacement birth rates.)

The changes driven by immigration are exposing major political divisions in most liberal, Western societies. The question facing liberalism today is how members of these societies’ ethnic majorities will face this existential challenge. 


Ethnic change and immigration are particularly good at drawing out the disagreements between negative and positive liberals. Negative liberals accept that conservative, communitarian whites may want less diversity and believe they should be free to pursue this goal through politically legitimate means so long as the rights of other citizens are respected. But positive liberals believe whites should escape the confines of their ethnicity and embrace a cosmopolitan conception of the self. Morally, they should prefer diversity over homogeneity and novelty over cultural continuity.

Positive liberals believe whites should escape the confines of their ethnicity and embrace a cosmopolitan conception of the self.

Positive liberal cosmopolitanism has deep roots in American political thought. The first consistent advocate of this position was the Progressive philosopher and old-stock New Englander John Dewey. Around 1910, Dewey argued that the role of Anglo-Protestant Americans, then the dominant U.S. ethnic group, was to lead the country’s other ethnic groups toward cosmopolitan interchange and, subsequently, to universalize themselves out of existence. Around the same time, Randolph Bourne, an Anglo-American Greenwich Village bohemian, began advancing a more radical version of Dewey’s universalism. Rebelling against his own group, Bourne argued that American WASPs should embrace a double standard, rejecting their own puritanical culture in favor of cosmopolitanism while encouraging minorities such as Jews to retain their ethnic affiliations.

Both of these strands—Dewey’s melting pot and Bourne’s multiculturalism—shaped the form that positive-liberal cosmopolitanism would take in the West. The social upheavals of the 1960s added a new moral imperative, antiracism, to this mix. Although early twentieth-century U.S. Progressives had opposed racism, their cosmopolitan millenarianism concerned European immigration, which they saw as unrelated to the domestic struggle for black rights. It was not until the 1960s, after the civil rights movement and the start of large-scale non-European immigration, that pro-immigration and antiracist positions fused in both North America and Europe. Positive liberals used the new idiom of antiracism to reframe their prior appreciation of diversity, adding the emotional heat of an emerging social taboo to older appeals to generosity. Activists now spurned Dewey’s melting-pot cosmopolitanism, which envisioned all groups dissolving, in favor of Bourne’s multiculturalist position that the majority throw off its traditions while minorities maintain theirs.

As the white majority share of Western nations declined after the 1960s, a white backlash might have been expected. Yet just as immigration diversified their countries, more and more Westerners were exposed to cosmopolitan perspectives via newly expanded university systems and television media. These two developments offset one another to produce a polarized response: white communitarians viewed rising diversity negatively while cosmopolitan multiculturalists elevated it to the status of a moral imperative. Today’s culture war is a product of this split.

Children are sworn in as U.S. citizens on the deck of the USS Iowa Museum in Los Angeles, July 2017.
Mario Anzuoni / Reuters


In this polarized climate, individual value differences—themselves heavily shaped by genes and life experience—largely determine whether a person adopts the “closed” communitarian or “open” cosmopolitan viewpoint. According to the political scientist Karen Stenner, two sets of value orientations, authoritarianism and conservatism, are particularly important. Authoritarians prefer order over dissent and diversity. Conservatives favor the status quo over change. The two are not the same: a conservative who grows up in a multiethnic society that becomes more homogeneous, such as interwar Istanbul, will wax nostalgic about the loss of diversity, while an authoritarian will celebrate it. But when a homogeneous society becomes diverse, as is true in today’s West, authoritarians and conservatives are on the same page.

Authoritarianism and conservatism are more predictive of right-wing populism than are social categories such as age, income, class, gender, or even education. In the British Election Study (BES), for example, support for the death penalty—a proxy for authoritarianism—predicts 11.5 percent of the variation in a person’s likelihood of having voted to leave the European Union, whereas income, age, gender, region, and education combined account for only 4.9 percent. Some groups, such as the elderly, the non–college educated, the working class, and men are more authoritarian on average, but 90 percent of the variation in authoritarianism and conservatism lies within social categories, not between them. This is explained in part by individual psychological and personality differences: the Understanding Society survey finds that 23 percent of two-person British households, whose members generally have comparable social positions, are divided on Brexit. 

Similar patterns can be found in data from the United States. For example, in a Policy Exchange-YouGov survey I conducted in August 2016 during the primaries, whites who said that it is more important for children to be “well-mannered” than “considerate” (another proxy for authoritarianism) rated Trump two points higher on a zero-to-ten scale than those who answered in the reverse—a massive effect. The same is true for conservatism. American whites who strongly agreed that “things in America were better in the past” scored Trump a 6.5 out of 10, compared to 0.5 among those who strongly disagree.

Among whites, these differences strongly influence how an individual views immigration. People who prefer order over diversity and, in David Goodhart’s phrase, who view “change as loss,” back right-wing populism mainly because of immigration. Not only did virtually all the BES respondents most opposed to immigration vote Leave, but more than 40 percent of white British Leave voters in the Policy Exchange data said that immigration was the most important issue facing the country. By contrast, only five percent of Brexit voters said inequality was their most important issue. In the United States, 25 percent of whites rating Trump ten out of ten said immigration was their top concern, compared to virtually none among those scoring him zero out of ten. Furthermore, British and U.S. data show that immigration is significantly more important for whites living in neighborhoods that have recently undergone rapid ethnic change than for whites in demographically stable places. This echoes the wider academic literature: although established local minority populations reduce white hostility to immigration through inter-ethnic contact, my research has found that local minority growth increases white opposition in around 90 percent of studies.

The equation of white identity with racism has emerged as a powerful emotional regime in the upper echelons of Western societies.


The divisions thrown up by immigration are significant and have proved difficult to resolve thanks to deep moral disagreement between the two camps. Since the 1960s, positive liberals have come to view opposition to immigration through the prism of antiracism. They believe it is racist for whites to wish to maintain their proportion of the population by limiting immigration. Whites, in their view, have a moral obligation to eschew group partiality. The equation of white identity with racism has emerged as a powerful emotional regime in the upper echelons of Western societies. By contrast, negative liberals (white as well as minority) do not consider it racist for whites to act in their collective self-interest if this does not abridge the rights of their fellow citizens.  

This thesis is borne out by survey data. In small sample surveys conducted using the Mechanical Turk platform, I uncovered little partisan difference of opinion as to whether it’s racist for someone to want a separate country for their race or to not have a boss of a different race. Over 90 percent of both Clinton and Trump voters view these preferences as racist. At the other end of the scale, few of any political stripe think it is racist to feel more comfortable among one’s co-ethnics. On questions that pertain to “tribalism,” however—wanting one’s child to marry a co-ethnic, wanting to live in an area where one’s group is in the majority, or wanting to limit immigration to maintain the population share of one’s group—people are strongly polarized. Divisions between whites and minorities are modest, but the distance between Brexiteers and Remainers, or Trump and Clinton supporters, is immense.

In effect, white liberals consider white clannishness to be racist. But the more conservative portion of whites, between 60 and 80 percent of the total, perceive group partiality on immigration as a form of what Shadi Hamid terms racial self-interest—that is, an expression of collective self-regard rather than prejudice. To measure these divisions, in December 2016 I asked a representative sample of 2,600 Americans and 1,600 Britons the following: “A White American [Briton] who identifies with her group and its history supports a proposal to reduce immigration. Her motivation is to maintain her group’s share of the population for cultural reasons. Is this person a) racist, b) racially self-interested, which is not racist, c) don’t know.”

Results in Figure 1 reveal that almost all white Clinton voters with advanced degrees believe the person in question is racist. But among white Trump voters without degrees just 5.5 percent concur, while 94.5 percent say this is racial self-interest, which is not racist. The 62-point difference between white Clinton (73 percent) and Trump (11 percent) voters on whether the statement is racist dwarfs the modest nine-point (45 to 36 percent) gap between minorities and whites. Similarly, in the United Kingdom nearly half of Remain voters say a white British person who wants to reduce immigration to maintain group share is racist. This rises to 80 percent among university-educated Remainers who prefer current or higher levels of immigration. Among Leave voters, just six percent agree the statement is racist, dropping to zero among white British Leavers with less than a high school education. 

Figure 1. Birkbeck-Policy Exchange-Yougov survey, 7-8 December, 2016.
Figure 1. Birkbeck-Policy Exchange-YouGov survey, December 7–8, 2016.
Eric Kaufmann

In March 2017, I teamed up with Ipsos-Mori to run the same question in an 18-country study. The share of Americans who said the statement is racist was almost identical to that in the Policy Exchange survey, at 36 percent. Tellingly, it is in Western countries that we find the strongest relationship between a person’s support for immigration and his belief that it’s racist to oppose immigration for ethnocultural reasons. Figure 2 shows that in India, Japan, South Korea, and Turkey, many who say the statement is racist want fewer immigrants while plenty who say it isn’t racist want more immigration. By contrast, in Europe and countries settled by Europeans, positive liberalism and immigration opinion are more tightly connected, as illustrated by most Western countries’ proximity to the red line. The suggests that antiracism has more deeply reframed the immigration question in the West than elsewhere in the world.

Figure 2. Ipsos-Mori Global @dvisor Survey, 18 countries, 15 March 2017. N=14,014. Aggregated data.
Figure 2. Ipsos-Mori Global @dvisor Survey, 18 countries, March 15, 2017. N=14,014. Aggregated data. 
Eric Kaufmann

Pooling individual responses together in Figure 3, I find that Westerners who say that the statement isn’t racist have a 72 percent chance of favoring immigration restrictions, compared to just 30 percent among those who believe that it is racist. Outside the West, that 42-point difference drops to 21 points—still statistically significant, but a much smaller effect. In other words, when it comes to immigration attitudes, Western publics are more value-polarized than non-Western ones.

Source: Ipsos-Mori Global @dvisor Survey, 18 countries, 15 March 2017. N=14,014. Non-European countries: India, South Korea, Japan, Mexico, Turkey, South Africa. Controls for individual’s education, income, gender and age.
Source: Ipsos-Mori Global @dvisor Survey, 18 countries, March 15, 2017. N=14,014. Non-European countries: India, South Korea, Japan, Mexico, Turkey, South Africa. Controls for individual’s education, income, gender and age. 
Eric Kaufmann

Interestingly, in Europe the clash between communitarian ethno-traditionalists and positive liberal cosmopolitans is less direct than it is in the United States: British supporters of right-wing populism are motivated by the actual effects of immigration more than by disgust at leftists who they see as stifling their identity. When I asked a small sample of white British Leave voters on Prolific Academic whether “immigration, putting pressure on public services” or “political correctness” was a more important motivation for their vote, two-thirds chose immigration and only one-third chose political correctness. But when I asked white Trump voters, two-thirds chose political correctness, even though I also included a third, “other” option for American respondents. This may be due to a more vocal liberal activism on U.S., as compared to U.K., campuses, as well as the American left’s long-standing emphasis on racial redress, which is not as prominent on the British left.   


Divides over immigration force liberals to confront the question of whether diversity is a moral imperative or a matter of taste. In Europe, conservative white pressure has steered the political conversation toward reducing immigration. But immigration skeptics, most of whom affirm liberal antiracism norms, are generally unwilling to couch their position in ethnocultural terms, pretending instead that they are concerned about pressure on public services or jobs. These are post-hoc rationalizations—what social psychologist Jonathan Haidt terms the “rider” rather than the unconscious “elephant” that largely determines the path the rider takes. The consensus in the academic literature is that people’s economic situation does not explain their opposition to immigration. Unfortunately, the sidelining of cultural concerns in the name of antiracism forces anti-immigration politicians to pursue damaging policies, such as leaving the EU or cutting off immigrants’ access to welfare, which address the rationalizations instead of the root causes.

Resolving this tragic state of affairs will require give-and-take on both sides. The right needs to stop amplifying irrational fears and stereotypes of out-groups such as Mexicans and Muslims, which is clearly racist. Tributes to noxious causes such as Nazism or the Confederacy have no place on public land. Yet it is also time to shelve the conceit that Western whites who prefer the ethnic status quo over more diversity do so primarily out of racism. The American National Election Study (ANES) shows that warmth toward whites is not correlated with coolness toward blacks and Hispanics—quite the opposite. And Duke University political scientist Ashley Jardina reports that whites who identify with their group are no more hostile to minorities than whites who do not.

Even if immigration to the West stopped, which it won’t, demographic change will continue. To slow group decline, white ethnic boundaries would have to open wider to mixed-race individuals, emphasizing European ancestry rather than unmixed white appearance. We see this with many black and Hispanic ethnic groups, which accept the offspring of mixed marriages as full group members, although Asian groups by and large do not.

As the ethnic composition of Western countries continues to transform, it is inevitable that some form of white identity will seek political expression. This is not necessarily a bad thing—a confident, loosely bounded, and moderate white identity is a prerequisite for overcoming polarization and advancing a politics of the common good. The problem is when more extreme forms of identity are allowed to fester in the shadows. Avoiding this requires recognizing that it is wrongheaded for positive liberals to ask conservative whites to celebrate their group’s demise, just as it is misguided for right-wingers to insist that minority ethnicity be erased in favor of one-size-fits-all nationhood. Instead, communitarians and cosmopolitans should strive to understand the world from each other’s point of view. Immigration and national identity should reflect a Berlinian spirit of compromise, not a Millian crusade to banish communal yearnings that is likely to end in failure.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now