Identity and Belonging
Many Western societies have become far more multi-ethnic within two generations. The data on composition change, the measurable effects on trust and cohesion, the political fights it has fed, and the strongest version of each serious position.
(up from about 17% in 1970)
(about 2.5% in 1971)
(about 18% in 2005, when first counted)
A note on framing. Identity is one of the most charged topics in modern political life. The numbers are real, the underlying social changes are real, and the political conversation around them in many countries is unusually distorted - in different directions on different sides. The page below walks through the data plainly and presents the strongest version of each serious position. Where research findings are uncomfortable, this page does not edit them. Where political claims go beyond the research, it marks the difference.
What has changed
Most Western countries today are substantially more multi-ethnic than they were two generations ago. The composition change has been driven primarily by migration, and secondarily by differences in birth rates between groups. The pace varies by country, with the United Kingdom, the United States, Canada, Australia, and Germany seeing especially fast change since 1990, and the Netherlands and Sweden seeing large change since the 1970s, while France's change began earlier and has been slower.
The United States in 1970 was about 83% non-Hispanic white, 11% Black, and 4% Hispanic, with smaller shares of other groups. By mid-2025, on Census Bureau estimates, it was about 56% non-Hispanic white, 20% Hispanic, 14% Black, and 7% Asian, the rest multiracial or of other groups. The Bureau's 2023 projections place the non-Hispanic white share below 50% in the second half of the 2040s, depending on assumptions about migration and inter-group identification.
England and Wales in 1971 had about 1.2 million residents from non-white minorities, on later estimates - roughly 2.5% of the population. At the 2021 census, 81.7% of residents were white (including white minorities such as Polish and Irish migrants) and 18.3% belonged to Asian, Black, mixed, or other ethnic groups. Several major cities, Birmingham and Leicester among them, now have non-white majorities. In London, the case most often cited and most often stated loosely, white residents of all origins are still a small majority, at about 54%, while the white British share has fallen to about 37%.
Germany built its postwar prosperity partly on guest workers from Turkey, Yugoslavia, Italy, and others. Reunification, EU expansion, the 2015 refugee inflow, and Ukrainian refugees from 2022 added further composition change. In 2025 about 31% of the population had a "migration background", meaning they or at least one parent were not born German citizens (federal statistics office). The country does not collect data on race or ethnicity in the way English-speaking countries do.
Sweden, the Netherlands, France, Belgium, and Austria all show similar patterns of significant composition change, with migration concentrated in major cities. Spain, Italy, and Portugal experienced rapid migration changes after 2000 from a much lower baseline.
Canada and Australia have been actively shaping migration policy to attract skilled migrants, and have generally absorbed substantial composition change with less political disruption than several European countries.
What is unusual about this period is not the existence of immigration. The United States in particular has been a country of immigrants for its entire history. What is new is the speed and scale of composition change in countries that previously had relatively homogeneous populations. England and Wales going from over 97% white to under 82% in fifty years is a faster and larger change than most precedents in modern history outside of major conquests or the breakup of empires.
What the research finds
Research on how composition change affects social trust, political stability, economic outcomes, and wellbeing is mixed and contested. Several patterns are robust enough to be worth noting.
Diversity correlates with lower social trust in the short run. Robert Putnam's controversial 2007 paper, drawing on a survey of more than 29,000 Americans in 41 communities, found that ethnic diversity was associated with lower social trust both within and between groups - even after controlling for income, education, age, and many other variables. A 2014 review of 90 studies (van der Meer and Tolsma) found consistent effects only on cohesion within neighborhoods, more often in the United States than elsewhere, and no link to weaker cohesion between ethnic groups. A 2020 meta-analysis of 87 studies (Dinesen, Schaeffer, and Sonderskov) found a statistically significant negative relationship, strongest for trust in neighbors and when diversity is measured locally. Why the link exists is disputed. The longer-run effect, as societies adapt and develop new shared identities, is less clear.
Immigration also brings measurable economic gains, unevenly shared. Most studies find positive effects of immigration on national economic growth, innovation, entrepreneurship, and labor-market dynamism. Migrants are more likely than native-born workers to start businesses and to take essential service jobs that local labor markets struggle to fill. The US National Academies found in 2017 that first-generation immigrants cost governments more than the native-born, while their children are among the strongest net contributors. Wage losses, where they occur, fall mainly on earlier immigrants and workers without a high school diploma.
Integration outcomes vary substantially by group, generation, and country. Some immigrant groups have converged rapidly with host populations across most measures (Asian Americans in the US, several Eastern European groups in Western Europe). Some have integrated more slowly (parts of the Maghrebi community in France, parts of the Turkish community in Germany, parts of the Pakistani community in the UK). The reasons are varied - educational starting points, religious differences, residential concentration, host-country policy, and the receiving political environment all matter.
Pace of change matters. Most research on integration suggests that slower demographic change is generally easier to absorb than faster change. Communities that experience large composition shifts in a short period tend to show more political backlash and slower trust recovery than communities that experience similar total change over longer time periods. Pace is something policy can influence; absolute level usually is not.
The class dimension is large and often underweighted. The communities that experience composition change most directly tend to be working-class urban neighborhoods. Wealthy suburbs and rural areas often see less direct demographic change. Critics of high immigration argue that this lets better-off voters favor migration policies whose direct effects fall mostly on lower-income communities, and that it feeds a backlash elites tend to underestimate. William Julius Wilson and others argue that race and class together shape these dynamics in ways single-axis analyses miss.
How countries compare
Patterns of composition change and political response vary widely across countries, with implications for what policies have worked and what has not. Foreign-born shares below are UN estimates for 2024; the notes use the latest national data.
The takeaway: composition change is large, fast in historical terms, and uneven across countries. Even Canada and Australia, with their selective systems, saw opposition to current intake levels rise sharply after 2022. The low-migration countries (Japan, much of Central Europe) face labor shortages as their populations age; Japan's foreign population is now growing fast. There is no costless option, and the trade-offs are visible in every direction.
Why this is politically difficult
Identity politics is uncomfortable to discuss honestly, partly because three things that are usually held separate become entangled.
The data is the data. Composition change is real, large, and historically fast. Effects on trust, neighborhood patterns, and politics are measured. Playing these facts down and dramatizing them are both common moves in the political conversation, and neither helps anyone - including the migrants and long-settled residents whose interests each side says it protects.
The values question is contested. Reasonable people, looking at the same data, reach different conclusions about how much composition change is desirable, how much should be controlled, what the obligations of host societies and newcomers should be, and whether policy should take group membership into account at all. None of these is simply a question of fact. The US dispute over diversity, equity, and inclusion (DEI) programs is a current example. Executive orders in January 2025 ended federal DEI programs and revoked a 1965 order requiring affirmative action by federal contractors. Supporters say this restores equal treatment regardless of race or sex; opponents who sued argued that the orders' vague terms chill lawful speech and programs. An appeals court lifted an injunction against the orders in February 2026, though narrower challenges continue. Among S&P 500 companies, use of the term "DEI" in annual reports fell 68% from 2024 to 2025, which the Conference Board reads as reframing rather than abandonment.
The framing is heavily strategic. Every political faction has incentives to describe the data in ways that support its preferred conclusions. Pro-immigration framings tend to under-emphasize short-term integration costs. Restrictionist framings tend to under-emphasize long-term economic benefits and the human reality of migrant lives. Claims about race and ethnicity are often used as rhetorical weapons in fights that are about something else. A reader who follows only one frame is consistently given a partial picture.
The serious work of citizenship in this area is to hold the data, the values, and the framing strategies separately and apply judgment.
The paths from here
Dated September 2026, with net migration sharply lower in the United States, the UK, and Germany, and parties campaigning for much lower immigration leading the polls in Germany and France. The composition change already underway in most Western countries has substantial momentum. The shape of the next thirty years depends partly on policy choices about immigration levels and integration support, and partly on what kind of national identities are being built.
Continued composition change with growing political backlash
Migration continues at roughly the current pace. Composition continues to change. Political backlash continues to grow alongside it, with restrictionist parties gaining further ground in many Western countries. The political environment becomes more polarized on identity questions; coalition-building becomes harder.
Will it happen? Still the base case in much of Europe, where such parties lead the polls in Germany and France, though the Sweden Democrats lost ground in September 2026. Migration itself has slowed sharply in the largest Western destinations (see path 2), weakening this path's first premise.
Tighter migration policy reduces the inflow
Public pressure produces meaningful tightening of migration policy in major receiving countries. Net migration falls. Composition continues to change but more slowly, and the political tension reduces. The cost is a worse demographic and labor-market position over the next twenty years.
Will it happen? Under way in the three largest Western destinations: net migration roughly halved in the United States and Germany in 2025 and has fallen by more than 80% from its 2023 peak in the UK. Evidence on whether tension then eases is mixed: the share of Americans wanting less immigration fell from 55% in 2024 to 29% in June 2026 (Gallup), while Reform UK made large gains in English local elections in May 2026. The economic and demographic costs are real too, and many countries have struggled to maintain restrictive policies once they hit their labor-market consequences.
A new shared civic identity emerges
Major receiving countries succeed at building multi-ethnic shared identities that bind newcomers and long-established residents to common institutions, language, history, and rituals. The American experience after the 1965 immigration act is the modern reference case; whether others can replicate it is open. National identities are reinvented rather than abandoned, with explicit room for citizens of every background.
Will it happen? Possible but hard. Successful examples (parts of US history, parts of the British Commonwealth experience) suggest it can be done. Failed examples (parts of European integration, parts of post-imperial transitions) suggest it can also fail.
Identity tribes harden into separate political camps
The groups within multi-ethnic societies organize more strongly along ethnic, religious, or racial lines. Politics becomes increasingly a competition among identity coalitions rather than a contest of ideas across them. The civic-national frame weakens. Political polarization deepens.
Will it happen? Partly, and the recent US evidence points both ways. Voters sorted less by ethnicity in 2024: Hispanic voters split 51% to 48% between Kamala Harris and Donald Trump, after backing Joe Biden 61% to 36% in 2020 (Pew). But views of diversity have split by party: in late 2025, 87% of Democrats and 40% of Republicans said it is important for companies to promote racial and ethnic diversity, against 86% and 61% in 2019 (Pew). Whether the trend stabilizes or intensifies depends partly on choices that institutional design and political leadership can influence.
Convergence across generations softens the picture
Second and third-generation migrants increasingly identify with their host country, intermarry across groups, and follow educational and economic trajectories that converge with the broader population. The visible composition change of the parents becomes less visible in the lives of the grandchildren. The "minority" categories themselves blur as multi-ethnic identification becomes more common.
Will it happen? Has been happening visibly in the US, Canada, and parts of Europe. The 2020 US census counted 33.8 million multiracial people, up from 9 million in 2010, a jump the Census Bureau attributes largely to changes in question design and coding. Whether the convergence continues at the same pace, slows, or reverses depends on factors that are not fully predictable.
The cases usually cited as successes combine selective immigration policy, strong integration support, confident host-country civic identity, and political leadership that takes both the concerns of newcomers and of long-established residents seriously. None of those conditions are universally present in the West today. Whether they can be built in the next two decades is the deepest political question of the period.
Where serious analysts disagree
Identity is one of the topics where the gap between political slogans and careful research is widest. The named voices below represent serious positions held by working scholars; reading them directly is more useful than reading commentary about them.
Diversity creates short-run challenges that careful policy can mitigate but not erase
Robert Putnam's 2007 finding - that ethnic diversity correlates with lower social trust in the short run - was uncomfortable for the political and academic environments that received it. Putnam himself is a longtime liberal and explicitly supportive of immigration; his finding was that the data showed something more complicated than the slogan that diversity is strength. Holders argue that ignoring the finding does not help build better policy.
Held by: Robert Putnam (Harvard) himself, plus a wider body of trust research, including the 2020 meta-analysis described above. Critics, among them Maria Abascal and Delia Baldassarri (2015), re-analyzed Putnam's own data and attribute the association to residential sorting and inequality rather than to diversity itself.
Integration generally works over time and most current concerns are misframed
Looking at multi-generational data, the children and grandchildren of migrants tend to converge with native-born populations on educational outcomes, economic outcomes, language, identity, and political participation. The integration story is one of largely successful absorption, with specific failures usually traceable to specific policy choices rather than to the existence of migration itself. Most of the public anxiety reflects short-run friction rather than long-run trajectory.
Held by: Richard Alba (CUNY, "The Great Demographic Illusion"), Roberto Suro, and a substantial body of immigration research. Their data on multi-generational outcomes is extensive; critics ask whether the future will resemble the past well enough for those patterns to continue.
The political backlash is rational and is being underestimated
Citizens of high-immigration societies who feel their countries' character is changing in ways they did not vote for are not all racists or fools. Many of them have legitimate concerns about wages, housing, public services, schools, and cultural transmission that are being dismissed by elites who do not bear the direct costs. The dismissal feeds the backlash. Honest engagement with the concerns - even when one disagrees with the political conclusions - is more useful than treating restrictionism as illegitimate by definition.
Held by: Eric Kaufmann (Buckingham, "Whiteshift") and a body of political-science research on immigration backlash. His position is contested in academic circles, where critics argue that treating ethnic-majority attachment as a legitimate interest lends cover to ethnic nationalism. It has gained traction as parties campaigning against high immigration have grown.
Race and class cannot be separated
Discussion of identity politics in the United States that focuses only on race or only on class consistently misses the picture. The lives of working-class Americans of every race share more with each other than with the upper classes of their own race. Working-class Black, Hispanic, and white Americans are subject to similar economic pressures, similar housing stresses, similar healthcare gaps, and similar political marginalization by elites of various races. Building political coalitions across race within class would change politics; the way the conversation is currently structured prevents that.
Held by: William Julius Wilson (Harvard) and the broader tradition of race-and-class scholarship. Holders conclude that single-axis identity politics misses much of what the data show. Critics point to gaps that persist within classes: Raj Chetty and colleagues (2020) found that, at the same parental income, Black boys grow up to earn less than white boys in 99% of US census tracts.
Civic-national identity can be reformed inclusively
The failure of many Western multi-ethnic societies to handle composition change well is not a failure of multi-ethnic society as such. It is a failure of civic-national identity to be reinvented for the new reality. Successful multi-ethnic societies (parts of Canadian, Australian, and US history; some city-state examples) have rebuilt national identity to explicitly include all citizens, not as a fiction but as a worked-out civic project. The work is hard but is more promising than either ethno-nationalism or post-national cosmopolitanism.
Held by: Yascha Mounk (Johns Hopkins, "The Great Experiment") and a strand of liberal-democratic political theory. Their position is contested and prescriptive; the supporting evidence is mixed but real.
What the evidence supports firmly: composition change is real, large, and historically unusual. Diversity is associated with lower trust, especially among neighbors, though scholars dispute why. Most groups converge over generations, at different speeds. What stays contested is how much weight the backlash's concerns deserve, how far race and class can be separated, and whether an inclusive civic identity can be built fast enough. None of this maps cleanly onto the slogans of either political side.
What this means for you
Identity touches almost every reader directly: where you live, who you live among, who your children go to school with, what your country's politics looks like. A few practical observations that work regardless of one's political position:
If you are thinking about where to live
The pace of demographic change varies dramatically across cities, neighborhoods, and regions. Some neighborhoods have been in slow steady change for decades; some have been transformed in five years. Whether either pattern is comfortable for you and your family is a personal question. Going in with eyes open about the demographic trajectory of a place - not just where it is today - is more useful than being surprised later.
If you are raising children
Most countries' children today will grow up in more ethnically diverse environments than their parents. Whether you welcome this, want to control it, or are indifferent to it, the practical reality your children will live in is one where they will work, marry, and form friendships across boundaries that did not exist in earlier generations. Equipping them with both confidence in their own background and respectful curiosity about others is more useful than choosing between defensive identity politics and dismissal of identity altogether.
If you are an immigrant or descendant of immigrants
Your experience of integration depends on factors largely outside your control - the receiving country's policies, the local communities you encounter, the broader political environment. It also depends on choices you can make: language acquisition, professional networks, civic engagement, willingness to engage across community lines. The most successful immigrant trajectories tend to combine genuine engagement with the new country with confident retention of what is valuable from the original culture. The choice between assimilation and identity-preservation is rarely as binary as it is sometimes framed.
If you have lived in your country for generations
The demographic change of your country is real and large. Both the political tendencies that play it down and the ones that frame it apocalyptically offer distorted pictures. The realistic posture is to take the data seriously, engage thoughtfully with newcomers as individuals rather than as representatives of categories, and support the policies that produce successful integration rather than the ones that produce neither effective restriction nor effective integration.
If you are voting and engaging politically
Identity politics in most countries is more useful as background than as foreground. The politicians who deliver results on immigration, integration, education, and civic life are usually the ones who treat it pragmatically rather than the ones who use it primarily as rhetorical fuel. Voting for specific policies (selective migration, integration funding, language education, anti-discrimination law, equal-opportunity hiring) is more useful than voting for slogans about identity. The slogans are designed to mobilize; the policies are what change outcomes.
The mechanics behind this
The identity story sits on top of three deeper mechanisms covered elsewhere on this site. If the analysis above depends on ideas you want to understand first, these fundamentals make the conversation more legible:
