In 2000, Harvard political scientist Robert Putnam - the man who wrote *Bowling Alone*, the most cited work in social science of its generation - launched the largest study of civic life in American history. He surveyed nearly 30,000 people across 41 communities, from Los Angeles to rural South Dakota.
He was looking for the effects of diversity on social trust. He expected to find some short-term costs but long-term benefits. Instead, he found something that made him sit on his own research for six years.
What Putnam Found
The data was unambiguous. In the most ethnically diverse communities in America, residents:
- Trusted their neighbors about half as much as those in homogeneous communities
- Had lower confidence in local government, community leaders, and local media
- Were less likely to vote
- Volunteered less
- Gave less to charity
- Worked on fewer community projects
- Had fewer close friends
- Reported lower quality of life and lower happiness
- Spent more time watching television
The specific trust numbers tell the story:
| Community Type | "Trust neighbors a lot" |
|---|---|
| Homogeneous (Dakotas, rural New England) | 70-80% |
| Diverse (Los Angeles, San Francisco) | ~30% |
Putnam's own summary of the findings:
> "Those in more diverse communities tend to distrust their neighbors, regardless of the color of their skin, to withdraw even from close friends, to expect the worst from their community and its leaders, to volunteer less, give less to charity and work on community projects less often, to register to vote less, to agitate for social reform more but have less faith that they can actually make a difference, and to huddle unhappily in front of the television."
The crucial word in that quote is "regardless." Putnam didn't find that diversity made people distrust those who were different. He found it made them distrust *everyone* - including their own neighbors who looked like them. Diversity didn't create in-group/out-group conflict. It created isolation. He called it "hunkering down."
The Cover-Up
Putnam collected the data in 2000. He released a preliminary summary in 2001. Then he went quiet.
For six years, one of the most important findings in modern social science sat unpublished. When the *Financial Times* asked him about the delay in October 2006, Putnam said he had "delayed publishing his research until he could develop proposals to compensate for the negative effects of diversity."
He said publishing the raw data without solutions "would have been irresponsible."
The paper finally appeared in June 2007 in *Scandinavian Political Studies* - not in a top American journal, but in an obscure Scandinavian one. When the *Boston Globe* covered it in August 2007, it made national news.
As *City Journal* noted: "Academics aren't supposed to withhold negative data until they can suggest antidotes to their findings."
Putnam himself is a political liberal who supports immigration. He wasn't hiding his data because he had an agenda against diversity. He was hiding it because the data contradicted what he believed, and he didn't want it used by people he disagreed with.
That tells you everything about how this topic is handled in American intellectual life.
The Variables That Don't Explain It Away
Critics immediately tried to explain away Putnam's findings. He anticipated this and tested for every alternative explanation:
Income inequality? No. The trust deficit remained after controlling for income.
Crime rates? No. The trust deficit remained in low-crime diverse areas.
Age differences? No. People born in the 1920s and the 1970s showed the same patterns.
Community wealth? No. Rich diverse communities showed the same trust deficit as poor ones.
The diversity-trust connection was robust across every demographic variable Putnam could test.
Assimilation: Then vs. Now
If diversity reduces trust, the obvious policy question is: can assimilation fix it? Can diverse communities become cohesive over time?
The historical answer is yes - but only under specific conditions. And those conditions no longer exist.
The Pause That Made Assimilation Work
From 1924 to 1965, America effectively closed the door on mass immigration. The foreign-born share of the population fell from 13% to 4.7% - the lowest in American history.
During those 40 years, something remarkable happened. The 14 million immigrants already in America in 1920 assimilated. Their children and grandchildren learned English, intermarried, and became simply "American." By 1970, the Census Bureau stopped asking which language people spoke at home because the answer was overwhelmingly English.
Recent research confirms the mechanism. A 2024 study found that immigrants from countries most affected by the 1924 quotas were more likely to become naturalized citizens and more likely to marry someone born in the United States. The authors concluded: "The Immigrant Exclusion Act hastened the assimilation of already-landed immigrant men."
The pause worked because it removed the one thing that prevents assimilation: continuous reinforcement of the old identity through new arrivals.
What's Different Now
Today, there is no pause. Immigration runs at over 1 million legal admissions per year, plus millions more unauthorized. The foreign-born population has risen from 9.6 million in 1970 to over 51 million today.
The language data shows the effect:
| Year | Non-English speakers at home |
|---|---|
| 1980 | 23.1 million |
| 2019 | 67.8 million |
Nearly 68 million Americans speak a language other than English at home - triple the 1980 figure. About 25 million are classified as having Limited English Proficiency.
By the third generation, language shift does happen. Third-generation Mexican-Americans away from border areas are unlikely to be bilingual. Third-generation Asian-Americans speak only English at home at rates above 90%.
But here's the problem: continuous immigration means there's always a massive first generation that hasn't made that shift yet. And in areas of concentrated immigration, English can become the minority language entirely. In Miami-Dade County, only 28.1% of residents speak English at home. Spanish is the language of business, media, and daily life.
That's not assimilation. That's transplantation.
The Technology Problem
Previous waves of immigrants were cut off from their home countries. An Italian arriving in New York in 1905 might write letters. He might send money by postal order. But he couldn't video-call his family in Naples, scroll through Italian-language social media, or watch Italian television.
Today's immigrants can maintain a near-complete connection to their homeland while living in America. WhatsApp groups with family back home. Satellite TV in their native language. Social media feeds that reinforce the old identity rather than building a new one.
As *National Review* documented in 2025, the internet has "eased the pressure to integrate, allowing immigrants to keep one foot in the old world." Research shows social media functions as both a bridge and a barrier - connecting immigrants to their new country but also enabling permanent cultural separation.
The pressure to assimilate has always come from necessity. You learned English because you had to. You adopted American customs because your neighbors expected it. You married outside your ethnic group because that's who was available.
Technology removes that necessity. You can now live in America while never really leaving home.
Eric Kaufmann and the Pace Problem
University of London professor Eric Kaufmann has studied immigration and social cohesion across the Western world. His central insight is about pace.
> "Deeper integration - intermarriage, shared norms, trust between neighbours - unfolds across generations. When immigration levels outpace those slower processes, societies don't collapse. They thin. Trust weakens, social attachment erodes, and communities become more transactional, less rooted."
Kaufmann's research shows that when people are told assimilation works - that the ethnic majority's culture is preserved and strengthened by integration - their opposition to immigration drops significantly. The fear isn't of immigrants. It's of unassimilated immigration - of permanent demographic change without cultural convergence.
His survey experiments found that leave voters in Britain's Brexit referendum were driven primarily by identity concerns, not economics. They were "less price-elastic" than remain voters - meaning they'd accept economic costs to reduce immigration. That's not xenophobia. That's a rational preference for social cohesion over GDP growth.
The Numbers That Matter
Intermarriage: A Key Assimilation Metric
Intermarriage rates show assimilation working - but slowly, and unevenly:
| Group | Intermarriage rate (newlyweds) |
|---|---|
| Asian (US-born) | 46% |
| Hispanic (US-born) | 39% |
| Black | 18% |
| White | 11% |
| Asian (foreign-born) | 24% |
| Hispanic (foreign-born) | 15% |
The gap between foreign-born and US-born tells the story. US-born Asians intermarry at 46% - nearly double the foreign-born rate. US-born Hispanics intermarry at 39% - more than double. Assimilation through intermarriage works, but it takes a generation, and it requires the first generation to stop growing.
Civic Participation
Naturalized citizens vote at lower rates than the native-born overall (54% vs 62% in 2016), though this varies by group. Structural barriers - navigating institutions, party mobilization - blunt participation even among educated immigrants.
Military service tells a more positive story. About 20% of all Medal of Honor recipients were born abroad. Immigrant soldiers have lower dropout rates than native-born soldiers (18.2% vs 31.9% after 48 months). But immigrant veterans are only 4.5% of the total - far below the foreign-born share of the population.
Second-Generation Mobility
The economic data on second-generation immigrants is genuinely positive:
| Metric | 1st Generation | 2nd Generation |
|---|---|---|
| Median household income | $46,000 | $58,000 |
| College degree | 29% | 36% |
| Homeownership | 51% | 64% |
| Poverty rate | 18% | 11% |
Second-generation immigrants earn more, own more homes, and are better educated than their parents. This is assimilation working.
But sociologist Alejandro Portes has documented a counter-pattern: "segmented assimilation," where some second-generation groups assimilate downward - into disadvantaged communities rather than the middle class. Risk factors include racial discrimination, disadvantaged neighborhoods, and lack of economic mobility ladders.
Upward assimilation is not guaranteed. It depends on the receiving community having the social infrastructure to support it - and that infrastructure erodes when communities lose trust and cohesion, exactly as Putnam documented.
Denmark's Warning
Denmark has gone further than any Western country in trying to engineer assimilation. Its 2018 "Ghetto Plan" designated neighborhoods where over 50% of residents were non-Western immigrants and required children from those areas to attend 25 hours per week of "Danish values" immersion starting at age one.
The results are instructive. Denmark's restrictive integration policies actually improved employment outcomes for recent non-Western immigrants. But even with aggressive state intervention, a "significant and persistent income gap" remained between non-Western immigrants and native Danes across generations.
Denmark's lesson: forced assimilation produces better outcomes than no assimilation policy at all, but it cannot fully substitute for the natural assimilation that happens when immigration is low enough for communities to absorb newcomers organically.
What Assimilation Actually Requires
The research points to specific conditions that make assimilation work:
1. A pause in new arrivals
The 1924-1965 restriction proved that immigrants assimilate faster when new arrivals stop reinforcing the old identity. Continuous immigration at current levels makes full assimilation mathematically impossible - the first generation is always larger than the assimilating generations.
2. Geographic dispersal
Ethnic enclaves slow assimilation. When immigrants cluster in communities where their language and culture dominate, the pressure to adopt English and American customs disappears. Miami-Dade's 28% English-speaking rate is the result.
3. Economic opportunity in the mainstream
Immigrants assimilate into the middle class when economic mobility is available. The 1924-1965 period combined immigration restriction with a booming economy - tight labor markets, strong unions, rising wages. Today's economy offers less mobility for working-class immigrants.
4. Social expectation
Assimilation requires the host society to expect it. When multiculturalism replaces the expectation of assimilation - when schools teach "cultural preservation" instead of "becoming American" - the social pressure that drives integration disappears.
5. Time
Stanford research shows immigrants erase about half the cultural gap with natives after 20 years. Full assimilation takes two to three generations. At current immigration levels, America never gets that time - each decade brings millions more first-generation immigrants before the previous wave has fully integrated.
The Core Problem
Putnam's research reveals a fundamental tension in immigration policy. Diversity reduces social trust and civic participation. Assimilation restores it - but assimilation takes generations and requires conditions that current policy makes impossible.
America is adding diversity faster than it can assimilate it. The result is exactly what Putnam found: people hunker down. They trust less. They participate less. They withdraw into their own lives and watch television.
This isn't an argument against immigration. It's an argument against the *pace* of immigration. The 1924-1965 pause didn't end immigration permanently - it created the conditions for the most successful assimilation in American history. The result was a cohesive, high-trust society where Italian-Americans, Irish-Americans, Polish-Americans, and Jewish-Americans became simply Americans.
That can happen again. But only if America does what it did before: slow down, absorb, and integrate the people who are already here before adding millions more.
Robert Putnam found the data. He tried to hide it. The data doesn't care.
Sources
Putnam's Research
- Putnam, Robert D. "E Pluribus Unum: Diversity and Community in the Twenty-first Century." *Scandinavian Political Studies* 30(2), 2007 - Original study
- Harvard Weatherhead Center: "The Downside of Diversity" - Institutional summary
- City Journal: "Bowling With Our Own" - Analysis of Putnam's delay in publishing
Language Data
- US Census Bureau: Languages We Speak in the United States - 67.8 million non-English home language speakers
- Census QuickFacts: Miami-Dade County - English minority at 28.1%
- Migration Policy Institute: Bilingualism Persists, English Still Dominates - Generational language shift data
Assimilation Research
- Stanford SIEPR: What History Tells Us About Assimilation of Immigrants - Historical vs modern assimilation comparison
- ScienceDirect: The Immigrant Exclusion Act and Assimilation Outcomes (2024) - 1924 Act accelerated assimilation
- Pew Research: Second-Generation Americans (2013) - Self-identification and economic mobility
Intermarriage
- Pew Research: Intermarriage in the US (2017) - Rates by race, ethnicity, and nativity
Kaufmann's Research
- Kaufmann, Eric. *Whiteshift: Populism, Immigration, and the Future of White Majorities* (2019) - Immigration and social cohesion
Denmark
- Oxford Review of Economic Policy: Immigration and Integration in Scandinavia - Danish integration policy outcomes
Technology and Assimilation
- National Review: How the Internet Broke Assimilation (2025) - Technology reducing assimilation pressure
Civic Participation
- Migration Policy Institute: Immigrant Veterans in the United States - Military service data
- Pew Research: Naturalized Citizens as Voters (2020) - Voter turnout data
Image Credit
- Photo by Nathan Dumlao on Unsplash