When Americans hear the term “melting pot,” they often think of New York City: tenement blocks, Ellis Island, the Lower East Side, a montage of pushcart vendors becoming small business owners. Our national mythology about assimilation is rooted in the urban and Northeastern experience, suggesting that large, multicultural cities best assimilated immigrants.
Census data from 1880–1930 tell a different story. Northeastern states with many immigrants, such as New York, Massachusetts, and New Jersey, underperformed on the most concrete measure of assimilation: “outmarriage,” or marriage across ethnic or immigrant-generational lines. By contrast, states with fewer immigrants, such as Wyoming, Montana, and Mississippi, had higher outmarriage rates. The melting pot worked, but not where or how we expected—and today, the factors that made that assimilation possible are weakening.
Finally, a reason to check your email.
Sign up for our free newsletter today.
In a recent Niskanen Center report, I examined the marriage patterns of the melting-pot generation. Using IPUMS census microdata, I tracked outmarriage rates of second-generation Americans living between 1880 and 1930, breaking down the data by ethnic group and state. I also adjusted for each group’s “concentration”—its share of the state population—to reflect how groups were not distributed evenly across the country. A German American in Wyoming had far fewer potential co-ethnic spouses than did a German American in Wisconsin.
My report focuses on marriage patterns because they better demonstrate assimilation than other metrics, such as English fluency or occupational mobility. Children of immigrants could have learned English or held jobs alongside third-generation Americans without having meaningfully assimilated. Marriage outside of one’s ethnic group, by contrast, is a hard test: it proves intimate relationships have crossed ethnic lines and ensures children grow up in a fundamentally different culture.
My results inverted the Ellis Island narrative. It was the Western and Deep Southern states—not those in the Northeast—that led the nation in outmarriage rates. Wyoming, Oregon, and Washington had rates 15 to 27 percentage points above the concentration-expected rate. Georgia, Mississippi, and Alabama outperformed concentration expectations by between 10 and 16 percentage points. New York, Massachusetts, and New Jersey, by contrast, recorded outmarriage rates between 2 and 8 percentage points below the expected threshold.
Regional factors shaped these trends. In the West, thin ethnic institutional networks and a “new beginnings”–oriented culture loosened communal pressures and left second-generation Americans freer to choose their spouses. In the South, the children of immigrants could assimilate into a region deeply rooted in place and history. In both cases, the result was faster integration into the American mainstream than in the Northeast, the region most associated with the immigrant experience in American memory.
The upper Midwest’s culture, by contrast, stifled integration. Scandinavian and German communities in Wisconsin, Minnesota, and North Dakota maintained low outmarriage rates even at moderate concentration levels—11 to 15 points below what demographics predicted. These were not simply states where co-ethnics happened to live near each other. These Scandinavian and German communities built ethnic ecosystems, with Lutheran churches conducting services in Norwegian, widely circulated German-language newspapers, and ethnic social halls organizing community life from cradle to grave.
New England featured its own version of this dynamic. Massachusetts, Rhode Island, and the northern New England states had relatively low outmarriage rates. Irish parishes, Italian mutual aid societies, and French-Canadian parochial schools gave these communities the institutional infrastructure to maintain ethnic boundaries even when they were not especially large in population terms.
Across all regions, concentration alone explains only about 7 percent of the variation in outmarriage. The decisive factor was institutional power: whether an ethnic community had the churches, schools, newspapers, and cultural organizations to sustain a separate social world. Where those institutions existed, ethnic boundaries held. Where they did not, assimilation happened within a single generation.
Some groups were more likely than others to marry outside their community. The children of Scottish immigrants, for example, had an outmarriage rate of 87 percent. Italian Americans, by contrast, had an outmarriage rate of 25 percent.
The most important event to increase Italians’ and other groups’ likelihood of marrying outside their community was an act of Congress. The Immigration Act of 1924 curtailed immigration from Southern and Eastern Europe, which in turn cut off the replenishment pipeline that had sustained ethnic enclaves for decades. Without fresh arrivals, enclaves thinned. The children of immigrants found themselves with fewer co-ethnics to marry, fewer institutions operating in their parents’ language, and greater incentives to orient toward the English-speaking mainstream.
Strong non-ethnic local institutions accelerated the process further. Utah was the number-one state for cross-ethnic second-generation marriage, scoring 16 percentage points above what concentration predicted. One reason for that was the presence of the Church of Jesus Christ of Latter-day Saints, which recruited converts from across Northern Europe and placed them in communities where religious identity superseded national origin. The results are visible in the data: Swedish Americans’ outmarriage rate was 85 percent in Utah but just 45 percent in Minnesota, despite their being less than half as concentrated in Utah. German Americans’ outmarriage rate was 78 percent in Utah compared with 37 percent in Minnesota and 29 percent in Wisconsin. A shared identity that transcended ethnicity was more powerful than any other factor in the data.
Geography mattered independently of institutions. Mexican Americans outmarried at just 16 percent nationally—the lowest rate of any group in the dataset. But 30 percent of Mexican Americans in California married outside their community, compared with just 12 percent in Texas. This pattern recurs across virtually every group: outmarriage rates climbed as second-generation Americans moved west, away from the established enclaves and social pressures of their parents’ communities. The West offered thinner ethnic networks and fewer constraints on individual choice, conditions that, whatever their cultural causes, produced measurably faster assimilation.
Three conditions emerge from the data as essential to the melting pot’s success: enclave thinning through reduced immigration, a strong shared regional or local identity that transcended ethnicity, and regional cultures that loosened ethnic-communal pressures. All three have weakened since 1930.
Continuous inflows from Latin America have replenished enclaves, with each successive wave resetting the figurative assimilation clock. Each new wave of arrivals reestablishes the institutional infrastructure that the historical data show delays integration, a dynamic that Manhattan Institute president Reihan Salam identified in Melting Pot or Civil War? as central to explaining why some groups assimilate slower than others. That infrastructure is visible today in Univision, a leading American broadcast network that operates in Spanish, and in the more than 1,500 Mexican hometown associations registered across the country, organizations whose purpose is maintaining ties to specific municipalities back home.
While settlement patterns continue to cluster immigrants in established co-ethnic communities rather than scattering them, American culture has also become more nationalized in ways that undercut the regional and local factors that once encouraged assimilation. The children of today’s immigrants no longer benefit from the frontier culture that produced rapid assimilation in the Northwest or the staunch regionalism that encouraged identification with the mainstream in the Deep South. Instead, the children of immigrants everywhere have been thrust into a culture that downplays sources of identity not tied to ethnic origin and race.
This is the result of a decades-long campaign by the Left to recast assimilation as cultural erasure rather than a precondition for equality. Ironically, many of our present problems stem not from the children of immigrants refusing to assimilate but from those children assimilating into the wrong thing: a victim-centered worldview, cultivated in universities and professional institutions, that treats foreign and preferably oppressed origin as a form of social currency.
Consider New York City Mayor Zohran Mamdani’s wife, Rama Duwaji, who apparently lists herself being as “from Damascus” on social media despite having been born and largely raised in the United States. Duwaji is almost certainly not doing so under pressure from her Syrian family. She is doing so because, in the upper-middle-class circles where the most successful children of immigrants tend to land, signaling distance from the American mainstream confers more status than embracing it.
The melting pot was a hard-won achievement that rested on conditions conspicuously absent from our present approach to immigration. To be sure, assimilation is almost certainly possible for today’s immigrants—the example of Italian Americans alone proves that even highly insular groups can become part of the American mainstream.
But assimilation is a choice. If we want the children of today’s migrants to assimilate, we will have to choose dispersal over concentration, civic identity over multiculturalism, and a national culture worth insisting upon.
Photo by CHARLY TRIBALLEAU / AFP via Getty Images