The Melting Pot Is Broken

How slowing down immigration could help us build a more cohesive and humane society.

Illustration by Mark Stamaty

I’m obsessed with the idea of America as a melting pot. We owe the term to Israel Zangwill, who wrote a mostly forgotten play of the same name in 1908. (I remember it like it was yesterday …) The play was forgettable, but the snazzy metaphor was not. “Yes, East and West, and North and South, the palm and the pine, the pole and the equator, the crescent and the cross—how the great Alchemist melts and fuses them with his purging flame!” How indeed!

Much to my chagrin, the melting pot metaphor has been out of style for decades. Way back in 1963, Nathan Glazer and Daniel Patrick Moynihan famously lamented that in New York City, at least, we were already “beyond the melting pot,” as various ethnic groups maintained their distinctiveness generation after generation. In 1972, Michael Novak wrote Rise of the Unmeltable Ethnics, a quirky defense of ethnic particularity and lower-middle-class cultural conservatism against the (alleged) homogenizing influence of academic liberalism. Then, of course, there was the rise of multiculturalism, which held that diversity ought to be celebrated, and that there was something both tragic and unjust about the expectation that members of ethnic minority groups ought to surrender what separates them from the cultural mainstream. America is no longer a melting pot, we’re told. It’s more like a salad bowl, full of ingredients that retain their unique flavors.

As delicious as this multiethnic salad sounds—more croutons, please—I continue to be the melting pot’s biggest fan. As a consequence, I’ve gone from being a rah-rah enthusiast for mass immigration to one who is more skeptical of its virtues. That’s because I think the melting and fusing of different ethnic groups is essential to building a more cohesive and humane society, and that slowing down immigration would help this process along.

The biggest challenge we face in the United States, in my view, is a lack of togetherness. I realize that this sounds like a line from a cheesy love ballad, but bear with me. There is a vast gulf separating those who belong to networks of family members, friends, and acquaintances that grant access to valuable social goods and those who don’t belong to these networks. What kind of social goods do I have in mind? Money is the obvious one, but so is know-how. Most people learn to make their way in the world from parents or peers. Think about every job opportunity you’ve ever had, or how you learned to navigate any big, impersonal bureaucracy. The insider knowledge that was imparted to you was not available to just anyone—you, and those close to you, had privileged access to it, and whether you like it or not, this privileged access has played a central role in your successes.

This isn’t simply a racial inequality. There are plenty of ultra-connected people from minority backgrounds and there are plenty of whites who are outsiders in every socially relevant respect. But race is often used as a rough proxy for this kind of inequality because it is generally true that an American who can trace her roots back to the Mayflower will have a bigger, richer, and let’s say more privileged network than a descendant of slaves, or a recent immigrant with little in the way of formal education.

Race also allows us to identify this phenomenon in the data. Last summer, for example, Elizabeth Ananat, Shihe Fu, and Stephen L. Ross released an important working paper, which found that the gap in average wages between whites and blacks gets bigger as the size of the city in which they live gets bigger. Their (fascinating) explanation for this finding is that while workers benefit from the knowledge spillovers that come from living and working in a place where there is a higher concentration of people doing a certain kind of job, these spillovers tend to be bounded by race. That is, blacks have fewer same-race peers than whites from whom they can learn new skills and suss out new opportunities, and so whites gain more insider knowledge with each passing year, which in turn allows them to earn more money. There are, of course, some blacks (and Asians and Latinos) who are plugged into “white” networks, which aren’t always lily-white, especially in places like Silicon Valley and Wall Street. As a general rule, however, race makes a big difference.

Instead of calling this racial inequality, you could call it roots inequality. Who among us has the firmest, deepest roots in American life, which can allow us to stumble and make mistakes while still being able to depend on loved ones who can lend a hand? This isn’t so much about how long one’s ancestors have been in the country. Rather, it is about the solidity of your connections to other people who themselves stand on solid ground, and who can afford to offer help of all kinds, monetary and otherwise, when you need it.

Earlier this week, Richard V. Reeves and Joanna Venator of the Brookings Institution observed that educated people tend to have educated parents. This news will probably not shock you. What is interesting is that the transmission of education advantage from one generation to the next seems to be even stronger than the transmission of economic advantage, at least for the children of the most educated parents.

Reeves and Venator suggest that a big part of the story is the rise of positive assortative mating, the phenomenon in which people with similar levels of education marry each other. This contributes to household income inequality, predictably enough. Yet it has another, subtler effect. When a college-educated adult has a child with a non-college-educated adult, she can pass along her tacit knowledge about what it takes to make it through the educational system even if her partner isn’t in the same boat. When two non-college-educated adults have a child, that young person is going to have a much tougher time working the system. It’s no coincidence that the high-achieving, low-income students who never make it to selective schools are disproportionately drawn from the children of non-college-educated parents, and from rural areas and small cities. These kids might have high grades and test scores, but they don’t belong to the networks that grease the wheels of upward mobility.

What does any of this have to do with immigration? Americans are, for obvious historical reasons, deeply romantic about the immigrant experience. More than one-tenth of Americans, like me, are the children of immigrants, and there are many more third- and fourth- and fifth-generation Americans raised with heady stories about flinty ancestors. The truth is that some immigrants are poised for great success in a society like ours, and others will have a tough time making their way into the middle class. If we accept that we have a collective responsibility for the well-being of every member of our society, as I think we should, it makes sense to select immigrants who have at least a fighting chance of making it.

When thinking about which immigrants do have that fighting chance, it’s important to recognize that the economic realities of the first decades of the last century are profoundly different from those of the first decades of this one. That earlier era was one of labor scarcity, when people with limited skills could climb the economic ladder by doing dangerous, strenuous work. In today’s economy, people with limited skills are finding that market wages are stagnant or even falling.

The educational gap between immigrants and natives in the 1900s was modest when compared to the yawning gap that separates them today. Earlier this year, the OECD, the consortium of the world’s rich democracies, released its latest assessment of adult skills. The survey divvies up people from age of 16 to 65 into several levels of proficiency in literacy, numeracy, and, my personal favorite, “problem-solving in technology-rich environments.” A disturbingly high 18 percent of Americans fell in Level 1 or below (the lowest proficiency levels) in literacy, 30 percent in numeracy, and 61 percent in problem solving. The numbers were far worse for U.S. immigrants, 40 percent of whom fell in Level 1 or below on literacy, 48 percent on numeracy, and 76 percent in problem solving. Given the transmission of educational advantage from one generation to the next, it is a safe bet that second-generation Americans like me are overrepresented among the lowest native-born performers.

The fact that our immigration policy has in recent decades tended to lower the average skill level of our workforce isn’t news, or at least it shouldn’t be. What we often fail to appreciate is how the fact that so many immigrants have such limited skills interacts with the rise of positive assortative mating and opportunity hoarding. College-educated people tend to marry other college-educated people, and we are all more likely to share opportunities with people in our networks than with random strangers. Those realities mean that newcomers to our society with below-average skill levels, as well as their children and even their grandchildren, are going to have a hard time getting past the bottom rungs of American society. If you believe Gregory Clark, an iconoclastic economist at UC–Davis, it might take even more than three generations for the descendants of less-skilled immigrants to reach an average level of social status.

Legalizing large numbers of unauthorized immigrants will definitely help them attain that social status. Yet it won’t change the fact that even under the best circumstances, the wages commanded by people with less than a high school diploma tend to be very low, and the social connections they can draw upon are usually limited to other people facing similar challenges. Moreover, while the best evidence we have finds that less-skilled immigration doesn’t have a negative effect on the wages of less-skilled natives, it does have a substantial negative effect on the wages of less-skilled immigrants already living in the U.S. These are precisely the people who have the weakest social connections to other Americans, and who need all the help they can get to put down roots in this country.

Which brings me back to the melting pot. There is an alternative to allowing today’s less-skilled immigrants and their descendants to form the bedrock of an ever-expanding underclass. There is a way to help poor members of our foreign-born population form the social connections they will need to move from the margins of American society to the mainstream. What we need to do is limit the future influx of less-skilled immigrants.

Zangwill wrote The Melting Pot in the 1900s, but the true heyday of cultural amalgamation, among whites at least, started in the 1920s, when the United States shut off the spigot of European immigration for three generations. Mary C. Waters, the Harvard sociologist best known for her work on Caribbean immigrant identity, has written that “in the absence of appreciable numbers of new arrivals successive generations of acculturated Americans, not unassimilated greenhorns, became the majority among the new ethnics.” This in turn meant that for second-, third-, and fourth-generation Italian and Polish Americans, “ethnicity became less intense and increasingly intermittent, voluntary, even recreational.” Ethnic enclaves slowly faded away and the descendants of European immigrants intermarried their way into the mainstream.

Now, in contrast, ethnic enclaves in our biggest cities are flourishing, and expanding their footprints—the booming Chinatowns of Brooklyn and Queens stand in stark contrast to Manhattan’s fading Little Italy. Those Chinese Americans who, in effect, “exit” their ethnic groups by assimilating and intermarrying and leaving Chinatown behind are, Waters explains, replaced by new arrivals who are content to lead a Chinese life on American soil. Distinct ethnic cultures are continually being replenished through immigration. For Waters, this means that the story of assimilation—the third-generation Chinese American who gets an MFA at Yale, marries an Anglo investment banker, and who hardly ever sets foot in Chinatown—is obscured by this parallel story of ethnic replenishment. It is also true, however, that ethnic replenishment will tend to limit assimilation and intermarriage, as most people prefer to marry and socialize with members of their own group.

So if we want the Mexican and Bangladeshi immigrants of our time to fare as well as the Italian and Polish immigrants of yesteryear, we need to do two things. First, we need to spend a considerable amount of money to upgrade their skills and those of their children, as the world has grown less kind to those who make a living by the sweat of their brow. Because public money is scarce, this is a good reason to limit the influx of people who will need this kind of expensive, extensive support to become full participants in American society. Second, we need to recognize that a continual stream of immigration tends to keep minority ethnic groups culturally isolated, which is yet another reason to slow things down. No, this won’t suddenly mean that poor immigrants will become rich, and that well-heeled insiders will stop hoarding opportunities. But it will give us the time we need to knit America’s newcomers into our national community.