Politics Foreign Affairs Culture Fellows Program

Harvard’s George J. Borjas

Adding up the cost of uncontrolled immigration

“Trump started the campaign with some comments about immigration, if you recall, last year. A really good question to ask, which nobody has really asked, is would he have gotten traction if he hadn’t shocked the system that way so early on? What he said, you can disagree with it strongly. But despite that, it really provided an incredible shock by introducing into the debate something people don’t usually talk about very often.”

Those are the accented, dispassionate words, spoken the Friday after the election, of a Hispanic immigrant and Harvard professor—specifically, George J. Borjas, an economist who specializes in immigration research. For decades Borjas has been shocking the system in his own way, arguing—carefully, with the support of intricate statistical analysis—that immigration comes with tradeoffs, particularly reduced wages for the native workers who compete with immigrants.

To most academics, that’s heresy. To many others it might sound like common sense—and at last we can experience Borjas’s ideas in a widely accessible form. After focusing for years on academic research, Borjas is bringing his findings directly to the public through a new book called We Wanted Workers, a rejuvenated blog (gborjas.org), and the occasional op-ed in mainstream publications such as Politico.

But for Borjas the elevated public profile extracts a price. “I’m not the kind of person that wants to be in the media every single day of my life,” he says. “I go home, I have a very quiet life. So for me to take the step is a little bit of a stretch.”


It is hard to call Borjas an outsider. He stands, after all, at the top of his field. He publishes in the best journals, presents his work at leading academic conferences, is consulted when major scientific organizations (e.g., the National Academies) summarize the state of knowledge in his field. But it’s equally hard to say he fits in. He has both a background and a perspective that are unusual among his peers.

Borjas was born in 1950 in Cuba, where his parents owned a garment factory. Nine years later, Fidel Castro took over the country and confiscated that factory. In 1961, Borjas’s father passed away, and early one morning shortly thereafter, he listened as planes attacked a military base near Havana: the failed Bay of Pigs invasion. Borjas was soon transferred from a Catholic school to a “revolutionary” one. There, he writes in We Wanted Workers, he proved adept at memorizing propaganda—and forgetting it once he had passed the test.

It was here that Borjas developed a deep distrust of ideology and expert opinion—and, almost paradoxically, where he found his own place on the ideological spectrum. “I’m a conservative,” he says bluntly when I ask. “Very conservative. Having grown up in Cuba, it’s impossible not to be. I’ve seen firsthand what political correctness does to the public debate. I’ve seen firsthand what bad economic management does. I’ve seen firsthand what political persecution is. Believe me, I’ve seen all of this as a kid.”

Borjas and his mother escaped to Miami in 1962, spending some time in a heavily Cuban neighborhood there before decamping to Hoboken, N.J. The immigrant experience, too, left its mark on his outlook. “Most economists would say the objective function of immigration should be to maximize wealth,” he says. “Maybe it’s my own immigrant background that makes me see things a little differently. Immigration has also played an important role in history in providing hope to people around the world. I actually believe that.”

But he continues: “At the same time, you don’t want to lose too much.”


Perhaps oddly for someone who gained immensely from moving from one country to another, Borjas has spent much of his career trying to answer the questions of who loses from immigration and how much. He didn’t go into the field to make a point; he says that when he began his work in the 1980s, there wasn’t much of a policy debate to speak of, anyway.

“If you look at my main papers, they’re really methodological papers: how do you measure this, how do you estimate that parameter, and things like that. That’s always been the thing that attracts my interest as an economist,” he says. “I hate to say this, but I’m a little bit of a geek. I like to play with mathematical models, I like looking at data, and that’s what I’m doing for a living.”

Immigration is a good topic for someone of this inclination—it is incredibly difficult to study, because immigrants, native workers, and economic conditions interact in all sorts of complex ways that can’t easily be captured through the brute force of math. But complexity is a bug, not a feature, for someone trying to write for a general audience.

For an early draft of We Wanted Workers, Borjas did his best to summarize what he does in plain English. It didn’t quite work. “People read it and said, ‘Can you do something like Mariel?’” he recalls. They were referring to one of the most widely cited immigration studies ever, a piece of research that Borjas still calls one of his favorites. He decided to reread it.

In 1980, when a group of Cubans drove a bus through the gate of the Peruvian embassy in Havana and demanded asylum, the Peruvian ambassador refused to return them to Cuban soil. Thousands more soon flooded into the embassy, and Fidel Castro abruptly announced that anyone who wanted to leave Cuba could do so through the Port of Mariel. More than 100,000 “Marielitos” ended up in the U.S. that year—mainly working-class males, most staying in Miami.

It’s a textbook example of a “natural experiment”: if immigrants drive down natives’ wages, natives’ wages should have gone down in post-1980 Miami. But the key study of the boatlift, conducted by the economist David Card ten years after the event, found that that hadn’t happened. “The Mariel immigrants increased the Miami labor force by 7%, and the percentage increase in labor supply to less-skilled occupations and industries was even greater because most of the immigrants were relatively unskilled,” the study noted. “Nevertheless, the Mariel influx appears to have had virtually no effect on the wages or unemployment rates of less-skilled workers, even among Cubans who had immigrated earlier.”

Looking back at the study a quarter-century later, Borjas started to notice aspects of it that he would have done differently—Card had failed to look specifically at high-school dropouts, for example, and he had chosen questionable cities with which to compare Miami. Borjas decided to “see what I would find if I started from scratch,” he says. “Within an hour I was getting the graph.”

This graph, that is:


What Borjas did was so simple that even a journalism major (such as yours truly) can retrace his steps. Using publicly available Census data, he calculated the average wage of the native workers most likely to suffer from competition with the Marielitos: non-Hispanic male high-school dropouts, age 25–59, in Miami. Then he calculated the average wage of similar workers everywhere else in the U.S. And then he compared the two trendlines, adjusting for inflation and averaging each three-year period (to “smooth” the random year-to-year fluctuations).

He revealed a jaw-dropping falloff in wages for the most economically vulnerable native workers, something that had been sitting undiscovered in the Census data for decades. Rather than saving it for his book, he wrote a 60-page academic paper with a slew of additional analyses and statistical tests, circulated drafts to his peers, and eventually published it in the same journal that had featured Card’s original study.


Not surprisingly, the response came swiftly. Nor was it surprising who wrote it: University of California-Davis economist Giovanni Peri. Borjas balked when I referred to Peri as his “rival,” but the two have a history of publishing quite different findings.

Borjas believes that immigration has severe “distributional consequences” for natives: those who compete with immigrants lose; those who employ immigrants benefit from cheaper labor; and everyone else enjoys slightly lower prices. Peri, by contrast, argues that immigration benefits virtually everyone. Immigrants and natives tend to have different skill sets and thus often complement each other, rather than competing, in the workforce.

The divide is best illustrated by looking at how the two men see immigrants at the extremes of the skill distribution. The complementarity argument is especially strong when it comes to highly skilled workers: it’s easy to imagine that foreign mathematicians and native mathematicians could learn a lot from each other, and become more productive together than they were apart. But Borjas argues that even in this ideal scenario, immigrant competition is a net loss for native workers. After the fall of the Soviet Union, for instance, hundreds of immigrant mathematicians came to the U.S.; native mathematicians saw their unemployment rate rise and published fewer papers.

The complementarity argument is especially weak, by contrast, in the case of low-skill workers: why on earth would native high-school dropouts benefit from an influx of foreign high-school dropouts? But Peri argues that even high-school dropouts have benefited from immigration on balance, at least in the modern United States. “Even at a low level of education, immigrants are concentrated in a small number of occupations,” Peri says, especially manually intensive jobs in fields like construction and agriculture. “These are the same jobs that young Americans have been leaving in massive numbers.”

In Peri’s telling, when companies hire cheap immigrant labor for these positions, they are able to expand—meaning that they need to hire more natives for other positions. “They are actually doing jobs a little bit more communication-intensive—like mediating or supervising—so knowing the language is important,” he says. “If a firm hires a worker, it then needs a supervisor. If it hires a landscaper, it needs someone to do the accounting.”

Perhaps it’s only natural, then, that Borjas and Peri see things differently even when answering a simple question like, “Did wages go down for Miami’s low-skill workers after 1980?”

The biggest weakness of Borjas’s study is that it’s based on relatively little data. The Census survey he used is quite sizable nationwide, but it’s not so big when restricted to non-Hispanic male high-school dropouts age 25–59 in Miami. In fact, it includes only about 20 people each year who fit that description, though that adds up to 60 when you use three-year averages like Borjas did. Peri and a graduate student, Vasil Yasenov, expanded the focus of the analysis to include all non-Cuban (as opposed to non-Hispanic) men and women without a high-school degree age 16–61, and relied on a larger Census survey. The clear decline in wages disappeared.

The upside to this approach is that there’s more data. The downside is that in broadening the sample, Peri weakens the focus on a consistent set of workers whom we’d expect to be affected by the new arrivals from Cuba. Particularly problematic is the inclusion of women (who tend to hold different jobs than men and were shifting into the workforce in great numbers in the 1980s) and teenagers (including those still in school).

Peri and Borjas have gone back and forth since then, each standing by his original findings, and each claiming that his results hold up even when plausible changes are made to the methodology. (Peri can remove high schoolers or analyze men and women separately without changing his results; Borjas can use a bigger survey.) It’s a pattern familiar to those of us who follow academic debates on hot-button issues.

The exchange also raised the issue of whether such academic debates should simultaneously play out in the media: though Borjas says he made no attempt to publicize his original paper, some news outlets reported his results, and Peri and his coauthor launched their academic response with a complementary Wall Street Journal article essentially saying Borjas had botched his analysis.

Borjas saw that as out of line. “Just imagine if the person writing the Mariel paper last year had been an economist coming right out of school,” he says. “Something like that could destroy their career entirely. They have no way to fight in the public arena against these claims.”

For his part, Peri says the WSJ approached him, not the other way around—and that Borjas, too, has made accusations through the press. “It’s certainly true that he has massively trashed my work on his blog and in magazines,” Peri says, noting an Atlantic article by David Frum accusing Peri of “data mining.” (“Soon after [Peri’s paper] appeared, I called Borjas for comment,” Frum wrote. “He had just finished a reply, he said, and it would be a doozy. And so it is.”)

“I don’t want this to become a battle in the press, but I do think research should inform what we write and say about the impact of immigration,” Peri says. “And the idea that I’m more involved in this than he is—I don’t think this is true.”


There is much more to immigration economics than the question of whether native workers take a hit to their wages. Economists have also invested much effort in quantifying the overall gains from immigration (including gains to the immigrants themselves), and in tracking immigrants’ assimilation over the years in their new countries. Borjas has studied all of this, and in We Wanted Workers he provides readers a glimpse behind the curtain—explaining not just what various studies find, but also the assumptions and data manipulations that are needed to produce each result.

There can be little doubt that immigration produces economic gains. A low-skill worker who moves from Mexico to the United States, for instance, becomes immensely better off, regardless of how he affects native wages. But what happens if we extrapolate from this fact and “imagine there’s no countries,” as one of Borjas’s favorite musicians once said “isn’t hard to do”?

This has been a favorite exercise of libertarian open-borders supporters, who claim—with studies to back them up—that there are “trillion-dollar bills on the sidewalk” that humanity can pick up merely by ending immigration restrictions and letting the third world flood into the first. That brings out the geek in Borjas. “What assumptions are these people making to get all these trillion-dollar bills?” he asks. “What would happen if you reconstruct the model they’re using?”

It turns out some key facts are left out. For one thing, even with open borders, many of the world’s poor wouldn’t move, and thus couldn’t benefit. Humans seem to place an immense amount of value on staying in place, among family, friends, beloved cities, and people who speak the same language. And for another, those who did move would be large enough in number that they might overtake the cultures and institutions that make rich countries rich, rather than magically becoming wealthy themselves simply by virtue of having moved to a wealthy nation.

This raises the question of how quickly immigrants assimilate under various conditions—something Borjas has been fascinated with since his earliest days as an economist. In 1978, a popular study noted that immigrants who’d been in the country for longer out-earned more recent arrivals, implying that immigrants gradually catch up to natives over time. In his very first study of the economics of immigration, published in 1985, Borjas challenged that idea, pointing out that more recent arrivals were less skilled than previous waves of immigrants had been. He argued it was better to track each immigrant wave over time, rather than comparing different waves with each other—and that method revealed far less assimilation.

Indeed, even the Ellis Island immigrants took about a century to fully catch up with natives, and over the last 50 years new immigrants have fallen further and further behind natives in terms of education and wages. Most troubling of all, assimilation—the speed with which they catch up—has slowed over time as well. Two major reasons: the United States is a less rewarding place for low-skill workers than it used to be, and newer immigrants are not improving their skills as much as their predecessors did. Newer arrivals are less likely to learn English, for example, especially if they live in ethnic enclaves.

Importantly, it’s fairly predictable which immigrants will succeed economically and which won’t. If immigrants from a given country arrive with more years of schooling, they also tend to see more wage improvement over time, for example. And the impacts persist over the generations: if immigrants from a given country have high wages, their descendants likely will too.

In other words, if we wanted to, we could select immigrants according to a very good guess at what they and their children will contribute to our country in the coming decades. But should we want to?


Borjas is the first to acknowledge that empirical facts get us only so far. Two people with the same understanding of the facts could very well design completely different immigration plans, simply because they think immigration policy should accomplish different things.

This was especially evident in “Two Immigrants Debate Immigration,” a Reason magazine exchange between Borjas and the libertarian commentator Shikha Dalmia. “What role, if any, does the need to protect individual liberty, mobility rights, and freedom of association play in your thinking on immigration policy?” she challenged him.

“I have never devoted much time to thinking about that,” he responded. Later in the exchange he was much more sweeping in his dismissal of high-minded idealism: “My thinking has never been guided by deep philosophical thoughts.”

Borjas takes a more practical, down-to-earth approach. “I am willing to say let’s have high-skill immigration and let’s have low-skill immigration,” he says. “We’ll have to think about the number, because it’s not clear that the number we have now is the correct number, but whatever the number is, let’s make sure that the gains don’t just go to one group and the losses aren’t just suffered by workers. Let’s make sure everybody has a stake in the system.”

So what does the George Borjas immigration plan look like? “I think if we’re going to talk about changing immigration policy, the first thing we have to do is be sure that the inflow of illegal immigration slows down dramatically,” he says. “Because if anyone can cross the border or overstay a visa, it really makes a mockery of legal immigration policy.”

Borjas suggests requiring employers to use the E-Verify system to ensure their workers aren’t in the country illegally, and otherwise supports a period of “benign neglect” toward those already here: no deportation, but no amnesty either, at least until the enforcement-first strategy proves successful.

He would make drastic changes to legal immigration as well, in particular the H-1B program, through which companies can bring in skilled workers. “Maybe the people who claim it’s terrific should bear some of the cost,” he says. “They’re laughing all the way to the bank. We have to set up a system where they appreciate what it’s like to be a worker who has to train their replacement.” Borjas has suggested charging companies thousands of dollars for each visa; if skilled immigrant workers are truly as valuable as these companies claim, he says, that will be a small price to pay. Another option would be to tax the industries that most benefit from immigration.

Underlying all of this is a sense of “responsible nationalism”: Borjas encourages us to ask ourselves, “Who are you rooting for?” In his view, American immigration policy should exist mainly to benefit Americans.

This isn’t to say that benefits to the immigrants themselves carry no weight. It is to say they are secondary.


Borjas’s impact on the academic debate can hardly be denied. “He’s tried to do what’s uncommon in immigration research—to try to emphasize that there are winners and losers. That’s his most important contribution,” says Steven Camarota, director of research at the Center for Immigration Studies, a think tank that has published Borjas’s work and brands itself as “Low-immigration. Pro-immigrant.” “He’s talking about the losers when few other researchers or policy analysts have given them much thought.”

Peri, despite his fundamental disagreements with Borjas, also doesn’t hesitate to concede the effect Borjas has had. “He’s clearly one of the most productive and one of the most interesting people writing about economics of migration,” he says, praising some of Borjas’s work as “classic.”

“I think he has raised the standard,” Peri adds. “Each one of us knows someone who thinks differently is going to scrutinize and check our papers very carefully.”

thisarticleappearsUntil recently, though, Borjas had not been very active in the public conversation, preferring the academic one. “I don’t see myself as an activist at all,” Borjas says. “I’m not registered a Republican, I’m not registered a Democrat, I’m completely independent. I used to be a registered Republican, and the George W. Bush administration convinced me I wasn’t one of them.”

But could Borjas see himself getting involved in policymaking directly? He already got the call once—from N. Gregory Mankiw, a fellow Harvard economist who headed George W. Bush’s Council of Economic Advisers for two years—and turned it down. “I had three kids and they were 7, 8, 9 years old,” he says. “As you know from my book, I grew up in a one-parent household because my father died early on, and I also moved around a lot. Honestly, I didn’t want my kids to have that kind of lifestyle, and I wanted to be a father who was present all the time.”

He doesn’t regret that decision, but more than a decade has passed. “Would I have done it if things weren’t like that?” he wonders. “I might do it. If I get a phone call this year, I might do it.” 

Robert VerBruggen is managing editor of The American Conservative.



Become a Member today for a growing stake in the conservative movement.
Join here!
Join here