The Immigration-Wage Myth

Subscribe here: Apple Podcasts | Spotify | YouTube | Overcast | Pocket Casts

Why are people frustrated by high levels of immigration? As refugee crises proliferate, this has become a central political question. In order to justify anti-immigration policy or rationalize restrictionist sentiment, commentators and elected officials have repeatedly returned to one hypothesis: Immigration must be bad for American workers.

There’s just one problem: This hypothesis is wrong. Economists have studied this question repeatedly, in a variety of contexts and in every segment of the population, and they have found that the demand effect consistently outweighs the supply effect. Simply put, when immigrants come to a place looking for jobs, they also demand goods and services—thus creating jobs for native-born workers. Immigrants need legal services and taxi drivers; they need groceries and cars. The question has always been which effect is bigger. And the literature has resoundingly answered that the demand effect wins out.

This doesn’t explain away all immigration worries. But it should force politicians to seriously reckon with why xenophobia exists instead of resigning themselves to treating new immigrants as an economic burden, when, for example, they were actually the “sole source of growth in the U.S. working-age population in 2021 and 2022.”

On today’s episode of Good on Paper, I’m joined by my colleague Rogé Karma who recently dove into the economics literature, originally expecting to find some negative effects on wages, only to be repeatedly struck by the truth: Anti-immigration sentiment has no economic justification.

“I think there is a lot of this deep discomfort with non-materialist explanations,” Rogé argues. “I think one of the most revealing things here is that the demographic that is most opposed to immigration are older folks living in rural areas, many of whom are retired. And the people who tend to be most supportive of immigration are working-age people living in big cities where immigrants are more common. So if you thought, Okay, this is a product of the people who immigrants are directly competing with … you would think, Oh, this would show up where the immigrants are, and it doesn’t.”

The following is a transcript of the episode:

Jerusalem Demsas: It’s Election Day, and in place of any exit-poll astrology, we’re going to talk about something that’s been a driving force in the campaign: immigration, specifically the research about the relationship between immigration and wages.

A common line bandied about in politics is that immigration reduces the wages of native-born Americans. It’s most commonly been pushed by restrictionists on the right, like J. D. Vance and Donald Trump.

Demsas: However, I’ve noticed a growing openness to the idea that immigration hurts American workers, not just from longtime restrictionists, but also from Democrats and liberals who are scarred by their loss in 2016 and fretting over the possibility of losing the 2024 election. But sometimes a lot of smoke is just a smoke bomb.

[Music]

This is Good on Paper, a policy show that questions what we really know about popular narratives. I’m your host Jerusalem Demsas. I’m a staff writer here at The Atlantic, and today I’ve asked my colleague Rogé Karma to come on the show. We’re going to talk about a recent deep dive he did into the economics literature on the relationship between immigration and wages.

The common thinking goes: If you increase the supply of labor, then you’ll reduce the price of that labor. If immigrants simply weren’t allowed in, then companies would be forced to pay American workers high wages. It seems so obvious, so why does study after study find this to be so wrong?

[Music]

Rogé, welcome to the show!

Rogé Karma: Thanks so much for having me. It’s great to be here.

Demsas: This is one of those episodes where I’m actually having trouble deciding which narrative is the conventional narrative.

Karma: (Laughs.) It’s because you’ve been steeped in the economic literature for far too long.

Demsas: Exactly. But there’s the conventional wisdom in academic circles that immigrants do not reduce native-born wages. But that’s not, I think, the average person’s perception of this, especially if they’re listening to politicians who, on both sides of the aisle, will be kind of making these arguments.

So I want to walk through the evidence together here because, Rogé, you recently wrote a piece, and you’ve spent a big chunk of time this year diving into the research space and really trying to figure out what’s going on. Like, Where is the evidence actually leading us? And I want to start with the Mariel Boatlift. Can you tell us what that is and then what economist David Card found when he looked into it?

Karma: Of course. And the first thing I will say is: I do think there is a little bit of a man-on-the-street, common-sense view that goes something a little bit like, Well, given that there is a fixed pool of jobs in a country, if you add a bunch of foreign-born workers, they’re going to take those jobs from natives. And if you just apply Econ 101, as the supply of a good goes up, like labor, then the price of that good, i.e. wages, falls. And so I think there is a little bit of an intuitive sense that more immigrants would mean lower wages and lower employment prospects. And I think this was actually the conventional view on both sides of the aisle for much of the 20th century, in much of the economics profession for much of the 20th century, until this study came along and shattered the consensus.

And so what happened was: In 1980, Fidel Castro lifted Cuba’s ban on emigration. And that caused about 125,000 Cubans to migrate from Mariel Bay in Cuba to Miami, Florida, and about half of them settled there, which represented a 7 to 8 percent increase in the Miami workforce, which is 25 times the amount that the workforce expands due to immigration in the U.S. every year. So this is a huge change, an incredibly large change.

And years later, what the economist David Card—who will go on to win a Nobel Prize for his work in empirical economics—what he realizes is that this created a perfect version of what economists call a “natural experiment,” that because of this big one-time influx of immigrants to Miami, you could compare the trajectory of native-born wages in Miami to a variety of other cities that prior to the boatlift had similar employment and demographic trends. These include Atlanta, Los Angeles, etcetera. And I think the view was, Look—if there’s any place you’re going to see the negative effect of immigration on wages, it’s going to be with this unprecedented, large, random shock.

And that’s why the finding that Card comes to is so surprising, because he finds that the boatlift had virtually no effect on the wages of native-born workers, including those without a college degree. If you look at a chart of the wages of workers in Miami compared to most of these other cities in the U.S. at the time, there’s almost no difference. You can’t even tell the boatlift happened. And I think what that points to—and the big, overarching explanation that I think the common-sense wisdom got wrong—is that immigrants aren’t just workers. Immigrants are also consumers. They buy lots of things, like healthcare and groceries and housing.

And so at the same time that they are competing with Americans for jobs, they’re also creating more demand for those jobs. They’re creating more employment opportunities. And when you increase the demand for labor, that pushes wages up, even if you increase the supply of labor that pushes wages down. And we can talk about some ways in which this was later challenged and complicated, but I think that’s the big missing piece of the common-sense take.

Demsas: Yeah. I think there’s a level to which you have to really draw out how this works in the real world, because people come, and they’re like, Okay. Now there’s more people who want to eat at McDonald’s. You have to hire more people on shift to service that demand. There are more people demanding taxi cabs. There are more people who now need immigration-lawyer services, so that means you need more legal assistants. That means you need more paralegals. That means you need more janitors cleaning the buildings because they’re expanding into new office space.

There’s a level to which this positive flow is not intuitive to people, because it’s so downstream of the initial event, which is: People are here looking for jobs. It’s the immediate, first thing they see happening.

Karma: Exactly. But it’s funny because when we think about this in a slightly different context, it’s very intuitive. For example, you don’t see Republican politicians going to high-school or college graduations and yelling at graduates or complaining that all of these graduates are about to undermine the wages or employment prospects of adults in the labor force.

Demsas: (Laughs.)

Karma: And that’s because we understand, when it comes to native-born people, like, Wait—population growth doesn’t necessarily mean less for everyone. And so I think when you take this to a slightly different context, it’s like, Oh, wait. This actually does make a lot of sense.

Demsas: Well, I think there’s one thing that I really want to draw out here. Because if you’re an individual person who’s—let’s say you are a high-school graduate. You are working in the types of service-sector jobs that are usually competing with immigrants. Maybe on net what you’re saying makes sense for the entire labor market, but doesn’t it change when you look downstream at the people who are the most likely to be competing with new immigrants for jobs?

Karma: This is exactly the right question. When I mentioned the complications earlier, this is where they come in. There is an argument that has come up in response to the Card paper and its response to a lot of the natural experiments on this. And I should say, also, in addition to the Mariel Boatlift study, there were similar experiments in the subsequent years in Israel, in France, in Denmark that all came to very similar results.

But then there was a backlash set of critiques, which was just this: Okay. On average, wages might not be affected, but what about the least-skilled, the least-educated workers? And, particularly, what about those without a high-school degree who work in the professions that are most likely to be competing with these new immigrants, most of whom—if we’re talking about, at least, undocumented immigrants—are less skilled themselves?

And this was the critique made, and has been long made, by a Harvard economist, George Borjas. And in 2015, he went back to the Card study, and he looked specifically at this group of high-school dropouts. And he found—or, at least, at the time, it seemed like he found—that actually there was a sizable negative effect on this smaller group. And again, this was the explanation: Okay, maybe on average it works out, but the supply-and-demand effects of immigrants are asymmetric.

Immigrants who are unskilled, who come into a country—they compete only with a certain subset of the least-skilled workers, but they’re spending their money broadly. So they might get a job as a lettuce picker or construction worker, but they’re spending their money on a lot more than just housing and lettuce. And so on net, they end up hurting these less-skilled workers more.

Demsas: It’s an inequality story too. All of us get the benefits, especially those of us in high-skilled jobs that aren’t really experiencing this competition, but they’re not concentrated for the lowest income.

Karma: And this is the most, I feel, poignant critique because, yes, this makes higher-skilled workers better off, but it hurts the least of us. And what is really interesting, though, is that Borjas’s debunking of Card has since been debunked.

Demsas: Oh, my gosh. Recursive debunkings. (Laughs.)

Karma: I know. This is all the fun of an academic debate. It has all the titillating content we want.

Demsas: (Laughs.)

Karma: So if you look at what Borjas did, what’s interesting is he didn’t just look at high-school dropouts. He also excluded from his sample women, nonprime-age workers, and, most confusingly, Hispanics, which is sort of absurd. And his justification was that these exclusions left only the workers that were most directly competing with the Marielitos. But it left a total sample of just 17 workers per year.

Demsas: I find this fact so insane. It’s one of those things where I don’t understand what the point of extremely rigorous journal processes for econ journals are if they’re allowing something like this to go by unnoticed, unflagged.

Karma: And I think the reason is because at that point, it’s really hard to tell the difference between an actual empirical finding and just statistical noise.

Demsas: I mean, it’s 17 people.

Karma: It’s 17 people. It’s this extremely specific and hard-to-justify group. And then, what’s interesting is there’s a couple follow-up studies, one of which finds that the effect that Borjas found was because of a change in the way that the census counted Black workers and Black individuals.

Demsas: Oh, yeah.

Karma: And if you just take the way that they were measuring it before, the entire effect goes away.

Demsas: Even with those 17, the sample size?

Karma: Even with those 17.

Demsas: Wow. Okay.

Karma: But I think the broader critique is, Wait. This is ridiculous. Let’s just actually do what the critique says, which is: Let’s just look at all workers who don’t have a high-school degree. And when you actually look at that, Card’s original findings hold up. Actually, workers that lack even a high-school degree didn’t have their wages negatively affected.

And in the subsequent years after this debate, there have been other natural-experiment studies that have found the same thing. One that I really think was done quite well was on Puerto Rican immigrants after Hurricane Maria who came to Orlando. It found the same effect. It actually found, to your point, that while wages for construction workers, specifically, actually did become depressed a little bit, that was offset by a boost in the wages for leisure and hospitality workers. And so, actually, one wrinkle to this story is that maybe some sectors might experience a little bit of this. But on net, it won’t affect the entire skill group. The entire skill group ends up pretty well off. And I think this, for me, was a very counterintuitive finding. And when I asked economists about it, the leading explanation is what was described to me as “specialization plus scale.”

On paper, it looks like—and the assumption has long been—immigrants without a high-school degree are perfect substitutes for native workers without a high-school degree. But it turns out that that’s actually not true. And I think the restaurant industry is a good example of this. Take fry cooks: A bunch of new immigrants come in, and they take jobs as fry cooks working in restaurants. That might depress the wages of local-born fry cooks. But what also happens is: Because the cost of labor has gone down for fry cooks, and because now there’s all this more demand for restaurant services, you get restaurants expanding. You get more restaurants opening up.

And what happens when restaurants open up? They don’t just have to hire more fry cooks. They hire more waiters and waitresses and bartenders and chefs. And it turns out that a lot of new immigrants can’t fill those roles, because they don’t have the English skills or the tacit cultural knowledge to do so. And so, actually, if you were a native-born worker and you just stayed a fry cook, you might have seen your wages depressed, but you’re actually far more likely to have gotten a job in one of these professions that is now more common, that actually pays more, because immigrants have entered.

Demsas: So you get promoted.

Karma: Exactly.

Demsas: I think the other group of people that people often point to as being harmed by this are actually recent immigrants, right? It may be the case that there’s not a substitution effect between native-born workers and foreign-born workers. But if you are the first person off the Mariel Boatlift, and then the thousandth person is coming off, you guys are probably competing.

I always find this a bit of a weird argument because people usually talking about this are immigration restrictionists. Are they taking the position of the most-recent immigrants who’ve come to this country?

Karma: Yeah. Don’t you care about all the other immigrants?

Demsas: Yeah. Yeah. What’s going on there?

Karma: That is a really good point. And I should say, just because these studies don’t find much of an effect on native-born wages of natives of all skill levels does not mean that immigration has no cost at all. And I think this is actually one of the most-important, most-consistent findings, is we do see a pretty sizable effect on the wages of other immigrants, in large part because they don’t have the substitution effect.

Another cost is inequality. A lot of these studies find that, even though a lot of lower-skilled native workers aren’t affected, immigration ends up boosting the wages of higher-skilled workers, in part because immigrants are also demanding the services of, let’s say, architects or computer programmers, etcetera. And so it’s not a huge change in inequality, but it does slightly exacerbate inequality.

And then I would say the third one is what I talked about earlier, which is if you look sector by sector. It’s very possible that a construction worker or a worker in a specific sector where a lot of immigrants come in might experience some wage losses. That is very possible. Even if the aggregate effect on an entire skill group is not negative, you can see concentrated losses.

Demsas: But this is just true of all effects, right? If the average effect is positive, 50 percent of people are below the average of all things.

I think the thing that I’m getting at here is—and one of the things I really liked about your approach to this—that you were very, very careful to try and be as fair as possible to both sides of this debate. And what I’m hearing is that there’s so much reaching you have to do to really find serious costs to immigration. Even when you do, it’s like, Slightly exacerbate inequality. Maybe there are certain industries where you have some impacts, but those people are also benefiting from the growth, and they’re also benefiting from substitution effect, etcetera.

And it’s not to just pooh-pooh all that, but I think it’s really interesting to talk about why there’s such an intense desire to find this effect. And I don’t know if you have a thought on why this narrative is so important to people, because there are other reasons that people could say they’re anti-immigration, but there’s a real desire to make it about wages. There’s a real desire to make it about economics.

Karma: Yeah. I will say: I want to definitely talk about these different reasons. And I’d be very interested in your theory, too, and I have my own. I will say, one good-faith reason for this concern, one that I think will be brought up a lot is, Well, what about when we look at history?

And so it is true, and lots of folks, including David Leonhardt, liberal columnist at The New York Times, has pointed out that during this mid-century, quote-unquote, golden-age period—1940s, ’50s, and ’60s—you saw really high wage growth for the working class. You saw a really big reduction of inequality, really fast rise in living standards, and also very low immigration. And then from the 1980s on, you see much higher levels of immigration, and you see wage stagnation for the median worker. You see an exacerbation of inequality. And so I think one thing is, If we look at history, maybe these experiments aren’t capturing everything. They’re only looking at one city at a time. And when we look at the broad sweep of American history, it really does look like this is happening.

And I think that is a critique that’s important to take seriously. But at the same time, one of the golden rules of social science is “Correlation does not equal causation,” right? There were a lot of things happening starting in the 1970s and ’80s that also affected workers, also affected inequality—everything from technological change to globalization to the weakening of labor unions and concentration of corporations. And I think a lot of those other things were going on, and I think two data points are really instructive here.

Demsas: Well, before you get into that, I actually think you’re being super generous to this argument, which I think is your MO here. I think it’s important to be intellectually generous at the front part. But I want to be very clear here: This is not looking at the broad sweep of American history. This is looking at the ’40s and ’50s and ’60s, and going, Huh. This extremely transformative time in American history, where there’s tons of growth happening because World War II is ending. Also, the World War II production, in general—lots of stuff happens, of course, following that, anti-growth stuff that we’ve talked about in this podcast in the past.

And I think it’s kind of weird and, I think, feeds into the question I was even asking you earlier about, like: There’s such an intense desire to make this true, and when you look back at the foreign-born share of the population in the United States over, actually, our long term, in 2023, 14.3 percent of Americans are foreign-born, and that’s in line—and lower—than large parts of the 19th century. So what you see in American history, when looking at the foreign-born share of the population, is: You see we’re at roughly 14.8 percent, even throughout the 1800s. You see a massive dip start to happen during the Great Depression—normal. People kind of stop emigrating when that happens. And then you don’t really see a catch-up happening until very recently.

And so there’s a level to here where I’m like, If Leonhardt and others want to make this critique, they need to then explain the entirety of the 1800s in American economic history. And I think there’s a desire not to really wade into that debate, because they’re just pointing at a simple correlation and going, I’m sure this explains it. I actually don’t find this even minimally persuasive.

Karma: I know. I think you’re totally right. And also, you don’t even have to go back to the 1800s. You can just go back, I don’t know, the past four years, where we’ve had a huge, massive surge of undocumented immigration. And at the same time, we’ve had wages at the very bottom of the income distribution rise at their fastest pace since the 1940s, a huge reduction in wage inequality.

And so even if you’re going to make the correlation argument, it’s like, Wait. The last couple of years sort of disproved this. And even over this time period that Leonhardt and others are talking about, what you have is: The places that receive the most immigrants are the places that have the least wage stagnation. It’s Texas. It’s Florida. It’s the Acela corridor. And so I think you’re right. I really wanted to put that out there because I think it is a very common argument, but it’s not one I find remotely persuasive.

Demsas: There’s one other thing that I think other folks point to a lot, and I’m going to ask you to explain it, because you’re explaining all these studies for me so nicely. But the National Academy of Sciences has a study called the “Economic and Fiscal Consequences of Immigration.” I feel like this is a calling card for a lot of folks who are pro–immigration restriction. What does it say, and what’s important about that study?

Karma: This was a large report that looked at, or at least purported to look at, a bunch of different studies that claim to be a sort of meta-analysis of a lot of the immigration literature and tried to come to a conclusion on what it all says. And the big conclusion that they came to was, when we look on average, wages are not affected, especially when we look in the long term. But there was a disagreement within the panelists over, specifically, high-school dropouts. And there’s a chart that often is linked to or is often brought up by immigration restrictionists. It is table 5-2.

Demsas: Oh my gosh, you know exactly where it is.

Karma: It shows a lot of negative numbers. And the thing that I will often remind people is: That chart is basically talking about high-school dropouts. Okay, put that aside. There are a lot of studies on there that seem to show negative effects. One of them is the Borjas study that we talked about earlier. And George Borjas was actually one of the panelists on this report, which may indicate or may give you a hint of why it turned out the way it did. But when you actually go through and look at these studies, most of them are not the kind of high-quality, natural-experiment study we’re talking about. A lot of them just focus on Black men, Hispanic men—like, very particular subgroups.

And then, also, a whole bunch of these studies are in a category of, what they’re called, “skilled-cell studies.” And these studies are different, right? They’re not looking at a specific causal link created by a natural experiment. What they’re saying is, We’re going to just look at the entire group of unskilled workers in the United States. We’re going to look at immigration flows, and then we’re going to make a bunch of assumptions, and a bunch of assumptions about the substitutability of native-born and foreign-born workers, about how fast capital adjusts. And based on those assumptions, we’re going to make big claims.

And so there was this famous other Borjas paper in the early 2000s that made the claim that when you do one of these studies, it shows really intense negative effects. And so this report that is often used, and I know this is so wonky, is just—

Demsas: We love wonky here.

Karma: Yes. But I think it is just a case study, in that listing a bunch of studies with varying qualities, looking at varying different groups, is just not the most-accurate way to do things. And then, yeah, I could go on. There are lots of other problems with it, but I just think that that is one of the ones that frustrates me the most and frustrates a lot of the economists who I spoke to for this piece.

Demsas: I asked you a question earlier, and now I’m just going to give you the answer that I have to it, which is this question about why it’s so persistent, people desire it. One of my theories for this is that there’s a real desire to sane wash anti-immigrant sentiment.

When large parts of the population hold opinions, and particularly when they are different than the kinds of people who are in media or are in elite spaces—like, most people who work in media are living in cosmopolitan cities, have gone to college, have often maybe interacted with people who are from foreign-born countries repeatedly throughout their lives because they’re, like, living in New York or Chicago or L.A. or something like that. And as a result, like, they are not really in touch with some of the more common anti-immigration sentiments, and as a result, they feel kind of uncomfortable being like, Well, they’re all racist and xenophobes. They don’t want to sound like that. And so in order to look at this sentiment in the country and go, like, Well, I don’t want to call them a bunch of people who hate immigrants, I need to find some more material explanation for their opposition to it.

And I think it’s weird here, because I actually think it’s important to take very seriously what people are saying. Like, there are people who have serious cultural concerns with people coming into the country. And some of those things, I find not reasonable, and some of them I find—I don’t find really any of them reasonable, but I understand why someone would feel that way without having to be a bad person. Like, do I wish that people didn’t have those attitudes? Sure. But I think that they’re not lying when they tell you the things that are actually concerning them. And you write something really nicely on this, in your piece, and I’ll let you say it, but can you just talk to us a little bit about what surveys of public opinion actually find in regards to people’s opposition to immigration?

Karma: Well, first of all, I think that analysis is really spot on. I think sane washing is a good descriptor. The one that I haven’t had is often a veneer of respectability. And I think there is a lot of this deep discomfort with nonmaterialist explanations, in part, also, because—especially, if you’re thinking about, let’s say, center-left folks—if you’re part of a political party that you think needs to respond to people’s views on immigration, it’s much easier to say, Well, look—we already believe in raising worker wages.

And so, all of a sudden, if immigration gets looped into the set of values that we already believe, we’re then comfortable to give in to people’s instincts here. Whereas if it feels like pandering to darker forces, I think that makes liberals, especially, less comfortable with doing it. And I think to your point, though, it’s like: If we don’t acknowledge those darker forces, it’s not always great. And I think what you’re getting at, too, is in this piece, consistently what you find in sort of surveys of public opinion is that it’s not primarily material explanations that explain things. It’s things—a lot of them are about cultural difference, about violations of social norms, about crime, about national identity.

And I think one of the most-revealing things here is that the demographic that is most opposed to immigration are older folks living in rural areas, many of whom are retired. And the people who tend to be most supportive of immigration are working-age people living in big cities where immigrants are more common. So if you thought, like, Okay, this is a product of the people who immigrants are directly competing with are the most anti-immigrant, you would think, Oh, this would show up where the immigrants are, and it doesn’t. And so I think that really speaks to it.

I think the other thing that really speaks to it is, like, have you just listened to the Republican Party? Like, Donald Trump and J. D. Vance will occasionally mention wages, especially when it’s Stephen Miller talking to The New York Times, when it’s J. D. Vance in a vice-presidential debate or talking to a New York Times interviewer. That’s when they will bring up this wages argument. When they are speaking to an audience that they know is very center or left, they will, like, bring up this wages argument. But if you listen to the guy at the top of the ticket, right, it is, These folks are “poisoning the blood of our country.” It is portraying immigrants as a sort of psychopathic horde of murderers. It’s spreading conspiracy theories about pet-eating immigrants. Like, it’s very hard to take seriously that this is actually the main concern when the leader of the party who is anti-immigration is, like, so openly pointing to a very different set of issues.

[Music]

Demsas: After the break: why less immigration would mean a poorer quality of life in America.

[Break]

Demsas: I think one of the things that I also find reasonable for why people struggle is the Econ 101 explanation you give right at the beginning is intuitive. And it shows up in many different parts of our econ reporting, this question of thinking outside of partial equilibriums, right? If you were to just say, I’m looking at just the impact of immigrants on wages, holding equal all other effects on the population, on the economy, then you would see negative downward pressure on wages.

But economists don’t do that. Our lives aren’t lived in partial equilibriums. We live in general equilibrium. There are multiple different markets working together at all times. And you have to look at not just what’s happening with the effect on the labor markets, what’s happening in the effect on the housing market, the consumption of random household goods. And so I think that that’s really difficult to do in normal conversation. And what I think is really funny is that now—you kind of foreshadowed this—there’s a big push to blame immigrants for the housing market, and it’s like, Oh, the only time immigrants are consumers of anything is when they’re consuming housing. Otherwise they’re just competing with you for wages. They’re not buying anything else.

What I wanted to ask, though, is about this other argument that people also make, which is: Okay—maybe you’re right that in the world that we live in, given that immigration is always happening, companies can rely on there going to be some level of immigration. They’re used to a high level of immigration happening. You don’t see these negative effects on wages. But in a world where you were to just, like, really, really tamp down—really, really stop immigration from coming in—companies would have to reshape how they’re doing their hiring practices. The entire American economy would change if it wasn’t reliant on foreign labor.

And so this idea that there are these, quote, jobs Americans won’t do isn’t true. They would do them at a price. They would do them if the wages were better, if the working conditions were better, and that we should strive for these higher-quality jobs. And companies that can’t do that, well—they should just not do that anymore. They should just literally stop relying on foreign labor. And so I think that how you respond to that is really important because, you know, I do think a lot of people are starting to, like, fixate on that argument.

Karma: This has actually been one of the largest justifications for what Donald Trump has called the largest deportation effort in American history that he wants to enact when he’s in office. Any of his advisors are talking about this. They talk about: This is going to force employers to hire workers at higher wages, to give them better jobs, and that’s a big reason why we should do it. So I think it’s a really important one to address.

And what’s nice is we actually have some really good empirical studies on this. We don’t just have to guess as to what would happen and assume as to what would happen. My favorite of these studies, although there are a few, looked at the Secure Communities program, which was a DHS program that deported about 500,000 immigrants between 2008 and 2014, so during the Obama administration. And the way that this happened was: It happened sort of semi-randomly across communities, such that it created a sort of natural experiment where you could look at how it affected communities where it had happened and how it affected not-yet-affected communities.

And the findings were shocking even to me because I would think, Okay, maybe when you get rid of a lot of these workers, there’s just going to be more jobs available.

Demsas: It’s like a shock.

Karma: It’s such a big, immediate shock. But what the authors find is that for every 100 immigrant workers who are deported, there are actually nine fewer jobs for natives. And this isn’t just temporary jobs. This is, like, permanently, there are fewer jobs for natives in the community, unemployment goes up, and wages slightly fall. And I think this kind of finding is repeated across different examples through American history.

There’s another great study of the H2B program, which allocates low-skilled workers to companies, and also finds that when companies aren’t allocated those workers, they don’t hire a bunch of natives. They actually just produce less. And so what happens when immigrants are ripped away from these communities is the interconnected web of employment and workers whose jobs depended on each other all gets torn up, right? Businesses close. Businesses have to stop producing as much. There are just less child-care services. There are less meals served. There are less houses built—either for reasons of: Employers actually can’t have a viable business with higher labor costs, whether it’s because natives don’t always want to do these jobs, or whether it’s just because, for the reasons we talked about earlier, there’s just a lot of benefits from the specialization of labor that occur when immigrants are in a place.

One way I think about this is sort of the opposite of the story that we were talking about earlier. When we talked about immigrants coming in and creating the specialization plus scale, that just happens in reverse. Instead of businesses expanding, and therefore being able to hire more natives because they’re expanding, businesses are shrinking. They’re shutting down. They’re closing. And when that happens, native-born workers get caught in the crossfire. When there’s less demand for your services as a restaurant, and your costs are higher, and so you have to close down, you’re not just getting rid of your immigrant workers, you’re getting rid of all the native-born workers who are working there too. And so I think that’s what these studies are finding, is you just can’t neatly remove immigrants from communities without having huge backfiring effects on the native-born.

Demsas: Yeah. I think it’s even useful to play it out in the best-case scenario. I think the best-case scenario for the folks who are making this argument is that there’s this short-term harm, but then you just need to let the economy play out and develop new businesses to figure out new business models that work. And in the best-case scenario, you’re talking about a poorer country. You’re talking about a country where your output, your growth is literally less. And that sounds very abstract, but we’re talking about less stuff. You have less money. You can buy fewer things. You can buy a worse quality of life. Your housing is probably worse. Even the basic stuff where you’re talking about, Can you afford child care?—fewer people can do that.

A lot of things are worse when the output declines, when it’s harder for businesses to try new things, when there’s difficulty with dynamism in the economy, where you can’t start a bunch of different kinds of businesses quickly, see what works, and have that kind of churn. And so I think it’s even difficult to conceptualize: When people are making this argument, they’re saying, We should take on the costs of being a poorer country for the sake of national homogeneity of where you were born.

And so I think that that’s the trade-off we’re talking about here. It’s not that America would cease to exist, right? There are a lot of countries who follow the sort of principles we’re talking about here, where they are really strict on who gets to come in, and they’re just poorer than America. And I think that that’s the really clear trade-off that I think often restrictionists won’t make baldly.

Karma: First of all, even the best-case scenario you’re talking about is one that has actually no empirical evidence. It’s all theoretical. So that’s the first point. The second is that this gets back to, I think, your point about general versus partial equilibriums, too. Because when we’re even just looking at these wage or employment studies, they’re holding a lot constant. And everything they’re holding constant also changes in an actual scenario where you deport millions of immigrants.

So there is another great study from the economist Ben Jones and a few others looking at immigrants and entrepreneurship. And they looked at basically every single business that opened up between 2005 and 2010 and looked at basically the country of origin of the person who started that business. And they found that immigrants are 80 percent more likely to start new businesses than native-born individuals.

And when they actually did the math, they found that immigrants, by entrepreneurship alone, are creating far more jobs than they take. One response to that would be, Oh, okay. Well, maybe this is high-skilled immigrants. Maybe this is the Google-founderkind of effect. But actually, they found that there was no difference in the rate of entrepreneurship between individuals from OECD countries and from non-OECD countries. And if you just think about this for a second, think about the people who end up coming here, the amount of risk they have to take, the perseverance that it takes to, like, actually get to the U.S.—it wouldn’t be surprising that these people are, like, more intrepid and more entrepreneurial.

Demsas: Just huge selection effects. Like, if you can make it through the Darién Gap, what does that say about you?

Karma: Exactly. And so that’s one effect that is completely lost in a lot of these studies. One of my favorite studies of this is one that was done in Denmark, because in Denmark, what is interesting—unlike in the U.S., where you have to just look at a specific city—Denmark has data on individuals for the entire country. It’s a pretty small country. And so researchers can actually track what happens to every single individual worker when new immigrants come in. And that gives you a sort of accuracy that you don’t necessarily get with the natural-experiment studies in the U.S., at least at a countrywide basis. And what they found is that native-born—even less-skilled native-born—workers end up responding to immigration by entering higher-paying occupations, by moving to higher-opportunity cities, and by actually getting better education, such that they actually had higher wages as a result of it.

And so I could go on. You could talk about the amount of women who are able to be in the workforce because of immigrants providing child care. Like, you can list this out, and there are all these ways in which even these studies are missing the sort of beneficial effect that immigrants are having that you would be taking away if you just suddenly got rid of all these people, in addition to this atrocious humanitarian effects.

Demsas: I find that the Danes—like, I wanted some sort of poll on their privacy concerns. I’m just like, Do you guys not care? I mean, like, I think it’s great. I would be pro-this everywhere, but I’m just surprised that countries are able to do this. There would be a revolt in America.

Karma: Even if the data is anonymized, I’m like, The data the researchers got was anonymized, but the data the government has is not anonymized.

Demsas: We don’t even let the government share data like that. Like, the IRS can’t just send the Treasury Department, like, all the data they have on people’s tax returns.

Karma: But you know what? You know what? Great for the Danish for doing it, too, so we can learn more about immigration through them.

Demsas: So true. But so the thing that’s interesting is: We’ve made a bunch of arguments here about why this is actually really positive for the economy. But regardless of that, there’s been a backlash, and we’re seeing that right now. I mean, this is airing on Election Day, and so we’re, I’m sure, in the future, just pacing nervously to see what’s going on.

Karma: (Laughs.) Apologies to anyone who was listening to this looking for a soothing distraction.

Demsas: But this has been probably the most-important issue of this election. Maybe inflation is another one. But the two most-important issues. I did an episode earlier this year with John Burn-Murdoch where we talked about the sense that Americans are very xenophobic and this narrative that they hate immigration—they hate immigrants—and that’s, like, just a fact of the world, and that all immigration has basically been this plot by elites to shove it down our throats. And, of course, we explored how a lot of that narrative is really overblown and underestimates just how strong pro-immigrant sentiment is in America, particularly relative to other countries.

And I still stand by that analysis, but there has been a shift in public opinion, even in the past year. You’ve seen polls come out that have really indicated that there’s been a backlash effect to the high levels of immigration that are kind of returning us to the 1800s averages.

And so first, can you just talk us through that backlash? What are the numbers there? What are we seeing?

Karma: Totally. The thing that first drew my attention to this was, as you were saying: The way this has impacted the election is that you’ve just seen such a hard right turn in the rhetoric from candidates on both sides. And I remember listening to Trump in 2024, making 2016 Trump sound like JFK in just how crazy he was. And then looking at the Democratic side, where the message went from, in 2020, Joe Biden promising to restore moral dignity to our asylum system, and then in 2024 Kamala Harris saying that, Actually, no. She is the one that will fortify the border, not Trump.

Demsas:Do not come.

Karma:Do not come. And underlying this is quite possibly the most dramatic shift in public opinion that I’ve ever seen. So going back to the 1960s, Gallup asks Americans every year this question: Do you think immigration should be increased, kept the same, or decreased? In 2020, only 28 percent of Americans said that immigration should be decreased. Actually, more Americans said it should be increased. By 2024, just four years later, the percentage of those who wanted it decreased had nearly doubled to 55 percent, the first time that there had been a majority of Americans who wanted immigration decreased since the early 2000s.

And just to put this shift in context, I think when Americans think about big public-opinion shifts, they think of gay marriage. And they think of the increasing support for gay marriage. Support for gay marriage, according to Gallup, increased about 20 points over the course of around a decade, maybe eight or nine years. This shift we’re talking about was nearly 30 points in four years. It makes gay marriage look gradual and small by comparison. And this immigration shift is most concentrated among Republicans, but it’s also Democrats. It’s also independents. And it’s especially been sharp in the past year.

Demsas: This is one of those things where I think it’s really important for people who, like myself, are in favor of high levels of immigration, first of all, to accept that, at some level, if you get that, you will have some negative effects, but I think also to really narrow in onto people’s specific concerns.

So I did an article earlier this year. It’s called “Something’s Fishy About the ‘Migrant Crisis.’” And basically, I was just like, Okay, there are high levels of immigration in a lot of places in this country. But not every place in this country is experiencing backlash, right? Like, you’re hearing these stories in New York and Chicago about people sleeping on the floors of police stations in Chicago. In New York, I had, like, an affordable-housing lawyer tell me—she was a very liberal person telling me that she was kind of concerned because there were migrants in the street in midtown Manhattan who were just, like, lying on the ground.

And there’s a lot to which I’m like, You know, if these people who are talking to me are some of the most-liberal people on immigration are expressing kind of like, Well, we can’t handle this. Like, we obviously can’t handle this, it indicates a very specific problem, right? Like, people—these New Yorkers, Chicagoans—they’re not afraid of immigrants or foreign-born people. There’s huge levels of foreign-born share in New York and Chicago. And the number of people that were entering we’re talking about, you know, that were coming in and demanding services from local government were a very small fraction of this.

So I was trying to understand what was going on there, and I really came down to the specific concerns people have. People don’t want to see local tax dollars being spent on newcomers to their city, if they feel like they need things that the city’s not actually taking care of. They don’t want to see their schools being used as shelters instead of being used in order to service, you know, their kids. And there’s just kind of general sense that, like, Now there are people sleeping on the streets. There’s nowhere for them to be housed. Like, It’s actually reducing my quality of life a bit. Clearly, there’s a sense of it being overwhelmed.

But then when I looked in places like Miami and Los Angeles, and in Texas and in Houston—Miami and Houston, in particular—I was like, There are way more immigrants who have come through a Houston, in Texas, than have come through a Chicago. So why are we seeing such backlash?

And I came to like two reasons. One is that many of the immigrants were not able to get work permits. The other thing that’s really important here is the Greg Abbott busing program. Because, most people, they come into the country, and, you know, what happens? They have networks that they’re following. Like, either they have populations of people that they’re able to get help from, or there are even people who are kind of recruiting them as they’re kind of coming over, like, Oh, we need work. We need people to come do this. And so there’s a level to which, like, there’s a natural flow to where they end up.

Greg Abbott has, I think, maybe the most-effective political stunt in American history—I genuinely think, like, reshaped the entire conversation on immigration by doing this. And then he says, Okay, I’m gonna bus people—effectively breaking these kind of natural shifts—to Democratic cities. And when people kind of show up randomly, there’s, like, of course, this massive transaction cost that’s enacted. And, you know, Texas is a border state, and I think, at some level, I kind of understand. They’re like, Oh, everyone should have to experience what it’s like to have all this kind of flow of people coming in. But Texas has boomed as a result of this too.

So anyway, I think the real thing that’s important here is that people who are in favor of immigration have to address these specific concerns. You have to make sure that there is, like, actually a clear, orderly process by which people are being resettled. If there’s not, I mean, that’s going to lead to backlash, even from people who are in favor of immigration. And the most frustrating part of my reporting is learning the Biden administration had basically abdicated their responsibility to try and help with the resettlement process of people across the country, because they were afraid of being blamed. And to me, I’m like, Well, you were still blamed. So I’m not sure it worked out for you.

Karma: Everything you’re describing here, I think, falls under the banner of what has been called either, colloquially, “chaos theory” or, more academically, the “locus of control theory” of immigration, which is that populations tend to be able to handle high amounts of immigration if they think the process is orderly and fair, and they become much more likely to oppose immigration when they see the process is chaotic and unfair and disorderly.

You know, you have a great example in that piece of the U.K., post-Brexit, having very high levels of immigration and anti-immigrant sentiment decreasing. Something you see in the U.S. is that even as you have these massive shifts in the amount of immigrants people want let in, when you ask questions like, Do you believe undocumented immigrants make a contribution to society? or, Do you support a path to citizenship for nondocumented immigrants? and even, Should it be easier to immigrate to the U.S.? people’s views actually haven’t changed nearly as much. And they remain more pro-immigrant than they were in 2016, which speaks to the fact that what people are upset about here—they’re not suddenly xenophobic. They don’t suddenly hate immigrants. A lot of what’s happening is that they’re responding to the chaos of the process.

I think my favorite part of that piece that you wrote was this point that you made about how there’s a way of looking at Greg Abbott’s busing program as working, in the sense of, like, Look—didn’t this prove his point? He said that liberal cities should have to handle this, and he proved that they couldn’t. But the process by which he did it was engineered to achieve that outcome, right? You point out in that piece that there are 3 million foreign-born people in New York City, in a city of 8 million, and they’ve been bused a couple tens of thousands. And it has led Eric Adams to say, like, New York is falling apart. That does not mean that New York can’t handle that amount of immigrants. The specific process by which this happened was engineered to achieve an outcome of chaos. And that’s what people are responding to.

Demsas: Well, let’s leave it on an optimistic note. Always our last question: What is something that you originally thought was good on paper but didn’t pan out in real life?

Karma: So this is quite a pivot from what we were talking about earlier. Last year I got engaged to my girlfriend.

Demsas: Oh my gosh, yes! Congratulations!

Karma: That was not the thing that was—

Demsas: Oh god.

Karma: That would be bad. But the way I did it was: It was our five-year anniversary, in Rome, very romantic. I knew I wanted to propose in front of the Pantheon, which was my partner’s favorite building in Rome.

Demsas: She’s an architect.

Karma: She’s an architect, yes. But I didn’t want to do it when there were a bunch of crowds around, so I was like, How can I figure out a way to get us there, like, early in the morning? And so I decided, in a decision that looked very good on paper, to book a Vatican tour for, like, 9 a.m. And so I was like, Oh, I’ll propose, and then we’ll go on this tour of the Vatican, and it’ll be, like, really cool. And it’ll be, like—we’ll see the Sistine Chapel. Sounded great. Looked good on paper.

It turns out that immediately after you have one of the most emotionally riveting experiences of your life—

Demsas: (Laughs.) You don’t want to go on a tour!

Karma: The last thing you want to do is go on a three-hour tour of the Vatican where you have to wait until the last 15 minutes to see the Sistine Chapel. We’re just, like, so badly just wanting to get out of there and be with each other, and we were just in such a great mood, only to have, like, the biggest buzzkill in the world be the Vatican.

Demsas: This is so funny, Rogé. I didn’t know the story. Wait. I can’t believe—so you went on the tour?

Karma: We went on the tour. I wish so badly I would have said, Let’s just forget the tour. But we were in such good spirits after. We’re like, This is going to be so great. Like, actually, looking back, I’m like, Was that even good on paper? I don’t think so.

Demsas: I was actually with you. I was like, Okay, yeah. Then you had a fun tour.

Karma: Like, a nice walking tour, architecture tour—probably great. When you’re, like, confined to the Vatican and just looking at, like, our guide—she was great, but she was just explaining every little thing, and we’re just like, We don’t want to be in public with a million people. We just want to be with each other. This is very strange.

Demsas: I think this is my favorite “good on paper” yet. This is unreal. (Laughs.)

Karma: I put a lot of thought into this, and I was like, This one was bad. This is not my best. It’s a funny story now.

Demsas: Yeah.

Karma: You know, I look back on it—I’ve looked back on it very fondly. So yeah, that’s my “good on paper” story.

Demsas: Thank you so much, Rogé. Thanks for coming on the show.

Karma: It’s been a pleasure being here. Thanks so much for having me.

[Music]

Demsas: Good on Paper is produced by Jinae West. It was edited by Dave Shaw, fact-checked by Ena Alvarado, and engineered by Erica Huang. Our theme music is composed by Rob Smierciak. Claudine Ebeid is the executive producer of Atlantic audio. Andrea Valdez is our managing editor.

And hey, if you like what you’re hearing, please leave us a rating and review on Apple Podcasts.

I’m Jerusalem Demsas, and we’ll see you next week.