Michael Clemens on the Trillion-Dollar Question of Immigration

On this episode, Nathan Goodman is joined by Michael Clemens to discuss why immigration policy matters not just for migrants themselves but for broader economic growth. Drawing on his influential work, including “Economics and Emigration: Trillion-Dollar Bills on the Sidewalk?” (JEP, 2011) and “The Place Premium: Bounding the Price Equivalent of Migration Barriers” (REStat, 2019), Clemens explains how even modest liberalization of migration can create enormous gains, why exchange is positive-sum, and how complementary skills across the workforce drive production. Together they assess the claim that immigration undermines culture and institutions and revisit historical panics ranging from the Chinese Exclusion Act to the Dillingham Commission.

Dr. Michael Clemens is a professor in the Department of Economics at George Mason University and a Non-Resident Senior Fellow at the Peterson Institute for International Economics. He helped build the research program on international migration at the Center for Global Development.

Show Notes:

• Samuel Bazzi, et al., “The Confederate Diaspora” (NBER, 2025)
• Timur Kuran’s book, Freedoms Delayed: Political Legacies of Islamic Law in the Middle East (Cambridge University Press, 2023)
• Chloe N. East, et al., “The Labor Market Effects of Immigration Enforcement” (Journal of Labor Economics, 2023)
• Mexican Migration project

Transcript:
 

NATHAN GOODMAN: Hello and welcome to the Hayek Program Podcast. I'm Nathan Goodman. Today I'm speaking with Professor Michael Clemens. Michael Clemens is a professor in the Department of Economics at George Mason University and a Non-Resident Senior Fellow at the Peterson Institute for International Economics since 2023. He studies the economic causes and effects of migration all over the world. For 20 years prior to his role at GMU, he built the research program on international migration at the Center for Global Development in Washington, D.C. Professor Clemens, thank you so much for joining me.

MICHAEL CLEMENS: Thank you.

NATHAN GOODMAN: So, the first thing I want to ask is sort of a biographical question of sorts. How did you get interested in working on the economics of immigration?

MICHAEL CLEMENS: Wow, well, one of the most striking economic facts about the world to me is that the same person in different places has a very different economic life. And I remembered literally the day I recognized that. I was 12. My dad's a professor and was visiting the National Autonomous University of Mexico in Mexico City for a summer and took me with him and that was my first exposure to a developing country in living in memory. And our toilet broke in the apartment we were staying in. We had to bring a plumber in. And I overheard how much he earned and noticed that it was way less than plumbers in where I was growing up in Salt Lake City made. And so I had a little conversation with my dad about why would that be? That doesn't make any sense. It's the same guy, same wrench, same toilet … and the power of just moving between places to radically alter people's economic lives along with many other things sort of was planted in my memory, but I only started to work on migration as a development economist many years later.

NATHAN GOODMAN: Fascinating, yeah, I mean, think that's a really important fact. And it's a fact that you've explored in greater depth in your work, that vast difference between people's economic lives depending on where they're able to be. And so one paper that you wrote that touches on this point and that had a big influence on my thinking was called “Trillion Dollar Bills on the Sidewalk” in which you estimate the global costs of immigration restrictions. So, what did you find in this paper?

MICHAEL CLEMENS: Thank you. Yes, that strange and striking fact about the world economy that the same person in a different place can have a radically different economic life and economic productivity has an implication, which is that just moving between places for a worker, for a unit of labor in an economic sense is an arbitrage opportunity. It's not just taking a fixed quantity away from one person or group of people and giving it to another; a reallocation of resources, it creates value. It adds to aggregate prosperity. And when you have a world in which the exact same Haitian worker, to take a concrete example from a project I was directly involved in, the exact same Haitian worker in Cap-Haitien can have a labor whose value is about $10 a day at home, but has a value of $10 an hour in the United States. That means that the same person over the course of a year could go from adding value that is valued by the world market at $3,000 a year to $30,000 a year. And those kind of large changes add up quickly with even modest movements of people. It doesn't take a large movement of people across borders to add quite a lot to the world economy. 

So in that paper, I just did a back of the envelope calculation that was not a prediction of the future or a serious general equilibrium model of the whole world economy, which certainly can't fit in any economist's head or in a human being's head, but just a scenario to illustrate how large those gains are at the global level. And the bottom line is that even the movement of one in 20 people from poorest countries to much richer countries would add more value in aggregate to the world economy than the total elimination of, when I wrote that paper, all remaining barriers to goods trade and all remaining barriers to cross-border capital movement put together. So that paper is often discussed in terms of total elimination of borders and what would it mean. Neither I nor any other economist can calculate what would be the economic effects of total elimination of borders. That was a back of the envelope calculation to suggest what is the impact of a marginal change in border regulations and that impact is just vast.

NATHAN GOODMAN: And it's important to emphasize not just for the immigrants, although they would get a massive gain by being able to increase their wages substantially, but for everybody else who wants to purchase the products of their labor, right? Because you're emphasizing that they're creating value through mutually beneficial exchange. So, I think that's an important point to really drive home. When we talk about creating value, we're talking about creating value for people all over, not just for people who choose to move, but for everyone who benefits from the products and services and goods that they create. Is that fair to say?

MICHAEL CLEMENS: That's such a deep insight and it's one of many areas where the economic schools of thought represented at George Mason really make a crucial contribution in thinking about the system of exchange as a system rather than a mechanism. It's really whenever a person who is an immigrant takes a job in a migrant destination country; there are two sides of that contract. And the only reason that contract exists typically is because both people benefit. So, there is a wide range of benefits that spread beyond the migrants and in the public imagination and in punditry and memes and other pablum that floats around online, you often see the benefits of migration as being goodies that are taken by migrants. That's a... that's not an accurate representation at all about the economic benefits that arise from exchange between people.

NATHAN GOODMAN: Yeah, absolutely. I think that emphasizing the positive sum nature of exchange is a really important part of the economist's task when zero-sum thinking becomes so prominent. Now, several years after you published “Trillion-Dollar Bills on the Sidewalk,” you published a related paper with Montenegro and Pritchett on “The Place Premium,” which relates to that very first concept you introduced to us. So, what did you find in that paper?

MICHAEL CLEMENS: Yes, well, you heard me assert before that the same person can have radically different productivity when they cross borders. And we just put together for the first time micro data, individual level data on individual workers in a large number of countries with basic control variables that you would see in a study of the most fundamental determinants of wages in a labor economics paper, like level of education, years of experience, that sort of thing, so that we could just control for stuff and ask, let's look at a US worker and a Haitian worker and see what is their relative productivity as assessed by the labor markets in their countries. And there's a vast difference there. And how much of that goes away when we control for important things, like being Haitian born is an important thing. So, let's compare Haitian born people in the United States to Haitian born people in Haiti. Education is important thing. Let's compare people at the same level of education, et cetera. And there's still a very large difference in the productivity of workers between countries. And it's just enormous. 

It's much larger than the returns to education in many developing countries. It's enormous even for kinds of workers and occupations where the returns to any unobservable skills shouldn't be extremely high, such as the most basic manual jobs, suggesting that there really is a very large place premium or effect of where you are rather than who you are on your productivity. And that's been debated since then, the magnitude of the place premium has been debated in the literature, but the idea that it is very large for poor countries is, I think, something that is not questioned or really questionable by reasonable inquiry.

NATHAN GOODMAN: Excellent. So far what we've been discussing, the fact that you can become much more productive by moving and that that increase in productivity means that there's substantial gains from exchange, mutually beneficial exchange between migrants and those that they sell their labor to and those who benefit from and purchase the products and services that arise from that labor, suggests that liberalizing immigration could substantially enrich the world. But some economists have a counter argument to this, which you've referred to as the new economic case for migration restrictions. So, what is this new economic case for migration restrictions and how do you respond to it?

MICHAEL CLEMENS: There's this joke about the statue of John Harvard in the middle of Harvard Yard, which is that it's the statue of three lies because it says John Harvard founder 1638 and it's not a statue of John Harvard. He wasn't the founder, and it wasn't founded in 1638. The new economic case for immigration restrictions is not new. So, I'll just criticize myself in retrospect for using that moniker. It's a very old argument. You see it in the very popular bipartisan policy of Chinese exclusion that the US instituted in 1882 and maintained de facto for 83 years. Subsequently, there's an extensive record of people's reasoning for instituting this policy. John Stuart Mill weighed in, Democrats and Republicans debated it at length in the Congress. And a major concern was exactly this, that somehow Chinese people had a culture that was incompatible with American culture, would dilute or corrupt American culture in some way, not raised to respect democratic values and democratic rule, et cetera, et cetera, or even were prone to crime or were not able to maintain stable families or something like that.

Just a lot of nonsense that was not based on fact. Similar arguments were made by the Dillingham Commission in the early 20th century, whose arguments eventually won the day and led to the largest heretofore immigration shutdown in US history, which was the 1921 and then 1924 acts, the latter the Johnson-Reed Act that really drastically curtailed not just immigration, but specifically immigration from particular places and explicitly in the Dillingham Commission that provided the written motivation for those laws, there was this concern that Catholics are loyal to the pope and not to the US president, and a lot of Jews are anarchists or communists and aren't compatible with our institutions and are somehow going to dilute or corrupt the institutions that the whole society and the whole economy rests on. And in retrospect, those concerns were just wildly overstated. They were adopted by numerous high level figures, including some of the founders of the American Economic Association, the president of MIT, prominent figures at the California Institute of Technology, the Massachusetts Institute of Technology, Harvard professors. 

These were mainstream views at one time, not by people who would be thought of at the time or should be thought of now as crazy or deluded, but really just not based on fact. You know, we ended the Chinese exclusion experiment in 1965 when the quota for Chinese immigration went from 105 to about 20,000. And since then we've had a whole lot of Chinese immigration and I'm not aware of any evidence at all that that has had a systematically detrimental effect on any US institution, on any US labor market, on US innovation, on US productivity in any sense whatsoever. So the idea that if you were sitting in 1882 and you put forward the notion that when Chinese people come, they are going to bring something other than just labor and railroads and mines and other things, they might bring ideas, they might bring ways of interacting with each other, they might bring values from other traditions, is not at all the same as there is any sort of quantitative reason to believe that those things will undermine US culture and values to a degree that will impoverish Americans. Those are radically different claims. 

And the idea that migrants bring culture with them or bring some predilection for certain institutions with them is true, is verified by various studies. My old friend and colleague, Sammy Bazzi, at UC San Diego, has a paper called “The Confederate Diaspora,” showing that southerners that moved around the US brought a worldview and values with them where they went. And the idea that somehow the modern version of concern about Chinese, concern about Catholics, concern about Jews, which is of course concern about Muslim immigrants among others, is widespread in America and there is economic research suggesting that institutions and culture matter, Timur Kuran is a brilliant prominent Turkish economist who has argued that traditional financial institutions, among others in Islamic culture, have played a role in the relative levels of economic development that we see now today between predominantly Muslim countries and other countries. So, the idea that culture can come to some non-zero degree with migrants and that culture matters for economic development are things that I think have been upheld by a lot of the most serious economic research we have. 

I don't think Daron Acemoglu and Simon Johnson, Nobel laureates who have studied exactly this issue, would disagree with either of those statements. But to go from there to saying there is even remotely a quantitative reason to believe that all of the enormous benefits from exchange and specialization that come from marginal relaxations of the extremely high barriers to migration that we see in the world now are somehow overwhelmed by that effect at the margin or anywhere even in the same solar system as the margin is not one that I see any empirical support for. In my paper with Lant Pritchett in the Journal of Development Economics, we just try to give intuitive examples about this. Toronto, Canada is, last time I checked, 50% foreign born. It might be over 50% foreign born now. So vastly greater than the United States, which is about 16.5% foreign born now. The migrants in Toronto, Canada are predominantly from developing countries. And if there was something that was bringing the low total factor productivity of those countries relative to the US, where Canada, with them when they went to Canada, we should observe that walking around in Toronto right now. Toronto is a lovely place to live. It's a vibrant, diverse city. It's a center of industry and innovation. There's no sign at all that we're anywhere near a margin at 50% of the population that would be systematically degrading Toronto's ability to produce valuable goods and services for the people who live there. So I think it's—the interesting conclusions there are that culture affects development and that culture does travel with migrants. I find the policy conclusions regarding migration are totally uninteresting because I don't think they're policy relevant.

NATHAN GOODMAN: That makes sense. So, what we've covered so far really goes from sort of the most fundamental truth of economics. Exchange is mutually beneficial. People can move to places where they can be more productive and have more enriching mutually beneficial exchanges, both for buyers and sellers. While it is true that immigrants affect culture, there's no reason to think that their effect on culture or institutions could lower productivity or would lower productivity to the extent that it would reduce those gains from trade. And so, so far this seems like a really compelling case for liberalizing migration restrictions being something that could enrich the world and enable a lot of people to have much better lives and enable us to have a much more prosperous society. But some people might say that they're really not concerned about these economic questions. They're concerned more about whether immigrants come to the US legally, whether they followed the law. And you've done some research that's relevant to that, specifically looking at how the ability to cross legally impacts rates of illegal border crossings. So what did you find in that research?

MICHAEL CLEMENS: Yes, thank you. You know, illegal activity is a major concern of lots of people, and it is a function of policy. But economists have thought carefully about exactly how it's a function of policy. So, let's just step back from migration and think about a totally non-immigration example. Ivory is illegal. I've never bought any ivory. I don't own anything made of ivory. It's generally banned to trade ivory internationally or to buy and sell ivory in the United States. There are ethical ways to get ivory. When an elephant dies, it's possible to harvest that ivory and make stuff out of it and not contribute to poaching or endangerment of species or anything with any other negative externality that I'm aware of. There have been pilot programs to create small channels of ethical trade in ivory around the world. And a notable paper evaluated those and found that they had actually increased the trade in illegal ivory. And you might think to yourself, well, if you're a person who wants ivory and you have a choice between buying it on the black market or in the little pilot store where it's being sold ethically, of course you're going to go to the ethical store, and the legal trade should displace the illegal trade. But when you think about it just a little more, you can imagine channels why, and these are documented in the paper I'm referring to, you can imagine channels why more legal trade in ivory could actually complement illegal trade in ivory by familiarizing people with stuff that's made out of ivory, by building a taste for ivory. When I get a present for somebody, I would never think of getting ivory, but if somebody showed me a piece of legal ivory and I really liked it, then my demand for ivory might rise and the aggregate demand, some of which passes through legal channels, some of which doesn't, might go up. And it becomes an empirical question whether marginal legalization of ivory displaces or encourages illegal ivory trade.

Coming back to migration at the US border over the last several years, there was a huge number of arrivals at the border without any prior authorization to enter the United States. In some months, 10,000 people a day arriving at the border seeking entry, no US passport, no US visa, no prior arranged permission to enter. And there was enormous concern, bipartisan concern about that situation. And as you mentioned, many people concerned about chaos at the border, about the rule of law in the United States and certainly at the border. And what I was interested in that paper is, well, there were substantial numbers of people allowed to enter and remain in the US under a legal authority, which is that the 1965 Immigration Act, which is still the fundamental framework for regulating immigration in the United States, creates a parole authority for the president to use pretty arbitrarily to allow any person to enter for humanitarian reasons and remain for a certain period. And that conveys lawful presence in the United States. To be lawfully present in the United States, you must be either admitted or paroled. Admitted means you have a prior arranged document to enter. It's what most people experience at the airport. They're admitted by Customs and Border Protection. You can also be paroled and you're lawfully present. So that was a kind of legalization of a flow that was, until parole is given, in violation of law. 

So it's a real question of whether allowing people to enter with parole and thus remain lawfully in the United States displaces unlawful crossing and unlawful entry, unlawful presence, or actually encourages it. It's possible that people come to the border and if they can get parole, fine. And that causes them to be less likely to cross unlawfully. But you can also imagine indirect ways, kind of like in that ivory example, where letting more people in lawfully gets communicated back to people in a migrant origin country thinking, well, I talked to so and so, more and more people are finding ways to get in, I'm just gonna go try. And if they can't get in lawfully, they might be more likely to go through the desert, to go through the river. And thus it's possible, at least on paper, that in that indirect way, allowing more lawful entry and lawful remaining in the US could cause more unlawful entry on net, net even of the displacement.

So what I did was just do some very basic time series econometrics to ask: there was substantial variation in the number of people over time who were being allowed to enter and remain lawfully by the previous administration. When those surges happened, were they followed by more people coming to the border and thus more people entering both lawfully and more people entering unlawfully? Do we have evidence that on net those lawful entries were complementing or displacing unlawful entries? And I found that they on net displaced them, that an increase in the number of lawful entries was followed by typically a lasting decline of about a third in percentage terms in the number of unlawful crossings. And this doesn't mean that those lawful entries were a major determinant of the unlawful crossings that were happening. I estimate that just the decision of how many people to allow to lawfully enter explains about 9% of the variation over the last several years in the number of people crossing unlawfully. But it does suggest that an important ingredient of regulating unlawful behavior, if that's your goal, alongside enforcement and the other things that affect people's incentives or whether or not to break the law, is the legal alternatives. And that's rarely a part of the discussion when pundits and policymakers are deciding how to regulate the Southwest border. 

Right now, we have a very strict enforcement-only policy. About 10,000 active military troops deployed at the Southwest border as we speak on top of and assisting the standard complement of Customs and Border Protection officers, and the number of crossings of the border without prior authorization has gone from at some times, as I mentioned, 10,000 a day in December 2023, for example, to about 10,000 a month. So just total collapse. I think it's important in setting a future and perhaps more sustainable border regulation policy to take account of the joint roles of enforcement and the availability of lawful channels.

NATHAN GOODMAN: So legal pathways to immigrate could be an important aspect of restoring order at the border is sort of what you're saying here.

MICHAEL CLEMENS: They're crucial and we know that from various strands of evidence. One of them is there's this fascinating project at Princeton University called the Mexican Migration Project that since 1983 for decades has interviewed systematically across Mexico people who migrated to the US, capturing their whole migration history. Like if they crossed 19 times, they ask about all 19 times where they did it? Where did they go? What did they do? How much did they earn, et cetera? And if you look in the last 10 years of that data set, there are people crossing with no visa to work in the US. There are people crossing to work with a visa, like a seasonal farm work visa, the H-2A visa. And if you ask the question, well, we have the whole migration history here. 

If you just take the people who in the last 10 years crossed with an H-2A visa, they're crossing because there are hundreds of thousands of H-2A visas available now that didn't used to be available. What fraction of those had previously come as an unauthorized migrant? The answer is more than a third. That is, hundreds of thousands of people a year crossing on a lawful visa to add value to the US economy by engaging in an exchange relationship with US farmers. And these are people who are now doing it on a visa who literally the same person used to do a similar thing on the black market and they have been shifted from the black market into a more regulated market by the availability of a lawful channel. In my time in government, I tried to emphasize the role of lawful channels alongside enforcement and the other tools that the government has as a means to achieving the ends that the public has clearly wanted—greater order and greater rule of law at the border.

NATHAN GOODMAN: Excellent. So, the current administration of course, is not expanding legal channels generally, but contracting legal channels. And instead, they're choosing very much an enforcement-only approach. And a big part of their enforcement approach is implementing a program of mass deportation. Now, economists have studied prior mass deportations. So what does their research tell us about the effects of such mass deportation efforts?

MICHAEL CLEMENS: Yes, thank you. I mean, let me just mention briefly that the public image of the policy, which has been cultivated by the administration, is that of enforcement and crackdown and deportation. I'd recommend a term more like exclusion, because as you mentioned, we're seeing the mass elimination of lawful channels to immigrate to the US, too. The refugee program has been almost entirely shut down; well in recent years that's been some eight, nine, 10% of immigration to the US. We're seeing many steps against international students, which is definitely going to reduce the number of people coming on F-1 and J-1 visas next year, lawful immigrants. We're seeing people who have parole—which as I mentioned is a reason under law that you are lawfully present in the United States. If you have parole, by definition, you are not unlawfully present—and parole is being eliminated en masse, which turns a person who is lawfully present into a person who is unlawfully present. So alongside all of those things, as you mentioned, there is a mass campaign targeting the removal of about a million people who aren't lawfully present from the country. Those two alongside each other—crackdown is not the right word because a large number of the people being targeted are lawfully present and lawfully immigrating. We're talking about mass exclusion.

Getting finally to what you asked, which is about mass deportation of removal of people who are unlawfully present. That is something that has been studied extensively by economists. There we have recent experience with large scale deportation. A particular program that began right at the end of Bush 43, but was mostly executed under the Obama administration is called Secure Communities, which has given us a lot of evidence about the effects of mass deportation. And it's because of how the program was designed. Secure Communities was a program of agreements between local law enforcement officials, county by county, and federal immigration enforcement officials to share data and hold people for certain periods so that ICE could come get them if wanted. And the crucial thing for economists is that it resulted in a lot more deportations, as you can imagine, but it was not rolled out across the country at the same time. If Secure Communities began in every county all at once and something happens to the labor market, it would be hard to distinguish it from something else that might have affected the labor market. But what happened instead was that this slow rollout in a kind of piecemeal way created lots of situations where there was a county where this mass deportation effort had arrived and was starting to remove lots of people, otherwise similar counties or even neighboring counties where it hadn't arrived and was not yet affecting the local labor market. 

So, Chloe East of the University of Colorado and a group of co-authors published a paper in the Journal of Labor Economics, which is the top journal of labor economics, extremely highly vetted for separating causation from correlation. And what they were testing was, well, how many jobs for natives did this open up when mass deportation comes to town and a lot of unauthorized immigrants who are supplying their labor are removed—what happens to native employment? And the surprising answer that they come to, which has been replicated several times, is negative. It actually eliminates employment of native workers. And that might seem to make little sense. How could it be, you know, even if natives are not interested in doing those jobs, you might expect effective zero, but how could it be negative? How could it actually eliminate jobs for Americans? And the answer, which they spend much of the paper on, is that the mechanism is by deterring business activity, by deterring the formation of new businesses and by encouraging the exit of existing businesses. That is, small landscaping firms that just never get founded, and all of the US jobs that would be associated with that within the firm or outside the firm—to do the books for that landscaping firm, to deliver fuel or supplies to that landscaping firm—and all the ripple effects that would generate US employment in both inside and outside that firm are gone. 

That finding doesn't at all mean that while the jobs that were opened up by mass deportation are jobs that no Americans want to do. That finding is perfectly compatible with some of those jobs being of interest to some Americans. It's just that the net effect of deterring business activity, which is the only thing that creates jobs, alongside whatever willingness there was for some Americans to do those jobs, the net effect was strongly negative. So that's why if the administration achieves its targets of roughly a million people removed a year, and we were to suppose that maybe half or 60% of those people are workers, employed workers in the labor force, this suggests that they could be eliminating hundreds of thousands of jobs for Americans each year by that mechanism. And that's really extraordinarily unfortunate. That finding to me does not at all mean that, well, laws shouldn't be enforced because Chloe East found that mass deportation eliminates jobs for Americans. That would be a really ludicrous conclusion in my view. 

What it does suggest is that enforcement-only could be welfare reducing for the United States relative to alternative policies—reforming the laws that we have, which as I mentioned, were created by my grandparents and great grandparents’ generation in 1965 and have not been importantly revised in 35 years. That is the legal stricture under which we're laboring now and altering that stricture for this century to create lawful channels for some of those workers to engage in mutually beneficial exchange with Americans could be a vastly better policy that is totally different from arguing that somehow we should not enforce laws. Although you see so many pundits now, anybody who says a word to question whether mass deportation is in the economic interest of the United States is immediately accused of being against the enforcement of law. And that's just disingenuous garbage. It's not true.

NATHAN GOODMAN: Yeah, absolutely. And so, I think one big upshot of this is when we think about means and ends, right, if the stated end is protect and create jobs for Americans, the chosen means of mass deportation are an inappropriate means to that end. And so, thinking about ways to expand legal pathways for immigrants rather than engaging in this enforcement-only approach is going to be better for not just immigrant workers, but American workers.

MICHAEL CLEMENS: Absolutely. And there are many studies showing that people who arrive lawfully and can remain lawfully are more productive in the labor market. That means they are more specialized and when they're more specialized that means that Americans have more opportunity to specialize in things that complement them and everybody benefits. That is really possible. There's just … zero sum thinking dominates the airwaves now. But a major contribution that economists can make now and have always made and will continue to make is to point out the indirect and sometimes hard to see, but tremendously powerful benefits of mutually beneficial exchange.

NATHAN GOODMAN: Absolutely. So, another aspect of immigration that is important to think about is how it relates to our aging population. So, can you tell us what the demographic trends look like currently and how does immigration policy relate to the possibilities for the future of America's labor force?

MICHAEL CLEMENS: Yes, I'm a really awkward person who's not great at making friends and you're going to see why when I say that a trivia question that I like to ask people at cocktail parties and things is the following question. The US Census Bureau does projections of the population. You can look them up on census.gov right now. They do it by age and the projections for the near term are—unlike some of economist’s projections—are really quite accurate because it's pretty well known what is fertility, what is the death rate at certain age groups, et cetera. They always do a zero migration scenario because that is the component of population change that is least certain. And so the trivia question is: in what year did the zero migration scenario for the growth of the US labor force turn negative? Is that something that's going to happen in 2040, 2050, 2070? Is something that already happened? And the answer is that it turns negative around 2010. That is, if we're talking about the 1990s when I was in college, we were in an America where with zero immigration, even without all the Mexicans who were coming in at that time and many others, we would have had a robustly growing labor force just due to demographic change among the native population. Right now in 2025, not at some future date, but at the moment that we're speaking right now, if there were zero immigration in this year, and this is from the 2023 Census Bureau projections, in 2025, the working age population of the US falls by 1 million workers per year, gone permanently.

That's the number of people aging out of the labor force at 65, of the working age population at age 65, minus the number of people aging in at 18. And in that zero-immigration scenario, the over 65 generally retired elderly population grows at 500,000 a year at the same time that the labor force is shrinking 1 million per year. So, thinking about that for a second, if we just say, well, not all immigrants are labor force participants, let's suppose that something like two thirds of them are. That means one and a half million immigrants per year just to keep the labor force from shrinking. And it would take a lot more than one and a half million immigrants per year to keep the labor force at parity with the elderly population, most of whom don't sell labor anymore, but depend exclusively on capital to buy other people's labor. That's a very drastic economic situation. It means that a large share of all economic growth in America, which has traditionally depended on a growing labor force, depends on immigration. 

So, the economic question is not, well, why do we need these immigrants? Why aren't we investing in our own people? Our own people are not creating enough babies and have not for a long time created enough babies to create a growing labor force. Now, all of that said, the historic relationship between economic growth and a growing labor force is changing and could change to a substantial degree over the next couple of generations. There's a lot of magical thinking now emanating from Silicon Valley and blogs and memes and that sort of thing that the marginal product of labor is going to fall to zero. It's really—it's great to have a shrinking labor force because nobody's going to need labor anymore.

There are extremely serious economists like Nobel laureate, Daron Acemoglu, that have been quite skeptical of those claims. Certainly, in the near term, the occupations that the Department of Labor projects to grow most in labor demand in the US economy over the next decade include many, many occupations that are just not going to be automated. Care work is not a thing that is going to be automated. And if somebody comes up with a robot that can help elderly people take a bath, most elderly people are not going to want that robot, frankly. The idea that we're living in a post-labor world or that any time soon we're going to be living in a post-labor world I think is extremely unfounded. And certainly, at the margin where we're living now, the economic growth in the United States depends crucially on immigration. So, the question is, is it gonna be lawful or unlawful? Is it going to be high skill or low skill? Is it going to happen in some parts of the country or other parts of the country are important questions. But if your approach to this problem is “why don't we just do it with our own people,” you need to take a hard look at the demographic choices that “our own people” have been making and whether those are compatible with the growing economy that most of us want.

NATHAN GOODMAN: Yeah, I think that's a really important point. So, as we think about the future of the American labor force, you mentioned that we can think about, for instance, how many immigrants are high skilled versus low skilled. So, what does economics teach us about the impacts of both high skilled and low skilled immigration?

MICHAEL CLEMENS: Thank you. Yes, this is another area where there's just a lot of magical thinking. If you read a lot of public discussion about immigration, you could get the idea that while the US economy runs on computer engineers from India and scientists from Poland and other highly skilled innovators with PhDs who are pushing out the technological frontier and ... Sure, Silicon Valley runs on programmers and highly educated entrepreneurs. What I wish people understood more is that Silicon Valley runs crucially on farm workers. It runs crucially on delivery workers and security workers and construction workers and childcare workers and many other essential inputs to that production process. The example I often give to students is a surgeon and a cleaner. There are zero people who want to have surgery in a dirty surgery room. There are essentially zero surgeons who are going to clean the surgery room. So there needs to be somebody to clean that surgery room if there's going to be any surgery and it's not going to be the surgeon. That doesn't mean that the surgeon isn't essential to surgery, but the cleaner is equally essential to the surgery because my demand is zero for surgery in a dirty, bloody surgery room. And the necessity for workers at all levels of formal education and tacitly acquired skill to complement each other is something that I wish were a greater part of the public discussion of immigration. 

When you look at the wages of migrants, which reflect their marginal productivity relative to their best option, the surgeon earns more and the cleaner earns less. That doesn't mean that at the margin the economy only benefits from admitting marginal surgeons and never marginal cleaners. They are inputs to a joint production function that complement each other. And for that reason, an immigration system that is built exclusively for high skill, which believe it or not is our immigration system now, we have no work visa substantially, there are immaterial exceptions to this, but we have no work visa for non-college work. There is a minor employment-based green card category for workers who don't have a college degree, but it's little used. 

Almost all of the employment-based green cards that we give are to people with a college degree and specialized knowledge. We have temporary work visas for people doing non-college work, but those are exclusively seasonal—again with immaterial exceptions—seasonal farm work, seasonal non-farm work for hotels and landscaping companies and forestry and seafood packing and that sort of thing. But for year-round work that doesn't require a college degree, we have no meaningful non-immigrant visa, we have no meaningful immigrant visa. We have the large majority of green cards going to family sponsorship, so we have that indirect channel of hope for the best—hope they arrive with the skills and occupations that we need without any employer sponsorship or indication that employers demand the specific skills that they have—program. But that's really just inadequate to the needs of the economy.

This idea that somehow people with high education and earning high wages are the only thing the economy needs is really dominant worldwide. The UK, at many points in its history, has had a salary minimum for immigration to the UK, as if the idea that somebody earning less than quite a high annual wage, which by the way is higher than a lot of actual UK citizens earn, can't be beneficial and can't engage in a mutually beneficial exchange and specialization that produces a joint product such as surgery that is beneficial to Britons, is not even on the radar screen of a lot of politicians and the public. I wish that there was just a much broader understanding of the potential for complementarity.

NATHAN GOODMAN: Absolutely. Yeah, people often seem to think about the labor market just in simple partial equilibrium supply and demand where they're thinking of a single supply curve for labor and more workers would mean driving down wages for everybody instead of thinking about the intricate way in which we all fit together into a vast network of cooperation and exchange, where often our labor complements each other. You mentioned the example of the cleaner and the surgeon. We could also think of somebody who's working as a waiter and then someone who's working as a chef or a dishwasher in the back of a restaurant. I mean, even thinking about our own work, right? Like there are people involved in producing this podcast and my labor is complimented by their labor, right? And so we can think about this in any sector of the economy. No one is an island working alone or solely in rivalrous competition with their fellows. We all benefit from a vast amount of cooperation. And that means that often our labor is complimentary rather than merely a strict substitute.

MICHAEL CLEMENS: Thank you for that and you know, this is an area where people often rightly criticize pointy headed economists for talking in these abstractions and stepping away from the real-world experience of normal folks. Anybody can understand these things. You just mentioned restaurants. We've all been in restaurants where there's a person who's waiting the table. There's a receptionist out front. There's a chef making, designing cool dishes, but ... Absolutely none of us are ever going to walk in that restaurant again if the plate that comes to the table isn't clean, meaning a crucial necessary input to production is somebody's got to be there washing those dishes and the chef ain't going to do it. And the waiter can't do it. The idea that workers complement each other and the real-world experience of walking around the United States and seeing that many immigrants are washing dishes in restaurants and therefore creating the entire possibility to eat at a restaurant for every American is not a foreign concept. It's not an abstract thing. It's not hard to understand. Americans have seen construction sites, either been involved in construction sites or observed construction sites where it takes a team of people to produce a house. 
And there have got to be people who are specialized in framing and concrete pouring and roofing and tiling and also in electrical work and plumbing and sales and management. And if you have been around construction in the US enough, you'll notice that the propensity of different kinds of people to do those tasks are not identical. 

A lot of immigrants are specialized in framing and tiling and roofing and concrete pouring, not exclusively, but many are. And a lot of natives are specialized in sales and management and other tasks that are necessary to construction. Again, not at all of them, but many natives are specialized in those things. There's a propensity for natives to specialize in those things. There's a propensity for immigrants to specialize in other tasks. And a hundred percent of them are necessary to make those houses. You don't want a house that doesn't have a foundation that's been poured properly. You don't want a house whose framing has been done with substandard nails or wood that's gonna catch fire, but you're also not gonna get a house where the management and back office and sales operation is not making that house available to you at a price you can afford. This is part of the everyday bread and butter experience of Americans. So, I don't think it's an abstract economic concept and I don't think it's difficult to understand.

NATHAN GOODMAN: Yeah, absolutely. I think it's a reality of living in a world where people work together to produce things and where we're social creatures that benefit from a vast network of cooperation and exchange with each other. And that's something that we've all lived. So, throughout your career, you've worked to understand these concepts, to do research on these issues. And you've done it not just in academia, but in nonprofit and policy circles. So, you've also done a lot of work to communicate the insights that we've been talking about today. So how do you approach the different roles that you've taken throughout your career and how has your work been received?

MICHAEL CLEMENS: On the assumption that anybody cares about Michael Clemens, I mean, I never sought a role. My dad's a math professor and there's probably a part of me that'll always feel inadequate as a mathematician because I can never equal him. But I didn't set out to say, well, I'm gonna be a professor or I'm gonna be a government official or I'm gonna be a think tank person. My whole life I've just wanted to be a person who learns about stuff and tries to figure stuff out and figure out how to explain that to people who are making important decisions. That's it. And there are opportunities to do that as a professor—walking into a classroom and teaching is a chance to take stuff that you hope you have learned and figure out a way to explain it to people, a wide variety of people, in a way that could be useful to them. Being in government is a way to sit in meetings. In the White House or in the CIA, which I've done, and try to explain things to people that are based on fact. And it's very much always been demand driven rather than supply driven for me. When people asked me to serve in government, I did that. When people said, how about if you apply to this professor position, I did that. And they're all just fascinating ways to learn about the world and try to explain it in useful ways. But it's incredibly fraught to be involved in policy making and decisions that make a difference to people, and you anger everybody. 

There are tons of people on the political left who don't like my research, people on the political right who don't like my research. I've been attacked for questioning the value of foreign aid when I pointed out that successful foreign aid, which increases the general prosperity of developing countries, tends to encourage migration—because migration is not a thing that is deterred by economic growth and rising prosperity and rising interconnectedness in otherwise isolated and poor developing countries. It actually complements more exchange and more interconnectedness with the world, which is why in most countries over the last two generations, rising prosperity has not been associated with more people saying, things are great now at home, I'll just stay there. Economic growth and the accompanying demographic change, for example, in Mexico, were associated with much more migration to the United States—not everybody deciding to just stay in Mexico. And that's been a pattern that's been repeated across the world. But I've been panned on TV and many other places for being against the left's enterprise of eliminating global poverty for that reason.

But I didn't set out to find facts that would justify somebody's position. I just note that extremely obvious empirical regularity and I believe that should guide policy. And of course, there are people on the political right who see me making the point that there are all kinds of ways to regulate activity that you believe to be harmful. And one of them is to change what it is, not just enforce it out of existence, but to enact policies that shape migration. If there's something about migration that you don't like, like that it's chaotic or uncontrolled or that people are arriving who have not been properly vetted or people who are arriving without the proper skills, those are all concerns that can be directly addressed by better regulation. That does not necessarily involve tens of thousands of troops pointing guns at people and telling them that absolutely no one can come in. That's not the only answer. When you pursue facts, those are often at odds with political agendas and you're gonna make people upset. I've always found ways to make a contribution that have been really satisfying. And I've been just incredibly fortunate to be able to make any kind of contribution to any of the above institutions. But really, it's always difficult. I think anybody who tries seriously to be effective in research and policy is never going to be great at both of them. There are a few people who manage it, a handful of people who are just superb at both naturally, but for mere mortals like myself, it's always going to be difficult.

But it's also highly rewarding because working in policy disciplines your research in a way that keeps it from veering off into questions about how many angels dance on the head of a pin and vice versa. Being involved in research means you can often be the only person in the room who has a knowledge of what the basic facts are when extremely weighty decisions are being made. And that happened several times in my career. So, it's a great place to be, but a difficult place to be. It's also another thing that makes it difficult is the incentives involved. You know, people in academia don't tend to respect policy work. They see it as a source of bias and factionalism and taking you away from pure science, some of them, and vice versa. People who are interested in policy are very often interested in the details of law and what it would take to get different political factions to support this or that. And those conversations can often spin very far away from very basic facts like, are you sure that fostering greater economic growth in Honduras will deter people from leaving Honduras? Because if that happened, it would be one of the first times it ever happened in the history of developing countries. And that often means that people in the worlds where you have each foot discount what you're saying. And one just has to sort of be okay with that in order to pursue a joint career. But I find the advantages greatly outweigh those disadvantages.

NATHAN GOODMAN: I think that's a really thoughtful answer. And I think for early career scholars who are listening, I think some of the key lessons that come from this are one, pursue the truth and always be attentive both to the concrete facts on the ground and to the basic insights of economic theory. And then try to communicate what you've found and what you've learned clearly. And you do that whether you're in the classroom or whether you're speaking to policymakers. And I really admire the way that throughout your career, you've communicated such important findings and done such rigorous scholarship to try and learn true things about the world and communicate them on issues that have such great importance to the lives and prosperity of so many people. So, thank you for that.

MICHAEL CLEMENS: That's generosity I don't deserve but thank you.

NATHAN GOODMAN: I really appreciate you taking the time to speak with me today, Dr. Clemens. I've learned a lot from the conversation, and I hope that our listeners have too.

MICHAEL CLEMENS: Thank you so much.