Why skilled immigration (usually) benefits both countries

When to worry about brain drain, and when not to worry.

"Satya Nadella" by n.bhupinder, CC BY 2.0

I’ve been a staunch and relentless advocate for increasing skilled immigration to the United States. This is not because I don’t value other immigration — I do! — but because skilled immigration is simply such a high priority for the U.S. Not only is skilled immigration crucial for maintaining and extending U.S. technological dominance, but skilled immigrants are also the most highly beneficial in terms of paying taxes and raising wages for other Americans.

However, a question I often get is: If skilled immigration helps the U.S., doesn’t it hurt the countries that send us the immigrants? For example, Samo Burja recently posted the following:

Some friends of mine asked me why this statement isn’t obviously correct. And to be fair, I myself have raised the alarm over Canada poaching skilled immigrants from the U.S. So I think it’s worthwhile and important to write a post that tries to answer the question of whether brain drain is something we should be concerned about.

The short version here is that there are both theoretical and empirical reasons to think that skilled emigration can help the economies of developing countries like India, the Philippines, or Indonesia that send skilled immigrants to the U.S. (Nor should we actually be worried about Canadian competition; it’s really just a wake-up call for our own policy mistakes.) But there might be a few cases in which we should actually be worried. And when two countries are on unfriendly terms, there are national security reasons to worry about brain drain.

Some very simple models

Simple economic models don’t always describe the real world very well, but they can help us to understand some basic concepts about the things we’re arguing about. So let’s do a couple of those.

First, let’s talk about what it means for immigration to be “good for” or “bad for” a country. “Good” or “bad” could mean several things. It could refer to economic living standards, or it could refer to national power. When people leave a country, it gets smaller, which might reduce its national power. As an example, take Ireland; a huge percentage of Irish people left the country in the 1700s and 1800s, which probably made it less able to resist British aggression and dominance. But let’s set aside the question of national power for now, and just focus on living standards.

Model 1: Composition effects

Consider the simplest possible model of production: Everyone just farms for themselves and eats what they grow, and there’s no trade at all. Each person has a certain personal productivity level, which is fixed. The “skilled immigrants” are just the people who are better at farming.

Now imagine there are two countries, America and Smallonia. Smallonia consists of just three people: Mia, Mark, and Sofia. Here’s how much each of them produces per year:

Mia: 100

Mark: 60

Sofia: 50

The per capita income of Smallonia is 70. The median is 60.

Now suppose Mia, who is obviously the most skilled of the bunch, moves to America and does her farming there instead. Smallonia now just includes Mark and Sofia, who keep on producing the same amount as before. Smallonia’s per capita income is now 55, and its median is also 55. So Smallonia was hurt...right?

Well, no, it wasn’t. If you think of Smallonia as just Mark and Sofia — i.e., the people who actually live in Smallonia after the population transfer — then their living standards are completely unchanged when Mia moves! After all, this is a world where everyone just farms for themselves, so how could their incomes change? The folks left behind in Smallonia are no better or worse off than before.

Alternatively, if you think of Smallonia as Mia, Mark, and Sofia — i.e., the set of people who lived in Smallonia before the population transfer — then their living standards are also unchanged, for the same reason.

So what this ultra-simple model teaches us is that even if the official per capita income of a country as a political unit goes down, that doesn’t mean the living standards of any human in that country go down. Composition effects don’t hurt anyone. (Of course in this example, the immigration doesn’t help anyone in America either.)

What this also means is that it’s wrong to think of immigrants purely as producers. They’re also consumers. Sending an immigrant to another country isn’t just like sending a deposit of oil; you also have one less mouth to feed. Obviously people don’t consume only what they produce, but I think this example is a useful corrective to the idea that immigrants function like a natural resource.

Model 2: Natural resource constraints

Let’s do a slightly modified version of the previous model, in which productivity depends not just on people’s personal ability, but also on the amount of land they have to farm. So production in Smallonia will be:

Mia’s production = 100 * Mia’s land

Mark’s production = 60 * Mark’s land

Sofia’s production = 50 * Sophia’s land

All the land amounts start off at 1, so total production is just 210, and per capita production is 70, like in the previous example. But now when Mia moves to America, Mark and Sophia divide up her land. So now, Mark’s land is 1.5, and Sofia’s land is 1.5. So the total production of Smallonia is now 60*1.5 + 50*1.5 = 165. Meaning that per capita income is now 82.5, which is higher than it was before, even though total output is smaller than before (because the best farmer moved away).

This example helps us understand natural resource constraints. Sending immigrants to another country means more natural resources for everyone who stays. This is basically the story you’ve heard about wages going up for farmers in Europe after the Black Death. It might sound silly, but in fact lots of developing countries in the present day make a lot of their money by exporting natural resources, and sending immigrants away means more resource rents to go around.

Model 3: Gains from trade

OK, let’s bring trade into the equation. Suppose that instead of just growing food for themselves to eat, Smallonians make software and sell it to America. America, which is a much much bigger country, makes everything else, and Smallonia imports everything except software from America. Because Smallonia is…well…small, the price for its exports is pretty much determined outside its borders. (This is called a “small open economy” model.) Assume that each Smallonian consumes however much they personally export to America.

When Mia immigrates to America, Smallonia loses her export revenue, and its imports drop by a corresponding amount (no trade deficits are allowed here). This is just like the situation in Model 1. But in this case, Mia produces more in America than she did in Smallonia (otherwise she wouldn’t have moved there). So America gets a little bit richer. Assuming that Smallonian software is a normal good, this raises American demand for Smallonian exports by a little bit, meaning that Mark and Sofia get a slightly better price for their exports…meaning that they’re able to afford more imports than before.

In case you like supply-and-demand graphs, here’s what that looks like:

Smallonian per capita GDP goes down thanks to the composition effect, just like in Model 1. But the per capita GDP of Mark and Sofia — the people who remain in Smallonia — goes up a little bit. Thanks to gains from trade, everyone gains.

(Note: If Smallonia has redistributive taxation, then Mark and Sofia might still lose out, because their highest earner just left.)

Anyway, the point of this example is to demonstrate another way that the sending country can gain when its talented people move out — through increased trade with the receiving country.

Model 4: Trade barriers

Now let’s consider a case where high-skilled emigration is definitely bad for Smallonia. Assume that America has very high trade barriers, so that trade isn’t possible between the two countries; America makes everything it consumes, as does Smallonia.

In this case, the loss of Mia is a pure loss to Mark and Sofia. When she moves to America, she can no longer interact with Smallonia in any way; she might as well have been launched into Venus. Previously, Mark and Sofia were able to benefit from their gains from trade with Mia; they were all able to specialize, which boosted total Smallonian income. With Mia gone, that will be less possible, and Mark and Sofia will get poorer.

This example, like the others, is extreme and unrealistic. But it illustrates an important general principle. High-skilled immigration benefits the sending country to the degree that the immigrants continue to affect their old country.

It also suggests another principle, which is that big countries can more safely send high-skilled immigrants than small ones. The 1,408,000,001st Indian person doesn’t contribute much to specialization within India, but the 10,877th Nauruan very well might contribute meaningfully to specialization within Nauru.

Complicating factors

Of course, those very simple models were just for the purpose of explaining principles — composition effects, natural resource constraints, gains from trade, and so on. But in the real world, a whole lot of other things can happen when high-skilled individuals move from one country to another. For example, here are some potential benefits that didn’t even come up in the models above:

• High-skilled immigrants might transfer technologies back to their old country after learning about them in the America.

• Immigrants might influence American businesses to do more deals with businesses in their old country. (This is called “brain gain”.)

• Immigrants might direct American investment dollars to their old country.

• Immigrants’ success might inspire more people back in the old country to get an education, so they can succeed in a similar way.

• Immigrants might influence Americans around them to have more of a taste for the exports of their old country.

• Immigrants tend to send remittances to their relatives in the old country.

• Immigrants might move back to their home countries, bringing expertise with them. (This is called “brain circulation.”)

And here are some potential costs the simple models didn’t include:

• A loss of local high-skilled individuals in a country might hamper a country’s innovative capacity.

• Fewer high-skilled individuals might mean less entrepreneurship.

• Fewer high-skilled individuals might mean a degradation in the quality of government, leading to worse policy.

It’s very hard to know, on balance, which of these are happening, and how much. So it’s very hard to get a total quantitative accounting of what high-skilled emigration does to a country. But we do have some evidence on some of these effects. I highly recommend the book The Gift of Global Talent, by William Kerr, which goes over many of these ideas.

Some evidence on brain gain

A fair amount of research has been done on the topic of whether brain drain is outweighed by “brain gain”. Over time, the observed effects have tilted more toward brain gain, though it’s not a slam-dunk case yet. Here are a few papers:

1. “Medical Worker Migration and Origin-Country Human Capital: Evidence from U.S. Visa Policy”, by Paolo Abarcar and Caroline Theoharides (2021)

These authors study the results of a U.S. reform that made it much easier for foreign nurses to come to America in the years from 2000 to 2007. Looking at the Philippines, they find a big increase in the number of people getting nursing degrees, which far more than canceled out the number lost to immigration:

Combining data on all migrant departures and postsecondary institutions in the Philippines, we show that nursing enrollment and graduation increased substantially in response to greater U.S. demand for nurses. The supply of nursing programs expanded to accommodate this increase. Nurse quality, measured by licensure exam pass rates, declined. Despite this, for each nurse migrant, 10 additional nurses were licensed. New nurses switched from other degree types, but graduated at higher rates than they would have otherwise, thus increasing the human capital stock in the Philippines.

Of course it’s not clear that the new nurses are as high quality as the ones who left, but a 10 to 1 ratio is…pretty darn good.

2. “The IT Boom and Other Unintended Consequences of Chasing the American Dream”, by Gaurav Khanna and Nicolas Morales (2021)

Nursing is a pretty specific field. But IT is a far more general one. Khanna and Morales studied U.S. H-1b law changes and their effects on demand for Indian high-skilled immigrants in various occupations. They find that the ability to send skilled immigrants to the U.S. IT industry sparked a major IT boom back in India, through much the same channels that Abarcar and Theoharides found.

We study how US immigration policy and the Internet boom affected not just the US, but also led to a tech boom in India. Students and workers in India acquired computer science skills to join the rapidly growing US IT industry. As the number of US visas were capped, many remained in India, enabling the growth of an Indian IT sector that eventually surpassed the US in IT exports…The H-1B program induced Indians to switch to computer science occupations, and helped drive the shift in IT production from the US to India[.]

Now, there is one big caveat here. Khanna and Morales find that the Indian IT boom was enhanced not just by Indians getting more education and switching to IT professions, but also by Indian H-1b workers who returned to India after their visas expired. This is called “brain circulation”, and it’s an important additional way that skilled immigration can help a sending country. But this benefit, unlike the others, does depend on some of that immigration being temporary.

3. “Human Capital Investment under Exit Options: Evidence from a Natural Quasi-Experiment”, by Satish Chand and Michael A. Clemens (2019)

This paper uses a very different and much darker natural experiment — pair of coups in Fiji that resulted in policies that discriminated against South Asian Fijians. The South Asian Fijians ended up investing a lot in education, in the hopes of escaping to the U.S. via our high-skilled immigration policies. In the end, most ended up not leaving Fiji, and their increased education contributed to the country’s economy.

4. “Globalization, Brain Drain and Development”, by Frédéric Docquier and Hillel Rapoport (2011)

This paper documents several different examples of skilled immigration, including the Indian IT industry (which had positive effects for India) and European researchers leaving for better pay in the U.S. (which might have had negative effects on Europe if Europe wasn’t receiving an inflow of researchers from the developing world). It also reviews the literature on African medical workers who go to work in the U.S., and finds little effect one way or another.

5. “Brain drain and human capital formation in developing countries : winners and losers”, by Michael Beine, Frédéric Docquier, and Hillel Rapoport

This study is just about correlation; it doesn’t have a natural experiment or policy experiment to isolate causation. It basically just correlates high-skilled immigration with subsequent changes in the level of human capital in the sending country. The authors find a positive correlation, supporting the idea of “brain gain”. For countries where human capital is already high (e.g. Europe), brain gain presents less of an opportunity. And for countries where a very high fraction of their skilled individuals emigrate (basically, small countries), brain drain can be so costly that it outweighs brain gain quantitatively. But for big, poor countries like India, brain gain wins over brain drain hands-down.

Overall, the literature pretty strongly supports the “brain gain” idea. There’s less research on other possible benefits to the sending country, like increased investment, business linkages, and tech transfer. I did find one paper by Kugler, Levintal, and Rapoport (2013) that finds that international bank loans to a country tend to increase after that country sends out a bunch of skilled immigrants. In general, a lot more research is needed on this topic. But keep in mind that just one single benefit — brain gain from increased education — is often enough to outweigh brain drain in its entirety, at least according to the educational measures we have available. The additional increased benefits from increased trade, investment, and tech transfer come on top of that.

When to fear (or take advantage of) brain drain

In general, both simple theory and evidence suggest some guidelines for how we should structure our high-skilled immigration policies so as not to hurt other countries. The basic lesson is that we should take more of our skilled immigrants from large poor countries like India, Indonesia, and Nigeria. The best way to make this shift would be to eliminate the country cap on employment-based green cards, which privileges immigrants from small countries over big ones.

As for whether the U.S. should worry about countries like Canada brain-draining us by poaching our skilled immigrants, the answer is “Yes, but only very slightly.” The U.S. is much too big to be affected by Canadian poaching; I highlighted the issue more to provide a contrast between Canada’s rational skilled immigration policies and our own highly politicized, dysfunctional mess.

But there’s also one more important case to talk about, which is how we might use brain drain intentionally in order to compete with another country. Operation Paperclip, the effort in the 1940s and 1950s to capture Nazi scientists and make them work for us, was partially intended to keep them out of the hands of the USSR. The Soviet Scientists Immigration Act of 1992 was not intended to weaken Russia, but probably ended up having that effect once the Ukraine War cut off most trade, investment, and research links between the U.S. and Russia. So there are precedents here.

There’s a question of whether we might be able to pull something similar with regards to China. My view is that this is essentially impossible; although we could potentially lure a few strategically important geniuses into defecting, China is simply far too big for brain drain to have a major negative effect on it in the aggregate.

More worrying, though, is the exodus of Chinese scientists from the United States, following a U.S. crackdown on spying and on Chinese government funding for U.S. researchers. Obviously it’s good if spies leave, but many of these scientists are probably not spies, simply researchers worried that they’re no longer welcome. If the decoupling of trade, investment, and technology transfer between the two countries continues, skilled immigration will become more of a zero-sum game; Chinese scientists leaving the U.S. will then constitute a costly brain drain. So we should be thinking hard about how to balance concerns over espionage with the need not to scare away critical talent.

---