Knocking on heaven's door

Immigration isn't just not bad. It's good.

Aug 31, 2021

Un grupo de inmigrantes contempla la Estatua de la Libertad desde Ellis Island, 1910. — A group of immigrants look at the Statue of Liberty. From Rue des Archives / Cordon Press.

I have previously made the case that immigration doesn't reduce wages or employment. Now I want to make a different case: that it is good. I’ll go over some other arguments about its “badness” as well, then focus on the positives.

Immigrants won’t take your job

The best way to examine the labor market effects of immigration is to focus on large refugee waves, or large and sudden inflows or immigrants. Why? Because most of these aren’t economic migrants, meaning they can’t self-select for labor market suitability, tend to just leave suddenly, and are such an extreme case that effects can be most clearly identified. Again, I have covered this in detail in a different post.

The paradigmatic case is the Mariel Boatlift (see the video above for a quick summary). In 1980, in a handful of months, over 125,000 Cuban immigrants arrived to Miami from Mariel Bay, Cuba. Most of them were low-skilled and many didn’t even speak English. Strikingly, a 1990 study by David Card found that the inflow had no effect on native wages or unemployment.

Now, another economist named George Borjas (get used to his name) published a series of papers that seemed to conclude the opposite, that the Boatlift hurt American workers - Blacks and Hispanics the most. But Borjas reached his conclusion by trimming the samples to meet certain criteria, which ended up making them so small no conclusions could be reliably drawn. Plus, the surveys used by Card, Borjas, and Michael Clemens’ refutation of Borjas all had a large change in their racial composition: post-Boatlift surveys had more Black people in the samples, which worsened the average labor market result because Black people are unemployed at higher rates and earn lower wages.

Generally, the first mistake immigration opponents make is to just look at macro-level variables: if wages go down and/or unemployment goes down after a big migration wave, it means that immigrants caused it. But this ignores that immigrants actually earn lower wages and have higher unemployment rates in general, due to their lack of connections and often mismatched skills, so the change could be due to composition and not actual outcomes for natives.

You can see this in the massive migrant wave from the USSR to Israel in the 1990s: more Russians in an occupation meant lower wages for Israelis, but only if the Russians joined the profession - meaning that the Russians worked in professions that didn’t utilize their skills properly, so their average wages (and those of all Israelis, considering Russians made up 15% of the labor market) went down. An example of this is that, among doctors, Russians tended to take lower paying positions, while native Israelis received promotions.

Immigrants simply don’t compete with average American workers - they’re either very low skill or very educated, so instead they complement American labor of the opposite kind. This is partly because immigration systems are actually biased in favor of highly-skilled immigrants, who are very productive.

Another potential explanation is that immigrants tend to push out native workers from the labor market - so the negative effects immigration restrictionists talk about don’t materialize because the amount of people in each labor market category doesn’t change. Well, if you actually look at the growth pattern of cities, it just doesn’t add up. Plus, it’s pretty likely that immigrants don’t move to places at random, but rather tend (or at least tended) to move to places with plenty of other immigrants. So this “pushing out” theory just doesn’t add up.

Finally, immigrants don’t tend to bring their own food, clothes, or housing - they consume things in their new home countries. Another study of the Mariel Boatlift found that employers tended to just hire the migrants at the same wages as natives anyways, because the higher demand for their goods meant they could afford it. People forget that, in a labor market, there’s a demand side as well!

The United States of Latin America

A new case against immigration popped up: the countries immigrants come from are really really bad. They are very poor countries, with lots of crime and violence. Governments are very weak and prone to corruption and authoritarianism. So what if taking in a ton of immigrants who have been shaped by those ideas makes their hosts countries more like them?

George Borjas, persistent bad guy of this piece, makes this case (at roughly the same time as his other case against immigration crumbled). It goes like this:

As the important work of Acemoglu and Robinson (2012) suggests, ‘‘nations fail’’ mainly because of differences in political and economic institutions. For immigration to generate substantial global gains, it must be the case that billions of immigrants can move to the industrialized economies without importing the ‘‘bad’’ institutions that led to poor economic conditions in the source countries in the first place. It seems inconceivable that the North’s infrastructure would remain unchanged after the admission of billions of new workers. Unfortunately, remarkably little is known about the political and cultural impact of immigration on the receiving countries.

Basically, Borjas makes the case that immigrants might carry the reason their home countries are failures with them, like a Trojan Horse of underdevelopment.

A first way to look at this is to wonder if immigrants actually differ substantially from natives, or if prospective immigrants do either. It doesn’t seem like they do! And natives don’t shift towards wanting a different sort of economy out of hatred for migrants either. From Clark, Lawson, Nowrasteh, Powell, and Murphy (2015):

That is, countries with more immigrants in 1990, experience stronger private property rights and less regulation over the ensuing two decades. (…) Immigrants do not appear to bring a desire with them for the corrupt, highly regulated environment from which they often emigrate. Nor do the native born respond to greater immigration by implementing a more stringent regulatory environment in order to preclude immigrants from participating in the economy.

But you can even assume that Borjas is initially right and reach the opposite conclusions - that is, that migration is good. If you take it as granted that immigrants are actually substantially different in their institutional beliefs than Americans, and knowing that immigrants tend to be able to transmit their norms, then you could make a different point - that these cultural norms will lower productivity.

Firstly, there’s plenty of evidence that immigration increases productivity - for example, because it makes natives specialize in tasks that mostly require communication skills, and take over jobs that don’t. Plus, how would even “cultural norms” affect productivity, which does have components such as social norms, institutions, or management - but also others like technology or knowledge. If El Salvador doesn’t have a software industry because it has too little know-how and too few computers, then Salvadoran computer programmers won’t turn the US into an authoritarian narcostate.

Also, migrants aren’t “normal people” - they tend to be the cream of the crop, the best and brightest. So maybe their cultural values are very different than those of other people in their home country. It would be laughable to say that Cuba would be a leader in e-commerce if Miguel Bezos took his family back to Cuba.

And institutions are socially determined. You don’t “bring” them with you, but rather, you adapt to them. To take an obvious example, very few countries drive on the left side rather than the right side. So it would make sense for immigrants to the United Kingdom to demand that the side of the road they drive on be changed. But instead, they don’t, and simply accept the new norm. If this assimilation is tied to the impact on productivity, then this impact (if it ever exists) is reduced over time - because norms take hold of the new people.

Basically, even if you accept the idea that migrants could have very different ideas than natives, then you actually have to prove that those ideas would take hold of their host countries - and there’s barely any evidence in favor of that. An obvious example are Indian migrants: India is a very poor country, and it’s pretty disfunctional in many ways (for example, it’s teetering closer and closer to being an actual autocracy). And yet, Indians are one of the highest income groups in most OECD countries, and they don’t appear to be frothing fascists either.

Welfare fridge magnets

Another anti-immigration argument is made by George Borjas (who else?): that countries with large welfare states attract immigrants who are intent on mooching off their host countries by taking as much free money as they can. This happens due to a variety of reasons: for example, that immigrants are in general poorer than natives, and that they plan their migration to countries with more generous safety nets - or, within the US, choose more generous states over less generous ones.

The most obvious reason this isn’t true is that immigrants frequently can’t even qualify for them, which is by definition true of illegal immigrants. Plus, they’re frequently much younger than established residents, since immigrants are almost always working age, which means they usually pay into welfare systems far more than they take out. In addition, comparing immigrants to average native citizens misses, once again, that immigrants are much poorer - immigrants aren’t especially different with regards to public services than similarly poor natives. Finally, this analysis doesn’t even apply to high-skilled migrants, who are pretty unlikely to ever need government benefits in general.

Additionally, his case doesn’t even work on its own terms. Immigrants tend to move to states where people of their same national origin live, and they stopped moving to high-welfare states like New York and California right after Borjas made his case - and instead moved to states like Texas or Arizona - that aren’t precisely Nordic. The rationale appears to be economic growth rather than generosity of welfare.

All in all, immigrants in the US actually use social services at far lower rates than natives. For example, the government spent about $40,000 dollars less on benefits on immigrants than natives, and if natives used Medicare and Medicaid at the same rates as immigrants do, the programs would be nearly 40% smaller. From Barrett & McCarthy (2008):

Regarding the question of whether generosity in the welfare system leads to an inflow of likely welfare recipients, the evidence is both thin (in the sense of a very limited number of studies) and weak (in the sense of results being either on the margins of significance or quantitatively minor). It could well be the case that other factors, such as the prior existence of networks or favourable labour-market conditions, have much stronger influences on location decisions

Job makers, not job takers

There’s another incredibly awkward fact immigration restrictionists have to grapple with: immigrants tend to become job creators, i.e. have extremely high rates of entrepreneurship. In the United States, 13.7% of people are immigrants - but they make up 20% of self-employed workers and 25% of startup founders. There are some very obvious examples of this - like second generation Cuban-American Jeffrey Bezos.

There are three main factors to this. The first is labor market discrimination. Many people are unwilling to hire immigrants, for a variety of reasons: no recommendations, lack of knowledge about qualifications or experience, plain old bigotry, etc. As a result, immigrants frequently gravitate towards “being their own bosses”, which doesn’t pose the same problems.

Secondly, many countries have very restrictive immigration policies, which tend to select for highly skilled workers or for entrepreneurs who secured funding in the host country (the start-up visa, which the US does not have). This means that the channel for immigrant entrepreneurship might simply be that immigrants are selected for their entrepreneurial abilities before even migrating.

Thirdly, there’s the fact that the kind of person who tends to pack up their whole life and move to a different country is probably more risk-tolerant than most. 40% of companies fail in the first three years. If higher willingness to take risks is correlated with both chances of emigrating and chances of entrepreneurship, which it seems to, then immigrants are simply disproportionately the type of person to try their luck.

Plus, given the fact that immigrants are frequently discriminated against, and that they have mismatched skills and fewer connections, then communities of immigrants might tend to cluster in certain industries. Ethnic groups already having cumulative experience and accumen for a certain type of business might allow other members of those groups to break into them more easily than other labor markets.

Immigrate, invent, repeat

Immigrants, being disproportionately highly skilled, are also very important for scientific progress and for market-centered inventions. For example, about 30% of Nobel Laureates made their contributions while living outside their home country. To take another example, German emigrants to the United States prior to the second World War increased US patent numbers by nearly a third, controlling for previous trends and for the fields they went into.

Currently, the main instrument for highly skilled immigrant “inventors” is the H-1B visa, whose recipients make up 24% of US workers with bachelor’s degrees and 47% of PhDs. More presently, for Indian and Chinese inventors who went into the united States through an H-1B visa, workers are much more productive in the US than in their home countries. Additionally, there isn’t much crowding out of native-born inventors, and no reduction of native inventors’ output. Put another way, it’s a win-win

Worldwide, and in general terms, 10% of patent inventors are immigrants (and emigrants, obviously), and 57% of them move to the United states, making up 18% of all patent holders who live in the country. Inventors generally pick countries with good economic opportunities for them, and where they speak one of the languages spoken.

Given that various scientific discoveries and technical innovations have proven to be exceedingly important for not only the US economy but for the global economy as well, it makes sense to allow immigrants in - they’re good for science!.

Good for their host countries…

Let’s just cut to the chase: immigrants are good for the economy even if they don’t invent groundbreaking technologies, or start businesses, or aren’t bad.

One of the main reasons is that, due to lower fertility, a smaller population, and higher educational attainments, low-skill immigrants tend to be very much employed - because they do jobs native-born Americans simply won’t do. Because those jobs are the kind that many higher-skilled workers rely on, it reverberates across the economy - to the tune of $5 to $10 billion dollars a year, per (shockingly) George Borjas. And, additionally, because immigrants have a level of skill that many Americans don’t, they’re also very mobile - that is, they move around a lot, to break into niches in the labor market. According to Borjas (2001):

The empirical evidence, therefore, suggests that immigrants may play an important—and neglected—role in the U.S. economy: They make up a disproportionately large fraction of the marginal workers whose location decisions arbitrage differences across labor markets. Put differently, immigration improves labor market efficiency. Moreover, it turns out that part of this efficiency gain accrues to natives, suggesting that existing estimates of the benefits from immigration may be ignoring a potentially important source of these benefits.

Actually, Borjas also makes the same case in a different way: because immigrants already incurred all of the costs of moving, they can move to wherever they like, which means they can exploit economic opportunities as much as possible. And not only they have different sets of skills than natives, and take different types of jobs - but their skills complement those of natives and allow for specialization. He does have a caveat, but it is that the effects are “too small” - which is, bluntly, a massive cope.

There is another effect to consider: slower population growth seems to be driving slower economic growth. I have gone over this some in my competition and concerntation post. The explanation is, revealingly, labor supply: because it’s growing so slowly, fewer firms are created because there are fewer entrepreneurs. Plus, bigger firms can scoop up workers by paying them comparatively more than the smaller firms. There are no new firms being created, and many smaller ones can’t compete with bigger ones and go out of business. This leads to an ossified economy, with massive ancient firms paying a baffling combination of bloated salaries and uncompetitive ones. Plus, large companies are investing less and less, even on things like R+D, so the economy has less innovation. As a result, output grows very slowly. Allowing immigrants, or more immigrants, could result in a virtuous cycle of firm rejuvenation, competition, and innovation.

Even without big brained firm demographics changes, and without greasing the labor market, immigration is just good. From Ortega and Peri (2001):

We find that, already within one year, employment responds to new immigrants one for one, and capital adjusts in order to maintain the capital labor ratio. We do not find any significant effect of immigrants on total factor productivity. These results, taken together, imply that immigration has no negative impact on average wages, or on income per worker in the short run (one year) or in the long run (five years). The inflow of immigrants only increases the overall size of the economy without altering the distribution of income between workers and capital owners. This is due to the fact that capital owners respond efficiently to a larger labor pool by investing more.

Part of this can come down to what kind of labor market institutions are present: if they are very restrictive, so hiring and firing are costly, entry to professions is difficult, and wages are consequently very rigid, then immigrants can have negative effects on the outcomes of natives. If the labor market is competitive, however, there wouldn’t be a “dog eat dog” market.

Plus, many immigrants are highly skilled, so if they come from STEM and similar backgrounds, may contribute valuable knowledge to their home nations even after emigrating. And because the capital/labor ratio gets lower wen there’s more labor, added on top of the higher demand for goods, then this incentivizes investment, which is generally considered to be a good thing.

This is very much in line with a series of ideas about the labor market, where enhanced competition for labor incentivizes investment in technology that makes labor more productive (and raises wages). I won’t get into it, but it’s a plausible enough idea.

… and for their home countries

We usually leave this out, but when immigrants move, they move from somewhere as well. We can obviously believe that moving from poor to rich countries is benefitial for the migrants themselves, because otherwise they wouldn’t do it. But what about their compatriots who don’t?

Well, for starters, immigrants are far more knowledgeable of their home countries than natives, which means that they can ease transactions and investment simply by providing their know-how. And they might pour some of their savings (especially if they’re planning on returning) into their home countries, which is generally a positive for poorer nations. This is counteracted, however, by lackluster records of financial openness and poor allocation of capital in developing countries. Additionally, migrants may increase productivity in their home countries too: by introducing their compatriots to new technologies or new forms of organization and doing business.

A common counterargument is the famous “brain drain” - that developing countries would lose all of their most promising minds to the rich nations and thus not be able to develop. This seems like an eminently reasonable concern.

It’s very obvious that development only matters insofar it benefits people and not vague polities - so unless there is compelling evidence that emigration of skilled workers harms other people in the countries they come from, then it’s just unconscionable to prevent specific individuals from improving their wellbeing. And, since people from poor countries have all as much agency as anyone else (nobody would say that Doctors Without Borders is brain drain), then you need to make an affirmative case that is solidly grounded that the choice of emigrating is an inherently harmful one.

There is also an intereting point to consider: people don’t just emigrate because of money. They might belong to prosecuted minorities, or feel unsafe from crime, illness, or war, or might want better outcomes for their children. Migration is itself strngly correlated to quality of life, which results in people not wanting to live somewhere if it’s a bad place. Movements of skilled people from poor countries to rich countries follow remarkably similar patterns to movements of skilled people within rich countries, from rich to poor areas. Nobody in their right mind would claim that an engineer moving from an impoverished Native American reservation to Boston causes the reservation to remain poor. There seems to be a “migration curve”, that countries expel more and more migrants as they get closer to upper middle income status, and then stop. Exactly why isn’t that well known.

But going back to the question, it’s not even clear that letting skilled professionals leave poor countries is bad. They certainly don’t make the full product of their labor, and saying “Mozambique would have better health outcomes if it had more doctors” is a laughably naive statement - underdeveloped healthcare systems are a reason why doctors and nurses leave the third world, not a consequence of it. Within African nations, for example, most physicians are concentrated in small urban areas that are much richer than the rest of the country - so migration isn’t exactly harming underserved populations. Migration doesn’t cause underdevelopment in the same sense that inflation isn’t caused by decisions to raise prices - you’re just looking at the surface level. Tautological statements like “there are few high tech firms in developing countries because all the workers skilled in those disciplines are leaving” don’t actually help to understand how to develop those industries - because the causation is precisely backwards.

And even if it were true that skilled workers had large externalities and that their loss was primarily caused by emigration, you’d still have to defend the fact that restraining their choices so much is valid. There is some evidence that women’s integration with the labor force led to a reduction of the quality of education, because the most skilled women pursued other careers. It wouldn’t be very hard to make a case for sexism based on “the greater good”, but reversing the gains in women’s rights is obviously a non-starter. Cuba’s medical missions based on forced labor and coercion are considered a human rights abuse, so it’s not clear why coercing third worlders to serve their own nations isn’t too.

Conclusion

Lifting all restrictions on the movement of people would be one of the most efficient ways to boost growth in the planet. Lifting all barriers to international trade would boost global output by, at most, 4% (most estimates are closer to 1% or 2%). Lifting all restrictions to capital flows would raise it by another point or so. The smallest increase in global GDP from reducing the barriers to labor mobility is 67%, with estimates all the way to 150%. There are differences of one-thousand times between the earnings of people in rich and poor countries - you could scarcely make a case that people in rich nations are one thousand times more productive.

Allowing emigration from rich to poor countries would be the most effective way ever devised to reduce global poverty and improve world economic output. Only 14% of people in developed nations live on an income below the global average ($16 dollars per day PPP), but 82% of the global population does. Rich countries are incredibly rich - and the losses from migration are primarily due to policy decisions from rich nations themselves, and not from inherent qualities of migration.

Now, we can’t just ignore the current context: Afghanistan. As many as 200,000 people might want to flee. And I say, take them! Besides from the West actually bearing a lot of responsibility over the situation (either by entering, staying, leaving, take your pick), it’s also not bad for them, and good for them. Many of the studies mentioned above, starting with the Mariel Boatlift, were of refugee waves. Israel took in double digit percentages of its population every year when the USSR fell and not much happened.

We also can’t ignore something else: immigration restrictionism has an incredibly racist history. The US, like most Western countries, basically had a policy of open borders until the 1920s - except for Asians, because of, once again, racism. Then, in 1924, the border was slammed shut for basically everyone, and only a trickle of people were let in - and no Jews, no Eastern Europeans, and no Southern Europeans. The effects were not good: manufacturing cities suffered from the job losses caused. Meanwhile, internal migrants (who are literally as well inserted as possible) still didn’t cause any losses - many of them were Black, and fled from the rural and violent South to the prosperous North and Midwest.