Increasing Skilled Immigration is a Terrible Idea
A Plea to the Tech-Right
When asked, 71% of Trump supporters want to increase high-skilled immigration to the United States. This isn’t a priority for most of Trump’s coalition, but one wealthy and disproportionately influential pro-Trump faction has consistently and publicly advocated for increasing high-skilled immigration, to the point that Trump himself has endorsed giving green cards to all foreign students. This is the libertarian-adjacent tech-right, whose support for Trump is principally motivated by regulations, freedom of speech, and wanting to avert California-style political dysfunction in the rest of the country. They are making a fatal mistake.
The case for high-skilled immigration is simple. National IQ is the best predictor of economic growth and development, explaining roughly 70% of the variance in GDP per capita by itself1. Combine this with the allocative benefits underlying mainstream economists’ support for immigration2 and the potential innovative benefits of getting more smart people into cognitive clusters like San Francisco, and the argument for admitting high-skill (high-IQ) immigration is strong. But it is wrong. Skilled immigrants are not just labor and innovation machines. They, and their children, are influential political and cultural actors whose effect in those realms is to destroy those things that make America exceptional.
America is Exceptional
The tech-right appreciates some of America’s virtues, particularly our economic dynamism and freedom of speech, often contrasting the United States favorably with the rest of the Anglosphere or with Europe. They are right to do so.
Economics
National IQ may explain most of the variance in wealth between countries (with a history of Communism and large resource endowments accounting for most of the remainder), but the tails come apart at the high end. Among rich countries without large resource rents or a history of Communism, the United States stands out as being much wealthier than IQ alone would predict. Meanwhile, the non-US Anglosphere (United Kingdom, Australia, New Zealand, and Canada) underperforms, being poorer than either the United States or continental Western Europe, despite its higher IQ. The gap between the Anglosphere and the United States is even more distinctive if you look at productivity rather than GDP per capita3.
The comparison with the non-American Anglosphere is particularly instructive, because these countries are very similar to the United States culturally, genetically, and institutionally. Britain, Australia, and Canada were about as wealthy as the United States as recently as 2008.
All three of these countries have embraced a policy of enormous amounts (far more, per capita, than the United States) of skilled, legal immigration (Australia, Canada, Britain), particularly from India and China. All three have seen stagnant (per capita) economies, personal incomes, and skyrocketing housing prices. One might object on the grounds that immigration to Canada, Australia, and Britain isn’t really skilled (though it is among the highest IQ in the world in Canada and Australia), but this is what “increase skilled immigration” looks like when refracted the bureaucracies that will inevitably be tasked with implementing it. We should treat these countries as a cautionary example; we could very easily end up in their shoes.
Freedom of Speech
American speech is the freest in the world, a fact recognized by much of the tech-right. A comparison to our Anglosphere cousins is instructive4. The United States is not jailing tens of thousands of people or sending police door-to-door for political speech as the United Kingdom does. We are not calling on speech to be regulated as a weapon of war, as the (now former) Prime Minister of New Zealand did in 2023. We do not have Canada’s hate speech laws5 or widespread government subsidies of pro-government media outlets. American scientists are not prohibited by the government from communicating their results to the public, as in Australia. We can justifiably be proud of our strongly pro-free speech legal environment.
But we can’t rest easy. America’s free speech maximalism comes from Supreme Court decisions based on the First Amendment. And the Constitution6 does not enforce itself. The Soviet Union infamously had constitutional guarantees of freedom of speech; without a strong pro-free speech elite culture the letter of the law is meaningless. All it takes is the wrong judicial appointment from a pro-censorship Democratic Party and it could be removed. And, when the bureaucracy and tech company employees are in alignment, the government can and has worked with tech companies to censor without tripping legal safeguards. It’s no surprise that, despite the First Amendment, nearly half of Americans do not feel free to speak our minds. Maintaining American freedom of speech requires that tech workers, judges, lawyers and bureaucrats are broadly OK with it. If the composition of those groups becomes sufficiently opposed, it will be disappear. Right now, the United States has one of the most pro-freedom-of-speech populations in the world7—but change Americans and you will change America.
Disaggregating “Skilled Immigrants”
Just as “immigrants” can be useful decomposed into skilled and unskilled, “skilled immigrants” conceals an important divide. Using innovation as a proxy for skill8, we can reasonably divide skilled immigration to the United States into two groups: white and Asian. Excluding the “other” category for lack of information, we can estimate that European immigrants make up 11.9% of US immigrants, but 39.3% of US immigrant innovators. Meanwhile, Asian immigrants make up 30.9% of non-other US immigrants and 55.6% of non-other US immigrant innovators. Combined, these two groups make up 42.8% of non-other US immigration and 94.9% of non-other US immigrant innovators. Skilled immigration from elsewhere is a rounding error.
Of the two groups, Europeans are more likely to be innovators (8.2 times more represented among innovators than among the US population vs 4.5 times for Asians), but both are very disproportionately inventors and comprise a significant chunk of American innovation. This makes sense; immigration from both places is highly selective9, but more so from Europe (and Europe is more innovative per-capita to begin with).
As a rule, European immigrants are more accomplished than their Asian counterparts10, as are their kids, but both make significant contributions. However, neither skilled immigrants nor their children confine themselves to laboratories. They become professors, activists, lawyers, judges, politicians, bureaucrats, and journalists. In other words, they get involved in politics. Unlike economics or innovation, which is positive-sum, politics is zero-sum. More influence for group A means less for group ~A. And unless stopped, an Asian elite will destroy what makes America exceptional.
The Politics of Asian-Americans
Economics
Asian-Americans are an incredibly left-wing group11, so much so that despite their economic success, Asians are much more pro-government-intervention than whites, in line with blacks or Hispanics (who have much more to personally gain). For instance, a supermajority of Asians (66%) believe that the government should do more to solve problems, as opposed to doing too many things better left to businesses and individuals (differential: +32, vs +10 for whites).
Similarly, a supermajority (70%) of Asians choose government regulation as necessary to protect the public interest over government regulation of business usually doing more harm than good12 (differential: +41, vs +8 for whites).
And when explicitly asked, Asians are much less pro-capitalist (+10 over socialism, +5 among those with a strong impression) than whites (+31 over socialism, +20 among those with a strong impression) than whites.
Pro-capitalist boosters of Asian immigration might note that they are still slightly more supportive of capitalism than socialism, but what matters is the difference from the status quo. The balance of power among the American elite determines the current equilibrium between a market economy and state intervention. Mass immigration of an elite group 20+ points more favorable to state over private action than the people they are competing with will shift things in the direction of intervention. The tech-right is rightfully concerned with creeping regulationism and socialism. They should take heed.
Freedom of Speech
Americans, as a group, are very pro-freedom-of-speech by world standards. But that conceals considerable interethnic variation. Descendants of northwest Europeans are, broadly speaking, the most pro-freedom-of-speech, while Asian groups are much less so.
The tech-right is rightfully dismissive of “misinformation studies” for being an exercise in partisan redefinition of truth. But not everyone shares their views. When asked about protecting press freedom vs curbing false info, Asians are 12 points more likely than whites to favor government censorship.
There is a connection between increased Asian dominance of tech companies (particularly Google) and these same companies’ shift from a broadly libertarian ethos of freedom-of-speech to focal points of global censorship. It is not a coincidence that the woman who banned Trump from Twitter, Vijaya Gadde, is an Indian immigrant.
Affirmative Action
The tech-right rightfully despises Affirmative Action/DEI as an anti-meritocratic institution-wrecking racial spoils system. It may surprise those familiar with the Asian-focused fight to eliminate Affirmative Action in elite college admissions to know that Asians support Affirmative Action by large margins.
The resolution to this seeming contradiction is that Asians oppose Affirmative Action in college admissions, where it hurts them, but support it in other areas, like government or corporate hiring and promotion and above all taxpayer-subsidized low-interest Small Business Administration loans, where it benefits them. There is no universal principle involved.
Views on America
We only have one America. If you value America’s unique freedom and economic and technological development and dynamism, the wellbeing of the country is paramount. There is no America II to flee to. To that end, the American elite must be patriotic, that is, acknowledge America’s virtues13, seek to maintain and improve them, and have a stake in the country.
As it happens, Asians feel the least positive about America of any racial group. Asians are least likely to believe that America is better than most other countries, to feel proud or grateful to be American, to feel that life in the United States is better than in the rest of the world, or to feel attached to the United States (More in Common, 2020). I initially found this counterintuitive, since Asian-Americans are majority foreign-born, and so personally moved to America because it was better than their home country, but it’s true. As engineers, workers, or scientists, this is fine, but it’s not the profile of a group you want running your institutions.
Change Over Time
This might be bearable if it were a transient phase that faded as Asians accustomed themselves to the United States. Unfortunately, the reverse is true. Asian-Americans in fact de-assimilate over time. Or rather, Asians assimilate to the leftist, anti-American norms14 of Asian-Americans15.
Every political position noted here will get worse with time rather than better. The left-wing views of present Asians are well to the right of the views of future Asians.
Conclusion
“The supreme function of statesmanship is to provide against preventable evils.” - Enoch Powell
Skilled immigrants and their children will, if permitted to participate in the political process, form a disproportionately influential elite16. Under the current system, skilled immigrants are majority Asian, and if the results of similar experiments in the rest of the Anglosphere are any guide, any increase in skilled immigration will be even more Asian. Asians are, relative to whites, pro-regulation, pro-socialism, pro-Affirmative Action, and anti-freedom of speech. The predictable effect of ramping up skilled immigration will therefore be to shift the US elite, and thus the country, even more in the direction of regulationism, DEI, and censorship.
You can have capitalism, meritocracy, freedom-of-speech, and patriotism, or you can have enormous amounts of (implicitly Asian) legal, skilled immigration. But you cannot have both, not under an immigration system remotely similar to the existing one17. To support increasing skilled immigration while viewing American leftism as a civilization-wrecking scourge, as Elon Musk and many others in the tech-right do, is shooting yourself in the foot18. Rather than lobby to expand skilled immigration, the tech-right should either join broad-spectrum immigration restrictionists and accept losing the benefits or work to create a different skilled immigration system that does not lead to the Asianization of the United States elite19.
If you’re not familiar with this, or don’t believe that causality goes from intelligence to development rather than the other way around, I recommend reading Cognitive Capitalism, National Intelligence Really Is the Best Predictor of Economic Growth, and Garret Jones’ book Hive Mind, in that order.
Which in practice do not pan out for low-IQ immigration, for the reasons underlined in the linked texts on national IQ and economic development. People are not labor machines; genetics, culture, politics, and (slowly-growing) capital and land per person all matter.
In my opinion, GDP per capita is the more relevant metric for this article, because throughput isn’t free. Continental Europe is close to the US on productivity, but doing more with more, as the United States does, is impressive and noteworthy, especially for the tech-right, which is full of very hard working people who want to accomplish great things.
The rest of the world is generally as bad or worse.
To give you an idea of the Overton Window for freedom of speech up north, the government of Canada recently attempted to amend these laws so that a judge could order house arrest on suspicion that you would commit hate speech in the future.
Those wishing to rely on Constitutional protections alone should recall that nearly half of their countrymen do not view the document as sacrosanct. Per Cato:
“A significant minority (44%) of Americans would be open to "writing a new American constitution to reflect our diversity as a people," while 56% would oppose writing a new constitution. This builds upon the work of political scientist Eric Kaufmann who found a similar pattern of results.
Democrats stand out with 63% who favor writing a new constitution, compared to 16% of Republicans and 37% of independents. A majority (54%) of Americans under 30 also favor designing a new constitution. However, support drops among older cohorts including among those aged 30-44 (47%), 45-54 (40%), 55-64 (27%), and 65 and above (25%). Black Americans (73%) and Asian Americans (60%) also support designing a new constitution. In contrast, majorities of Hispanic Americans (56%) and White Americans (68%) oppose designing a new constitution.”
Given how difficult amending or abolishing the Constitution is, this doesn’t mean the text itself will be altered. But people who want to get rid of the Constitution will not interpret its text as intended or written; reinterpreting the Constitution into requiring the ideological cause du jour is something that’s happened many times before.
This has the partisan, racial, and sex loadings you’d expect. Whites, men, and Republicans are more pro-freedom-of-speech. White Americans are probably the most pro-freedom-of-speech ethnic group in the world by a significant margin.
Other proxies, like education, are bad because foreign degree holders are not as capable as American degree holders. Immigrants with foreign post-secondary degrees are less literate and worse with computers than the median native-born American, and only barely more numerate. Given how low the median American is, that’s very bad. And this doesn’t disaggregate by race; native-born whites would be much higher.
Can’t emphasize this enough that increasing immigration will reduce selectivity. That Asian immigration is selective is not a fact of nature and massively expanding it will, with certainty, make it less so. But that’s not the main point of the article.
Another example. 2/3 of US Nobel Prizes in 2023 were won by immigrants. All of them were European.
It should go without saying, but I’ll include the disclaimer here that group-level statements don’t necessarily apply to individuals. But the immigration bureaucracy can’t possibly evaluate all the individual details of each applicant; broad, group-level characteristics are what’s relevant to policy.
I believe both of these statements are true and there’s no contradiction between them. The point of the poll, however, was for respondents to pick which one they think is more true.
Because without knowing that America is better than other countries in specific ways and for specific reasons, you can’t keep it that way.
It’s not just politics. Asians have the lowest bastardry rates of any race in the United States, but this is only because foreign-born Asians have children in wedlock. American-born Asians have kids outside of marriage at about the same rate as whites.
A fun alternative way to view this is to look at those surnames for which partisanship correlates most strongly with age (ie, the difference in support for Democrats vs Republicans between young and old people with that name is largest). As it turns out, these are all Vietnamese surnames; first-gen Vietnamese refugees fleeing Communism are fairly right-wing, but their kids are extremely left-wing Asian-Americans.
To get an idea of the speed of this, the incoming Harvard class is 37% Asian. Unless we change course, that’s a pretty good estimate for what the US power elite will look like in 20 years. Up from near-0 at the start of the century. And it will keep going up.
An explicitly race-based system that only allowed in skilled whites or a labor visa system which allowed people in but did not grant them or their descendants citizenship or the opportunity to participate in the political process would not have the same problems. But neither is in the cards.
There’s no good place for this in the main article, but a common argument among the AI-aware (which includes much of the tech-right) is that the United States should massively ramp up skilled immigration (of AI researchers) from China to "win the AGI race." Most of the arguments in this article don't apply here. If you (1) believe that the Singularity is imminent and (2) strongly prefer it begin in the United States over China, it makes sense not to care about the long-term societal outcomes (it takes a few years for immigrants to become politically relevant, by which point human history will have ended one way or another) of skilled Chinese immigration. But this argument is still wrong on its own terms.
To the extent that the US-China AGI race is even real (the low price of cutting-edge GPUs in China, despite low supply thanks to US export controls, suggests demand is low, which would imply that all serious hyperscalers are American), the US advantage is in algorithms and ideas. The Chinese advantage is in actual industrial production; China can build out its data centers and electricity supply several times faster than the US can. The obvious Chinese path to victory is to steal American algorithms or model weights, then out-build us in the physical world. And Chinese researchers in important American labs is by far the easiest way for the PRC to do this (for reasons that should be obvious: not only is there national loyalty, but Chinese thieves have the option of returning to China, guaranteeing physical safety and the opportunity to get rich using American secrets in the enormous Chinese market. Non-Chinese can't do this.). China has successfully stolen American secrets to the tune of hundreds of billions of dollars across dozens of industries, overwhelmingly through the Chinese diaspora. It’s wishful thinking to expect AI to be different, and ramping up Chinese AI researcher immigration makes this attack surface much wider. There's sometimes a bad analogy made with heavily-Jewish scientific refugees from Nazi Germany in the 1930s. This group had strong ideological reasons to hate the Nazis and thus were not a serious espionage risk with respect to Germany. This is not the case for Chinese immigration to the US, which is mostly economically driven.
Any plan to use Chinese immigration to help America technically compete with China (not just in AI) is doomed by this.
Aside from explicit racial restrictions, 1924-style ethnic quotas, or a guest worker program, one possible route would be to index a country’s immigration quota to its GDP per capita. This would naturally favor skilled immigration from Europe, which is already better (more innovative and entrepreneurial) than its Asian equivalent.
The subtext of this article is that very specific political views are genetically determined, which is nonsense. Yes, it has been shown that where you are on the left-right spectrum is partly hereditary, but views on specific policies are highly malleable over time within this spectrum.
I'm still not convinced that it's practical to cut off *all* Asian immigration.
Surely, it should be feasible to only permit high IQ, right-leaning Asians to immigrate to the West. That'd be a great way to gain the best of both worlds. https://zerocontradictions.net/faqs/immigration#unfavorable-races