Asian immigration and the signaling model of education
The elephant in the room of the college admissions grind
Written by Arctotherium.
There’s a small cottage industry of articles, blog posts, and essays pointing out how much more difficult, time-intensive, and “grindy” upper-middle class American childhoods have become in the 21st century [1][2][3][4][5][6][7][8][9][10][11]. All of these identify the same cause: elite university admissions are far more competitive. This has led to much more intensive helicopter parenting to stop kids from falling behind in the increasingly important education-status race. It’s not just money. The social decay afflicting America since the 1960s is far worse among the non-college-educated (as Charles Murray showed in Coming Apart) so avoiding downward mobility is much more important than it once was.
Helicopter parenting has gotten so intense that it even continues into university, turning what was once the start of independent adult life into a further extension of childhood. The effects of this are widely lamented: the pressures of Ivy League admissions are said to be crushing kids into status-obsessed zombies. And the effort it requires from parents is even blamed for falling birth rates.1 Setting aside the broader societal implications and long-term effects, isn’t making kids unnecessarily miserable for 13 years of their lives bad enough in and of itself?
One might reply that it’s all worth it. Feeding 30% of kids into the college admissions meatgrinder might be bad, but stagnation is the default state of humanity and avoiding that is imperative. If it takes heroic efforts on the part of 12-year-olds to develop the skills needed to keep technological civilization running, then that’s a sacrifice that must be made. But is that really what’s going on?
The signaling model of education
The signaling model of education states that the labor market returns of education2 are primarily due to signaling, rather than human capital (education making students more productive) or pure selection (highly educated people being paid more because they are more productive for reasons that have nothing to do with education). That is, about 80% of the education premium is attributable to education revealing you have traits that make you more productive than your peers—as opposed to anything you actually learn in school.
There are a few big lines of evidence for this. First: “sheepskin effects”. A graduation year (in high school or college) is worth 6-7 times more on the labor market than a non-graduation year, and graduates (of both high school and college) earn significantly more than otherwise identical non-graduates. (This is not just selection—it applies to identical twins too). If labor market returns to schooling were mostly about human capital, it’s hard to explain why graduation years should be worth so much more than any other year.
Second, when tested, most students retain very little of the specific skills or knowledge they “learned” in school (and if asked, they tell you as much). Unschooled kids, who received no formal instruction at all, score only a grade level behind their schooled peers on standardized tests. If students don’t learn much in school, then it’s hard to see how it can significantly improve their labor market skills.
Third, the skills and knowledge students learn in school tend not to be related to whatever jobs they do afterwards. This is often used to attack the humanities—knowing the proper structure of a sonnet is irrelevant to essentially all jobs—but it applies to STEM subjects too.3 Only an infinitesimal fraction of people will need to factor cubic equations or understand the water cycle in their professional lives.
Fourth, countries occasionally implement education reforms where the curriculum is increased or decreased by a year. Yet these have no effect on the earnings of students in the relevant cohorts. This is compatible with both signaling and selection explanations, but not with the human capital account.
Fifth: credential inflation. As education expands, the same jobs, requiring the same skills, tend to demand higher credentials from applicants. Jobs that used to have no education requirements require high school graduation; jobs that used to require high school graduation require college; and so on. People have been pointing this out since the 1970s!
Sixth, international comparisons. Conditional on highly g-loaded test scores, additional years of education do not predict economic growth. Conditional on years of education, highly g-loaded test scores do predict economic growth. This strongly suggests that national intelligence, and not education, is primarily responsible for the widely-acknowledged link between educational attainment and economic development.
Likewise, because of foreign aid, sub-Saharan Africa has a more comprehensive schooling system than would be expected from its widespread poverty and weak states, but this has not led to economic development. (However, education does tank fertility, so it may be beneficial given how important natural resources and arable land are in sub-Saharan Africa).
A more comprehensive argument for the signaling model of education is laid out in Bryan Caplan’s book, The Case Against Education.4
Implications
Education is very expensive in both dollars and time (it takes 17 years of a typical American’s life, from kindergarten to college graduation, to have a decent shot at middle-class status) and on top of that appears to causally reduce fertility (by delaying both pairing and childbearing) and to increase parents' divorce risk.5
The chief implication of the signaling model is that investing in education is individually rational, but collectively destructive. You, personally, can greatly increase your lifetime earnings by getting more education than your peers. Yet if everyone does the same, the signal degrades and to stand out you need even more education. There’s no upper limit: the amount of education required can keep increasing and consume ever more money and time. This is analogous to an arms race (or if you prefer evolutionary biology, a Red Queen’s Race). Hence more investment in education makes everyone worse off.6
In the long term, the only beneficiaries of expanding education are people within the education apparatus itself.
In the modern7 context, this implies that the “cultures that value education” so beloved of conservatives are actually bad. Here’s Thomas Sowell:
Lebanese immigrants to various countries have, in their early stages, included many who were illiterate and few who were highly educated. Nevertheless, they—like the Chinese, the Jews, the Armenians, and others—came from a culture that valued education, even when most of them had very little education themselves. Nor was education the key to their initial rise. Typically it was after becoming established economically as entrepreneurs that middleman minorities could then afford to dispense with their children’s labor in order to let them go to school instead and, still later, pay for them to continue on into higher education.
The signaling model suggests that when groups prioritize education and invest heavily in it, life gets worse for everyone else. This—in addition to envy and sour grapes—is why “strivers” and “grinds” are often hated.
Goodhart’s Law
Goodhart’s Law states that when a measure becomes a target, it ceases to be a good measure.
Whenever there is a signal for desirable traits, prospective signalers can focus on either (1) improving those traits or (2) optimizing for the signal itself, making it a worse signal of the underlying traits (i.e., “Goodharting”). Educational attainment has been a target for a very long time, so it’s not surprising that it has been aggressively gamed. From a student’s perspective8, there are various ways to optimize for the signal, rather than for actual learning:
Cheating. This could involve copying homework, using LLMs to write essays, plagiarism, paying College Board employees for early access to the SAT, or even paying people to take tests for you (a common practice in India).
Taking standardized tests again and again to grind out a few extra points with superscoring and chance. The same applies to intensive test prep such as memorizing thousands of vocabulary words or learning various tricks to move faster on that SAT or ACT (without any difference in knowledge or ability). Test prep, on average, doesn’t do much… but that average includes dozens of people who receive a few hours of instruction for every person like my high school friend who took 60 practice SATs during junior year.
Memorizing everything needed right before the test, then forgetting it all immediately afterwards.
Sacrificing an extra-scholastic life to study harder. Some studying may contribute to real-life skills that are retained (though, given the signaling model of education, most of it doesn’t).
Many, many things related to extracurriculars in the college admissions process. Starting a business as a high schooler isn’t gaming the system, but starting a fake one to look good on your college application is. Joining clubs out of genuine interest isn’t gaming the system, but joining them to get into college is. Rinse and repeat with sports, music, charity work, and so on. Another of my high school classmates boasted about entering impressive-sounding but uncompetitive science fairs to pad his application.
With the exception of cheating, none of these is exactly immoral. Students didn’t create the system they operate in, and kids are told to do well in school from a young age by almost every authority figure (and they are right to listen). Nevertheless, all of the practices listed above reduce the prospects of peers who don’t engage in them, and once they become widespread, make everyone’s lives worse.
Asian grinds?
If you’ve ever spent time tutoring, attended a college admissions prep course, gone to a selective institution like Stuyvesant High School, or done STEM at a selective college, you might have noticed a glaring omission in all of the articles linked in the introduction.9 Not one of them mentions Asian10 immigration—except in the context of Asians being harmed by affirmative action at elite colleges.
Stereotypes suggest that Asian immigrants put much more effort into Goodharting education (and other zero-sum status signals) than other groups in the United States. Don’t take my word for it: Yale Law professor Amy Chua wrote an entire book11 about how she and other Chinese immigrants aggressively (some might say abusively) parented their daughters to maximise status.
As is usually the case in the social sciences, stereotypes are backed up by data. Although Asian Americans (though not every subgroup) have a small but consistent intelligence advantage over their white peers, this cannot explain their enormously greater academic achievement. What does explain it? Academic effort. This might seem laudable, but investing more in a negative-sum signaling contest is a collective vice, not a virtue.
Everyone loses from Asians putting more effort into academics. It’s easy to see how white kids shut out of opportunities by lower GPAs (that are totally unrelated to those opportunities) are hurt, but in turn Asian kids feel worse about themselves, spend less time with friends, and are not as close with their parents as their white peers.
Of the portion of the academic gap that can be explained by easily-measurable socio-cultural factors (which is about a third), 69-80% is explained directly by immigration status. This grind culture is found in first- and second-generation immigrants, and I would expect it to dissipate by the third generation. (Sample sizes are too small to check, but Jews had a similar reputation in mid-20th century America and don’t any more). Pro-immigration conservatives often use this focus on education status-signaling as evidence of immigrant moral superiority, but it is in fact destructive and wasteful.
International evidence
As you’d expect from the persistence of traits, these findings from the US match the international evidence.
Korean private tutoring schools or “hagwons” are infamous. About 78% of Koreans between first and twelfth grade attended a hagwon in 2022, as did 83% of five-year-olds in 2017, and about 95% of Koreans do at some point in their student lives. The average hagwon student attends for 7.2 hours a week, in addition to regular studies and homework, and as a consequence the average South Korean student works 13 hours a day. South Korea spends three times the OECD average on private schooling as a percentage of GDP, the highest in the world. These thousands of hours of studying are all to get high scores on the CSAT, the standardized test that determines most college admissions in South Korea. Government regulations and crackdowns to try to stop South Korean parents from spending so much time and money on wasteful zero-sum signaling have thus far failed.
South Korea is the most extreme case, but it isn’t alone. About 73% of junior high schoolers in Taiwan attend some form of cram school, for an average of 6.24 hours per week. About 70% of Singaporean students do the same. China is much poorer than South Korea, Taiwan and Singapore, and therefore has far fewer resources to spend on costly signaling. Yet the Chinese education industry grew at 11.3% per year between 2019 and 2023. This provoked a massive government crackdown in 2023–24 that banned people from offering classes in English, Chinese, or mathematics for profit. As with South Korea, demand is so high that the ban led to an explosion in underground quasi-legal tutoring. Note that China also relies primarily on standardized tests for college admissions.
China and India are both infamous for cheating—to the point that there have been riots by students in both countries when students were prevented from cheating by investigators. (China now threatens students with jail time for cheating.) International Asian SAT takers are also notorious for cheating1213, with common methods including impersonation14, purchasing tests from insiders at College Board, buying questions and answers from test takers in other time zones, and smuggling in vocabulary lists. The persistence of traits would suggest that this doesn’t stop when they enter the US, and indeed anecdotes from teachers suggest that recent Asian immigrants are dramatically overrepresented in cheating rings.15
The SAT
Some would-be education reformers—trying to fight massive affirmative action and the rise of extracurricular activities sucking up ever more of children’s time—have proposed relying exclusively on standardized test scores for college admissions. Korea, Taiwan, Singapore, China and India do rely much more heavily on test scores, and the result is that the focus of grinding and cheating shifts to the tests, as opposed to the extracurriculars. It’s not that standardized tests are impossible to game16; it’s that the effort isn’t usually worth it in the US because the rewards are too low. But this has started to change, with Asian test prep practices becoming increasingly common (sometimes in the form of overseas extensions of existing companies in Asia).
If you follow SAT scores, you’ve probably already seen the chart below. The SAT score gaps between every major race in the United States have been roughly constant since the late 1970s (Native Americans have small samples), with all trending up and down together in line with test changes, external factors such as the COVID lockdown, the rise of the test prep industry, and other things that might affect scores—with one glaring exception. Asian-Americans have gone from testing approximately equal to whites to breaking away from the pack like Secretariat at Belmont, to the point that they are now about 100 points ahead on average.
By itself, a 100 point gap on the SAT wouldn’t be shocking. 100 SAT points corresponds to 5.24 IQ points assuming a standard deviation of 229 and an IQ–SAT correlation of .80. And in high-quality samples the white-Asian IQ gap in the United States is 2-5 points, depending on the measure. But the SAT has a fairly low ceiling, and as such the Asian score distribution is truncated and not normally distributed.
If you move backwards from SAT scores to IQ scores, this would correspond to a 10 point IQ gap between Asians and whites in the US, a full 5–8 points larger than direct measurements show. The result is that Asians absolutely dominate the upper reaches of the SAT. A remarkable 25% of Asians in Michigan (which forces all high schoolers to take the SAT and hence is more representative than other states) scored between 1400 and 1600 versus 4% of white students.
This is not a criticism of the SAT. The SAT is not intended to measure intelligence—it’s intended to measure college readiness, and Asians get just as good grades in college as you would expect based on their SAT scores. That’s because the same traits and practices (a mix of ability, genuine learning and Goodharting) that enable students to do well on the SAT also enable them to get high college GPAs. With that said, I suspect the reason for the changing pattern over time is that the changes to the SAT designed to reduce the gaps between men and women and between whites and blacks have inadvertently made the test more gameable and preppable, and Asian test-preppers have figured out how to exploit that.
Proponents of standardized testing maximalism defend the tests against the charge that they are structurally racist thanks to test prep by pointing out that test prep gains are small and test prep isn’t actually that common. But there’s one exception.
When measured, East Asians are much more likely to take test-prep courses than other groups, and gain much more from them (although they’re not broken out here, I bet Indians would be intermediate between East Asians and the rest). But this is only for explicit test prep courses (and doesn’t take into account superscoring). The SAT tests specific skills learned in school, not just innate ability, and hence more general academic grinding will also enhance SAT scores.
This suggests that high-scoring Asian students are not, on average, as capable as you’d expect them to be compared to their peers of other races.17 Anecdotes bear this out. From the pseudonymous teacher and test tutor known as Education Realist:
I’ve known a lot of high scoring students of every ethnicity over the years–and by high scoring, I mean 1400-1600 on the 1600 SAT, and 2200-2400 on the 10 years with the three tests. 5s on all AP tests, 700+ on all Subject tests. Until that conversation, I would have said kids that had high test scores were without exception tremendously impressive kids: usually creative, solid to great writing, opinionated, spotted patterns, knew history, knew the underlying theory of anything that interested them. I could see the difference, I’d say, between these kids and those slightly lower on the score scale–the 1200s, the kids who were well rounded with solid skills who were sometimes as impressive, sometimes not, sometimes a swot, sometimes a bright kid who didn’t see much point in striving. (…)
Since that first real awareness, I’ve met other kids with top 1% test scores who are similarly…unimpressive. 98+ percentile SAT scores, eight 5 AP scores, and a 4.5 GPA with no intellectual depth, no ability to make connections, or even to use their knowledge to do anything but pick the correct letter on the multiple choice test or regurgitate the correct answer for a teacher. Some I could confirm their high scores, others I just trusted my gut, now that I’d validated instinct. These are kids with certainly decent brains, but not unusually so. No shame in that. But no originality, not even the kind I’d expect from their actual abilities. No interest in anything but achieving high scores, without any interest in what that meant.
It probably won’t come as a shock to learn that all the kids with scores much higher than demonstrated ability were born somewhere in east Asia, that they all spent months and months learning how to take the test, taking practice tests, endlessly prepping.
Checking the thesis
If it’s true that Asians systematically Goodhart the education signal more than other groups do, it should be evident somewhere—though not necessarily in personal income (because a signal gained through Goodhart’s Law is still a signal) or even proxies for innovation18 like patenting or paper writing (which are also gameable and dependent on positions that can themselves be gained through Goodharting). And it is.
It’s no surprise that blacks and Hispanics, who benefit from massive affirmative action and who regress to lower means, are not as skilled as whites, conditional on education. Yet when tested in circumstances where there’s no incentive to Goodhart, Asians are also less skilled, conditional on education, than their white counterparts. (This is consistent with higher Asian IQ, since Asian educational attainment is so much higher).
This is exactly what you’d expect if Asians systematically optimize the signal more than other groups do, relative to actual skills learned in school, and explains why educated second-generation Asians earn less than their white peers. Something similar happens with men and women: girls get better grades in school (and have as far back as there are records) and are much more likely to attend college. Despite this, men have better general knowledge (d = 0.5) and are probably slightly more intelligent (d = 0.2). Women’s educational attainment advantage over men is analogous to the Asian advantage over other groups (though Asians actually are slightly smarter, just not nearly as much as the education gap would imply).
Vivek Ramaswamy
On 26 December 2024, Vivek Ramaswamy said the same thing about Asian immigrant culture that I’ve been saying, only with the valence reversed. It’s worth quoting in full:
The reason top tech companies often hire foreign-born & first-generation engineers over “native” Americans isn’t because of an innate American IQ deficit (a lazy & wrong explanation). A key part of it comes down to the c-word: culture. Tough questions demand tough answers & if we’re really serious about fixing the problem, we have to confront the TRUTH:
Our American culture has venerated mediocrity over excellence for way too long (at least since the 90s and likely longer). That doesn’t start in college, it starts YOUNG.
A culture that celebrates the prom queen over the math olympiad champ, or the jock over the valedictorian, will not produce the best engineers.
A culture that venerates Cory from “Boy Meets World,” or Zach & Slater over Screech in “Saved by the Bell,” or ‘Stefan’ over Steve Urkel in “Family Matters,” will not produce the best engineers.
(Fact: I know *multiple* sets of immigrant parents in the 90s who actively limited how much their kids could watch those TV shows precisely because they promoted mediocrity…and their kids went on to become wildly successful STEM graduates).
More movies like Whiplash, fewer reruns of “Friends.” More math tutoring, fewer sleepovers. More weekend science competitions, fewer Saturday morning cartoons. More books, less TV. More creating, less “chillin.” More extracurriculars, less “hanging out at the mall.”
Most normal American parents look skeptically at “those kinds of parents.” More normal American kids view such “those kinds of kids” with scorn. If you grow up aspiring to normalcy, normalcy is what you will achieve.
Now close your eyes & visualize which families you knew in the 90s (or even now) who raise their kids according to one model versus the other. Be brutally honest.
“Normalcy” doesn’t cut it in a hyper-competitive global market for technical talent. And if we pretend like it does, we’ll have our asses handed to us by China.
This can be our Sputnik moment. We’ve awaken from slumber before & we can do it again. Trump’s election hopefully marks the beginning of a new golden era in America, but only if our culture fully wakes up. A culture that once again prioritizes achievement over normalcy; excellence over mediocrity; nerdiness over conformity; hard work over laziness.
That’s the work we have cut out for us, rather than wallowing in victimhood & just wishing (or legislating) alternative hiring practices into existence. I’m confident we can do it.
Vivek is wrong about America19, which is unusually hard working for a wealthy country and also has the world’s highest race-adjusted educational test scores.20 However, that’s not the main point.
According to him, superior foreign education culture (he doesn’t say Asian, but he’s Indian and Hispanics are not renowned for their STEM performance) is the reason tech companies demand foreign workers21 and the reason immigrants and their children are so prominent in STEM. The reason American kids can’t measure up is that they supposedly value athletics and social life over academics, while their parents foolishly allow them sleepovers, time hanging out with friends, and TV—rather than forcing them into math tutoring and extracurriculars. To Vivek, South Korea’s hellish scholastic grind is the ideal that American kids need to adopt.
It would be one thing if this really did produce incredible results, but it doesn’t. It just Goodharts the education signal and makes both childhood and parenthood worse for everyone. Because schooling is mostly signaling, education competition is negative-sum and the correct policy is to minimize it. Why on Earth should we force American middle schoolers to work 13 hour days to become accountants?22
Is this practically significant?
My argument can be summarized as follows:
Education is mostly signaling, so increasing competition among students and investment in education is collectively wasteful, while individually rational.
Asian immigrants, through a combination of grinding and cheating, Goodhart this signal for cultural reasons, thereby attaining more education than expected from their abilities.
Given (1) and (2), Asian immigration to the US makes life for aspiring upper-middle class children and their parents significantly worse—by worsening the college admissions grind that has come to dominate childhood.
The arguments for points (1) and (2) are strong. However, they are impossible to rigorously quantify because measuring Goodharting is very difficult (almost by definition) and because the ways in which education signaling affects people’s choices are so complex and opaque.23 Based on the evidence I’ve presented so far, it’s entirely plausible that (3) is real but the effects are too small to matter. After all, Asians are only 5.5% of US K-12 students.
I’ve argued before that the acid test of whether or not immigration is beneficial to the existing population is which direction members of that population move. Do they move towards the immigrants—to take advantage of the new economic opportunities, technologies, food or anything else they have to offer (as in the case of French Huguenots in Germany or Europeans in Africa)? Or do they move away—to avoid job competition, expensive housing, crime, public noise, dysfunctional politics, culture clash, or whatever other problems they bring (as in modern London, New York, Berlin, and Paris)? There is no way to precisely weigh every possible effect immigrants have, so it’s better to look at revealed preferences than try to come up with a single number to judge them on.
As it turns out, Asian immigrants to the United States cause white flight. Unlike Hispanics or blacks, this is not because of crime or strain on public services: Asians are a low-crime, high-income group. It is because of education competition in public schools. This effect is huge: the arrival of one Asian student leads to 1.5 white departures on average (there is no effect on black or Hispanic students, presumably because of how little they compete with Asians). From a signaling perspective, this is entirely rational. Asian arrivals don’t reduce white test scores (and hence don’t affect the human capital or selection aspects of the education premium). But they do reduce whites’ relative rank. Unlike market competition24, this is zero-sum, so there is no corresponding gain.25
The effects aren’t just local. They’re even more dramatic at the pinnacle of national education competition—entry into top colleges. Asians are enormously overrepresented at these institutions relative to their population. The Harvard class of 2028 is 37% Asian (versus 31% white, 16% Hispanic, 14% black), while MIT—which takes test scores even more seriously—is 47% Asian (versus 37% white, 11% Hispanic, 5% black). For reference, Asians are 39% of top decile SAT scorers, whites 46%.
In practice (and despite this having been repeatedly banned on both the state and federal level) essentially all selective colleges in America have a certain fraction of spots reserved ahead of time as racial privileges for blacks and Hispanics. But colleges also face pressure from conservative anti-affirmative action crusaders to rely more on test scores and to stop discriminating against (specifically) Asians. The combined effect is to squeeze white applicants from both directions, freezing out non-athletes without connections from elite universities and hence elite networks.26
This numerical dominance of future elite networks matters because Asians (especially Indians) are extremely left-wing and often hostile towards other groups in the US, especially whites.27 Their leftism is even more pronounced in younger generations: at the 2024 election, when Trump massively expanded Republican appeal to both the young and nonwhites, young Asians went Harris +49 (by contrast, young Hispanics went Harris +17 and young whites went Trump +10).
Conclusions
Conservatives who lionize Asian immigrants for using education as a means of upwards mobility are making a mistake. Because education is mostly signaling, using education in this way involves throwing more resources into a negative-sum competition. The Asian emphasis on education does not benefit the United States; instead it results in overworked helicopter parents, miserable kids, fewer white Americans in the future US elite, and a more powerful higher education complex. Goodharted educational attainment is then used as an argument for even more Asian immigration.
The good news is that Goodharting education really is cultural. If Asians are anything like Jews in 20th century America, it will disappear by the third generation. The bad news is that 80% of Asian Americans are first or second generation, and given high immigration and low fertility, this number will remain high indefinitely unless legal immigration is slashed.
Anti-affirmative action crusaders and higher-education reformers, such as the University of Austin, often advocate for test-first admissions policies on the belief that this will limit the admissions rat race because tests can’t be gamed. They’re wrong. As South Korea, India and China and their diasporas in the US show, standardized tests can be Goodharted (via cheating or countless hours of prep) and this can easily consume far more of kids’ and parents’ time than America’s holistic admissions nightmare. The reason this isn’t yet a huge problem is that, because test scores are only one part of the admissions process, the reward-to-effort ratio of Goodharting tests is too low to be worth it for most parents and kids. Plus, America doesn’t have much of a culture of test prep or cheating. But if standardized tests become overwhelmingly important, this will change—especially if large-scale Asian immigration continues.
There are two ways to fix the college admissions grind:
Reduce college’s status as a gatekeeper. One way to do this is eliminating disparate impact as a doctrine, which will allow employers to test for proficiency more directly (contra Caplan, disparate impact really does matter). Another way is for governments to stop demanding credentials for their own employees. A third is to stop publicly subsidizing colleges, which will lower educational attainment and hence the importance of the signal.
Restricting immigration (including student visas) to the US. This article has focused on Asian immigrants Goodharting the US education system, but black and Hispanic immigrants obviously benefit hugely from affirmative action, which has a similar result. Something like half the black students at elite universities are first or second generation immigrants (it was 41% in 2007 and is almost certainly higher now).
Implementing either of these measures would require a nasty political fight, but both have the benefit that they’d actually work—unlike test-based admissions.
Alleviating the education signaling spiral is not the main reason to restrict Asian immigration to the United States, or even in the top three.28 But giving aspiring middle class kids (including the Asian ones) miserable childhoods for no greater purpose, as Vivek Ramaswamy would prefer, is in and of itself a bad thing, and right-wingers should stop celebrating it.
Arctotherium is an anonymous writer interested in demographics and the future of civilization. You can find more of his writings at his blog Not With A Bang or at his Twitter.
Support Aporia with a paid subscription:
You can also follow us on Twitter.
Wrongly, but it’s understandable why someone might draw that conclusion.
I’m choosing my words carefully here. Education can do a lot more than just affect labor market returns, and school is about more than education (school advocates often mention daycare for parents and “prison-alternative” for potential delinquents). It obviously is possible for people to learn specific technical skills in school. It’s just that, overall, that’s not what school is about in 21st century America.
I consider this sort of attack on the humanities to be philistinism, though I can’t see a reason to force the vast majority of students (who don’t care) to pretend to learn history or poetry for a few years. This is not a defence of actually-existing humanities academia, which has been thoroughly captured by leftist ideologues and probably can’t be saved.
This should not be taken as a full-throated endorsement of The Case Against Education across the board. I think Caplan is basically right about the signaling model of education, but he badly misuses his evidence. In fact, his references are better than the book itself. For example, the low test scores he cites as evidence of the irrelevance of K-12 education are dragged down by blacks, Hispanics, and recent immigrants (Caplan, of course, supports open borders). In a book that advocates total defunding of public education, Caplan never once addresses the race problem—which has been at the core of education policy since the 60s.
On the other hand, the fact that marriage is increasingly the province of the college-educated is entirely due to selection (on intelligence). Education is not causal. I posit that this is because no-fault divorce has made marriage a comparatively worse deal for the less intelligent.
In prisoner’s dilemma terms: getting more education is defecting.
I specify “modern” because it wasn’t really true in the US until the 1960s. Many intuitions about the social value of education are stuck in the past.
From a school’s perspective, this looks like rampant grade inflation because students and parents (rationally) value the signal of a high GPA far more than what is supposed to be learned.
You may have guessed what I didn’t say about my high school classmates in the section on Goodhart’s Law.
By “Asian”, I mean first and second generation Chinese (including people from Singapore, Taiwan and Hong Kong), Korean, Indian, Pakistani, and Bangladeshi immigrants.
She also wrote World on Fire, which is a great book about market-dominant minorities and the common problems afflicting societies dependent on them—from ubiquitous affirmative action to pogroms to ethnic networking to repressive minority-backed dictatorships. Western countries (and East Asian ethnostates) are blessed by having market-dominant majorities. It’s insane to give that up through mass immigration.
There is also rampant fraud in other parts of the international student system, mostly in South Asia. In general, international students in the 21st century are a massive scam that transfers money from taxpayers to universities and employers using student visas as low-end labor visas, while screwing over aspiring domestic STEM grads—not to mention tech transfer to China. This is especially bad in Canada, Britain and Australia. Scaling it up in the US, as many immigrationists want to do, is nuts. The US should exclude China on national security grounds and the entire Indian subcontinent until they get scams and fraud under control, plus end the OPT visa to make sure foreign students are actually students (and birthright citizenship, to stop foreign students from doing an end-run around the US immigration system via anchor babies). See also.
For the record, birthright citizenship makes this much worse. It means that any sort of immigration at all, including supposedly-temporary labor migration and study migration, permanently affects the culture, politics and demographic composition of the country. Unbundling physical residency from political rights would allow for a huge win-win. And almost no-one who wants to massively expand immigration for economic or scientific (brain drain) reasons supports ending birthright citizenship.
I first became aware of how commonplace this particular method is when reading The Billionaire’s Apprentice, in which the author nonchalantly (and with zero shame) describes her socialist Indian father losing the opportunity to join the Raj civil service because he was caught paying someone to take the test for him.
This also accords with personal experience. I went to a heavily Asian high school. Cheating (usually in the form of circulating previous years’ or earlier class periods’ exams) was rampant and a regular subject of concern among parents, students and teachers. Several (Asian) students wrote poems or short stories in English class about the ubiquity of cheating and how difficult it was to avoid doing it themselves. I don’t want to exaggerate. I don’t know how common it was, but it clearly wasn’t universal (I didn’t cheat and was generally at or near the top in my classes), and many of my fellow classmates were exceptionally intelligent and didn’t need to cheat. But even so.
Neither are actual IQ tests. They work because nobody bothers to game them, as gaming them gets you nothing. If scoring well at Raven’s Matrices got you a spot at Harvard, you’d see a massive Raven’s Matrices prep industry pop up overnight, and it would be effective.
LLMs optimized for benchmarks underperforming headline numbers is analogous.
True innovation, of the zero-to-one sort beloved of Peter Thiel, is something else, but naturally resists quantification.
It goes without saying that truth is the ultimate defense. If Vivek were right, that would be that. Some criticisms of America, like our obesity or disgusting and violent cities, are correct and justified. But he’s wrong. Him being wrong, and not him criticising Americans, is the problem.
The US education system is very expensive, but as far as actual learning goes, it is world-class. But you can’t make gold out of dross, and you can’t turn people of average intelligence into calculus masters—which I think almost everyone would accept if not for the politically unacceptable intelligence gap between whites and blacks (Hispanics and Asians are less politically salient). So instead, every time TIMSS or PISA results come out, we get a bunch of reformers talking about how great the Finnish and South Korean school systems are—despite the Finnish one being low-intensity and child-centred, and the Korean one being the exact opposite.
Again, he’s wrong. If you look at H1-B breakdowns, it’s Europeans, not known for their grinding, who are the highest rated and compensated.
I think part of the reason the tech-right doesn’t get this is that in tech startups, there are often super-linear returns to effort and the potential payoff is almost unlimited. Working 80 hour weeks for a 99th percentile income is reasonable. Demanding ordinary Americans do so to become middle-class accountants or IT professionals, when there is no compelling need is ridiculous (it would be one thing if the alternative was Third World poverty, but it isn’t). The other part is a cynical grab for cheap and dependent foreign labor (with all the externalities foisted on their countrymen and the future).
How much does prospective college prestige affect parents’ decision making? What about decisions made on that basis of information percolating indirectly to other parents through peer groups? How do you quantify, let alone measure that? By inspection and by housing markets for school districts, the answer is clearly “a lot,” but you can’t put a dollar amount on it. And what about kids, who can do much to thwart their parents’ will about studying or extracurriculars?
Aside from obvious concerns like fraud.
Unlike, say, losing a job opportunity to a more capable applicant. This hurts the losing candidate but the winning candidate’s gain is equal to their loss and customers and/or employers also benefit.
Those who believe this is fine because Asians are better at STEM should consider the cases of Alec Radford (Bachelor’s at the Olin College of Engineering) or Luke Farritor (Bachelor’s at the University of Nebraska). Both are white male geniuses, and neither was picked up by the enormously competitive top universities in the country. They’ve made it work, but how many similar figures exist (or would have existed) in fields less newsworthy than machine learning? How many more who are not quite at that genius level, but would still make excellent researchers or engineers at top-tier institutions? These are the type of men who powered American STEM during its 20th century glory days, when Americans invented the modern world, won two World Wars and the Cold War, and went to the Moon. Identifying these men and giving them pathways to succeed was the original purpose of the SAT and most of the pre-Civil Rights education reforms in the 20th century.
Contrary to what might be expected by brain-drain fantasists, US STEM has gotten visibly worse since the massive Asian influx started in the 1990s (the H1-B visa was created in 1990). Rather than improving American STEM, the Asian influx has instead allowed it to continue functioning at all in the face of enormous political pressure to reduce the number of white men. A microcosm of this: Google was built by white men, but Asians came to dominate hiring in the late 2010s as the company’s major product became infamously enshittified.
This is not to claim that Asians are bad at STEM. That would be absurd. I just claim that they are less capable than legible metrics make them look, and that freezing white men out of top-tier STEM institutions like MIT on “meritocratic” grounds is bad.
I assume readers are sophisticated enough to understand the difference between individuals and averages and between individual behavior and group dynamics, but just in case I include this disclaimer here. I am not claiming all Asians are left-wing (though as voters, politicians, activists, judges, lawyers, journalists, and professors, they are in aggregate). I am definitely not claiming all Asians are hostile to other races, or even that most individuals are in their personal lives (intermarriage, coworker and peer relationships, and friendships all suggest otherwise). I am claiming that organized Asian political institutions, such as the Congressional Indian Caucus, are. This is what matters for the future of the American polity.
In order, I’d rank these as:
The harmful political effects of Asian, and especially Indian, immigration.
Ethnic conflict. Diversity per se is bad, and adding more diversity to the US is bad in and of itself. Restricting immigration wouldn’t turn the US into Japan (or the US in 1960 for that matter), but when you find yourself in a hole, stop digging!
Deep roots. The iron law of immigration is that immigrants make the receiving countries more like the sending countries. There are millions of talented Chinese, Indians and South Koreans. Nevertheless, I and most Americans, would strongly prefer the United States to remain the United States and not to become South Korea, China or India. At least one prominent Indian-American immigration advocate, Suketu Mehta, openly says that he is motivated by wanting the United States to become shittier than India so Americans are forced to beg for visas.
We didn't play the Ivy game. We targeted our flagship state university - the University of Washington. My kids found that ~ 75% of their peers in Honors / IB programs in high school were the children of highly educated South and East Asian immigrants - and they were very much grinds. Most of the Anglo students were unwilling to study that hard.
I work in tech and I told my kids - welcome to your workforce peers. You will have to reasonably match them. You will be working with and competing with them for the rest of your life. Even in a good school system I was not impressed with the standards - so I supplemented them.
My daughter's comments was that her friends had Chinese dragon mothers, Hindu elephant mothers, and she had an American Eagle father.
My daughter dropped out of high school after 10th grade to do early admission and her Civil Engineering degree. My son did Running Start and then did Business - MIS. A lot of my son's classmates were Chinese students. My son was not that impressed.
When I went to high school there was a similar situation - but I was competing with the children of the Holacaust survivors, who were also outsiders and very academically oriented.
My son works in tech now and his peers definitely are an international bunch. The civil engineering workforce is not as international.
I think the signalling function is particularily important outside of STEM. There is signalling within STEM, but it works somewhat differently snf strongly depends upon specifics - are you targeting academic reputation, industrial / startup, ... and there is strong field dependence as well.
Abandoning the vast elite status game is a reward unto itself. You might lose some money and status... but you'll show courage and gain resilience and integrity.
I just wish these people admitted that they were engaged in a status game. That is the ONE topic that none of them will ever discuss. It makes them uncomfortable...
https://jmpolemic.substack.com/p/its-not-real