For those of us without either kind of privilege, it seems pessimistically like a cruel joke and optimistically like a handy way of sorting who should go first to the wall when the revolution comes. (that's also a joke - the revolution isn't coming and the Ivy grads are safe)
Also look at it from the perspective of people who will actually teach these kids. Students in top schools do the readings and understand them better. It is easier to build on the basics and get to the interesting stuff. Students in poorer schools don't do the readings and even if they do have difficulty understanding them. The syllabus and standards have to be watered down. And you have to constantly sit on top of them because they often lack the integrity not to cheat. Teaching the first group is emotionally rewarding while teaching the second group is soul sucking. The difference is big enough that the switch from the second to the first was worth it for me even though I took a 15-20% pay cut. So, people with PhDs will compete to teach and research at where the smart hard-working students are. And given this dynamic, since faculty have say in admissions decisions, they are going to pick the ones they would prefer to teach
> that's Harvard's fundamental business? They admit actual smart kids on scholarship, alongside the kids of rich people who pay huge bribes. The smart kids get access to rich friends. The rich kids get to go to a college with a smartness rep as well as a richness rep.
What do you think would happen if we forced schools with low acceptance rates to increase their admissions year over year in order to maintain access to public grants and loan subsidies? Would this help them launder priviledge more effectively or dilute their brand? Can we infer anything from the fact that Ivies are so disinterested in growing their enrollment and increasing their impact?
Why are so few schools good at creating good signaling? It isn't hard to imagine a world with ten Harvards, yet that isn't our world. I kinda like how Harvard makes such fools of us all.
I think this conflates high-quality professors in the "renowned expert in field" and in the "really good at teaching" sense. It makes sense for the people best in their fields to teach the smartest students.
Partly because they often do get to some work at the edge of the field - see e.g. Scott Aaronson's occasional mention of smart undergrads he did research with - but also because maybe smart people think about the field in similar ways. Brilliant researchers aren't necessarily better teachers, but might be a better fit for brilliant students.
Conversely, what we actually want for the less smart students is to have professors who are better at teaching. They don't necessarily have to be top researchers (those students aren't doing undergrad research anyway), and selecting on top researchers for professors would limit our ability to get top teachers.
Since doing cutting edge research is cooler than being a good teacher, this does naturally lead to the places with the top researchers being more prestigious and having more resources. But this consequence can happen even in a system optimized to teach low performers well!
Interesting comment. I wonder what readers’ thoughts are on the opposing vantage point: what if the “laundering” aspect for the elite actually dampens the meritocratic prestige that the rest of applicants gain.
At what ratio does this become untenable? Or does the ratio not matter because of the chance to network with elites trumps the label of meritocratic prestige? (Cynical view)
(1) The first part, why send the top students to the best colleges?, misses one key aspect of this which is that the top students would like to be among other top students. So even if instructional quality, career opportunities, etc., were exactly the same at all universities, there would be some sort of impetus to sort. The question then becomes why should this have any connection to research quality, "elite" faculty, and other such things. It needn't, but there's no good measure of anything related to instructional quality, career opportunities, etc., and we tend to assume (not without reason) that quality in one area implies quality in others. Plus, students in many fields would actually like to work with top people in those fields -- whether or not this is likely to happen -- so there's some direct motivation for the link as well. (And conversely, the top people would like to work with or teach the best students.)
(2) Selection based on merit was always less than ideal, but it is getting worse. Currently, for example, the UC schools won't even *accept* SAT scores from applicants. Whatever biases standardized tests may have, they are less biased than, for example, recommendation letters, flowery essays, and other characteristics. If the trend towards subjectivity continues, will the signals described in the post remain as strong?
(3) As a faculty member at a fairly large public university, I am often annoyed at how discussions of higher ed focus disproportionately on small places like Harvard (edit: 7000 undergrads (I mis-typed 2000 earlier)). I don't particularly care about Arizona State, but its enrollment (112,000 undergrads) and that of similar places should warrant a lot more press than is the case.
(4) I've heard the concept of privilege laundering before; it seems obviously true.
I'm not at all sure that the quality of teaching is better at Harvard than at less prestigious places. I used to have a teaching affiliation with Harvard Medical School and have met some Harvard faculty, and I have had as psychotherapy patients about a dozen Harvard undergrads and people getting advanced degrees at the graduate and professional schools. My observation is that the professors who are so eminent in their fields that they are actually famous are utterly unavailable to undergrads and even to the grad students who are writing dissertations under their supervision, or carrying out their research in the lab. They are always flying off to conferences all over the world and cancelling their lectures & having grad students do them. They are preoccupied with their careers, and hate having to teach: They view teaching as the toilet-cleaning side of being an academic. They teach undergrads via lectures in giant groups in big halls, then have their teaching assistants run weekly discussion groups. The grad students all hate having to be teaching assistants, and try to get paid jobs doing other things, such as being advisor in one of the residence halls or working with other grad students to publish a journal. Out of the half dozen grad students I have seen, at least 2 have been so cruelly neglected and ignored by their famous advisors. that it seemed criminal to me to deal out that kind of treatment to someone so smart, who has worked so hard to get to Harvard, and wants so badly to do really good research that will launch their careers.
My own experience as an undergrad was at 2 places, both ivies. For the first 2 years I was at acollege, at the last 2 a university. I had no lecture classes at all at the college -- class size varied from maybe 10 to 30. At the university I was taking higher level courses, which are always smaller, and mostly avoided lecture classes there too. I only had one genuinely famous professor, and that person did in fact teach via a giant lecture. As for my many other teachers, whose smarts I enjoyed and revered -- they changed my life. They gave me wonderful, accurate flashes of insight into their vision of their field, and they took a lot of interest in me. I really do not know if they were less brilliant than the famous scholars at Harvard, or just much less interested in fame. But even if they were, say, only 80% as smart about Wittgenstein, calculus, history of religion, linguistics, 20th century poets, etc, I do not think I was a developed enough being at that point to benefit from the remaining 20% anyway. And besides, conveying to students your own sharp insights and convoluted intuitions about the field is only one part of teaching. The other part is mentoring the student, and these people were mostly excellent at that.
Quality of teaching is only mildly correlated with the teacher's degree of fame and glory. It's way too simple to say that the more prestigious the school you attend, the better the teaching you receive. I think it's likely I would have had as good an experience at many schools with fairly small class size and teachers who were not egomaniacs.
I have one very relevant question. You can pay a hefty donation and have your child admitted to Harvard even if it lacks intellectual ability to be admitted without big money. But are such students treated differently later, or is anyone treated equally once they are admitted?
Sticking to a less cynical perspective, mixing elites with talented people has networking benefits for both sides. The talented people connect with elites who can help them into more elite-level roles, and the elites connect with talented people who can help them fill whatever important roles they have open.
e.g., The rich kid who is going to inherit his father's successful business finds talented people who can help him run that business, and the talented kids get the opportunity to work in high-level roles in this successful business.
Here's a take on it: students' intellectual development depends in part on the environment in which they grow. Those who grow among smart classmates become smarter than they would be otherwise. Not only that, but we achieve the greatest overall societal gains by grouping the smartest ones together.
For a very simple mathematical illustration, consider:
(x - y)^2 > 0 , for x!=y
(x - y)^2 = x^2 - 2xy + y^2
therefore: x^2 - 2xy + y^2 > 0
therefore: x^2 + y^2 > 2xy
If x and y are unequal, then you get the best results by matching x with x and y with y.
Illustration: for x=5 and y=2, matching the best with the best gives us 5*5 + 2*2 = 29, whereas matching best with worst only give us 5*2 + 2*5 = 20.
This model only has two variables, but I'm fairly sure it has been generalized long ago.
So the utility of clustering the best students in the best schools stems primarily from maximizing the students' overall potential. It's not only about matching the best students with the best teachers. It is also, and maybe more importantly up to a certain age, about matching the best students with the best classmates.
> why do we send the top students to the best colleges? Why not send the weakest students to the best colleges, since they need the most help?
There's a weird assumption in the response (and, I guess, the question) that "the best colleges" provide more help than other colleges do.
In reality, "the best colleges" are, by definition, the ones with the best students. The young Yglesias' question is similar to the philosopher who asks whether it's OK to prefer dating people you find attractive: conceptually incoherent. It isn't possible to do anything else.
----
Schools arise as a way for elite children to acculturate to each other. When talented nonelites begin to be let in, we should assume that the reason is to benefit the elites who operate and use the school. There is an obvious symmetry in what happens: the go-getters who make it in on ability get connections to people with power. (As mentioned in the post.) The elites who make it in on pedigree get the same connections, but since they are the other side of those connections, we would describe them as getting connections to people with ability. Connections like that can be important if you're ever put in charge of something!
It isn't necessary to assume that smart people get admitted to Harvard so that elites who also go there can pretend to be smart.
----
It also isn't obvious that legacy admissions are of lower quality than "merit" admissions. If all you know about an applicant is their SAT score, you can admit them and watch them regress to the population mean. If you know (1) their SAT score and (2) that both their parents went to Harvard, the same thing will happen, except that the relevant mean toward which they will regress is much higher.
> “Fairly-earned privilege” means all the brilliant talented ambitious youngsters admitted on the basis of their SAT scores and grades and impressive accomplishments
“Fairly-earned privilege” sounds more like a contradiction than anything. It's not privilege if it's fairly earned.
Famous colleges have no doubt done studies of which applicants are likely to donate the most money over the next 60 years. Have any of these studies ever been published? Anybody have any links?
My impression from looking at the names on college buildings is that big donors tend to be legacies, jocks, men, business-oriented, frat boys, and Republicans or centrist Jewish Democrats (e.g., Michael Bloomberg).
That was precisely the theme of Daniel Markovits' book "The meritocracy trap". Remember that "meritocracy" wasn't initially invented as a positive word, but to denounce privilege laundering in the first place.
America and much of the west does spend a lot more on poor and bad students than on rich and gifted ones.
Inner city schools in the US are the best funded by a long way, in the UK I got more resources for being mildly dyslexic than for being smart. There are vast resources spent on every kind of learning difficulty even when the financial return is dubious.
The question is incredibly naive, as you would expect from a child. But the frightening thing is how common this naivety is among affirmative action supporters. Grown adults not being able to understand that sending the worst students to the "best" school will stop it being the best school. As if the 'bestness' of Harvard is some kind of intrinsic, durable property.
But I think all of this is rooted in the left-wing education myth. That is, to the extent that academic ability varies between people (and this isn't just white supremacy etc.), it is almost entirely explained by differences in educational "quality", which is obviously false. And I say this because I don't think even they are foolish enough to think that dumb kids graduating from harvard still dumb would be treated the way current harvard grads are treated. But we know clearly from the data that sending black people to college doesn't close the intellectual gap between them and white students. If anything, it grows i.e. the gap between black and white college grads is higher than for black and white high school grads, and this despite the survivorship bias resulting from a much higher dropout rate for black students (assuming dumber students are more likely to drop out).
A good friend of mine teaches maths at a mediocre university in the UK. This university recently slipped in rank from an already middling position and they had to lower the intake grade requirements one notch to maintain student numbers. Their list of coursee does not look any different from last year or from a list at a top university and it cannot really look very different as they would lose their accreditation if they dropped e.g. differential equations from their maths degree. Even a two page course content sheet does not look very different from one at a top university, at least for a layman. However, as these courses are now taught for lower level students, they need to spend more time at the basics and barely touch the advanced stuff. They still include the advanced stuff in the exams, but they are now just teaching it to ensure reasonable exam pass rates and do not actually achieve reasonable understanding of the subject for any bar a few top students. My friend is now really disillusioned and most of his skills of teaching advanced stuff are now wasted and these students barely get to the level of maths knowledge of an average graduate from a top high school after a 3 years degree. So even if course subjects look the same as ones at a top university , deep down below course contents is very different and require a different skill set to teach, and the first optimistic answer stands.
> That is: Harvard accepts (let’s say) 75% smart/talented people, and 25% rich/powerful people.
If we're talking about actually admitting people because their family is donating money, this would seem to be a substantial overestimate. There's actually a term for people admitted because their family donated money, and it's "development case": https://en.wikipedia.org/wiki/Development_case
At least going by that page (I haven't looked into it much further), the actual fraction of these is no more than about 5%. This linked article says that (at Dartmouth at least) it used to be more like 1% back in the 90s and 00s, and rose to about 5% in the 2010s: https://web.archive.org/web/20190316233913/http://www.dartblog.com/data/2014/09/011686.php So it's *possible* it's larger now? But if so that would seem to be a recent phenomenon.
This does leave the possibility of larger-scale "privilege laundering" via legacy admissions, essays, and activities and whatnot, so it's hard to say exactly what's going on -- and, if such a thing is occurring, it's actually *deliberate* -- but in terms of actual development cases they're much smaller. (With Harvard it does seem like this more general privelege-laundering is likely deliberate, from what I've read. Seems harder to say about other schools, though.)
On the less cynical side, advanced research has to happen somewhere and the people doing that want more people who can help (despite universities tilting further and further towards undergraduate education from research).
Also, instead of "best", you could say "most advanced". If we optimised for teaching we might put the most knowledgeable people teaching the most advanced students, and the best teachers teaching the students who most need to progress. What actually happens? A bit of both. Advanced universities often have more funding which naturally produces better teaching, but they don't prioritise "good teachers" as much as other things.
From the more cynical end of the spectrum, lots of institutions are "people with influence channelling people with ability". Sometimes the channelling is almost pure parasitism like modern academic journals which add almost nothing to the academic process other than a monopoly. Sometimes the channelling is the most valuable part (I don't know which, but I'm sure SOME big breakthroughs happened because someone had a vision and brought the right smart people together and wouldn't have happened otherwise). Universities used to be more patronage "richduke had money and chose to subsidise some random smart people". You can see that as generosity or monopoly depending whether you see the baseline as "everyone has enough to eat" or "everyone has as much as they can grab".
Like others in the comments here, I think it is as much about networking as signaling and education – but not networking in the shallow sense of “it’s all about who you know, not what you know”, but about meeting people who help you fulfill your potential.
My father’s favorite piece of advice for deciding where to go to college was that “you don’t need Lee Iacocca to teach you about supply and demand”. (Which usually led to quizzical looks, since very few college-bound Norwegians of this millennium know who Lee Iacocca was.)
The suggestion was that going to a top-tier university for your bachelor’s degree was a waste of money. And going for other degrees might also be a waste of money.
Since it influenced some of my own choices, I resented the naivety of that advice for years into my career – until I realized that I probably would have wasted an Ivy League education, and failed to grab the opportunities it afforded me, in much the same way that I wasted my actual education.
However, for those who are wicked smart and know how to take advantage, top-tier universities can offer you a peer group that can challenge you in a way very few others can, motivate you to reach higher that you thought was possible, and give you networking opportunities that open doors you might not even know existed and clear obstacles out of your way.
Genius and talent will often go stale and dull in a pool of mediocrity – unrecognized, unmotivated, confused for eccentricity or confusion, left to stagnate. But when pitted against other genius and talent, professors and peers who get it and challenge it and nurture it, it gets to sparkle and shine. You see it in music and literature and sports and science and business and culinary arts: The very best rarely appear alone.
And so, why do we send the top students to the top colleges? It’s a bit like asking why we send the best athletes to the best sports leagues. (“Send”, as if they don’t have a say in it themselves.) Because that’s where they can meet the people who can help them reach their full potential – not so much by teaching them, even though that is part of it, as by challenging them, supporting them, and putting them in the competition for the highest achievements.
And, when the best athletes and students congregate in one place, others dream of joining, and it becomes a self-perpetuating system for supporting excellence (which is strong enough to withstand some inevitable corruption and gaming of the system).
Aside from signaling, one of the functions of the university is acculturation. By admitting a mix of the brightest and the hereditary elite, the Ivies help to build a shared culture under their own influence.
(1) I am always highly suspicious of anecdotes involving cute moppets going "Papa, whither the course of educational attainment amongst the student body of our fair nation?" Maybe Matt's five year old is that precocious, but I do wonder about "what kind of discussion was Daddy having about such topics to set Kiddie up for that cute quote?" Much like all the women who go online about their six year old (and then some other woman goes on to cap it with 'well, my *four* year old did even more achingly woke thing') doing the Wakanda forever salute for Ruth Bader Ginsburg because of LGBT rights.
You thought I was joking about Rutkanda Forever?. Nope, I just got the details wrong:
(2) "Harvard students and faculty who are descended from people held in slavery in the United States"
Good God, how many more euphemisms can we come up with? I had just got used to seeing "enslaved" and "enslavers" used instead of "slaves" and "slave-owners" (which had been replaced first by "slave-holders") and now we're up to "people held in slavery". I now fully expect to see the unironic use soon of "prisoners with jobs" because now "slave" is a slur? How much more can we dilute terms in order to keep the grievance hamster wheel turning?
I think in Germany in certain universities there is a bit of an opposite effect: some of the big technical universities accept all applicants, but then are extremely rigorous in their sorting. Their degrees are prestigious precisely because you are expected to have "fought your way through" an anonymous, underfunded system with very poor tutoring and nobody actually caring whether you show up or not.
In the final paragraph you say "I expect that without such a system the elites would do their own thing without any concession to merit whatsoever - so maybe it beats the alternative"
So then why does the elite bother at all? If they control access to their ranks, why do they make this concession to merit at all? Certainly in most human societies the elite has never ruled on the pretense of being smarter, so why do it in ours?
Furthermore, does the elite in the US really control access to their ranks? Since more wealth in our society is newly created than is inherited, I would say they do not have this power (look how Zuckerberg vs Winklevoss turned out). If you look at a list of the richest Americans, you have to go outside the top 10 before you see inherited fortunes. And even then, it's only 1 generation inherited (Koch, Walmart). So it makes more sense that the reason for elite universities is that the current elites get to network with the next generation of elites (Musk, Bezos, Zuckerberg, not the children of past elites) than that they just want the aura of "merit" around them.
"Is is true? An oft-cited paper, Dale and Krueger, appears to find that, controlling for applicant characteristics, people who attend more selective college don’t earn more money later in life. "
A non-sequitur if I ever saw one, given the preceding paragraphs. What does income have to do with finding the cure for cancer? Maybe it ought to; if someone found the forever cure for all the cancers, they'd probably deserve to be among the top earners in the world. But I don't see that as an unavoidable consequence, because scientists in particular seem to be in the game for their personal satisfaction and prestige among peers rather than the big money. At least much more than your typical investment banker or what have you.
> But this seems false; most of the classes at top colleges are the same material that gets taught everywhere else; you don’t get into subjects that need world experts until postgrad.
At least in math and physics, the point of elite undergrad education is absolutely not the undergrad classes themselves. The leading incoming freshmen have already taught themselves at least half the undergrad curriculum and are doing graduate-level work by the second year, or sometimes even the first. (I wasn't at that level, but my first quantum field theory had 19 year olds, my string theory class had a 17 year old, and my PhD advisor was in the equivalents of those classes at 15.) Students of this sort clear out the graduate-level classes quickly and then progress to research, so they absolutely benefit from being at top colleges. The research opportunities and the network of peers makes an enormous difference to their lives.
My impression from within academia is that many STEM professors recognize this and therefore want to have such students around, but the admissions officers (except at a few enlightened places) block them. That's because the typical admissions officer is a freshly graduated English major who only views math as an instrument of torture. To these officers, such kids are uncharismatic and alien, and worst of all, their naively enthusiastic essays lack any sense of irony or pathos. It is an easy rejection. All the oddities of college admissions are not centrally planned or rationally chosen, they emerge from the biases and preconceptions of this class of people.
> but what does the Princess get? I think she gets the right to say she went to Harvard, an honor which is known to go mostly to the meritorious.
She gets to know lots of presumably smart people to give oil contracts to? (and, like, the very extended interview sort of know, instead of just "ah yes they went to Harvard".)
There's a missing aspect here, at least as the stories always described it some years ago (I'm struggling to find a source). Reportedly, Princeton would admit about 1/5 of the class from rich-but-not-super-rich families that were smart, but clearly a step below the top standard.
Why? Well, in any class, some of the kids will have to be at the bottom; that's how a graded/ranked system works. Super-smart kids are smart and competitive enough to get into Princeton, so they would be unhappy to be at the bottom. The super-rich kids are too few in number and too entitled to form the bottom quintile easily/happily. But students who realize the others around them are smarter would feel lucky to have gotten in at all and have a baseline of gratitude to the school. They'd be perfectly happy to be at the bottom of a *Princeton* class, and after graduation, would form the nucleus of the alumni-jobs-and-donations network.
Being a legacy is probably a significant plus to be a bottom-fifth admit, for obvious reasons.
I don't know whether it's true, but Princeton is perennially #1 in alumni donations and engagement, so if it's true, then it works.
I highly recommend the book “40 colleges that change lives”. There are quite a few colleges which don’t make traditional top 10 lists, which are more likely to provide the an excellent education to a student. As most of us have experienced, having access to a good teacher, makes all the difference, both in terms of how much we learn and how much we enjoy the subject.
When my children were in middle school and high school, I would occasionally get asked by another parent whether I would recommend that their child try to take the harder level of a math class or the less hard level. I always recommended that they try the harder level because our experience was that the “harder” classes tended to be easier and much more beneficial for the students because they would get matched with the better teachers .
One of my friends teaches at the Naval Academy in a technical area and she says that there is a huge difference in her enjoyment and ability to make significant progress when teaching the “best” versus the “worst” students. The category “worst students” includes students that are least prepared, some students that are least motivated, and some students that have the worst study habits. She can make substantial progress with those who want to work hard and learn, but the combination of challenged students in the lowest level classes makes her job very hard. So most teachers enjoy teaching the advanced classes much more.
There are also schools that have attracted some of the best known professors but then don’t have those professors teaching the classes because they’re carrying out research. So doing your homework to try to figure out where you get the best education can really pay off.
My definition of best education for college included some sense of social responsibility and some sense of life work balance. So we were put off by hearing from MIT students that they were overwhelmed by the amount of material and many of them were depressed. We were also put off when Harvard told us that their libraries were open 24 hours a day. My three children picked three amazing schools: McDaniel College, Williams College, and Olin College of Engineering, all of which had good reputations for teaching, and for a balance of education and other life values.
That said, if you wanted to send your child someplace where they would make connections with powerful people, Harvard might still be the best choice.
If the elites control the money supply, they can print money to reward themselves for whatever they deem good and worthy.
If the money supply is not politically controlled, and most wealth is ephemeral (networks, brands, and technologies, not land and physical resources), now the elites _have_ to compete on merit because they have no other means on which to compete.
If you just want a cynical explanatory rather than a justificatory answer, there's no need for most of the interesting speculation about signaling theory.
Top-level professional academics want to work with each other and with top-level students, and top-level students want the same. The people who run Harvard (and most universities) are a mix of idealists who want their university to do the best scholarship possible and cynics who want to attract as much alumni and grant money as possible; therefore, they try to recruit good scholars to make the virtuous scholastic synergism work, while slipping in some proportion of reasonably bright rich elites who can be counted on for automatic prestige and generous alumni donations.
The scholars get to intellectually inspire each other, the reasonably bright rich elites get to socialize with each other in final clubs, and the university gets paid in proportion to its success at making the scholars wealthy geniuses and the elites wealthy well-connected elites. Non-Harvard universities are generally trying to imitate this model as closely as possible, but they're less prestigious (self-fulfillingly) because they are less able to compete for the limited pool of brilliant faculty and brilliant and/or rich students.
Even in undergrad it's not the same material which is a huge part of the explanation that got dismissed here. When I was studying abroad at University of Queensland my girlfriend visited and sat in on a biochem lecture about the electron transport chain. She was amused because they did the standard vague "redox reactions happen and protons cross the mitochondrial membrane" explanation but the Amherst College biochem class went in on the chemical details of how various heme structures actually allow that to happen.
As a funnier example my physics professor told us a story about an acquaintance describing learning Ohm's Three Laws. Impressed, my professor asked what they were since he only knew the one. The three laws were: V=IR, I=V/R, and R=V/I.
More generally everything is like this, knowledge is fractal! You can learn about compact spaces as "closed and bounded" or as "space where every open cover has a finite subcover". Complex impedance helps you understand circuits but you don't strictly need it in an E&M class. And so on. Grouping students of similar ability together allows you to match talent to the appropriate depth and pace; if you randomly sorted students into colleges then for any given class ~40% of students would be overwhelmed and 40% would be bored out of their minds.
A subtlety: you don’t get the full signal just by being accepted to Harvard, you do generally need to actually graduate from Harvard, and Harvard has to put on all the trappings of actually being a highly elite education. In that sense it’s the same (useful, if overpriced) signal as every other college - a graduate has shown some minimum degree of seriousness and studiousness. Not saying there aren’t easy pathways or ways the very tippy top elites might pull the levers to get their most dullard progeny through, but I think that’s more rare than the preference at the admissions level.
Of course they most meritorious thing of all is to drop out of Harvard to go do something earth changing and extremely lucrative. You’re literally too smart for Harvard! But for that to work you have to actually be objectively talented, or at least have one really good idea and the ability to execute it.
I'm literally a math professor specializing in topology (though not "ten-dimensional hypertopology"). When I was at MIT as an undergrad I took the most advanced topology courses on offer in my sophomore and junior years, and jumped into research. Being around MIT professors was essential to my mathematical development, and so I strongly disagree with the second paragraph of this essay.
Who's this "we" that's sending the best students to the best schools? Nobody "sent" me to an elite school. I chose to go there, and the school chose to allow me to go there. My parents would have contributed financially wherever I chose to go.
Why wouldn't I prefer a good school, as good a school as I could get?
Why wouldn't a good school prefer good students, as good as they could get? The better the students, the better for the professors, the research, and the quality of classroom discussion. (Apart from the handful like legacies and athletes who are admitted for money reasons. Heck, even the good students are admitted partly for money reasons; they're more likely to be rich donors down the road.)
In case anyone hasn't already mentioned it, Bryan Caplan's "The Case Against Education" is very much worth reading if you're interested in learning more about the signaling theory of education. I'm almost finished with it, and I find it to be fairly thorough and comprehensive.
I think debates about college always get confusing if you lump together the "learning actual necessary knowledge for the future" functions (For say, mechanical engineers) with the signalling/approval/privilege functions (here is your communication major from Bad College X or Harvard). Yglesias' son (unironically) thinks school (where he is learning to read/write) is about learning necessary things for Life (like reading and writing), and so assumes college is like that.
"Best" college has the same issue. Best at what? The 5 year old assumes it means the best at teaching/improving student's knowledge, but we all know that is not what Harvard is best at. Step 1 in any difficult issue is to clarify what we're actually talking about!
Your analysis overlooked the role of competition. I am drawing on my experience as an Ivy League university professor. The most important reason for selecting for student rank is to increase the overall intellectual level of the student body. I am always impressed to watch the transition of students, who were top of the class in high school, and then find themselves in the middle of the pack in college. Ambitious students are also driven by an environment where the professors are at the top of the game in terms of research.
This is too cynical. Highly intelligent take pleasure in each other's company, and universities are the traditional place where they can congregate. The top faculty teach the brightest students because that is more agreeable job than teaching struggling students and they are not primarily motivated by charity.
The problem is that the economic returns to being perceived as intelligent have gotten too high, and this has lent illicit edge to the otherwise innocent and wholesome pleasure of spending time reading and studying with those that share your interests.
The child (and Matt as well) are mistaking "top college" with "has the best teachers for basic material". The two are very different. This might be true for elementary school, but not really above that.
From elsewhere: yes the top colleges use a lot of the same textbooks. They cover a lot more of the text and do it a lot faster. If you need help you don't go to the lecturer to tell him to slow down (he won't), you get your own help (tutors generally). Or you drop the course, or drop out.
This is of course confounded in the States by the athletic "scholarships". This depends heavily on the specific sport, but e.g. the top colleges known for scholarship aren't generally getting lots of money from college football games.
From earlier: I don't know much about Eton etc. apart from that IIRC they are boarding schools, which puts a pretty high lower limit on how poor your family can be if you don't have a "full ride" scholarship. And "full ride" is likely to be a large amount, so they wouldn't hand out a lot of them. So yeah the students will tend to be from rich families.
No idea how many of the French elite schools are boarding (if any).
Why do the best students go to the best colleges? Because that's what makes them the best colleges. To a first approximation, good students make good schools; good schools don't make good students.
If next September, the entering class of Harvard was sent down to Bridgewater State and the entering class at Bridgewater State was sent up to Harvard, Bridgewater would immediately become more intellectually active, along with a lot of extra-curricular activity created by the go-getters who had been admitted to Harvard. Meanwhile, in Cambridge, there would be an immediate drop in energy and intellectual excitement. If you could somehow do the same thing for the next three classes, Bridgewater would now be considered one of the best colleges in the country and Harvard would be considered mediocre on the college level.
Something I haven't seen mentioned is that rank-matching is not just more effective, it's also more efficient. Anecdote: my school district was 404th in per-capita funding (out of 407 in the state). However they used that money to hire a small handful of really excellent teachers, and ruthlessly tracked their students. Year after year, they produced more National Merit finalists than any other school district in that state.
The reason for undergraduates go to one of the Ivy Leagues is to receive a credential, not an education. It’s a lifetime ticket to the Big Club that George Carlin described. The one you and I are not members of, and never will be, regardless of our capabilities or accomplishments.
One of my Development Econ. profs would always tell us that you'll never be rid of your elites, and so long as they control resources/capital your best bet is to set-up a structure where they're incentivized to act in a beneficial way. From that perspective, there seems to be a potential benefit in letting elites network with the best/brightest students of the next cohort, and of course providing access to capital for the best/brightest students is itself a benefit
It is true that some skills can be learned but cannot be taught. Other skills can be taught but only to those who are capable of learning them.
No amount of instruction will teach a dog to speak Latin; it's not a matter of forcing enough teaching units from enough really talented and motivated teachers and you can get Fido through first year Latin grammar, while Joe "Publius" Brainimax could teach himself as much and get to the same place.
I think the obvious answer (at a system level) I haven't seen mentioned yet is that it aligns the incentives for students in school. If doing really well in school gets you into a top college, then you are incentives to do really well. But if, as the tweet goes, your ticket to a top school is being a really poor student, then you have terribly aligned incentives.
> Earnings are a poor proxy for “teaches better” - it would be great to have something like value-add to GRE scores - but AFAIK no study like that exists.
It seems like for highly ranked students they're already beyond what the SATs/GREs/etc are designed to measure, so it's hard to judge the value of teachers by those metrics. In math, you need a specialized test sequence like the AMC -> AIME -> USAMO to meaningfully distinguish between smart students, and I predict that _could_ pick up meaningful teacher effects.
Of course, the major error in the question "Why do we ...?" is that it assumes that there is a "we" that is deciding this. There are lots of independent actors in the situation and each has its own "incentives" (interests) and the result is some sort of game-theoretic equilibrium. And the sooner the kid learns that the world is not "We are all happily cooperating to advance a common good." the better.
One factor I haven't seen mentioned that I think is important is the use of college admissions as a signal in status competition among high-SES families. The obsession over getting one's child into an "elite" university seems to be much in excess of the educational, financial, and status benefits the child is going to get.
This gets interesting when the family's high SES comes from a high income from running a small business. Successfully running a business requires some intelligence, of course, but it depends a lot more on personality traits like diligence and ability to do well in win-lose negotiations. (Even in the largest companies, the CEOs' educational credentials are mediocre.) So on the average their children are not the highest-IQ despite that elite colleges officially select on things dependent on IQ.
Yep. A related issue is: Why is the high-status education done at the end of a student's time in education?
Clearly the teacher with the biggest leverage over anyone's educational success is a nursery (kindergarten) teacher. But they're not the ones getting the big bucks, or the guest columns in the NYT.
In regard to gifts and admissions, I'm reminded of Matt Levine:
"Analysts don't say to companies, "we will give you favorable research coverage if you will meet with our clients." (Sometimes companies do say that they won't meet with clients of banks who give unfavorable coverage, though.) And they don't say to investors, "we will bring you to a meeting with this company if you pay us a million dollars." An economy exists, but it is a gift economy, one in which it is crass to ascribe precise dollar values or to demand quid pro quos too
I think this idea really underscores that what elite colleges are selling (the merit) is the admissions process and not the education. The fact is for even the children of the rich/powerful-- the students who've been admitted on the basis of prestige and not academic strength-- the undergraduate classwork at every one of these schools just isn't very hard. Especially coming from a rigorous prep school (which leaves many such people with the impression high school was significantly more academically challenging than college!) and especially in our era of grade inflation where even a vague gesture at trying can receive an A-.
>most of the classes at top colleges are the same material that gets taught everywhere else
I see this claim repeated a lot, and I think it's just false. Most of what follows is anecdotal, but 1. I believe my experience is somewhat representative and 2. I think my position agrees with the simple, naïve understanding of the world.
I'm currently a grad student at a top university in a field easily confused with "hypertopology". My trajectory was (high school dual enrolled) small, not bad college -> (undergrad) good, non-elite state school -> (grad) elite school, and at each step, the difference was large and noticeable.
Subjective differences: At each step, everyone (students, professor) are more engaged, invested and knowledgeable. People were smarter, and it showed in the rigor of the classes, discussion and the questions.
Objective differences: (mostly about course content) This may not be only for math-y people, but I think it's my best argument:
Intro calculus: here are formulas, do it -> actual math, applying to situations -> teaches epsilon-delta to history majors.
Multivariate calc: literally no greens theorem -> grad/div/curl/greens theorem, etc -> differential forms (wtf).
Intro proofs: didn't even get to induction -> induction, equiv relations, basically everything you need -> everything on the left for the first half, then analysis for the second half
Analysis I & II: topology of R (limits, etc), sequences, infinite sums -> rigorous Riemann integration, existence/uniqueness of ODEs -> (the advanced one at least) the left plus Hilbert spaces/function spaces (wtf)
... (I could go on)
And these were just the undergrad classes! The grad classes (which undergrads typically have access to in their later years) showed similar difference.
The small college wasn't scraping the bottom of the barrel either--their SAT scores were around the 50th percentile. My general impression is that math graduates from my undergrad have a respectably solid foundation, those from my hometown were not at all prepared for higher math and those here (or the top students at least) are future professionals.
Moreover, I think this state of affairs just makes sense. Like, are you surprised that the smarty pants are over there learning more together? They can and do learn material 3 or 4 times as fast, what do you think they do with that extra time? more studying! Professors are going to pitch to the middle of their class, so a higher average skill will mean more and more rigorous material.
Why put all the smart people in one room together? Because they learn more that way! How else are you going to populate a class on measure theory and Hilbert spaces?
Well it's clear what we need to do then: regard Harvard graduates as contemptible losers. Leaning much too heavily on one or two personal experiences, this is what I've been doing for years. If more people were like me, the "signal" of having a Harvard degree would have negative value, and they'd be forced to drop the self-undermining privilege-laundering business.
On the mixing theory, what's the optimal ratio of merit to total? Now that Harvard HAS the prestige, why does it need to keep admitting Princesses of Whaterevstan at all? Does it all come down to who pays full tuition/has the highest expected value of becoming a future donor?
Those Matt Y blog commenters ;-). They know a lot about "what they've experienced" and they have some fluent ability to argue-amid-near-total-agreement, and to find stats - but their weird long foray into discussing evangelicals demonstrates that they have *no* clue how close of relatives evangelicals and Protestants more generally - are to the progressives who dominate Harvard.
The privilege smuggling makes more sense to me than hand-wavy "excellence" as a reason to want to go to one of these places. If they had just put that in their promotional material, I might have tried harder to go somewhere elite.
I earned an AB at Columbia in 1962. The degree was granted by Columbia College, one of at least three undergraduate operations at Columbia University. It is a truth generally acknowledged that if an organization operates under multiple aliases, that organization must perforce be criminal.
Nevertheless my education was honest and valuable in ways I didn't know. When you attend a selective university you very quickly learn. how good you really are. I was the product of a rigorous selective school system, but I soon met classmates who were smarter than me.
I had to learn coping mechanisms, for the English Department was a formidable place. I became something of a specialist in lesser known but reputable writers and had a knack for inserting them into my many, many papers. A life skill for a future academic.
Had I gone to a less selective school, I might have had an inflated view of my abilities and so suffered both disappointment and helplessness.
So, yes, the selective schools are crooked, at least in America. Secondary education is controlled by local school boards and supervised by state politicoes. The national exams are an embarrassment. Nevertheless talent will out and at least serve as a measuring rod for the bulk of the students accepted by less than honest means.
The mediocre legacy/sports graduate inheriting a business should have at least a sense for ability in his employees. So maybe that suffices.
All of this is correct and well-known. The missing follow-up question is: why, then, don't colleges compete ruthlessly with each other for top students, trying to displace the Ivies as top-rank signalers--and reap the rewards--by offering spectacular incentives to high-ranking students, the way, say, college athletic teams do (often in violation of NCAA rules)?
The answer is that they're all part of a massive cartel zealously guarding their collective position as granters of signaling credentials, and raking in obscene revenues in the process by charging cartel rents. If they began competing for students in an open market for signaling credentials instead of following the cartel's rules, they'd end up both massively reducing the cartel's overall income and revealing the market's true nature, thus opening it up to lots of outside competitors willing to undercut them. So instead they play by the cartel rules, which includes maintaining the fiction that they're providing education rather than signaling credentials, and using that pretense to justify their collusive control of the market and very limited, almost performative competitive practices that never threaten the cartel regime.
I tend to think this is correct, but that then raises the question: why are the Ivies so intent on maintaining race-based affirmative action policies in admission? Does that not undermine the meritocracy and provide little financial reward, at least in the short run, compared to legacy admits? Maybe the idea is you need to keep a certain number of minorities around to guard against the possibility of those minority groups forming alternative institutions? IE, maybe if Harvard doesn't consider racial backgrounds in admission, black enrollment declines fairly precipitously, black students start thinking of Harvard as kind of a white/Asian school, don't want or try to go there, and head off to HBCU's instead, and the result is that Howard University becomes known sort of as the Black Harvard, with the same status-conferring ability as Harvard, only limited to black students, but which nevertheless represents serious competition for Harvard & Co.
Any other reasons? I'm sure Freddie De Boer has written something about this, but I'm at work and don't have time to look for it.
In addition, it serve as daycare for pre-adolescent (elementary school) and day prison for adolescents/ young adults (high school), which most organized societies seems afraid to let roam freely :-).
It also serve as a big sieve, it ranks students into bad/average/good. The western world equivalent of Chinese imperial exams.
And finally, it also act as indoctrination stage to try to keep future citizen in line with the official narrative, and avoid possible troubles ahead/promote societal stability.
Telling which is the most important mission is left to individual appreciation, but my personal belief is that (1) is certainly not top of the list. With possible exception of writing/reading and basic arithmetic, I believe (1) is only a distant fourth goal. The 3 others are far more important, but sorting them out is difficult and tend to change from time/places to time/places...
Using good schools (in the sense they have a good track record of student achievement) to teach the weakest students would maybe make sense for teaching.
But not for daycare/dayprison, not for indoctrination, and certainly not for sorting students.
> (Is is true? An oft-cited paper, Dale and Krueger, appears to find that, controlling for applicant characteristics, people who attend more selective college don’t earn more money later in life. Here’s a gesture at a challenge to these results, apparently supported by Dale and Krueger themselves, though I can’t find any more information. Earnings are a poor proxy for “teaches better” - it would be great to have something like value-add to GRE scores - but AFAIK no study like that exists.)
Bryan Caplan's "The Case Against Education" seems to have a lot of evidence that students learn very little of anything (at least on average) during most education, including college. This includes both specific facts and "critical thinking skills." (His conclusion, of course, is that the primary reason students make more money after getting a degree is signaling; he disagrees that controlling for pre-existing characteristics completely eliminates the college wage premium).
> what does the Princess get? I think she gets the right to say she went to Harvard, an honor which is known to go mostly to the meritorious.
A couple of other ideas:
1. She gets an excuse to relax and party for 4 years while pretending to do something productive, relatively out of the Whereverstan public eye
2. The royal family of Whereverstan might be sufficiently disconnected from what American rationalists know that they're still laboring under the impression that a Harvard poli sci degree or something similar would teach information that is useful for a future world leader
3. She gets to meet the kids of other world leaders in a different setting than normal
4. It gives them an air of legitimacy back in their home country (similar to what you said but with more specific motivation)
> This is the wrong question: the right question is why they ever give spots based on merit at all.
It might be interesting to look into the history of how this all came to be. As you pointed out in an old SSC post about college admissions, it used to be relatively easy to get in once you cleared a few (mostly class-related) hurdles. I have no idea what kind of prestige (or much) a Harvard degree got you in 1880--my vague impression is that most people didn't really care and it was just sort of expected for certain families, but I don't actually know.
I'm genuinely impressed that Scott just signal-boosted Matt Christman, one of the most prominent Marxist podcasters. There's hope yet. (Capital M Marxist, too, not a crypto-Marxist).
Another prosocial reason for elite schools is the fact that smart students simply do better work when surrounded by their intellectual peers. They can bounce ideas off each other, and since they tend to be competitive, they work harder to beat their peers. In fact, I would say that the actual teacher of any given classroom matters far, far less than the quality of its students. The latter is almost always the limiting factor. Top-notch researchers also benefit from having a smart undergraduate population nearby, and vice versa.
I went to a small, private, engineering-focused university. There, I was unamiguously one of the top two students in a class of 50-100 students in my degree program, and tended to be well ahead of course material with relatively little effort. However, when I went to grad school where there was a tighter selection criterion, the courses covered material at a pace that was far more challenging to me.
A few years later, I took a couple of MIT's open courseware practical mathematics courses aimed at undergrads. I was impressed with how quickly they expected students to grasp the fundamental ideas. It was a lot more similar in pace to grad school than to my undergrad.
When I see someone say:
> most of the classes at top colleges are the same material that gets taught everywhere else
I can only interpret that as either
1) Engineering undergrad is notably harder than whatever degrees they're thinking of, or
2) They're ignoring how high up on Bloom's Taxonomy a course gets.
I can't speak to 1. I have only engineering and mathematics degrees, and both seem to have this pattern. But 2 is worth a few words.
Let's take the fundamental mechanical/civil engineering physics course 'Static Mechanics', which has a required equivalent in most or all universities. A local private university has the following course description:
> Subjects cover includes; force and moment vectors, equivalent systems, trusses, frames, and machines, equilibrium of particles and rigid bodies, static friction, centroids and moments of inertia.
(Why do they use a semi-colon instead of a colon? Who knows! Hopefully they're more rigorous in their engineering than in their punctuation.)
And lets take the MIT course description:
> Introduction to statics and the mechanics of deformable solids. Emphasis on the three basic principles of equilibrium, geometric compatibility, and material behavior. Stress and its relation to force and moment; strain and its relation to displacement; linear elasticity with thermal expansion. Failure modes. Application to simple engineering structures such as rods, shafts, beams, and trusses. Application to biomechanics of natural materials and structures.
Those sound really similar. The only concrete difference I can take from those course descriptions is that the MIT one definitely includes deformable solids, and the local one doesn't necessarily. How is that compatible with one covering the material so much faster than the other (suiting their better incoming student class?) It's really simple; if your students can grasp the basic facts and mathematics faster (Bloom's Remember and Understand tiers), you can then actually spend time on how to apply these mathematical facts to problems.
After all, problems do not come to professional engineers neatly giftwrapped with a 'USE ES201 FORMULA ON ME, LIKE WE TALKED ABOUT IN CLASS LAST MONTH' label attached to them. Learning to understand an engineering model is hard work, but it's not sufficient to actually solve problems (outside of very special cases.) You need to be able to apply it, and to analyze how appropriate of a fit the model is to the problem in reality you're trying to solve. Less time spent on learning the model is more time spent on it's limits, when and how to use it, etc.
I guess, in engineering at least, my model isn't any of Scott's. It's not that only the best teachers can teach these incredibly arcane topics. Or that we get the best payoff from matching best teachers and best students. It's that the best universities are those that attract the best students, and because they have the best students they can teach them more and better than they could if they had to slow down for worse students. And because they have more money (due to turning out better engineers who donate more to them, etc.) they can pay better teachers as well; but this is a small part of the improvement. The good teachers at a middling school could do as well as an average teacher at MIT, given the same students and a few years to tune their curriculum to those students.
How far this model extends to non-engineering is a real question; it's worth noting that MIT does not do legacy admissions, and so are not engaged in the proposed privilege laundering. I don't have the experience required to compare, say, HIST 30094: Modern India and Pakistan at a middling university and a top one.
The privilege-laundering idea seems kind of far-fetched. I suppose Americans do care enough about 'deserving' to inherit your father's business that it might seem plausible, but isn't the simple theory that colleges need money enough to explain why they continue to accept donations/rich kids?
Would be interesting to see a bit of a comparison to other countries where this happens far less/not at all.
For instance in the UK, Oxford and Cambridge entry is almost entirely based on exams and aptitude tests and so they raise far less in donations, partly because it’s basically impossible to buy your way in.
Would say this enables courses that are materially different/more challenging than other universities. Does in theory lead to a UK elite that is more academically inclined, but for lots of things outside of academia being a hard working smart nerd isn’t necessarily the most useful skillset to have in charge.
>most of the classes at top colleges are the same material that gets taught everywhere else
I disagree with this, particularly for STEM. The amount of information that faculty are able to productively spew at MIT undergrads is vastly higher than it would be at Midwit State. As a concrete example, Caltech uses Jackson's _Classical Electrodynamics_ as its undergrad E&M text. That's a book explicitly aimed at grad students. Only at Caltech/MIT/Harvard could it be used as the starting point.
Basic ability matters. When you select for it, you really can do more.
Among elite-college true believers, it's typically believed that the advantage of [elite school X, let's say Harvard] is not in the institution, general curriculum, and staff, but in the (1) classmates, (2) tails of the curriculum, and (3) other-than-classes academic opportunities. So by the steelmanned perspective of the pro-Harvard side, "most of the classes at top colleges are the same material that gets taught everywhere else" and "[elite colleges have] the world’s best writing teacher[s]" are missing the active ingredient of "best schools".
Example of 1: You can teach the students the same subject, but if you've selected a population of more-capable students, they'll have a higher rate of starting world-changing projects when they leave (or by dropping out). So the active ingredient (say the pro-elite-colleges partisans) is matching best students with best students.
Example of 2: "you don’t get into subjects that need world experts until postgrad" is true in the median case, but not true at the 90%tile of eg Harvard students studying technical subjects.
Example of 3: "access to research opportunities w/ faculty (selected for something like their research output)", covered in other comments.
The privilege laundering take is overly reductionist. It assumes that wealth and social capital have no relevant value and so it needs to be laundered with the smart/talented set to hide its inadequacy. But in reality, wealth and social capital are independent factors for the attainment of career prestige and so are valuable as predictors of future success (measured as high prestige leadership roles in society). Harvard is first and foremost an institution to signal prestige, and so their goal is to find the student population that will most likely end up in high prestige roles in society. There is no laundering going on, they are simply maximizing the relevance of a Harvard education in signaling prestige. The key point is to recognize that admission to Harvard isn't a reward for being the top-whatever percent in academic merit, its about what makeup will grow Harvard's prestige brand. This also has explanatory power for why the Asian admission rate is artificially suppressed. Asians tend to not have as many paths to attaining prestige positions in society, and so the value to Harvard's future prestige from admitting an Asian student is lower than a white student of lesser academic merit.
Another somewhat optimistic answer would go along the lines we want to maximize the chance of a genius with a clever idea actually pursuing it successfully and changing the world by researching a new tech or making a startup that uses new tech or something. This is much more likely if instead of having a bunch of average friends who say that's a cool idea and go about their day, they have a bunch of friends who are very competent or ambitious or rich who hear their idea, say let's make it happen, and help them either as hyper competent early employees or by giving them early funding either of which helps get the idea of the ground and gain momentum. Not sure how much I endorse this as accurate but it's more than zero and probably more than either of the optimistic takes you present.
In addition to providing "merit washing" by association for the children of the hereditary "nobility", the "commoners" admitted to Harvard and other "Ivies" graduate to staff the administrative (i.e. reasonably competent, well-compensated servants of the nobility) backbone of the Anglo empire. This is absolutely required on account of the well-known "regression to the mean" effect -- the nobles no longer have a dozen+ children each and cannot produce competent successors for themselves through purely biological means. So they outsource the "brains", while passing on their estates to biological heirs (often, as discussed, merit-washed for popular legitimacy) who will command the "brains", rewarding the latter with mid-6-figure wages (and sometimes a "lottery ticket" chance of a small share in capital ownership, via e.g. the startup circus.)
"Brain" isn't the primary selection criterion for that servant caste, however: the selected people must first and foremost be unshakeably loyal to their hereditary aristo masters. And so the primary criterion remains political reliability (as proxied by e.g. "affirmative action", aka "Кто был ничем, тот станет всем" ("who was nothing, will become everything", as seen in Ru version of "L'Internationale"), the correct "extracurriculars" (elite bonding rituals: rowing, excursions for imperialist "aid" to the Third World, etc.)
One can compare with the old USSR: as soon as a hereditary elite caste (re)appeared there, that country started behaving like a cheap clone of the Anglo empire, in a number of comically-obvious ways. It turned its Moscow State University into something quite like a copy of Harvard (in this particular respect) -- complete with "affirmative action", "easy majors" for aristo children, guaranteed employment for graduates in the mandarinate, and so on.
Every sclerotic bureaucratic empire must have "a Harvard".
"why do we send the top students to the best colleges? Why not send the weakest students to the best colleges, since they need the most help?"
Because educational resources are wasted on the weakest students.
It makes sense to give the most resources to those who can take advantage of them. Compare to athletics. You want world class facilities, doctors, trainers, etc. to be available to Olympic level athletes, so that they can go win medals for their countries, not to overweight people using the electric scooters at Walmart. The latter group might benefit from some of the resources, but their goal is so crude (lose weight) that they can do it with very basic resources (a street to walk on and some sneakers.) Likewise, the weakest students just need to be able to do the most mundane of tasks. No fancy anything needed to teach that. If you want new genius ideas/products/whatever to be created, you want to match the brightest students with the brightest people in their fields of study and the best equipment, facilities, etc.
How about like admission only high schools, the top schools write their own admissions test and fill the class only based on test score results? The test can cover what might be on an SAT, but also some subject material, but merit admissions would be the result. Yep, the endowments would suffer, but probably not all that much over time as networking and signaling would come in to play.
I think you're wrong to assume that this is designed in rather than naturally arrising.
In any situation where you're trying to find a matching based on mutual acceptance - i.e. where students can choose which universities to apply to and accept offers from, and Universities can choose which students to make offers to and which to reject - and where most of the students have similar rankings of how desirable they find different universities and most of the universities have similar rankings of how desirable they find different students, you're going to end up with the two rankings roughly matching - the most-desired students will end up at the most-desired universities, and so on down.
Most universities would rather have more able than less able students. So, whatever we mean by "best colleges", the most able students are likely to end up there.
'Privilege laundering' sounds a bit paranoid- as though the privileged are actually ashamed of themselves and are trying to pass as meritorious because... the privileged are actually shape-shifting lizards from Planet X! Wake up sheeple!
It may have been true that the sons of robber-barons wished to mingle with clerics and thus to come across as cultured and pious. Universities, after all, were originally ecclesiastical in origin. Similarly, the sons of legal shysters or crooked bureaucrats may have wanted to rub shoulders with the descendants of robber-barons under the pretense of devotion to Humanistic scholarship. Napoleonic France went in a different, supposedly more meritocratic direction. However, it was the intense competition between the Universities of German States which created the Seminar System and the notion of the Doctorate as a Research Degree. In the STEM subjects, this incremental approach meant that there were external economies of scale for Marshallian industrial districts more particularly where Finance partnered with Industry to create a system of cross-holdings.
America's extraordinary rise to Academic pre-eminence was based firstly on expansion of a more academic type of High School to virtually the entire population in the period from 1914 to 1938. In Europe, the assumption was that only a small percentage of scholars would attend 'Grammar Schools' or 'Gymnasiums' as opposed to Technical Schools of some type. Thus Americans were more ready to take advantage of the GI Bill, than any European country.
The second factor was the reputational benefit of creating a University, or endowing an Institute- e.g. that of Advanced Study at Princeton- and the networking benefits of alumni associations. The Land-Grant Universities could emulate this type of success, though, no doubt, some had already attained excellence in some fields- e.g. agronomy.
The third factor was the Cold War and 'Think Tanks' like RAND. Universities were competing for political influence. Chicago was at war with Harvard. This Saltwater vs. Freshwater battle defined the ideological space when I was growing up. Then, in the same manner that 'Independent Expenditure' campaigns disintermediated Political Parties- or, indeed, ideological Punditry- and gave salience to wedge-issues on the one hand, and straightforward, utterly blatant, special interest rent seeking on the other- so too did the run-away success of some, highly technical, research programs which, so to speak, 'unbundled' the Liberal 'idea of the University'.
It is this 'unbundling'- or lack of cross-elasticity, or synergy- which has made the 'matching problem' acute. On the one hand, you have dead-in-the-water research programs which have to fools some seemingly smart kids into signing up for them so as to keep their Ponzi scheme trundling on, while, on the other hand, the opportunity cost for smart, well connected, kids of investing in 4 year College has been rising. Essentially, there is a 'laundering' of worthless nonsense which however does not appear to be any such thing because smart kids of diverse backgrounds are signing up for it and- hoping to get tenure- being very vocal in their support for it, while, on the other hand, there are more and more veils being draped over the truth- viz. kids nowadays need to be like medieval scholars who moved from one campus to another. Indeed, this was a feature of German university life in the late eighteenth and early nineteenth century. That type of mobility is what put Germany ahead by the 1880s which is when John Adams spent a little time at a Gymnasium. However, he came to the conclusion that the future would lie with the Engineer-Savant- guys like Pareto and Barone.
What is the actual 'matching problem' currently faced by students and Colleges? Firstly, the set of students is only well-ordered for particular, quite narrowly defined, research programs- which however are very lucrative and probably vital to our national security. There has to be an unbundling event even if it means killing the sacred cow of tenure. Otherwise, as Vivek Ramaswami seems to be saying, the Chinese will eat our lunch.
A separate issue, the one this post and the comments on it focus on, has to do with coalition formation and stability. Essentially, the claim is that alumni of certain Colleges are dictatorial over the core of particular games in particular industries or particular contexts. However this should only concern us if there is something extraordinary about those recruited by the Colleges in question or, less plausibly, that the Colleges impart extraordinary traits to its students. But, in that case, the relevant 'Structural Causal Model' would enable us to improve outcomes for everybody. That's the opposite of paranoia or the politics of envy or spite.
Still, one might say that there is something egregiously wrong with our Social System if Harvard alumni write the jokes in our favorite TV programs. But, surely, that was a joke created by TV joke-writers to raise their own prestige while acknowledging that White America was white-collar and College educated.
Why do 'elites' want to let their kids play tennis with kids who are really good at tennis rather than the paralytic scion of inbred Aristocrats who can't lift a racquet? The answer is that their kids may benefit by playing more and better tennis. Sports, unlike almost all non-STEM instruction in H.E, is actually good for kids. Universities with good sport facilities- or one's located in big Cities where such facilities are easily accessible to the very rich- are a good place to gain a bit of polish and perhaps- like the Prince of Wales- pick up an acceptable spouse of good breeding stock. Assortative mating at Uni beats caste-based inbreeding. Or so it now appears. Once genetic engineering really takes off, the elite may indeed elect to become a different species. Till then, all we can do is pretend they are shape-shifting lizards who are using Harvard and Yale to camoflauge themselves.
The premise of the question is incorrect: we do not send the top students to the top universities. Instead, we provide equal opportunities to all students, and the schools select whom they accept. What is the motivation for Harvard to accept someone with less-than-stellar grades? There are financial, influential, and prestige reasons as the author describes. But beyond that, the risk of harm to Harvard's reputation is too great. And yes, there's risk: if you meet 6 Harvard graduates in a row and they're each individually morons, Harvard's reputation takes a hit among you and those whom you influence. University quality is based entirely on reputation.
I say we fix that. Here's a plan that's a step too far, so we'll moderate it afterwards but what if: State legislatures require universities to post all graded student work online, as a condition of maintaining their business licenses. Suddenly a grand corpus of college student work is built. After a few years, artificial intelligence agents (and natural intelligence agents) can assess those bodies of student works to determine the actual quality of a degree program at one institution compared to similar programs at other institutions. Prospective students can already determine the cost of education and they can now determine the quality of education being offered, so they are able to make value decisions. This drives cost down and quality up, if the number of seats at universities is collectively more than the number of customers. (It is.)
This proposal, of course, also allows each student's demonstrated knowledge to be evaluated. The up side is that employers (supported by agents that can evaluate the job candidate's artifacts in whole) can weight what they consider to be important. Need a business major with international finance expertise and who speaks Chinese? Need a scientist who understands issues in a global context? Done and done. The knock-on effect is that universities finally learn and become forced to react to what industry actually values. If no engineer from Fictional State University can get a job because their engineers know everything about DEI but nothing about engineering, that sends a message!
But this proposal also means anyone can read the paper that the professor required you to write about how great Slavery is. You had to write it or you would fail the class, but you never believed what you wrote and now you're running for office. So there has to be a way to shield the data from those who should not be viewing it. Fortunately, this is easy and mature. Data is available anonymized and in aggregate, and then students receive keys so they can grant individuals / organizations access to their work for time-limited periods up to one month (for example).
I read the title and thought "incentives". Students mostly prefer to have good teachers, and teachers, good students. It's really hard to create a performance test that someone can't fail on purpose. If you reward people for failing, the test ceases to be meaningful. Better to have a system where people mostly WANT good grades.
My naive understanding is there is considerable value in the connections made at elite institutations compared to connections made at eastern state university. Connecting the smartest with those of means and influence seems reasonable if goal is to create high value connections and catalyze higher value results.
Have you considered that this analysis fails to take into account DEI? For example, if Harvard was 75% merit and 25% nepo babies, that would still be a decent ratio. But in the modern day, I think the ratio for Harvard would be more like 50% merit, 25% nepo babies, and 25% diversity admissions. In that situation, it's much harder to launder privilege because nowadays, hiring a Harvard graduate gives you a 50% chance of them being a moron. I'm not exaggerating here: in fact, the situation is so bad that if I owned a company then when it came to hiring I would not view Harvard as a prestigious school at all and would be weigh a Harvard degree roughly equally to a degree from a state school. The only Ivy League degree I would strongly value would be MIT.
Wow, education is one of those topics where everyone is an expert because they went to school and yet we can't just blindly trust the experts as the system we have doesn't match the values of the common person who wants to improve their life. In this way experts are goal aligned and incentive aligned to come up with the system we currently have as a clear tautology. Despite the verbal only and public protestations otherwise, the hoarding and laundering of privilege with some admixture capture of the best commoners is the standard playbook for a successful aristocracy.
The TL;DR - is in line with Scott's final point about why there is a meritocracy at all. I don't think Scott has gone broad enough with his thinking. It isn't a simple matter of reputation or prestige or validity which these commoners lend to the elite/aristocracy. It is an entirely different process where patrons are present to meet promising commoners whom they can extract value and benefits from.
Would we be confused if we saw lesser and fallen nobility along with the top merchant families sending their best and brightest to a fancy meeting ground with the children of the higher aristocrats and nobility? Change a few words and the truth is clear. They don't want to extract privileged status from commoners, they already have that by birth. They want to extract connections, value, and to both cultivate and harvest the best from the herd of commoners.
I.
I find much of the confusion, hand wringing, and democratic discussion goes away if you just replace the word elite with aristocrat and learn a bit of history.
The renaissance model was very very successful for the aristocrats in Europe. They had better living conditions, more technology, an age of colonialism fuelled by improved weapons, and were all around better off and all they had to do was expand the idea of patronage beyond the arts into the areas of learning about the natural world. This attitude of harvesting the minds of the commoners and lower nobility and creating positions for them to study the world to see what they can figure out...this was super super super successful.
Our education system is simply an extension of this model to sort the wheat from the chaff. It is a mining operation by the aristocrats to get all the commoners to sort themselves out in competition. Meritocracy is 100% fully at odds with the core nature of aristocracy and they never ever intended to apply those ideas to themselves. They don't see themselves as the same as the commoners, they really don't and the vast majority of commoners have never met nor been privy to the inner thoughts of the aristocrats.
II.
We are a breeding stock and are like cattle they own. If you had an intelligent flock or herd of animals who might rise up and kill you if you told them the truth or let them learn or know the truth on their own, then you'd have a very unstable system with a lot of dead elite.
The Russian and French revolutions were a wake up call to expand mass indoctrination beyond simple religious compliance of commoners to their 'betters'. A new intellectual edifice would need to be built to contain commoners within a stable framework. I think some aristocrats worked on this problem in a very intentional way, but a lot of this is simply an alignment of incentives. If driving down one road clearly takes you off a cliff and the other to a bridge to cross the gap, then it isn't a particularly difficult choice.
III.
The great incubator of trial and error took hold and many many many diverse attempts, approaches, and styles of commoner management (one of the core tasks of the aristocracy) emerged in this new age of educated commoners who could think and read. The simple printing press which emerged long before commoners could read led to several revolutions around the world. Later on the radio was used by people in the Caribbean to organise revolts and resistance to colonial rule.
But the aristocracy finds a way and from more totalitarian approaches with daily government radio broadcast to owning the airwaves as a matter of governmental proxy of who is allowed to speak all the way to the Twitter Files style active suppression and censorship of speech.
This is nothing new and is a process within the core problem of elites wanting to harvest commoner minds in an extraction operation and the problems that causes. By and large this problem has been solved through a mixture of violence like the selected murders/assassinations of key leaders within commoner movements which challenge aristocratic authority or goals.
IV.
And the other part is a huge mind control operation of values and the stage production of democracy. What makes it so very convincing is the fact that the stage production is somewhat real with a small percentage of choices allowed to the commoners through a process - especially on issues which DO NOT affect the bottom line or power of the aristocrats who call themselves elites now.
The successful formula has proven to match nature, the carrot and the stick. Rewards and consequences. A stable system with some elements of true meritocracy in a bloody, unfair, and arbitrary process which still ends up costing a lot of resources. Combined with a careful sculpting of 'values'. You can read about this from one of the founders of the modern education system in Rockefeller who wanted a compliant and cooperative sea of minds which could be shaped and directed to the purpose of enriching him and empowering him further. He was very very open about it with all those initial grants and bribes to get modern mass education started. He wanted smarter workers, but not too smart.
But hey...a lot of people who went to those schools and picked up 12 plus years of someone else's values and worldview being drummed into their heads are all experts on what happened to them and why it was good and everyone should join the cult. And yet even with all the benefits of integration you find the truly wild peoples of the world are very very very very very resistant and reject schooling due to those embedded values.
The focus in the first few paragraphs on the classroom teaching skills of the professors is misplaced. Professors at top universities are famously rewarded for research much more than teaching, and there are few courses even at the graduate level that you need a special expert to teach. But when you have a class full of talented students you can go faster and they give each other better insights. This is not to say that the system we actually have isn't a hodgepodge driven by all the forces in the rest of the post.
Incidentally, The Chosen by Jerome Karabel is indispensable reading on this.
If hanlon’s razor is never attribute to malice blah blah, what would we call a razor that says look first to emergence based on game theory to explain real life situations? (So I think the medium cynical is probably the center of gravity, with this proviso: you’d expect the universities to be exactly as merit based as necessary to maintain their rank while gifting the rest of the slots to those likely to give them donations. The college rankings based on selectivity probably also help maintain the equilibrium; otherwise there’d be a much bigger incentive to give away greater and greater percentages of spots while you traded on your name for a while.) maybe assuming it’s that and not some kind of top-down design is also hanlon’s razor?
By my reading of the link, Dale and Krueger don’t support the challenge, they support the author of that article’s interpretation (i.e. that character matters)
While I'm less sure about Harvard my experience with Caltech suggests to me thar:
1) Some students are more capable than others and society benefits if those more capable students are pushed to the limits of their ability.
2) The best way to push students to the limit of their ability is to put them in classes with similarly able students. It's human nature to slack off or grow uninterested if bored and to rise to the challenge if not.
3) Whatever place you concentrate the best students at will become highly desierable to attend so selectivity is necessary to preserve that feature.
4) Wealth and the other trappings follow because producing the most successful graduates draws alumni money.
--
Yes you often get the best professors at the best schools but, while that makes form a nice research opportunity they are usually not the best instructors so really everyone wins.
I'm sorry, but the expression "Maharaja of Whereverstan" so typically blends American Exceptionalism with White privilege. Not seeing it and/or not understanding why, reeks even more of it. Whoever coined that expression should be first in the line to get their privilege laundered. Other than that the play was great.
This is likely true. I think there are big nuances in the distribution of student quality and of teacher skill-sets though. You touch on the latter; most professors can teach the basics, but few can teach truly advanced skills. There's no point wasting the latter on students who can't learn what they can teach.
Assuming an equal ability to teach things (all teachers teach the same stuff, but some are better at it),
Matt's kid isn't entirely wrong. It *would* work to send the worst students to the best teachers. They'd be better off for the experience. But it would be ridiculously inefficient. The distribution of a class made up entirely of 'worst' students might range between, say, the 10th and 30th percentiles of ability. Teachers teach to the average, at best, and probably more likely to the minimum fixable level (maybe the 15th percentile in this example). So progress is extremely limited, even for the best of the worst. And the difference between 30th percentile ability and 10th is much bigger than it sounds (it's not linear in the tails).
Elite institutions work the same way. It's about teacher quality and class availability, yes, but it's also about the degree to which your classes and teachers are able to be the best versions of themselves. An elite school has elite students; the lowest fixable minimum is irrelevant, because everyone's smart. Teaching to the average still means teaching to a very high level, and because the left tail is truncated there isn't so much of a gap between average and worst. Plus, the best kids in a place like that have ample opportunities to excel; they won't be held back much. Calc at Harvard (or MIT) is not calc at Western Prairie State.
Your theory about laundering privilege still works. Harvard's incentives are to keep quality high to facilitate those dynamics by admitting mostly the best. And it's not like complete idiots get into Harvard anyway, no matter how big the donation. The left tail is still truncated; it's just a bit lower and longer than it would be otherwise.
Thank you l, this is refreshingly in the middle of the argument. On a slightly different note,I've been thinking about and consequently noticing how 'virtue-signaling' is a deception in which people engage no matter the context.
I went to a specialized high school in NYC. We had a 99% college acceptance rate. That was 1000% due to the smartest students testing in, and 0% due to enhanced teaching capabilities. The administration was known to be a joke (teachers got together and wrote a public letter saying the principal was both incompetent and placed by nepotism). If you just put the worst students at the school, scores would plummet immediately. Frankly attending wasn't worth it for me, would have been better to go to my local school.
You focus on rich people laundering their prestige by making themselves look talented/smart, but I think it goes a bit further.
Premier schools also launder the privilege of *being* talented/smart, to make those people look competent/hardworking/trustworthy/etc.
This is in service of the political idea of meritocracy, that being smart/talented is sufficient justification for some people to have control over the lives of others and over the direction of the economy/country, and for them to deserve richer and easier lives with less hardship and more pleasure.
If you know a bunch of precocious 'gifted' children who are smart/talented in some ways but also barely able to take care of themselves and often venal or self-centered, you might question the wisdom or justice of this situation.
Even if all the gifted people you know are saints, you may worry about value misalignment or lack of domain knowledge between them and the masses, or simply be unconvinced that their competence at running the system justifies an unequal system with coercive power.
But a Harvard degree transforms the privilege of smart/talented into an aura of 'competent to lead' and 'part of the natural ruling class'.
I was a Harvard undergraduate in physics in the early-mid 1970s, after attending public schools on the south shore of Long Island (Nassau County, not the Hamptons). The greatest benefit I got from meeting elites was being able to see with my own eyes that coming from money didn't have any correlation with being a decent person. OTOH, the people I knew who had come to Harvard from poorer backgrounds than mine tended to be pretty nice, and often probably a lot nicer than I was. Also, I met a lot of jerks, of various backgrounds.
Obviously this is anecdotal. But thereafter I wasn't ever impressed by people with money just because they had money. Nor did I ever feel that making a lot of money, even from one's wits rather than an inheritance, could excuse someone for being a jerk. This was very liberating. I'm not sure I'd have gotten the same feeling from going to another Ivy, since I might have fantasized that people from Harvard were exceptional in some way.
In the roughly 50 years since, I haven't gotten any contracts in Whereverstan. The only work I ever got from a college connection was from a friend who kindly asked me to do legal work for his venture-backed start-up, at a time when I really needed some work. BTW, he grew up with a single Latin American immigrant parent, and cleaned toilets for a couple of years as his campus job. The start-up failed within three or four years, as most of them do.
My experience with Harvard faculty was mixed. They ran the gamut from inspiring to mediocre to jerks, with most in the middle -- though to be fair, I was too self-involved at that age to know how to build relationships with any of them. There are a couple I really regret not paying more attention to at the time (mathematician Raoul Bott at the top of that list: he was a delightful person, I discovered after I'd cut his class for most of the semester).
The single most life-changing aspect for me was managing a student-run orchestra: so, yes, the peer argument about bright students liking to flock together is pretty cogent. And there were so many ideas floating around in my classes, conversations, in the libraries and bookstores in the area that I've spent much of the past half-century exploring them and seeing them branch into other things. I'd never say that my college years were my best years, but Harvard is why I have a rich intellectual life today.
Which is why I think posts like the one above are simply low-dimensional cartoons, created by people who didn't go to Harvard and who could never imagine that it might be just as complicated as much of the rest of reality.
With regards to the initial point, I went to a top-ranked STEM school and taught/was a grad student at 2 other much lower ranked schools and can confirm that what the two schools taught was completely different. Students at the lower ranked school learned far less theory and weren't exposed to a lot of programming techniques or algorithms that we learned in the first 2 years at the top-ranked school. The grad-level algorithms course I took at one school basically entirely covered material that students at the top-ranked school were required to learn as an undergrad.
Poking my head in to explore, not a hypothetical situation. Relatives of mine (identical twin girls) recently accepted admission to Harvard and a New England state uni. Situationally on the “commoner” strata, if not “poor”(mother’s chronic disabling physical illness, parents divorced, father financially somewhat stable). White female New Englanders. Nearly identical academic achievements, including athletically (although not recruited athletically) at different public high schools. Test well. Constant stream of accomplishments related to very high ability mathematically, also verbally. Adversity also includes chronic illness of their own that requires careful management. Recently witnessed a drive by shooting outside the store where they work part-time. Between the two, had admissions with full scholarships at UNC, Syracuse, etc. One applied to Ivies, one did not. The one who has decided to attend Harvard was reluctant to accept. Why? Partly because her high school friend who resembles her closely, in ability, demographics, struggled greatly with social acceptance there for her freshman year last year. Aside from the accompanying “you don’t belong there” from her sister who did not apply. These two almost instinctively grasp that attending Harvard for them has it’s complications, not being wealthy. Their mom and dad grew up and/or attended college with others of great wealth; their experience was one of sometimes being more beholden than they would have liked, with no obvious long term benefits, because they both chose not to put themselves in service to those with the wealth. One of the twin daughters put her trust in her abilities that would make attending Harvard worthwhile. The other, the far spunkier of the two, did not. Their mom has been supportive of each as they made their choices. It has been interesting so far, to say the least, to read the recent SCOTUS decision in light of their recent experiences.
This is a question that sounds adorable, because it comes from a five-year-old. The same is not true of intelligent adults asking the same question.
My wife did a rotation in admissions, and participated in undergrad admissions decisions, while working as an assistant director at an Ivy. Her career: university fundraising.
(1) “Why do we send” is a false premise, “We” have no agency on admissions. And the only people sending anyone to a university is a situation where a family or equivalent is helping someone pay for & be successful at college. Colleges and universities are registered business entities. “We” refers to society, American society is not granted authority over university/college managers to make decisions about what any given universities and colleges do.
(2). A legit really effective teacher working as a professor is a somewhat rare individual, and highly valued by students, colleagues, the parents of students, alumni, and highly valued and sought after by other colleges and universities. People are free to make their own choices about where to apply to work, and teaching professors or research faculty with really great skills often have a number of nice options available.
In most human endeavors that people are passionate about and deeply invested in, A-players want to work with/collaborate with/mentor other A-players. Less so, mentoring B or C-players. An A-player teaching professor deciding to *focus on* helping C-player students is a pretty rare, laudable situation precisely because they chose something noble and against the grain. The professor who spends her time helping highly intelligent, motivated undergrads or grad students unlock challenging knowledge is no less laudable, and should be equally free of coercion by “We.”
The question implicitly presupposes that the quality of education might be better at a 'better' school. This is not true. Professors at highly ranked schools do not teach better than those at low-ranked schools. There is not some magic pedagogical method that is known only to elite schools. In fact, competition for faculty positions is so fierce that there are many excellent people at quite low-ranked schools. So, there is no reason to want mediocre students to go to a highly-ranked school - they would not benefit at all anyway.
What students benefit from is other students. Scott mentions cynical reasons this is good - hobnobbing with other elites, but there are more virtuous reasons.
One is that tracking is good for students. If all the good students are together they can learn at an accelerated pace and the mediocre students can learn at a slower pace. In the end, both groups are receiving instruction at the optimal level. This is a stable situation. If you try to introduce mediocre students to a good school, they will do poorly relative to other students. I know of students who were asked to withdraw from a good school to go to one more suitable for them, and in the end they were actually thankful for the advice. They learned more at the slower pace and went from being bottom-of the class losers to solid performers, which boosted their confidence.
If you suddenly changed the situation so that the worst schools were where the best students went, in a few years all that would happen is to reverse peoples sentiments about which schools are best/worst.
Young Yglesias's question is interesting because it suggests that the best colleges are the ones accepting small class sizes of highly promising high school students. There's a reasonable argument that Arizona State and the University of Central Florida are the best colleges in America because they produce so many college grads capable of entering the workforce.
Of all the great authors of the world how many of them majored in English lit or creative writing? I tend to the Charles Dickens and Herman Melville stumbled on writing while at University studying law or accounting or something.
Listening to a podcast on the affirmative action in admissions problem, I wondered: why don't we just ask colleges to accept everyone by lottery? Why even bother with SAT/ACT/GPA/GDP/whatever? Everything meritorious is in some way racist; not only is a non-discriminatory lottery non-racist, it also allows colleges to compete on the quality of their product—if you can produce high-earning graduates out of a random grab bag of students, you can claim legitimate prestige.
I am late to this but in more meritocratic university admittance policies there would be elite courses not elite universities, not at undergraduate level anyway. Medicine or Engineering would be difficult to get into at any university - and standards within those courses would be fairly uniform across institutions.
This is now Ireland works, the state exam determines what courses you can do and the standard needed for different courses varies significantly within any university.
That said Irish universities are not setting the world alight at research and post graduate level so maybe that’s a reason to concentrate the best in one, or a few, universities.
Is there any hard evidence of the theory that diversity causes better educational outcomes? I always see this asserted without evidence as a rationale for affirmative action, but it seems to be on much shakier ground than mismatch theory.
That may be true of Harvard, but I think there's a simpler explanation for the broader phenomenon: the brightest students would prefer to be taught by the brightest teachers (they'd learn more), and the brightest teachers would prefer to teach the brightest students (it's more rewarding). The only reason this isn't going on earlier in the education system is that governments feel a duty of care for kids, and make rules to stop it (at least in the UK), but adults are on their own, so the obvious equilibrium follows.
The example with the writing teacher is a bit absurd. The vast majority of great writers never had *any* formal training in creative writing at all. (or at most, basic)
So I heavily doubt the great writing teacher is going to do anything for the genius writer. The genius writer doesn't need a teacher, probably doesn't even want one.
Writing is not a normal profession - it's very debatable if teaching does anything good for you if your skill level is above mediocre.
A shaky assumption underneath this question is the idea that Harvard admissions committees have highly accurate screening systems that can select for true academic merit. Revisionist History has had several great episodes calling that assumption into question. At an anecdotal level, it's important to remember that Einstein was repeatedly bounced away from undergrad by standardized entrance exams. Then, after he graduated, he was saddled with crummy academic reference letters saying he'd never amount to anything.
Judging academic merit is surprisingly hard. The assumption that Harvard admissions committees are especially good at it doesn't seem well justified. Another possible answer to a five year-old's starting question might be that elite institutions should admit some seemingly sub-par students because some of the sub-par students might actually be geniuses.
I think there are plenty of elites that can make it into our nation's top schools, and the smartest elites will continue to do so. It will also create a greater incentive for wealthy families to prioritize education.
Alternatively: a sufficiently visible pipeline to success through education socially legitimizes learning as a credible way to achieve wealth and/or status (and achieve it virtuously!). There must be a clear reward within sight of the k-12 slog to focus teenagers and their families on academic success, and ‘go to a great school and you’re set’ might be a more convincing and actionable heuristic than ‘get good at something, wherever and however possible’.
For this reason (and to stick my neck out), we should be inclined to make the pipeline more credibly within reach of traditionally underserved/underachieving groups, perhaps through affirmative action and hamfisted attempts to showcase diversity in higher education.
It seems to me that the better the student is, the worse the professor is able to be. A smart student can teach themselves the material. If you look on MIT opencourseware, their teachers are pretty boring and unengaging. Meanwhile, the teachers at the "Great Courses" (intended for the layman) are quite excellent. This gels with what one commenter wrote about Harvard in 1975 and all the famous teachers being awful.
So the reason we don't send the worst students to the best schools is be because the best schools have the worst teachers, they don't need good teachers precisely because they are selecting for good students. The IVY league is not about making you smart, it is about proving you are already smart.
"Let’s imagine a stylized world where each applicant is either rich or smart.
If a college lets in only smart kids, it can’t keep the lights on. It goes out of business in short order.
If a college lets in only rich kids, in a year or two, the parents of the next crop of rich kids won’t be interested in making donations or paying tuition to a so-so institution. The rich parents will send their kids and dollars elsewhere.
So, there is some optimal allocation of rich and smart kids. At this optimal point, the rich kids’ parents’ money keeps the institution afloat and builds the labs where the smart kids commercialize nuclear fusion, cook up next-generation antibiotics, etc. The smart kids boost the college up the rankings and provide the prestige that draws in the rich parents.
Colleges have been playing this game for many rounds. They’ve had a chance to learn and tinker. I bet many of them know what that optimal point is. Relatedly, I think private-school applicants are so appealing to prestigious colleges because they sometimes allow admission offices to sidestep the tradeoff between rich and smart kids.
Of course, in real life, colleges are optimizing for more than two traits, some applicants align with more than one of those goals, and the super-selective places have huge endowments and so don’t have to operate from cash flow."
The privilege-laundring concept only applies to the US. In Europe, elite universities do not favour legacies, donors, etc.
For those of us without either kind of privilege, it seems pessimistically like a cruel joke and optimistically like a handy way of sorting who should go first to the wall when the revolution comes. (that's also a joke - the revolution isn't coming and the Ivy grads are safe)
Also look at it from the perspective of people who will actually teach these kids. Students in top schools do the readings and understand them better. It is easier to build on the basics and get to the interesting stuff. Students in poorer schools don't do the readings and even if they do have difficulty understanding them. The syllabus and standards have to be watered down. And you have to constantly sit on top of them because they often lack the integrity not to cheat. Teaching the first group is emotionally rewarding while teaching the second group is soul sucking. The difference is big enough that the switch from the second to the first was worth it for me even though I took a 15-20% pay cut. So, people with PhDs will compete to teach and research at where the smart hard-working students are. And given this dynamic, since faculty have say in admissions decisions, they are going to pick the ones they would prefer to teach
It also speaks to why not many voices seem to advocate for the elimination of the legacy system?
Eliezer's summary:
> that's Harvard's fundamental business? They admit actual smart kids on scholarship, alongside the kids of rich people who pay huge bribes. The smart kids get access to rich friends. The rich kids get to go to a college with a smartness rep as well as a richness rep.
https://twitter.com/ESYudkowsky/status/1677885529491386368
What do you think would happen if we forced schools with low acceptance rates to increase their admissions year over year in order to maintain access to public grants and loan subsidies? Would this help them launder priviledge more effectively or dilute their brand? Can we infer anything from the fact that Ivies are so disinterested in growing their enrollment and increasing their impact?
Why are so few schools good at creating good signaling? It isn't hard to imagine a world with ten Harvards, yet that isn't our world. I kinda like how Harvard makes such fools of us all.
I think this conflates high-quality professors in the "renowned expert in field" and in the "really good at teaching" sense. It makes sense for the people best in their fields to teach the smartest students.
Partly because they often do get to some work at the edge of the field - see e.g. Scott Aaronson's occasional mention of smart undergrads he did research with - but also because maybe smart people think about the field in similar ways. Brilliant researchers aren't necessarily better teachers, but might be a better fit for brilliant students.
Conversely, what we actually want for the less smart students is to have professors who are better at teaching. They don't necessarily have to be top researchers (those students aren't doing undergrad research anyway), and selecting on top researchers for professors would limit our ability to get top teachers.
Since doing cutting edge research is cooler than being a good teacher, this does naturally lead to the places with the top researchers being more prestigious and having more resources. But this consequence can happen even in a system optimized to teach low performers well!
Interesting comment. I wonder what readers’ thoughts are on the opposing vantage point: what if the “laundering” aspect for the elite actually dampens the meritocratic prestige that the rest of applicants gain.
At what ratio does this become untenable? Or does the ratio not matter because of the chance to network with elites trumps the label of meritocratic prestige? (Cynical view)
(1) The first part, why send the top students to the best colleges?, misses one key aspect of this which is that the top students would like to be among other top students. So even if instructional quality, career opportunities, etc., were exactly the same at all universities, there would be some sort of impetus to sort. The question then becomes why should this have any connection to research quality, "elite" faculty, and other such things. It needn't, but there's no good measure of anything related to instructional quality, career opportunities, etc., and we tend to assume (not without reason) that quality in one area implies quality in others. Plus, students in many fields would actually like to work with top people in those fields -- whether or not this is likely to happen -- so there's some direct motivation for the link as well. (And conversely, the top people would like to work with or teach the best students.)
(2) Selection based on merit was always less than ideal, but it is getting worse. Currently, for example, the UC schools won't even *accept* SAT scores from applicants. Whatever biases standardized tests may have, they are less biased than, for example, recommendation letters, flowery essays, and other characteristics. If the trend towards subjectivity continues, will the signals described in the post remain as strong?
(3) As a faculty member at a fairly large public university, I am often annoyed at how discussions of higher ed focus disproportionately on small places like Harvard (edit: 7000 undergrads (I mis-typed 2000 earlier)). I don't particularly care about Arizona State, but its enrollment (112,000 undergrads) and that of similar places should warrant a lot more press than is the case.
(4) I've heard the concept of privilege laundering before; it seems obviously true.
I always thought you wanted to make a cluster of smart people so they could learn and riff off each others ideas. Kinda a Bell labs model.
There should be a science of self-fulfilling prophecies. They turn up in lots of interesting places.
Or is there, under another name?
I'm not at all sure that the quality of teaching is better at Harvard than at less prestigious places. I used to have a teaching affiliation with Harvard Medical School and have met some Harvard faculty, and I have had as psychotherapy patients about a dozen Harvard undergrads and people getting advanced degrees at the graduate and professional schools. My observation is that the professors who are so eminent in their fields that they are actually famous are utterly unavailable to undergrads and even to the grad students who are writing dissertations under their supervision, or carrying out their research in the lab. They are always flying off to conferences all over the world and cancelling their lectures & having grad students do them. They are preoccupied with their careers, and hate having to teach: They view teaching as the toilet-cleaning side of being an academic. They teach undergrads via lectures in giant groups in big halls, then have their teaching assistants run weekly discussion groups. The grad students all hate having to be teaching assistants, and try to get paid jobs doing other things, such as being advisor in one of the residence halls or working with other grad students to publish a journal. Out of the half dozen grad students I have seen, at least 2 have been so cruelly neglected and ignored by their famous advisors. that it seemed criminal to me to deal out that kind of treatment to someone so smart, who has worked so hard to get to Harvard, and wants so badly to do really good research that will launch their careers.
My own experience as an undergrad was at 2 places, both ivies. For the first 2 years I was at acollege, at the last 2 a university. I had no lecture classes at all at the college -- class size varied from maybe 10 to 30. At the university I was taking higher level courses, which are always smaller, and mostly avoided lecture classes there too. I only had one genuinely famous professor, and that person did in fact teach via a giant lecture. As for my many other teachers, whose smarts I enjoyed and revered -- they changed my life. They gave me wonderful, accurate flashes of insight into their vision of their field, and they took a lot of interest in me. I really do not know if they were less brilliant than the famous scholars at Harvard, or just much less interested in fame. But even if they were, say, only 80% as smart about Wittgenstein, calculus, history of religion, linguistics, 20th century poets, etc, I do not think I was a developed enough being at that point to benefit from the remaining 20% anyway. And besides, conveying to students your own sharp insights and convoluted intuitions about the field is only one part of teaching. The other part is mentoring the student, and these people were mostly excellent at that.
Quality of teaching is only mildly correlated with the teacher's degree of fame and glory. It's way too simple to say that the more prestigious the school you attend, the better the teaching you receive. I think it's likely I would have had as good an experience at many schools with fairly small class size and teachers who were not egomaniacs.
Awesome, now apply this same lens to religion
I have one very relevant question. You can pay a hefty donation and have your child admitted to Harvard even if it lacks intellectual ability to be admitted without big money. But are such students treated differently later, or is anyone treated equally once they are admitted?
Sticking to a less cynical perspective, mixing elites with talented people has networking benefits for both sides. The talented people connect with elites who can help them into more elite-level roles, and the elites connect with talented people who can help them fill whatever important roles they have open.
e.g., The rich kid who is going to inherit his father's successful business finds talented people who can help him run that business, and the talented kids get the opportunity to work in high-level roles in this successful business.
Here's a take on it: students' intellectual development depends in part on the environment in which they grow. Those who grow among smart classmates become smarter than they would be otherwise. Not only that, but we achieve the greatest overall societal gains by grouping the smartest ones together.
For a very simple mathematical illustration, consider:
(x - y)^2 > 0 , for x!=y
(x - y)^2 = x^2 - 2xy + y^2
therefore: x^2 - 2xy + y^2 > 0
therefore: x^2 + y^2 > 2xy
If x and y are unequal, then you get the best results by matching x with x and y with y.
Illustration: for x=5 and y=2, matching the best with the best gives us 5*5 + 2*2 = 29, whereas matching best with worst only give us 5*2 + 2*5 = 20.
This model only has two variables, but I'm fairly sure it has been generalized long ago.
So the utility of clustering the best students in the best schools stems primarily from maximizing the students' overall potential. It's not only about matching the best students with the best teachers. It is also, and maybe more importantly up to a certain age, about matching the best students with the best classmates.
> why do we send the top students to the best colleges? Why not send the weakest students to the best colleges, since they need the most help?
There's a weird assumption in the response (and, I guess, the question) that "the best colleges" provide more help than other colleges do.
In reality, "the best colleges" are, by definition, the ones with the best students. The young Yglesias' question is similar to the philosopher who asks whether it's OK to prefer dating people you find attractive: conceptually incoherent. It isn't possible to do anything else.
----
Schools arise as a way for elite children to acculturate to each other. When talented nonelites begin to be let in, we should assume that the reason is to benefit the elites who operate and use the school. There is an obvious symmetry in what happens: the go-getters who make it in on ability get connections to people with power. (As mentioned in the post.) The elites who make it in on pedigree get the same connections, but since they are the other side of those connections, we would describe them as getting connections to people with ability. Connections like that can be important if you're ever put in charge of something!
It isn't necessary to assume that smart people get admitted to Harvard so that elites who also go there can pretend to be smart.
----
It also isn't obvious that legacy admissions are of lower quality than "merit" admissions. If all you know about an applicant is their SAT score, you can admit them and watch them regress to the population mean. If you know (1) their SAT score and (2) that both their parents went to Harvard, the same thing will happen, except that the relevant mean toward which they will regress is much higher.
> “Fairly-earned privilege” means all the brilliant talented ambitious youngsters admitted on the basis of their SAT scores and grades and impressive accomplishments
“Fairly-earned privilege” sounds more like a contradiction than anything. It's not privilege if it's fairly earned.
Four of the most famous people in history had teacher-student relationships: Socrates->Plato->Aristotle->Alexander the Great.
Famous colleges have no doubt done studies of which applicants are likely to donate the most money over the next 60 years. Have any of these studies ever been published? Anybody have any links?
My impression from looking at the names on college buildings is that big donors tend to be legacies, jocks, men, business-oriented, frat boys, and Republicans or centrist Jewish Democrats (e.g., Michael Bloomberg).
That was precisely the theme of Daniel Markovits' book "The meritocracy trap". Remember that "meritocracy" wasn't initially invented as a positive word, but to denounce privilege laundering in the first place.
America and much of the west does spend a lot more on poor and bad students than on rich and gifted ones.
Inner city schools in the US are the best funded by a long way, in the UK I got more resources for being mildly dyslexic than for being smart. There are vast resources spent on every kind of learning difficulty even when the financial return is dubious.
As with many (most?) things, we should ask, “What would John Mulaney have to say?”
“I paid you $100,000 for you to tell me to read Jane Eyre . . . and I didn’t.”
The question is incredibly naive, as you would expect from a child. But the frightening thing is how common this naivety is among affirmative action supporters. Grown adults not being able to understand that sending the worst students to the "best" school will stop it being the best school. As if the 'bestness' of Harvard is some kind of intrinsic, durable property.
But I think all of this is rooted in the left-wing education myth. That is, to the extent that academic ability varies between people (and this isn't just white supremacy etc.), it is almost entirely explained by differences in educational "quality", which is obviously false. And I say this because I don't think even they are foolish enough to think that dumb kids graduating from harvard still dumb would be treated the way current harvard grads are treated. But we know clearly from the data that sending black people to college doesn't close the intellectual gap between them and white students. If anything, it grows i.e. the gap between black and white college grads is higher than for black and white high school grads, and this despite the survivorship bias resulting from a much higher dropout rate for black students (assuming dumber students are more likely to drop out).
A good friend of mine teaches maths at a mediocre university in the UK. This university recently slipped in rank from an already middling position and they had to lower the intake grade requirements one notch to maintain student numbers. Their list of coursee does not look any different from last year or from a list at a top university and it cannot really look very different as they would lose their accreditation if they dropped e.g. differential equations from their maths degree. Even a two page course content sheet does not look very different from one at a top university, at least for a layman. However, as these courses are now taught for lower level students, they need to spend more time at the basics and barely touch the advanced stuff. They still include the advanced stuff in the exams, but they are now just teaching it to ensure reasonable exam pass rates and do not actually achieve reasonable understanding of the subject for any bar a few top students. My friend is now really disillusioned and most of his skills of teaching advanced stuff are now wasted and these students barely get to the level of maths knowledge of an average graduate from a top high school after a 3 years degree. So even if course subjects look the same as ones at a top university , deep down below course contents is very different and require a different skill set to teach, and the first optimistic answer stands.
> That is: Harvard accepts (let’s say) 75% smart/talented people, and 25% rich/powerful people.
If we're talking about actually admitting people because their family is donating money, this would seem to be a substantial overestimate. There's actually a term for people admitted because their family donated money, and it's "development case": https://en.wikipedia.org/wiki/Development_case
At least going by that page (I haven't looked into it much further), the actual fraction of these is no more than about 5%. This linked article says that (at Dartmouth at least) it used to be more like 1% back in the 90s and 00s, and rose to about 5% in the 2010s: https://web.archive.org/web/20190316233913/http://www.dartblog.com/data/2014/09/011686.php So it's *possible* it's larger now? But if so that would seem to be a recent phenomenon.
This does leave the possibility of larger-scale "privilege laundering" via legacy admissions, essays, and activities and whatnot, so it's hard to say exactly what's going on -- and, if such a thing is occurring, it's actually *deliberate* -- but in terms of actual development cases they're much smaller. (With Harvard it does seem like this more general privelege-laundering is likely deliberate, from what I've read. Seems harder to say about other schools, though.)
Interesting questions. I have misc thoughts.
On the less cynical side, advanced research has to happen somewhere and the people doing that want more people who can help (despite universities tilting further and further towards undergraduate education from research).
Also, instead of "best", you could say "most advanced". If we optimised for teaching we might put the most knowledgeable people teaching the most advanced students, and the best teachers teaching the students who most need to progress. What actually happens? A bit of both. Advanced universities often have more funding which naturally produces better teaching, but they don't prioritise "good teachers" as much as other things.
From the more cynical end of the spectrum, lots of institutions are "people with influence channelling people with ability". Sometimes the channelling is almost pure parasitism like modern academic journals which add almost nothing to the academic process other than a monopoly. Sometimes the channelling is the most valuable part (I don't know which, but I'm sure SOME big breakthroughs happened because someone had a vision and brought the right smart people together and wouldn't have happened otherwise). Universities used to be more patronage "richduke had money and chose to subsidise some random smart people". You can see that as generosity or monopoly depending whether you see the baseline as "everyone has enough to eat" or "everyone has as much as they can grab".
Like others in the comments here, I think it is as much about networking as signaling and education – but not networking in the shallow sense of “it’s all about who you know, not what you know”, but about meeting people who help you fulfill your potential.
My father’s favorite piece of advice for deciding where to go to college was that “you don’t need Lee Iacocca to teach you about supply and demand”. (Which usually led to quizzical looks, since very few college-bound Norwegians of this millennium know who Lee Iacocca was.)
The suggestion was that going to a top-tier university for your bachelor’s degree was a waste of money. And going for other degrees might also be a waste of money.
Since it influenced some of my own choices, I resented the naivety of that advice for years into my career – until I realized that I probably would have wasted an Ivy League education, and failed to grab the opportunities it afforded me, in much the same way that I wasted my actual education.
However, for those who are wicked smart and know how to take advantage, top-tier universities can offer you a peer group that can challenge you in a way very few others can, motivate you to reach higher that you thought was possible, and give you networking opportunities that open doors you might not even know existed and clear obstacles out of your way.
Genius and talent will often go stale and dull in a pool of mediocrity – unrecognized, unmotivated, confused for eccentricity or confusion, left to stagnate. But when pitted against other genius and talent, professors and peers who get it and challenge it and nurture it, it gets to sparkle and shine. You see it in music and literature and sports and science and business and culinary arts: The very best rarely appear alone.
And so, why do we send the top students to the top colleges? It’s a bit like asking why we send the best athletes to the best sports leagues. (“Send”, as if they don’t have a say in it themselves.) Because that’s where they can meet the people who can help them reach their full potential – not so much by teaching them, even though that is part of it, as by challenging them, supporting them, and putting them in the competition for the highest achievements.
And, when the best athletes and students congregate in one place, others dream of joining, and it becomes a self-perpetuating system for supporting excellence (which is strong enough to withstand some inevitable corruption and gaming of the system).
Aside from signaling, one of the functions of the university is acculturation. By admitting a mix of the brightest and the hereditary elite, the Ivies help to build a shared culture under their own influence.
(1) I am always highly suspicious of anecdotes involving cute moppets going "Papa, whither the course of educational attainment amongst the student body of our fair nation?" Maybe Matt's five year old is that precocious, but I do wonder about "what kind of discussion was Daddy having about such topics to set Kiddie up for that cute quote?" Much like all the women who go online about their six year old (and then some other woman goes on to cap it with 'well, my *four* year old did even more achingly woke thing') doing the Wakanda forever salute for Ruth Bader Ginsburg because of LGBT rights.
You thought I was joking about Rutkanda Forever?. Nope, I just got the details wrong:
https://www.dailydot.com/unclick/ruthkanda-forever-memes/
(2) "Harvard students and faculty who are descended from people held in slavery in the United States"
Good God, how many more euphemisms can we come up with? I had just got used to seeing "enslaved" and "enslavers" used instead of "slaves" and "slave-owners" (which had been replaced first by "slave-holders") and now we're up to "people held in slavery". I now fully expect to see the unironic use soon of "prisoners with jobs" because now "slave" is a slur? How much more can we dilute terms in order to keep the grievance hamster wheel turning?
I think in Germany in certain universities there is a bit of an opposite effect: some of the big technical universities accept all applicants, but then are extremely rigorous in their sorting. Their degrees are prestigious precisely because you are expected to have "fought your way through" an anonymous, underfunded system with very poor tutoring and nobody actually caring whether you show up or not.
In the final paragraph you say "I expect that without such a system the elites would do their own thing without any concession to merit whatsoever - so maybe it beats the alternative"
So then why does the elite bother at all? If they control access to their ranks, why do they make this concession to merit at all? Certainly in most human societies the elite has never ruled on the pretense of being smarter, so why do it in ours?
Furthermore, does the elite in the US really control access to their ranks? Since more wealth in our society is newly created than is inherited, I would say they do not have this power (look how Zuckerberg vs Winklevoss turned out). If you look at a list of the richest Americans, you have to go outside the top 10 before you see inherited fortunes. And even then, it's only 1 generation inherited (Koch, Walmart). So it makes more sense that the reason for elite universities is that the current elites get to network with the next generation of elites (Musk, Bezos, Zuckerberg, not the children of past elites) than that they just want the aura of "merit" around them.
"Is is true? An oft-cited paper, Dale and Krueger, appears to find that, controlling for applicant characteristics, people who attend more selective college don’t earn more money later in life. "
A non-sequitur if I ever saw one, given the preceding paragraphs. What does income have to do with finding the cure for cancer? Maybe it ought to; if someone found the forever cure for all the cancers, they'd probably deserve to be among the top earners in the world. But I don't see that as an unavoidable consequence, because scientists in particular seem to be in the game for their personal satisfaction and prestige among peers rather than the big money. At least much more than your typical investment banker or what have you.
> But this seems false; most of the classes at top colleges are the same material that gets taught everywhere else; you don’t get into subjects that need world experts until postgrad.
At least in math and physics, the point of elite undergrad education is absolutely not the undergrad classes themselves. The leading incoming freshmen have already taught themselves at least half the undergrad curriculum and are doing graduate-level work by the second year, or sometimes even the first. (I wasn't at that level, but my first quantum field theory had 19 year olds, my string theory class had a 17 year old, and my PhD advisor was in the equivalents of those classes at 15.) Students of this sort clear out the graduate-level classes quickly and then progress to research, so they absolutely benefit from being at top colleges. The research opportunities and the network of peers makes an enormous difference to their lives.
My impression from within academia is that many STEM professors recognize this and therefore want to have such students around, but the admissions officers (except at a few enlightened places) block them. That's because the typical admissions officer is a freshly graduated English major who only views math as an instrument of torture. To these officers, such kids are uncharismatic and alien, and worst of all, their naively enthusiastic essays lack any sense of irony or pathos. It is an easy rejection. All the oddities of college admissions are not centrally planned or rationally chosen, they emerge from the biases and preconceptions of this class of people.
> but what does the Princess get? I think she gets the right to say she went to Harvard, an honor which is known to go mostly to the meritorious.
She gets to know lots of presumably smart people to give oil contracts to? (and, like, the very extended interview sort of know, instead of just "ah yes they went to Harvard".)
There's a missing aspect here, at least as the stories always described it some years ago (I'm struggling to find a source). Reportedly, Princeton would admit about 1/5 of the class from rich-but-not-super-rich families that were smart, but clearly a step below the top standard.
Why? Well, in any class, some of the kids will have to be at the bottom; that's how a graded/ranked system works. Super-smart kids are smart and competitive enough to get into Princeton, so they would be unhappy to be at the bottom. The super-rich kids are too few in number and too entitled to form the bottom quintile easily/happily. But students who realize the others around them are smarter would feel lucky to have gotten in at all and have a baseline of gratitude to the school. They'd be perfectly happy to be at the bottom of a *Princeton* class, and after graduation, would form the nucleus of the alumni-jobs-and-donations network.
Being a legacy is probably a significant plus to be a bottom-fifth admit, for obvious reasons.
I don't know whether it's true, but Princeton is perennially #1 in alumni donations and engagement, so if it's true, then it works.
The concept “best school” is complicated.
I highly recommend the book “40 colleges that change lives”. There are quite a few colleges which don’t make traditional top 10 lists, which are more likely to provide the an excellent education to a student. As most of us have experienced, having access to a good teacher, makes all the difference, both in terms of how much we learn and how much we enjoy the subject.
When my children were in middle school and high school, I would occasionally get asked by another parent whether I would recommend that their child try to take the harder level of a math class or the less hard level. I always recommended that they try the harder level because our experience was that the “harder” classes tended to be easier and much more beneficial for the students because they would get matched with the better teachers .
One of my friends teaches at the Naval Academy in a technical area and she says that there is a huge difference in her enjoyment and ability to make significant progress when teaching the “best” versus the “worst” students. The category “worst students” includes students that are least prepared, some students that are least motivated, and some students that have the worst study habits. She can make substantial progress with those who want to work hard and learn, but the combination of challenged students in the lowest level classes makes her job very hard. So most teachers enjoy teaching the advanced classes much more.
There are also schools that have attracted some of the best known professors but then don’t have those professors teaching the classes because they’re carrying out research. So doing your homework to try to figure out where you get the best education can really pay off.
My definition of best education for college included some sense of social responsibility and some sense of life work balance. So we were put off by hearing from MIT students that they were overwhelmed by the amount of material and many of them were depressed. We were also put off when Harvard told us that their libraries were open 24 hours a day. My three children picked three amazing schools: McDaniel College, Williams College, and Olin College of Engineering, all of which had good reputations for teaching, and for a balance of education and other life values.
That said, if you wanted to send your child someplace where they would make connections with powerful people, Harvard might still be the best choice.
If the elites control the money supply, they can print money to reward themselves for whatever they deem good and worthy.
If the money supply is not politically controlled, and most wealth is ephemeral (networks, brands, and technologies, not land and physical resources), now the elites _have_ to compete on merit because they have no other means on which to compete.
If you just want a cynical explanatory rather than a justificatory answer, there's no need for most of the interesting speculation about signaling theory.
Top-level professional academics want to work with each other and with top-level students, and top-level students want the same. The people who run Harvard (and most universities) are a mix of idealists who want their university to do the best scholarship possible and cynics who want to attract as much alumni and grant money as possible; therefore, they try to recruit good scholars to make the virtuous scholastic synergism work, while slipping in some proportion of reasonably bright rich elites who can be counted on for automatic prestige and generous alumni donations.
The scholars get to intellectually inspire each other, the reasonably bright rich elites get to socialize with each other in final clubs, and the university gets paid in proportion to its success at making the scholars wealthy geniuses and the elites wealthy well-connected elites. Non-Harvard universities are generally trying to imitate this model as closely as possible, but they're less prestigious (self-fulfillingly) because they are less able to compete for the limited pool of brilliant faculty and brilliant and/or rich students.
Even in undergrad it's not the same material which is a huge part of the explanation that got dismissed here. When I was studying abroad at University of Queensland my girlfriend visited and sat in on a biochem lecture about the electron transport chain. She was amused because they did the standard vague "redox reactions happen and protons cross the mitochondrial membrane" explanation but the Amherst College biochem class went in on the chemical details of how various heme structures actually allow that to happen.
As a funnier example my physics professor told us a story about an acquaintance describing learning Ohm's Three Laws. Impressed, my professor asked what they were since he only knew the one. The three laws were: V=IR, I=V/R, and R=V/I.
More generally everything is like this, knowledge is fractal! You can learn about compact spaces as "closed and bounded" or as "space where every open cover has a finite subcover". Complex impedance helps you understand circuits but you don't strictly need it in an E&M class. And so on. Grouping students of similar ability together allows you to match talent to the appropriate depth and pace; if you randomly sorted students into colleges then for any given class ~40% of students would be overwhelmed and 40% would be bored out of their minds.
A subtlety: you don’t get the full signal just by being accepted to Harvard, you do generally need to actually graduate from Harvard, and Harvard has to put on all the trappings of actually being a highly elite education. In that sense it’s the same (useful, if overpriced) signal as every other college - a graduate has shown some minimum degree of seriousness and studiousness. Not saying there aren’t easy pathways or ways the very tippy top elites might pull the levers to get their most dullard progeny through, but I think that’s more rare than the preference at the admissions level.
Of course they most meritorious thing of all is to drop out of Harvard to go do something earth changing and extremely lucrative. You’re literally too smart for Harvard! But for that to work you have to actually be objectively talented, or at least have one really good idea and the ability to execute it.
I'm literally a math professor specializing in topology (though not "ten-dimensional hypertopology"). When I was at MIT as an undergrad I took the most advanced topology courses on offer in my sophomore and junior years, and jumped into research. Being around MIT professors was essential to my mathematical development, and so I strongly disagree with the second paragraph of this essay.
Great post!
Who's this "we" that's sending the best students to the best schools? Nobody "sent" me to an elite school. I chose to go there, and the school chose to allow me to go there. My parents would have contributed financially wherever I chose to go.
Why wouldn't I prefer a good school, as good a school as I could get?
Why wouldn't a good school prefer good students, as good as they could get? The better the students, the better for the professors, the research, and the quality of classroom discussion. (Apart from the handful like legacies and athletes who are admitted for money reasons. Heck, even the good students are admitted partly for money reasons; they're more likely to be rich donors down the road.)
It's microeconomics, not macroeconomics.
In case anyone hasn't already mentioned it, Bryan Caplan's "The Case Against Education" is very much worth reading if you're interested in learning more about the signaling theory of education. I'm almost finished with it, and I find it to be fairly thorough and comprehensive.
I think debates about college always get confusing if you lump together the "learning actual necessary knowledge for the future" functions (For say, mechanical engineers) with the signalling/approval/privilege functions (here is your communication major from Bad College X or Harvard). Yglesias' son (unironically) thinks school (where he is learning to read/write) is about learning necessary things for Life (like reading and writing), and so assumes college is like that.
"Best" college has the same issue. Best at what? The 5 year old assumes it means the best at teaching/improving student's knowledge, but we all know that is not what Harvard is best at. Step 1 in any difficult issue is to clarify what we're actually talking about!
How bad of a student can you be and still be able to bribe your way into an elite university?
Your analysis overlooked the role of competition. I am drawing on my experience as an Ivy League university professor. The most important reason for selecting for student rank is to increase the overall intellectual level of the student body. I am always impressed to watch the transition of students, who were top of the class in high school, and then find themselves in the middle of the pack in college. Ambitious students are also driven by an environment where the professors are at the top of the game in terms of research.
This is too cynical. Highly intelligent take pleasure in each other's company, and universities are the traditional place where they can congregate. The top faculty teach the brightest students because that is more agreeable job than teaching struggling students and they are not primarily motivated by charity.
The problem is that the economic returns to being perceived as intelligent have gotten too high, and this has lent illicit edge to the otherwise innocent and wholesome pleasure of spending time reading and studying with those that share your interests.
I'd give the same answer for why the strongest athletes have the heaviest weights. Because they can lift them.
The child (and Matt as well) are mistaking "top college" with "has the best teachers for basic material". The two are very different. This might be true for elementary school, but not really above that.
From elsewhere: yes the top colleges use a lot of the same textbooks. They cover a lot more of the text and do it a lot faster. If you need help you don't go to the lecturer to tell him to slow down (he won't), you get your own help (tutors generally). Or you drop the course, or drop out.
This is of course confounded in the States by the athletic "scholarships". This depends heavily on the specific sport, but e.g. the top colleges known for scholarship aren't generally getting lots of money from college football games.
From earlier: I don't know much about Eton etc. apart from that IIRC they are boarding schools, which puts a pretty high lower limit on how poor your family can be if you don't have a "full ride" scholarship. And "full ride" is likely to be a large amount, so they wouldn't hand out a lot of them. So yeah the students will tend to be from rich families.
No idea how many of the French elite schools are boarding (if any).
Why do the best students go to the best colleges? Because that's what makes them the best colleges. To a first approximation, good students make good schools; good schools don't make good students.
If next September, the entering class of Harvard was sent down to Bridgewater State and the entering class at Bridgewater State was sent up to Harvard, Bridgewater would immediately become more intellectually active, along with a lot of extra-curricular activity created by the go-getters who had been admitted to Harvard. Meanwhile, in Cambridge, there would be an immediate drop in energy and intellectual excitement. If you could somehow do the same thing for the next three classes, Bridgewater would now be considered one of the best colleges in the country and Harvard would be considered mediocre on the college level.
Something I haven't seen mentioned is that rank-matching is not just more effective, it's also more efficient. Anecdote: my school district was 404th in per-capita funding (out of 407 in the state). However they used that money to hire a small handful of really excellent teachers, and ruthlessly tracked their students. Year after year, they produced more National Merit finalists than any other school district in that state.
The reason for undergraduates go to one of the Ivy Leagues is to receive a credential, not an education. It’s a lifetime ticket to the Big Club that George Carlin described. The one you and I are not members of, and never will be, regardless of our capabilities or accomplishments.
One of my Development Econ. profs would always tell us that you'll never be rid of your elites, and so long as they control resources/capital your best bet is to set-up a structure where they're incentivized to act in a beneficial way. From that perspective, there seems to be a potential benefit in letting elites network with the best/brightest students of the next cohort, and of course providing access to capital for the best/brightest students is itself a benefit
It is true that some skills can be learned but cannot be taught. Other skills can be taught but only to those who are capable of learning them.
No amount of instruction will teach a dog to speak Latin; it's not a matter of forcing enough teaching units from enough really talented and motivated teachers and you can get Fido through first year Latin grammar, while Joe "Publius" Brainimax could teach himself as much and get to the same place.
I think the obvious answer (at a system level) I haven't seen mentioned yet is that it aligns the incentives for students in school. If doing really well in school gets you into a top college, then you are incentives to do really well. But if, as the tweet goes, your ticket to a top school is being a really poor student, then you have terribly aligned incentives.
A lot of stuff is downstream from this.
> Earnings are a poor proxy for “teaches better” - it would be great to have something like value-add to GRE scores - but AFAIK no study like that exists.
It seems like for highly ranked students they're already beyond what the SATs/GREs/etc are designed to measure, so it's hard to judge the value of teachers by those metrics. In math, you need a specialized test sequence like the AMC -> AIME -> USAMO to meaningfully distinguish between smart students, and I predict that _could_ pick up meaningful teacher effects.
Of course, the major error in the question "Why do we ...?" is that it assumes that there is a "we" that is deciding this. There are lots of independent actors in the situation and each has its own "incentives" (interests) and the result is some sort of game-theoretic equilibrium. And the sooner the kid learns that the world is not "We are all happily cooperating to advance a common good." the better.
One factor I haven't seen mentioned that I think is important is the use of college admissions as a signal in status competition among high-SES families. The obsession over getting one's child into an "elite" university seems to be much in excess of the educational, financial, and status benefits the child is going to get.
This gets interesting when the family's high SES comes from a high income from running a small business. Successfully running a business requires some intelligence, of course, but it depends a lot more on personality traits like diligence and ability to do well in win-lose negotiations. (Even in the largest companies, the CEOs' educational credentials are mediocre.) So on the average their children are not the highest-IQ despite that elite colleges officially select on things dependent on IQ.
Yep. A related issue is: Why is the high-status education done at the end of a student's time in education?
Clearly the teacher with the biggest leverage over anyone's educational success is a nursery (kindergarten) teacher. But they're not the ones getting the big bucks, or the guest columns in the NYT.
In regard to gifts and admissions, I'm reminded of Matt Levine:
"Analysts don't say to companies, "we will give you favorable research coverage if you will meet with our clients." (Sometimes companies do say that they won't meet with clients of banks who give unfavorable coverage, though.) And they don't say to investors, "we will bring you to a meeting with this company if you pay us a million dollars." An economy exists, but it is a gift economy, one in which it is crass to ascribe precise dollar values or to demand quid pro quos too
explicitly."
I think this idea really underscores that what elite colleges are selling (the merit) is the admissions process and not the education. The fact is for even the children of the rich/powerful-- the students who've been admitted on the basis of prestige and not academic strength-- the undergraduate classwork at every one of these schools just isn't very hard. Especially coming from a rigorous prep school (which leaves many such people with the impression high school was significantly more academically challenging than college!) and especially in our era of grade inflation where even a vague gesture at trying can receive an A-.
>most of the classes at top colleges are the same material that gets taught everywhere else
I see this claim repeated a lot, and I think it's just false. Most of what follows is anecdotal, but 1. I believe my experience is somewhat representative and 2. I think my position agrees with the simple, naïve understanding of the world.
I'm currently a grad student at a top university in a field easily confused with "hypertopology". My trajectory was (high school dual enrolled) small, not bad college -> (undergrad) good, non-elite state school -> (grad) elite school, and at each step, the difference was large and noticeable.
Subjective differences: At each step, everyone (students, professor) are more engaged, invested and knowledgeable. People were smarter, and it showed in the rigor of the classes, discussion and the questions.
Objective differences: (mostly about course content) This may not be only for math-y people, but I think it's my best argument:
Intro calculus: here are formulas, do it -> actual math, applying to situations -> teaches epsilon-delta to history majors.
Multivariate calc: literally no greens theorem -> grad/div/curl/greens theorem, etc -> differential forms (wtf).
Intro proofs: didn't even get to induction -> induction, equiv relations, basically everything you need -> everything on the left for the first half, then analysis for the second half
Analysis I & II: topology of R (limits, etc), sequences, infinite sums -> rigorous Riemann integration, existence/uniqueness of ODEs -> (the advanced one at least) the left plus Hilbert spaces/function spaces (wtf)
... (I could go on)
And these were just the undergrad classes! The grad classes (which undergrads typically have access to in their later years) showed similar difference.
The small college wasn't scraping the bottom of the barrel either--their SAT scores were around the 50th percentile. My general impression is that math graduates from my undergrad have a respectably solid foundation, those from my hometown were not at all prepared for higher math and those here (or the top students at least) are future professionals.
Moreover, I think this state of affairs just makes sense. Like, are you surprised that the smarty pants are over there learning more together? They can and do learn material 3 or 4 times as fast, what do you think they do with that extra time? more studying! Professors are going to pitch to the middle of their class, so a higher average skill will mean more and more rigorous material.
Why put all the smart people in one room together? Because they learn more that way! How else are you going to populate a class on measure theory and Hilbert spaces?
Well it's clear what we need to do then: regard Harvard graduates as contemptible losers. Leaning much too heavily on one or two personal experiences, this is what I've been doing for years. If more people were like me, the "signal" of having a Harvard degree would have negative value, and they'd be forced to drop the self-undermining privilege-laundering business.
On the mixing theory, what's the optimal ratio of merit to total? Now that Harvard HAS the prestige, why does it need to keep admitting Princesses of Whaterevstan at all? Does it all come down to who pays full tuition/has the highest expected value of becoming a future donor?
Those Matt Y blog commenters ;-). They know a lot about "what they've experienced" and they have some fluent ability to argue-amid-near-total-agreement, and to find stats - but their weird long foray into discussing evangelicals demonstrates that they have *no* clue how close of relatives evangelicals and Protestants more generally - are to the progressives who dominate Harvard.
I'm not really sure they are Harvard material.
The privilege smuggling makes more sense to me than hand-wavy "excellence" as a reason to want to go to one of these places. If they had just put that in their promotional material, I might have tried harder to go somewhere elite.
I earned an AB at Columbia in 1962. The degree was granted by Columbia College, one of at least three undergraduate operations at Columbia University. It is a truth generally acknowledged that if an organization operates under multiple aliases, that organization must perforce be criminal.
Nevertheless my education was honest and valuable in ways I didn't know. When you attend a selective university you very quickly learn. how good you really are. I was the product of a rigorous selective school system, but I soon met classmates who were smarter than me.
I had to learn coping mechanisms, for the English Department was a formidable place. I became something of a specialist in lesser known but reputable writers and had a knack for inserting them into my many, many papers. A life skill for a future academic.
Had I gone to a less selective school, I might have had an inflated view of my abilities and so suffered both disappointment and helplessness.
So, yes, the selective schools are crooked, at least in America. Secondary education is controlled by local school boards and supervised by state politicoes. The national exams are an embarrassment. Nevertheless talent will out and at least serve as a measuring rod for the bulk of the students accepted by less than honest means.
The mediocre legacy/sports graduate inheriting a business should have at least a sense for ability in his employees. So maybe that suffices.
All of this is correct and well-known. The missing follow-up question is: why, then, don't colleges compete ruthlessly with each other for top students, trying to displace the Ivies as top-rank signalers--and reap the rewards--by offering spectacular incentives to high-ranking students, the way, say, college athletic teams do (often in violation of NCAA rules)?
The answer is that they're all part of a massive cartel zealously guarding their collective position as granters of signaling credentials, and raking in obscene revenues in the process by charging cartel rents. If they began competing for students in an open market for signaling credentials instead of following the cartel's rules, they'd end up both massively reducing the cartel's overall income and revealing the market's true nature, thus opening it up to lots of outside competitors willing to undercut them. So instead they play by the cartel rules, which includes maintaining the fiction that they're providing education rather than signaling credentials, and using that pretense to justify their collusive control of the market and very limited, almost performative competitive practices that never threaten the cartel regime.
I tend to think this is correct, but that then raises the question: why are the Ivies so intent on maintaining race-based affirmative action policies in admission? Does that not undermine the meritocracy and provide little financial reward, at least in the short run, compared to legacy admits? Maybe the idea is you need to keep a certain number of minorities around to guard against the possibility of those minority groups forming alternative institutions? IE, maybe if Harvard doesn't consider racial backgrounds in admission, black enrollment declines fairly precipitously, black students start thinking of Harvard as kind of a white/Asian school, don't want or try to go there, and head off to HBCU's instead, and the result is that Howard University becomes known sort of as the Black Harvard, with the same status-conferring ability as Harvard, only limited to black students, but which nevertheless represents serious competition for Harvard & Co.
Any other reasons? I'm sure Freddie De Boer has written something about this, but I'm at work and don't have time to look for it.
Teaching is only part of school missions.
In addition, it serve as daycare for pre-adolescent (elementary school) and day prison for adolescents/ young adults (high school), which most organized societies seems afraid to let roam freely :-).
It also serve as a big sieve, it ranks students into bad/average/good. The western world equivalent of Chinese imperial exams.
And finally, it also act as indoctrination stage to try to keep future citizen in line with the official narrative, and avoid possible troubles ahead/promote societal stability.
Telling which is the most important mission is left to individual appreciation, but my personal belief is that (1) is certainly not top of the list. With possible exception of writing/reading and basic arithmetic, I believe (1) is only a distant fourth goal. The 3 others are far more important, but sorting them out is difficult and tend to change from time/places to time/places...
Using good schools (in the sense they have a good track record of student achievement) to teach the weakest students would maybe make sense for teaching.
But not for daycare/dayprison, not for indoctrination, and certainly not for sorting students.
> (Is is true? An oft-cited paper, Dale and Krueger, appears to find that, controlling for applicant characteristics, people who attend more selective college don’t earn more money later in life. Here’s a gesture at a challenge to these results, apparently supported by Dale and Krueger themselves, though I can’t find any more information. Earnings are a poor proxy for “teaches better” - it would be great to have something like value-add to GRE scores - but AFAIK no study like that exists.)
Bryan Caplan's "The Case Against Education" seems to have a lot of evidence that students learn very little of anything (at least on average) during most education, including college. This includes both specific facts and "critical thinking skills." (His conclusion, of course, is that the primary reason students make more money after getting a degree is signaling; he disagrees that controlling for pre-existing characteristics completely eliminates the college wage premium).
> what does the Princess get? I think she gets the right to say she went to Harvard, an honor which is known to go mostly to the meritorious.
A couple of other ideas:
1. She gets an excuse to relax and party for 4 years while pretending to do something productive, relatively out of the Whereverstan public eye
2. The royal family of Whereverstan might be sufficiently disconnected from what American rationalists know that they're still laboring under the impression that a Harvard poli sci degree or something similar would teach information that is useful for a future world leader
3. She gets to meet the kids of other world leaders in a different setting than normal
4. It gives them an air of legitimacy back in their home country (similar to what you said but with more specific motivation)
> This is the wrong question: the right question is why they ever give spots based on merit at all.
It might be interesting to look into the history of how this all came to be. As you pointed out in an old SSC post about college admissions, it used to be relatively easy to get in once you cleared a few (mostly class-related) hurdles. I have no idea what kind of prestige (or much) a Harvard degree got you in 1880--my vague impression is that most people didn't really care and it was just sort of expected for certain families, but I don't actually know.
Yale graduates in the comments: "Nope, nope, nope. You're wrong, Scott! What a SHOCKER."
I'm genuinely impressed that Scott just signal-boosted Matt Christman, one of the most prominent Marxist podcasters. There's hope yet. (Capital M Marxist, too, not a crypto-Marxist).
Another prosocial reason for elite schools is the fact that smart students simply do better work when surrounded by their intellectual peers. They can bounce ideas off each other, and since they tend to be competitive, they work harder to beat their peers. In fact, I would say that the actual teacher of any given classroom matters far, far less than the quality of its students. The latter is almost always the limiting factor. Top-notch researchers also benefit from having a smart undergraduate population nearby, and vice versa.
I went to a small, private, engineering-focused university. There, I was unamiguously one of the top two students in a class of 50-100 students in my degree program, and tended to be well ahead of course material with relatively little effort. However, when I went to grad school where there was a tighter selection criterion, the courses covered material at a pace that was far more challenging to me.
A few years later, I took a couple of MIT's open courseware practical mathematics courses aimed at undergrads. I was impressed with how quickly they expected students to grasp the fundamental ideas. It was a lot more similar in pace to grad school than to my undergrad.
When I see someone say:
> most of the classes at top colleges are the same material that gets taught everywhere else
I can only interpret that as either
1) Engineering undergrad is notably harder than whatever degrees they're thinking of, or
2) They're ignoring how high up on Bloom's Taxonomy a course gets.
I can't speak to 1. I have only engineering and mathematics degrees, and both seem to have this pattern. But 2 is worth a few words.
Let's take the fundamental mechanical/civil engineering physics course 'Static Mechanics', which has a required equivalent in most or all universities. A local private university has the following course description:
> Subjects cover includes; force and moment vectors, equivalent systems, trusses, frames, and machines, equilibrium of particles and rigid bodies, static friction, centroids and moments of inertia.
(Why do they use a semi-colon instead of a colon? Who knows! Hopefully they're more rigorous in their engineering than in their punctuation.)
And lets take the MIT course description:
> Introduction to statics and the mechanics of deformable solids. Emphasis on the three basic principles of equilibrium, geometric compatibility, and material behavior. Stress and its relation to force and moment; strain and its relation to displacement; linear elasticity with thermal expansion. Failure modes. Application to simple engineering structures such as rods, shafts, beams, and trusses. Application to biomechanics of natural materials and structures.
Those sound really similar. The only concrete difference I can take from those course descriptions is that the MIT one definitely includes deformable solids, and the local one doesn't necessarily. How is that compatible with one covering the material so much faster than the other (suiting their better incoming student class?) It's really simple; if your students can grasp the basic facts and mathematics faster (Bloom's Remember and Understand tiers), you can then actually spend time on how to apply these mathematical facts to problems.
After all, problems do not come to professional engineers neatly giftwrapped with a 'USE ES201 FORMULA ON ME, LIKE WE TALKED ABOUT IN CLASS LAST MONTH' label attached to them. Learning to understand an engineering model is hard work, but it's not sufficient to actually solve problems (outside of very special cases.) You need to be able to apply it, and to analyze how appropriate of a fit the model is to the problem in reality you're trying to solve. Less time spent on learning the model is more time spent on it's limits, when and how to use it, etc.
I guess, in engineering at least, my model isn't any of Scott's. It's not that only the best teachers can teach these incredibly arcane topics. Or that we get the best payoff from matching best teachers and best students. It's that the best universities are those that attract the best students, and because they have the best students they can teach them more and better than they could if they had to slow down for worse students. And because they have more money (due to turning out better engineers who donate more to them, etc.) they can pay better teachers as well; but this is a small part of the improvement. The good teachers at a middling school could do as well as an average teacher at MIT, given the same students and a few years to tune their curriculum to those students.
How far this model extends to non-engineering is a real question; it's worth noting that MIT does not do legacy admissions, and so are not engaged in the proposed privilege laundering. I don't have the experience required to compare, say, HIST 30094: Modern India and Pakistan at a middling university and a top one.
The privilege-laundering idea seems kind of far-fetched. I suppose Americans do care enough about 'deserving' to inherit your father's business that it might seem plausible, but isn't the simple theory that colleges need money enough to explain why they continue to accept donations/rich kids?
Would be interesting to see a bit of a comparison to other countries where this happens far less/not at all.
For instance in the UK, Oxford and Cambridge entry is almost entirely based on exams and aptitude tests and so they raise far less in donations, partly because it’s basically impossible to buy your way in.
Would say this enables courses that are materially different/more challenging than other universities. Does in theory lead to a UK elite that is more academically inclined, but for lots of things outside of academia being a hard working smart nerd isn’t necessarily the most useful skillset to have in charge.
>most of the classes at top colleges are the same material that gets taught everywhere else
I disagree with this, particularly for STEM. The amount of information that faculty are able to productively spew at MIT undergrads is vastly higher than it would be at Midwit State. As a concrete example, Caltech uses Jackson's _Classical Electrodynamics_ as its undergrad E&M text. That's a book explicitly aimed at grad students. Only at Caltech/MIT/Harvard could it be used as the starting point.
Basic ability matters. When you select for it, you really can do more.
Among elite-college true believers, it's typically believed that the advantage of [elite school X, let's say Harvard] is not in the institution, general curriculum, and staff, but in the (1) classmates, (2) tails of the curriculum, and (3) other-than-classes academic opportunities. So by the steelmanned perspective of the pro-Harvard side, "most of the classes at top colleges are the same material that gets taught everywhere else" and "[elite colleges have] the world’s best writing teacher[s]" are missing the active ingredient of "best schools".
Example of 1: You can teach the students the same subject, but if you've selected a population of more-capable students, they'll have a higher rate of starting world-changing projects when they leave (or by dropping out). So the active ingredient (say the pro-elite-colleges partisans) is matching best students with best students.
Example of 2: "you don’t get into subjects that need world experts until postgrad" is true in the median case, but not true at the 90%tile of eg Harvard students studying technical subjects.
Example of 3: "access to research opportunities w/ faculty (selected for something like their research output)", covered in other comments.
The privilege laundering take is overly reductionist. It assumes that wealth and social capital have no relevant value and so it needs to be laundered with the smart/talented set to hide its inadequacy. But in reality, wealth and social capital are independent factors for the attainment of career prestige and so are valuable as predictors of future success (measured as high prestige leadership roles in society). Harvard is first and foremost an institution to signal prestige, and so their goal is to find the student population that will most likely end up in high prestige roles in society. There is no laundering going on, they are simply maximizing the relevance of a Harvard education in signaling prestige. The key point is to recognize that admission to Harvard isn't a reward for being the top-whatever percent in academic merit, its about what makeup will grow Harvard's prestige brand. This also has explanatory power for why the Asian admission rate is artificially suppressed. Asians tend to not have as many paths to attaining prestige positions in society, and so the value to Harvard's future prestige from admitting an Asian student is lower than a white student of lesser academic merit.
Another somewhat optimistic answer would go along the lines we want to maximize the chance of a genius with a clever idea actually pursuing it successfully and changing the world by researching a new tech or making a startup that uses new tech or something. This is much more likely if instead of having a bunch of average friends who say that's a cool idea and go about their day, they have a bunch of friends who are very competent or ambitious or rich who hear their idea, say let's make it happen, and help them either as hyper competent early employees or by giving them early funding either of which helps get the idea of the ground and gain momentum. Not sure how much I endorse this as accurate but it's more than zero and probably more than either of the optimistic takes you present.
In addition to providing "merit washing" by association for the children of the hereditary "nobility", the "commoners" admitted to Harvard and other "Ivies" graduate to staff the administrative (i.e. reasonably competent, well-compensated servants of the nobility) backbone of the Anglo empire. This is absolutely required on account of the well-known "regression to the mean" effect -- the nobles no longer have a dozen+ children each and cannot produce competent successors for themselves through purely biological means. So they outsource the "brains", while passing on their estates to biological heirs (often, as discussed, merit-washed for popular legitimacy) who will command the "brains", rewarding the latter with mid-6-figure wages (and sometimes a "lottery ticket" chance of a small share in capital ownership, via e.g. the startup circus.)
"Brain" isn't the primary selection criterion for that servant caste, however: the selected people must first and foremost be unshakeably loyal to their hereditary aristo masters. And so the primary criterion remains political reliability (as proxied by e.g. "affirmative action", aka "Кто был ничем, тот станет всем" ("who was nothing, will become everything", as seen in Ru version of "L'Internationale"), the correct "extracurriculars" (elite bonding rituals: rowing, excursions for imperialist "aid" to the Third World, etc.)
One can compare with the old USSR: as soon as a hereditary elite caste (re)appeared there, that country started behaving like a cheap clone of the Anglo empire, in a number of comically-obvious ways. It turned its Moscow State University into something quite like a copy of Harvard (in this particular respect) -- complete with "affirmative action", "easy majors" for aristo children, guaranteed employment for graduates in the mandarinate, and so on.
Every sclerotic bureaucratic empire must have "a Harvard".
"why do we send the top students to the best colleges? Why not send the weakest students to the best colleges, since they need the most help?"
Because educational resources are wasted on the weakest students.
It makes sense to give the most resources to those who can take advantage of them. Compare to athletics. You want world class facilities, doctors, trainers, etc. to be available to Olympic level athletes, so that they can go win medals for their countries, not to overweight people using the electric scooters at Walmart. The latter group might benefit from some of the resources, but their goal is so crude (lose weight) that they can do it with very basic resources (a street to walk on and some sneakers.) Likewise, the weakest students just need to be able to do the most mundane of tasks. No fancy anything needed to teach that. If you want new genius ideas/products/whatever to be created, you want to match the brightest students with the brightest people in their fields of study and the best equipment, facilities, etc.
Is Matt Yglesia's son volunteering to attend kindergarten in East St. Louis?
Sounds like Rational Altruism being put in practice.
How about like admission only high schools, the top schools write their own admissions test and fill the class only based on test score results? The test can cover what might be on an SAT, but also some subject material, but merit admissions would be the result. Yep, the endowments would suffer, but probably not all that much over time as networking and signaling would come in to play.
I think you're wrong to assume that this is designed in rather than naturally arrising.
In any situation where you're trying to find a matching based on mutual acceptance - i.e. where students can choose which universities to apply to and accept offers from, and Universities can choose which students to make offers to and which to reject - and where most of the students have similar rankings of how desirable they find different universities and most of the universities have similar rankings of how desirable they find different students, you're going to end up with the two rankings roughly matching - the most-desired students will end up at the most-desired universities, and so on down.
Most universities would rather have more able than less able students. So, whatever we mean by "best colleges", the most able students are likely to end up there.
'Privilege laundering' sounds a bit paranoid- as though the privileged are actually ashamed of themselves and are trying to pass as meritorious because... the privileged are actually shape-shifting lizards from Planet X! Wake up sheeple!
It may have been true that the sons of robber-barons wished to mingle with clerics and thus to come across as cultured and pious. Universities, after all, were originally ecclesiastical in origin. Similarly, the sons of legal shysters or crooked bureaucrats may have wanted to rub shoulders with the descendants of robber-barons under the pretense of devotion to Humanistic scholarship. Napoleonic France went in a different, supposedly more meritocratic direction. However, it was the intense competition between the Universities of German States which created the Seminar System and the notion of the Doctorate as a Research Degree. In the STEM subjects, this incremental approach meant that there were external economies of scale for Marshallian industrial districts more particularly where Finance partnered with Industry to create a system of cross-holdings.
America's extraordinary rise to Academic pre-eminence was based firstly on expansion of a more academic type of High School to virtually the entire population in the period from 1914 to 1938. In Europe, the assumption was that only a small percentage of scholars would attend 'Grammar Schools' or 'Gymnasiums' as opposed to Technical Schools of some type. Thus Americans were more ready to take advantage of the GI Bill, than any European country.
The second factor was the reputational benefit of creating a University, or endowing an Institute- e.g. that of Advanced Study at Princeton- and the networking benefits of alumni associations. The Land-Grant Universities could emulate this type of success, though, no doubt, some had already attained excellence in some fields- e.g. agronomy.
The third factor was the Cold War and 'Think Tanks' like RAND. Universities were competing for political influence. Chicago was at war with Harvard. This Saltwater vs. Freshwater battle defined the ideological space when I was growing up. Then, in the same manner that 'Independent Expenditure' campaigns disintermediated Political Parties- or, indeed, ideological Punditry- and gave salience to wedge-issues on the one hand, and straightforward, utterly blatant, special interest rent seeking on the other- so too did the run-away success of some, highly technical, research programs which, so to speak, 'unbundled' the Liberal 'idea of the University'.
It is this 'unbundling'- or lack of cross-elasticity, or synergy- which has made the 'matching problem' acute. On the one hand, you have dead-in-the-water research programs which have to fools some seemingly smart kids into signing up for them so as to keep their Ponzi scheme trundling on, while, on the other hand, the opportunity cost for smart, well connected, kids of investing in 4 year College has been rising. Essentially, there is a 'laundering' of worthless nonsense which however does not appear to be any such thing because smart kids of diverse backgrounds are signing up for it and- hoping to get tenure- being very vocal in their support for it, while, on the other hand, there are more and more veils being draped over the truth- viz. kids nowadays need to be like medieval scholars who moved from one campus to another. Indeed, this was a feature of German university life in the late eighteenth and early nineteenth century. That type of mobility is what put Germany ahead by the 1880s which is when John Adams spent a little time at a Gymnasium. However, he came to the conclusion that the future would lie with the Engineer-Savant- guys like Pareto and Barone.
What is the actual 'matching problem' currently faced by students and Colleges? Firstly, the set of students is only well-ordered for particular, quite narrowly defined, research programs- which however are very lucrative and probably vital to our national security. There has to be an unbundling event even if it means killing the sacred cow of tenure. Otherwise, as Vivek Ramaswami seems to be saying, the Chinese will eat our lunch.
A separate issue, the one this post and the comments on it focus on, has to do with coalition formation and stability. Essentially, the claim is that alumni of certain Colleges are dictatorial over the core of particular games in particular industries or particular contexts. However this should only concern us if there is something extraordinary about those recruited by the Colleges in question or, less plausibly, that the Colleges impart extraordinary traits to its students. But, in that case, the relevant 'Structural Causal Model' would enable us to improve outcomes for everybody. That's the opposite of paranoia or the politics of envy or spite.
Still, one might say that there is something egregiously wrong with our Social System if Harvard alumni write the jokes in our favorite TV programs. But, surely, that was a joke created by TV joke-writers to raise their own prestige while acknowledging that White America was white-collar and College educated.
Why do 'elites' want to let their kids play tennis with kids who are really good at tennis rather than the paralytic scion of inbred Aristocrats who can't lift a racquet? The answer is that their kids may benefit by playing more and better tennis. Sports, unlike almost all non-STEM instruction in H.E, is actually good for kids. Universities with good sport facilities- or one's located in big Cities where such facilities are easily accessible to the very rich- are a good place to gain a bit of polish and perhaps- like the Prince of Wales- pick up an acceptable spouse of good breeding stock. Assortative mating at Uni beats caste-based inbreeding. Or so it now appears. Once genetic engineering really takes off, the elite may indeed elect to become a different species. Till then, all we can do is pretend they are shape-shifting lizards who are using Harvard and Yale to camoflauge themselves.
Yud already wrote this article in < 280 chars
https://twitter.com/ESYudkowsky/status/1677885529491386368
The premise of the question is incorrect: we do not send the top students to the top universities. Instead, we provide equal opportunities to all students, and the schools select whom they accept. What is the motivation for Harvard to accept someone with less-than-stellar grades? There are financial, influential, and prestige reasons as the author describes. But beyond that, the risk of harm to Harvard's reputation is too great. And yes, there's risk: if you meet 6 Harvard graduates in a row and they're each individually morons, Harvard's reputation takes a hit among you and those whom you influence. University quality is based entirely on reputation.
I say we fix that. Here's a plan that's a step too far, so we'll moderate it afterwards but what if: State legislatures require universities to post all graded student work online, as a condition of maintaining their business licenses. Suddenly a grand corpus of college student work is built. After a few years, artificial intelligence agents (and natural intelligence agents) can assess those bodies of student works to determine the actual quality of a degree program at one institution compared to similar programs at other institutions. Prospective students can already determine the cost of education and they can now determine the quality of education being offered, so they are able to make value decisions. This drives cost down and quality up, if the number of seats at universities is collectively more than the number of customers. (It is.)
This proposal, of course, also allows each student's demonstrated knowledge to be evaluated. The up side is that employers (supported by agents that can evaluate the job candidate's artifacts in whole) can weight what they consider to be important. Need a business major with international finance expertise and who speaks Chinese? Need a scientist who understands issues in a global context? Done and done. The knock-on effect is that universities finally learn and become forced to react to what industry actually values. If no engineer from Fictional State University can get a job because their engineers know everything about DEI but nothing about engineering, that sends a message!
But this proposal also means anyone can read the paper that the professor required you to write about how great Slavery is. You had to write it or you would fail the class, but you never believed what you wrote and now you're running for office. So there has to be a way to shield the data from those who should not be viewing it. Fortunately, this is easy and mature. Data is available anonymized and in aggregate, and then students receive keys so they can grant individuals / organizations access to their work for time-limited periods up to one month (for example).
What do you think?
I read the title and thought "incentives". Students mostly prefer to have good teachers, and teachers, good students. It's really hard to create a performance test that someone can't fail on purpose. If you reward people for failing, the test ceases to be meaningful. Better to have a system where people mostly WANT good grades.
My naive understanding is there is considerable value in the connections made at elite institutations compared to connections made at eastern state university. Connecting the smartest with those of means and influence seems reasonable if goal is to create high value connections and catalyze higher value results.
George W Bush. BA from Yale, MBA from Harvard. Earned “Gentlemen’s C’s” throughout his academic career.
Commencement Address at Yale:
“And to you C students, I want to tell you that you that you too can be President.”
I guess the whole Skull and Bones thing might help a bit too.
Have you considered that this analysis fails to take into account DEI? For example, if Harvard was 75% merit and 25% nepo babies, that would still be a decent ratio. But in the modern day, I think the ratio for Harvard would be more like 50% merit, 25% nepo babies, and 25% diversity admissions. In that situation, it's much harder to launder privilege because nowadays, hiring a Harvard graduate gives you a 50% chance of them being a moron. I'm not exaggerating here: in fact, the situation is so bad that if I owned a company then when it came to hiring I would not view Harvard as a prestigious school at all and would be weigh a Harvard degree roughly equally to a degree from a state school. The only Ivy League degree I would strongly value would be MIT.
Wow, education is one of those topics where everyone is an expert because they went to school and yet we can't just blindly trust the experts as the system we have doesn't match the values of the common person who wants to improve their life. In this way experts are goal aligned and incentive aligned to come up with the system we currently have as a clear tautology. Despite the verbal only and public protestations otherwise, the hoarding and laundering of privilege with some admixture capture of the best commoners is the standard playbook for a successful aristocracy.
The TL;DR - is in line with Scott's final point about why there is a meritocracy at all. I don't think Scott has gone broad enough with his thinking. It isn't a simple matter of reputation or prestige or validity which these commoners lend to the elite/aristocracy. It is an entirely different process where patrons are present to meet promising commoners whom they can extract value and benefits from.
Would we be confused if we saw lesser and fallen nobility along with the top merchant families sending their best and brightest to a fancy meeting ground with the children of the higher aristocrats and nobility? Change a few words and the truth is clear. They don't want to extract privileged status from commoners, they already have that by birth. They want to extract connections, value, and to both cultivate and harvest the best from the herd of commoners.
I.
I find much of the confusion, hand wringing, and democratic discussion goes away if you just replace the word elite with aristocrat and learn a bit of history.
The renaissance model was very very successful for the aristocrats in Europe. They had better living conditions, more technology, an age of colonialism fuelled by improved weapons, and were all around better off and all they had to do was expand the idea of patronage beyond the arts into the areas of learning about the natural world. This attitude of harvesting the minds of the commoners and lower nobility and creating positions for them to study the world to see what they can figure out...this was super super super successful.
Our education system is simply an extension of this model to sort the wheat from the chaff. It is a mining operation by the aristocrats to get all the commoners to sort themselves out in competition. Meritocracy is 100% fully at odds with the core nature of aristocracy and they never ever intended to apply those ideas to themselves. They don't see themselves as the same as the commoners, they really don't and the vast majority of commoners have never met nor been privy to the inner thoughts of the aristocrats.
II.
We are a breeding stock and are like cattle they own. If you had an intelligent flock or herd of animals who might rise up and kill you if you told them the truth or let them learn or know the truth on their own, then you'd have a very unstable system with a lot of dead elite.
The Russian and French revolutions were a wake up call to expand mass indoctrination beyond simple religious compliance of commoners to their 'betters'. A new intellectual edifice would need to be built to contain commoners within a stable framework. I think some aristocrats worked on this problem in a very intentional way, but a lot of this is simply an alignment of incentives. If driving down one road clearly takes you off a cliff and the other to a bridge to cross the gap, then it isn't a particularly difficult choice.
III.
The great incubator of trial and error took hold and many many many diverse attempts, approaches, and styles of commoner management (one of the core tasks of the aristocracy) emerged in this new age of educated commoners who could think and read. The simple printing press which emerged long before commoners could read led to several revolutions around the world. Later on the radio was used by people in the Caribbean to organise revolts and resistance to colonial rule.
But the aristocracy finds a way and from more totalitarian approaches with daily government radio broadcast to owning the airwaves as a matter of governmental proxy of who is allowed to speak all the way to the Twitter Files style active suppression and censorship of speech.
This is nothing new and is a process within the core problem of elites wanting to harvest commoner minds in an extraction operation and the problems that causes. By and large this problem has been solved through a mixture of violence like the selected murders/assassinations of key leaders within commoner movements which challenge aristocratic authority or goals.
IV.
And the other part is a huge mind control operation of values and the stage production of democracy. What makes it so very convincing is the fact that the stage production is somewhat real with a small percentage of choices allowed to the commoners through a process - especially on issues which DO NOT affect the bottom line or power of the aristocrats who call themselves elites now.
The successful formula has proven to match nature, the carrot and the stick. Rewards and consequences. A stable system with some elements of true meritocracy in a bloody, unfair, and arbitrary process which still ends up costing a lot of resources. Combined with a careful sculpting of 'values'. You can read about this from one of the founders of the modern education system in Rockefeller who wanted a compliant and cooperative sea of minds which could be shaped and directed to the purpose of enriching him and empowering him further. He was very very open about it with all those initial grants and bribes to get modern mass education started. He wanted smarter workers, but not too smart.
But hey...a lot of people who went to those schools and picked up 12 plus years of someone else's values and worldview being drummed into their heads are all experts on what happened to them and why it was good and everyone should join the cult. And yet even with all the benefits of integration you find the truly wild peoples of the world are very very very very very resistant and reject schooling due to those embedded values.
The focus in the first few paragraphs on the classroom teaching skills of the professors is misplaced. Professors at top universities are famously rewarded for research much more than teaching, and there are few courses even at the graduate level that you need a special expert to teach. But when you have a class full of talented students you can go faster and they give each other better insights. This is not to say that the system we actually have isn't a hodgepodge driven by all the forces in the rest of the post.
Incidentally, The Chosen by Jerome Karabel is indispensable reading on this.
If hanlon’s razor is never attribute to malice blah blah, what would we call a razor that says look first to emergence based on game theory to explain real life situations? (So I think the medium cynical is probably the center of gravity, with this proviso: you’d expect the universities to be exactly as merit based as necessary to maintain their rank while gifting the rest of the slots to those likely to give them donations. The college rankings based on selectivity probably also help maintain the equilibrium; otherwise there’d be a much bigger incentive to give away greater and greater percentages of spots while you traded on your name for a while.) maybe assuming it’s that and not some kind of top-down design is also hanlon’s razor?
By my reading of the link, Dale and Krueger don’t support the challenge, they support the author of that article’s interpretation (i.e. that character matters)
While I'm less sure about Harvard my experience with Caltech suggests to me thar:
1) Some students are more capable than others and society benefits if those more capable students are pushed to the limits of their ability.
2) The best way to push students to the limit of their ability is to put them in classes with similarly able students. It's human nature to slack off or grow uninterested if bored and to rise to the challenge if not.
3) Whatever place you concentrate the best students at will become highly desierable to attend so selectivity is necessary to preserve that feature.
4) Wealth and the other trappings follow because producing the most successful graduates draws alumni money.
--
Yes you often get the best professors at the best schools but, while that makes form a nice research opportunity they are usually not the best instructors so really everyone wins.
I'm sorry, but the expression "Maharaja of Whereverstan" so typically blends American Exceptionalism with White privilege. Not seeing it and/or not understanding why, reeks even more of it. Whoever coined that expression should be first in the line to get their privilege laundered. Other than that the play was great.
This is likely true. I think there are big nuances in the distribution of student quality and of teacher skill-sets though. You touch on the latter; most professors can teach the basics, but few can teach truly advanced skills. There's no point wasting the latter on students who can't learn what they can teach.
Assuming an equal ability to teach things (all teachers teach the same stuff, but some are better at it),
Matt's kid isn't entirely wrong. It *would* work to send the worst students to the best teachers. They'd be better off for the experience. But it would be ridiculously inefficient. The distribution of a class made up entirely of 'worst' students might range between, say, the 10th and 30th percentiles of ability. Teachers teach to the average, at best, and probably more likely to the minimum fixable level (maybe the 15th percentile in this example). So progress is extremely limited, even for the best of the worst. And the difference between 30th percentile ability and 10th is much bigger than it sounds (it's not linear in the tails).
Elite institutions work the same way. It's about teacher quality and class availability, yes, but it's also about the degree to which your classes and teachers are able to be the best versions of themselves. An elite school has elite students; the lowest fixable minimum is irrelevant, because everyone's smart. Teaching to the average still means teaching to a very high level, and because the left tail is truncated there isn't so much of a gap between average and worst. Plus, the best kids in a place like that have ample opportunities to excel; they won't be held back much. Calc at Harvard (or MIT) is not calc at Western Prairie State.
Your theory about laundering privilege still works. Harvard's incentives are to keep quality high to facilitate those dynamics by admitting mostly the best. And it's not like complete idiots get into Harvard anyway, no matter how big the donation. The left tail is still truncated; it's just a bit lower and longer than it would be otherwise.
Thank you l, this is refreshingly in the middle of the argument. On a slightly different note,I've been thinking about and consequently noticing how 'virtue-signaling' is a deception in which people engage no matter the context.
I went to a specialized high school in NYC. We had a 99% college acceptance rate. That was 1000% due to the smartest students testing in, and 0% due to enhanced teaching capabilities. The administration was known to be a joke (teachers got together and wrote a public letter saying the principal was both incompetent and placed by nepotism). If you just put the worst students at the school, scores would plummet immediately. Frankly attending wasn't worth it for me, would have been better to go to my local school.
You focus on rich people laundering their prestige by making themselves look talented/smart, but I think it goes a bit further.
Premier schools also launder the privilege of *being* talented/smart, to make those people look competent/hardworking/trustworthy/etc.
This is in service of the political idea of meritocracy, that being smart/talented is sufficient justification for some people to have control over the lives of others and over the direction of the economy/country, and for them to deserve richer and easier lives with less hardship and more pleasure.
If you know a bunch of precocious 'gifted' children who are smart/talented in some ways but also barely able to take care of themselves and often venal or self-centered, you might question the wisdom or justice of this situation.
Even if all the gifted people you know are saints, you may worry about value misalignment or lack of domain knowledge between them and the masses, or simply be unconvinced that their competence at running the system justifies an unequal system with coercive power.
But a Harvard degree transforms the privilege of smart/talented into an aura of 'competent to lead' and 'part of the natural ruling class'.
I was a Harvard undergraduate in physics in the early-mid 1970s, after attending public schools on the south shore of Long Island (Nassau County, not the Hamptons). The greatest benefit I got from meeting elites was being able to see with my own eyes that coming from money didn't have any correlation with being a decent person. OTOH, the people I knew who had come to Harvard from poorer backgrounds than mine tended to be pretty nice, and often probably a lot nicer than I was. Also, I met a lot of jerks, of various backgrounds.
Obviously this is anecdotal. But thereafter I wasn't ever impressed by people with money just because they had money. Nor did I ever feel that making a lot of money, even from one's wits rather than an inheritance, could excuse someone for being a jerk. This was very liberating. I'm not sure I'd have gotten the same feeling from going to another Ivy, since I might have fantasized that people from Harvard were exceptional in some way.
In the roughly 50 years since, I haven't gotten any contracts in Whereverstan. The only work I ever got from a college connection was from a friend who kindly asked me to do legal work for his venture-backed start-up, at a time when I really needed some work. BTW, he grew up with a single Latin American immigrant parent, and cleaned toilets for a couple of years as his campus job. The start-up failed within three or four years, as most of them do.
My experience with Harvard faculty was mixed. They ran the gamut from inspiring to mediocre to jerks, with most in the middle -- though to be fair, I was too self-involved at that age to know how to build relationships with any of them. There are a couple I really regret not paying more attention to at the time (mathematician Raoul Bott at the top of that list: he was a delightful person, I discovered after I'd cut his class for most of the semester).
The single most life-changing aspect for me was managing a student-run orchestra: so, yes, the peer argument about bright students liking to flock together is pretty cogent. And there were so many ideas floating around in my classes, conversations, in the libraries and bookstores in the area that I've spent much of the past half-century exploring them and seeing them branch into other things. I'd never say that my college years were my best years, but Harvard is why I have a rich intellectual life today.
Which is why I think posts like the one above are simply low-dimensional cartoons, created by people who didn't go to Harvard and who could never imagine that it might be just as complicated as much of the rest of reality.
With regards to the initial point, I went to a top-ranked STEM school and taught/was a grad student at 2 other much lower ranked schools and can confirm that what the two schools taught was completely different. Students at the lower ranked school learned far less theory and weren't exposed to a lot of programming techniques or algorithms that we learned in the first 2 years at the top-ranked school. The grad-level algorithms course I took at one school basically entirely covered material that students at the top-ranked school were required to learn as an undergrad.
Poking my head in to explore, not a hypothetical situation. Relatives of mine (identical twin girls) recently accepted admission to Harvard and a New England state uni. Situationally on the “commoner” strata, if not “poor”(mother’s chronic disabling physical illness, parents divorced, father financially somewhat stable). White female New Englanders. Nearly identical academic achievements, including athletically (although not recruited athletically) at different public high schools. Test well. Constant stream of accomplishments related to very high ability mathematically, also verbally. Adversity also includes chronic illness of their own that requires careful management. Recently witnessed a drive by shooting outside the store where they work part-time. Between the two, had admissions with full scholarships at UNC, Syracuse, etc. One applied to Ivies, one did not. The one who has decided to attend Harvard was reluctant to accept. Why? Partly because her high school friend who resembles her closely, in ability, demographics, struggled greatly with social acceptance there for her freshman year last year. Aside from the accompanying “you don’t belong there” from her sister who did not apply. These two almost instinctively grasp that attending Harvard for them has it’s complications, not being wealthy. Their mom and dad grew up and/or attended college with others of great wealth; their experience was one of sometimes being more beholden than they would have liked, with no obvious long term benefits, because they both chose not to put themselves in service to those with the wealth. One of the twin daughters put her trust in her abilities that would make attending Harvard worthwhile. The other, the far spunkier of the two, did not. Their mom has been supportive of each as they made their choices. It has been interesting so far, to say the least, to read the recent SCOTUS decision in light of their recent experiences.
Except the proportion of rich/powerful isn't 25% - it is at least 36% (legacy) and certainly much higher (donations, "pull").
So really all it is, at this point, is a Veblen good - something rich and powerful do in order to show they are rich and/or powerful.
This is a question that sounds adorable, because it comes from a five-year-old. The same is not true of intelligent adults asking the same question.
My wife did a rotation in admissions, and participated in undergrad admissions decisions, while working as an assistant director at an Ivy. Her career: university fundraising.
(1) “Why do we send” is a false premise, “We” have no agency on admissions. And the only people sending anyone to a university is a situation where a family or equivalent is helping someone pay for & be successful at college. Colleges and universities are registered business entities. “We” refers to society, American society is not granted authority over university/college managers to make decisions about what any given universities and colleges do.
(2). A legit really effective teacher working as a professor is a somewhat rare individual, and highly valued by students, colleagues, the parents of students, alumni, and highly valued and sought after by other colleges and universities. People are free to make their own choices about where to apply to work, and teaching professors or research faculty with really great skills often have a number of nice options available.
In most human endeavors that people are passionate about and deeply invested in, A-players want to work with/collaborate with/mentor other A-players. Less so, mentoring B or C-players. An A-player teaching professor deciding to *focus on* helping C-player students is a pretty rare, laudable situation precisely because they chose something noble and against the grain. The professor who spends her time helping highly intelligent, motivated undergrads or grad students unlock challenging knowledge is no less laudable, and should be equally free of coercion by “We.”
The question implicitly presupposes that the quality of education might be better at a 'better' school. This is not true. Professors at highly ranked schools do not teach better than those at low-ranked schools. There is not some magic pedagogical method that is known only to elite schools. In fact, competition for faculty positions is so fierce that there are many excellent people at quite low-ranked schools. So, there is no reason to want mediocre students to go to a highly-ranked school - they would not benefit at all anyway.
What students benefit from is other students. Scott mentions cynical reasons this is good - hobnobbing with other elites, but there are more virtuous reasons.
One is that tracking is good for students. If all the good students are together they can learn at an accelerated pace and the mediocre students can learn at a slower pace. In the end, both groups are receiving instruction at the optimal level. This is a stable situation. If you try to introduce mediocre students to a good school, they will do poorly relative to other students. I know of students who were asked to withdraw from a good school to go to one more suitable for them, and in the end they were actually thankful for the advice. They learned more at the slower pace and went from being bottom-of the class losers to solid performers, which boosted their confidence.
If you suddenly changed the situation so that the worst schools were where the best students went, in a few years all that would happen is to reverse peoples sentiments about which schools are best/worst.
Young Yglesias's question is interesting because it suggests that the best colleges are the ones accepting small class sizes of highly promising high school students. There's a reasonable argument that Arizona State and the University of Central Florida are the best colleges in America because they produce so many college grads capable of entering the workforce.
Of all the great authors of the world how many of them majored in English lit or creative writing? I tend to the Charles Dickens and Herman Melville stumbled on writing while at University studying law or accounting or something.
Listening to a podcast on the affirmative action in admissions problem, I wondered: why don't we just ask colleges to accept everyone by lottery? Why even bother with SAT/ACT/GPA/GDP/whatever? Everything meritorious is in some way racist; not only is a non-discriminatory lottery non-racist, it also allows colleges to compete on the quality of their product—if you can produce high-earning graduates out of a random grab bag of students, you can claim legitimate prestige.
I am late to this but in more meritocratic university admittance policies there would be elite courses not elite universities, not at undergraduate level anyway. Medicine or Engineering would be difficult to get into at any university - and standards within those courses would be fairly uniform across institutions.
This is now Ireland works, the state exam determines what courses you can do and the standard needed for different courses varies significantly within any university.
That said Irish universities are not setting the world alight at research and post graduate level so maybe that’s a reason to concentrate the best in one, or a few, universities.
Is there any hard evidence of the theory that diversity causes better educational outcomes? I always see this asserted without evidence as a rationale for affirmative action, but it seems to be on much shakier ground than mismatch theory.
That may be true of Harvard, but I think there's a simpler explanation for the broader phenomenon: the brightest students would prefer to be taught by the brightest teachers (they'd learn more), and the brightest teachers would prefer to teach the brightest students (it's more rewarding). The only reason this isn't going on earlier in the education system is that governments feel a duty of care for kids, and make rules to stop it (at least in the UK), but adults are on their own, so the obvious equilibrium follows.
“Fairly-earned privilege”
Curious frame to hold on a (correctly ime) hereditarian blog.
This is one context in which I might suggest more utilitarianism would help your argument. It’s not about merit or earning just desserts
it’s about expected return on investment.
The example with the writing teacher is a bit absurd. The vast majority of great writers never had *any* formal training in creative writing at all. (or at most, basic)
So I heavily doubt the great writing teacher is going to do anything for the genius writer. The genius writer doesn't need a teacher, probably doesn't even want one.
Writing is not a normal profession - it's very debatable if teaching does anything good for you if your skill level is above mediocre.
A shaky assumption underneath this question is the idea that Harvard admissions committees have highly accurate screening systems that can select for true academic merit. Revisionist History has had several great episodes calling that assumption into question. At an anecdotal level, it's important to remember that Einstein was repeatedly bounced away from undergrad by standardized entrance exams. Then, after he graduated, he was saddled with crummy academic reference letters saying he'd never amount to anything.
Judging academic merit is surprisingly hard. The assumption that Harvard admissions committees are especially good at it doesn't seem well justified. Another possible answer to a five year-old's starting question might be that elite institutions should admit some seemingly sub-par students because some of the sub-par students might actually be geniuses.
I think there are plenty of elites that can make it into our nation's top schools, and the smartest elites will continue to do so. It will also create a greater incentive for wealthy families to prioritize education.
Alternatively: a sufficiently visible pipeline to success through education socially legitimizes learning as a credible way to achieve wealth and/or status (and achieve it virtuously!). There must be a clear reward within sight of the k-12 slog to focus teenagers and their families on academic success, and ‘go to a great school and you’re set’ might be a more convincing and actionable heuristic than ‘get good at something, wherever and however possible’.
For this reason (and to stick my neck out), we should be inclined to make the pipeline more credibly within reach of traditionally underserved/underachieving groups, perhaps through affirmative action and hamfisted attempts to showcase diversity in higher education.
It seems to me that the better the student is, the worse the professor is able to be. A smart student can teach themselves the material. If you look on MIT opencourseware, their teachers are pretty boring and unengaging. Meanwhile, the teachers at the "Great Courses" (intended for the layman) are quite excellent. This gels with what one commenter wrote about Harvard in 1975 and all the famous teachers being awful.
So the reason we don't send the worst students to the best schools is be because the best schools have the worst teachers, they don't need good teachers precisely because they are selecting for good students. The IVY league is not about making you smart, it is about proving you are already smart.
I wrote about this:
"Let’s imagine a stylized world where each applicant is either rich or smart.
If a college lets in only smart kids, it can’t keep the lights on. It goes out of business in short order.
If a college lets in only rich kids, in a year or two, the parents of the next crop of rich kids won’t be interested in making donations or paying tuition to a so-so institution. The rich parents will send their kids and dollars elsewhere.
So, there is some optimal allocation of rich and smart kids. At this optimal point, the rich kids’ parents’ money keeps the institution afloat and builds the labs where the smart kids commercialize nuclear fusion, cook up next-generation antibiotics, etc. The smart kids boost the college up the rankings and provide the prestige that draws in the rich parents.
Colleges have been playing this game for many rounds. They’ve had a chance to learn and tinker. I bet many of them know what that optimal point is. Relatedly, I think private-school applicants are so appealing to prestigious colleges because they sometimes allow admission offices to sidestep the tradeoff between rich and smart kids.
Of course, in real life, colleges are optimizing for more than two traits, some applicants align with more than one of those goals, and the super-selective places have huge endowments and so don’t have to operate from cash flow."
https://clarkecollegeinsight.substack.com/p/one-more-view-of-the-nyu-orgo-scandal