Skip to content

Are Ivy League schools overrated?

I won’t actually answer the above question, as I am offering neither a rating of these schools nor a measure of how others rate them (which would be necessary to calibrate the “overrated” claim). What I am doing is responding to an email from Mark Palko, who wrote:

I [Palko] am in broad agreement with this New Republic article by William Deresiewicz [entitled “Don’t Send Your Kid to the Ivy League: The nation’s top colleges are turning our kids into zombies”] and I’ll try to blog on it if I can get caught up with more topical threads. I was particularly interested in the part about there being a “non-aggression pact” outside of the sciences.

This fits in with something I’ve noticed. I know this sounds harsh, but when I run across someone who is at the top of their profession and yet seems woefully underwhelming, they often have Ivy League BAs in non-demanding majors (For example, Jeff Zucker, Harvard, History. John Tierney, Yale, American Studies). My working hypothesis is that, while everyone who graduates from an elite school has an advantage in terms of reputation and networks, the actual difficulty of completing certain degrees isn’t that high relative to non-elite schools. Thus a history degree from Harvard isn’t worth that much more than a history degree from a Cal State school.

And David Brooks graduated from the University of Chicago with a degree in history . . .

In all seriousness, I don’t know if I agree with the claim in the headline of that article Palko links to.

I was very impressed by some of the Harvard undergrads I taught. Then again, they were statistics majors. In the old days, statistics might have been considered the soft option compared to math, but I don’t think that’s the case anymore. If anything, math majors are sometimes the sleepwalkers who happened to be good at math in school and never thought of stepping off the track. Anyway, it’s hard for me to make any general statements considering that I don’t teach many undergrads at all at Columbia.

Palko responded:

Yeah, I don’t want to put down Harvard grads, even the history majors. I’m sure that a disproportionate number of the brightest, most promising young historians are working on Harvard B.A. What’s more, I suspect most of them are developing valuable relationships with some of the most important names in their field.

What I’m wondering about is the popular notion that Ivy League schools are hard to get into and hard to get through. The first part is certainly true and the second appears to be true for STEM (which also has an additional self-selection bias). I’m not just not sure if it holds for all fields.

I don’t think there’s any question that selection bias, networking opportunities and halo effects play a large role here. What if they account for most of the benefit of attending an elite school for most students? This is worrisome from both sides: students are twisting themselves into knots to meet artificial and frankly somewhat odd selection criteria; and we’re giving the students who meet these odd criteria huge advantages in terms of wealth, career, and influence.

That can’t be good.


  1. cugrad says:

    I think it’s person specific in ivy league. There are people who can’t speak English, but can program and prove very well. But to get in, you must have excelled in something.

    As for the education, it also depends. There are great professors, but if you fail to understand them (’cause of yourself or the prof), the quality of education does not exceed an non-ivy school.

    • Anonnn says:

      These non-English speakers in math are so annoying/ debilitating. They can write math within their limited dictionary but are wholly incapable of reading your work unless you purposefully dumb yourself down around them.

  2. numeric says:

    I think what is missing is the context about the Ivy League schools, which is, simply that as this society has become more unequal, people (and parents in the middle class) understand that if you aren’t in the top 10% in this society, your life is unpleasant (and to really enjoy life, you want to be in the top 1%). The most certain way to get there is to get the degree from the good school. Thirty years ago, you went to an Ivy League school and became a doctor or lawyer and ensured a good life. Now it’s finance, consulting, and, increasingly, programming apps to get into a start-up. Inequality creates distortions. To ameliorate distortions, reduce inequality. At that point an Ivy League education becomes less important (and they can go back to being doctors and lawyers).

  3. Matt says:

    Why would difficulty be of interest here. Seems trivial to make something difficult. One could make the crappiest program difficult to get through – e.g. ‘only the top 2% get through’. I would think the value of any process/system is the incremental value of its outputs wrt its inputs. Obviously that is difficult to assess, but difficulty seems like a lame proxy.

  4. j says:

    “My working hypothesis is that, while everyone who graduates from an elite school has an advantage in terms of reputation and networks, the actual difficulty of completing certain degrees isn’t that high relative to non-elite schools.”

    Obviously it’s possible to coast through college, no matter where you go or what you major in. If you want to slack, Harvard isn’t going to stop you. But for people who don’t want to slack, the quality and rigor of the coursework at an elite institution is NOT comparable to that of, say, a state school. I say this as someone who has taken lots of classes at a state school (STEM and non-STEM) and graduated from an Ivy. It is simply not comparable. You don’t get Andrew’s class at a state school.

    • numeric says:

      You don’t get Andrew’s class at a state school.

      On the contrary, you can get excellent courses in stats at state schools, since typically the professors there graduate from the Ivies.

      • j says:

        Okay, maybe you attended a better state school than I did. My first stats class was at a state school, and it’s a mini-miracle I decided to give it a second chance.

        • Bob says:

          Well, that experience is not limited to state schools. My first (frequentist) statistics course was at a top ten (USNEWs ratings) but not Ivy school. It was terrible! The most poorly taught course that I took at that university. In contrast, the random process course I took (in a different department) was excellent although rather demanding.


      • j says:

        numeric: Now you’ve got me thinking about how I can make my language more precise and less sweeping. I guess my claim is that if you pick an Ivy at random, and you pick a state school at random, and you look at the top 10% or 1% most challenging/rigorous classes available to freshmen at both institutions, you’ll find that those at the elite institution are much more challenging (in a substantive way, not the ticky-tacky way that Matt mentions above). And then if you repeat for sophomores, juniors, and seniors, you’ll find the same thing. I gave Andrew’s class as an example because his class was one of the highlights of my life, but maybe Math 55 (or its analogues at Yale, Stanford, etc.) would be a better example.

        • Anon says:

          But J, the hypothesis above refers to less exceptional individuals at Ivy schools and those who are not choosing a wider range of majors, including fuzzier and more subjective ones.

          • j says:

            Anon: I was trying to respond to the claim that “the actual difficulty of completing certain degrees isn’t that high relative to non-elite schools”. I think this is a silly criticism because, sure, the path of least resistance is probably equally non-resistant no matter where you attend school (something something horse to water), but at a more elite school you have considerably more freedom to make your coursework extremely challenging and rigorous if you so choose, right from the start. Harvard has >4 different math tracks for freshmen who have already taken AP Calculus — so much freedom!

            I will grant that it’s harder to compare fuzzy courses at different schools because the curriculum is less standardized. I would feel comfortable extending my claims to philosophy and economics, but perhaps not to women’s studies or art history.

  5. Andreas Baumann says:

    I would suspect that the ranking would be something like Harvard physics > Harvard history > State U physics > State U history, simply because I would expect the selection effects to dominate the attrition effect. I think Palko is obscured by the fact that when academically mediocre physicists go into business, they end up in “invisible” vocations like quants, etc. Thus you end up seeing the mediocre historians who went into business, but not the mediocre physicists.

    • Andreas Baumann says:

      (Besides, attrition rates are in general not that diverse over the field of majors, although of course this is over the universe of colleges).

    • HI says:

      Really? My guess was that most students who choose to major in physics or math and who actually graduate clear some minimal level of competency no matter which school they go, whereas history majors vary more.

      Ed Witten, the physicist, was a history major at Brandeis, so I guess it’s possible to be a history major and scary smart.

      • Anon says:

        Ron Unz did a secondary major in history, though his main thing was physics at Harvard.

      • Andreas Baumann says:

        Yeah, I was talking about ability contingent on school (and graduating). I’m not doubting that on average physics majors are smarter (or that there are more history majors who never should have graduated). I’m just expecting the selection effects on school to swamp the selection effects on major within the school. Does anybody know of any data, that allows us to investigate this?

      • Martha says:

        Anecdotal information: When I was advisor for the statistics “specialization” in the undergraduate math major at Texas, I found more than once that an “average” student in the specialization would get an internship, and wow their employer with their knowledge/competence/insight. (Typically other employees were from business or economics backgrounds.) The students themselves found that a little scary, since they weren’t very sure of their knowledge.

  6. Erin Jonaitis says:

    I thought that was already the reputation of Ivy League for undergrad — a lottery among the brilliant to get in, and then everybody gets an A.

    I like Matt’s point above, that difficulty is an awkward proxy for value. I work now in the context of medical education, which runs into a similar issue: most everyone who walks through the door is very smart and hardworking, so what should you expect your pass rate to be? With medicine, there’s at least reasonably high agreement about what medical students need to know and why it’s important. But with history (e.g.) I suspect the agreement is not as strong and the downside of a badly-prepared graduate not as straightforward to see. There’s almost certainly still some downside but most of the bad outcomes (poorly-informed policy decisions etc) could have a lot of antecedents.

    Andrew, do you know of any people who write about these issues (pass rate, grade inflation, titrating the difficulty of a curriculum, etc) from a statistical perspective? It seems to me that there’s a lot of meat there.

    • Andrew says:


      Val Johnson wrote a whole book on grade inflation from a statistical perspective. Eric Loken has written on related issues, as has Howard Wainer. And I’m sure that lots of psychometricians have looked at this.

      There’s also the question of how best to communicate the understanding that’s out there. If blogging and the internet had never existed, I’m pretty sure that I’d be editing a few statistics journals (I always say no when they ask, because I think I can serve the community better by blogging etc), in which case I’d organize a special issue of JASA or whatever on the topic. As it is, though, it’s not clear to me that journal issues get any particular readership. It’s tough to break through the clutter. Maybe I can ask some experts to guest-post here on statistical issues in setting a curriculum and evaluation of students. A few years ago I tried to organize this sort of thing on The Statistics Forum, a blog we set up under the auspices of the American Statistical Association, but it didn’t really work. It was too hard for me to get people to post anything. So I think the best bet right now is to do things through this blog.

    • Ben says:

      This was my first reaction as well – since when has anybody thought that Ivies were hard to graduate from once you got in?

      My viewpoint tends to be that prestigious universities exist to further their own prestige and influence, which they accomplish by 1) admitting the truly brilliant who will make a contribution to their field 2) admitting the savvy self-promoters who will succeed in business or politics. In either case, the academic “value” of the degree is irrelevant because it’s access to professors that matters for 1) and halo/network that matters for 2).

      • Andrew says:


        One reason I like to teach future leaders in science and society (or, as you call them, the students who are truly brilliant and the savvy promoters) is that then I think my teaching will have the largest effect.

        By teaching the “truly brilliant,” I am influencing (I hope in a good way) future developments in science and technology, e.g. by teaching hierarchical models to a future bio-researcher, I could (indirectly) cure cancer or reduce the impacts of climate change or whatever. Or even just influence the future writers of the next generation of statistics textbooks.

        By teaching the “savvy promoters,” I am influencing (I hope in a good way) future policies in business and government, e.g., by teaching hierarchical models to a future business leader or government official, I could (indirectly) help the development of more efficient and effective policies.

        I think both of these are worthy goals (even if not guaranteed to work or even to go in the right direction, for example perhaps my excellent statistical methods will teach a future Ed Wegman how better to cover his tracks).

        But they are not the only goals of education. I could, for example, have the goal of educating students who otherwise would not be exposed to modern statistical ideas. Or I might want to use my educational efforts to help the largest number of students, or to reduce educational inequality. Teaching at Columbia or Harvard doesn’t seem like the most effective way to attain those goals.

        In any case, none of what I’m talking about has anything to do with furthering the prestige and influence of the university. It’s all about using my university position to further other goals.

        If you’re talking about the president of the university, though, that’s different: I guess he or she probably should have a primary goal of preserving and extending the institution and its influence (assuming those influences are positive or at least benign).

  7. Erin Jonaitis says:

    Thanks, Andrew! My library has a copy of Val Johnson’s book.

    I agree with you about the importance of knowing which audience you want to reach and how best to find them. I’m interested in this particular issue for a mix of personal and professional reasons, so I could be found in a lot of places. But I would absolutely be interested in guest posts on the topic here.

  8. Patrice Boivin says:

    A graduate from Canada went to the US to finish his Masters in Geology.

    He discovered that there were two streams:
    Those who went to school in beat-up cars and worked like crazy.
    Those who went to school in Audis, BMWs, sprots cars and partied a lot.

    I keep wondering how the people in the second group ended up with good grades.

    • Kyle C says:

      Same was true in law school. The truth was that most of the fancy-car people were smarter, had gone to better colleges, had had more interesting preprofessional jobs, and knew how to study efficiently.

  9. Steve Sailer says:

    Here’s an excerpt in Salon from a new memoir by a former New York Times’ editorial staffer who is a graduate of William Paterson University in New Jersey. It makes an informative comparison to the work of NYT writers complained about above such as John Tierney and David Brooks:

  10. Steve Sailer says:

    When I was applying to college in 1975, Stanford was notoriously hard to get into and easy to get out of.

    Judging by metrics like alumni giving and prosperity of the surrounding community (i.e., Silicon Valley), it’s hard to say Stanford has been pursuing the wrong strategy.

  11. Howard Wainer says:

    IMost of my career was in industry and research institutions, what teaching I did do was at Chicago, Princeton and Wharton.
    Hence I am a remarkably poor judge of the quality of instruction at other institutions. But, to me, the big question relates to partitioning credit for success to the instruction vs. quality of the student — in sports terms, is it the coach or the general manager?
    The students I had when I taught statistics in the civil engineering department of Princeton would likely learn irrespective of the instructor.
    Can we tell much about Don Rubin’s quality as a professor by the success that Andrew Gelman had after studying under him?
    This is summed up by:

    Charles Dillon (Casey) Stengel had a long career as a major league manager, but in his early years, he was never particularly successful . His fortunes improved markedly after he was hired by the Yankees. Casey explained, “I only became a genius after Mickey Mantle joined the team.”

Leave a Reply