« Signaling in Economics | Main | Reveal Rejects? »

March 22, 2009


The most interesting thing to me: success, in practically all these studies, is measured by how much money people make. It's a more objective criterion than, say, how well they can write or the quality of their lifestyle, but it feels awfully naive, like it's missing a Grand Canyon-sized aspect of success to define it merely as "how much $ graduates make."

The real question is the value of an exclusive education vs the cost of such an education. ~100k by graduation requires a huge increase in lifetime earnings to balance it out.

"those who attended the most selective colleges would earn an average of $2.9 million during their careers; those who attended the next most selective colleges would earn $2.8 million; and those who attended all other colleges would average $2.5 million."

Having been to both, I'll just make some vague observations. I think the unmeasured confounder is individual student's drive/ambition. Some people just don't want to make as much money as possible or work super, super hard. These people would also be more likely to go to a less prestigious school over the more prestigious one. So the magnitude of difference in income by regression is likely to be an overestimate.

Besides financial cost, there is a psychiatric cost for certain people to attending Prestigious school X. You are more likely to be in the bottom half, or bottom 1/4 of students. Some people don't have the temperaments to tolerate this. These people develop maladaptive social behavior that isn't going to help them in the long run. (or well, make them nice people to be around)

I don't doubt the CV effects of having a prestigious school on your resume (and hence increased earning potential, maybe not as large as stated by regression) -- but I'd be curious about the educational effect. It would be interesting to see if people's subsequent standardized test scores were effected by their choice (MCAT, LSAT, GMAT, etc.) Or some other metric, rather than $.

He should go to the best ranked school that he gets into. Social networks are now global, not regional. The top ranked schools have the best social networks. The teaching is the same, the cost is higher, but the social networks are much more important now than a generation ago.

Unless they were controlling for major, they might miss things like those with higher SAT scores going into science instead of finance. That could explain the peak at 1100.

The discussion of the effect of major on earnings reminds me of a survey of UNC-Chapel Hill graduates. It concluded geography majors were the highest paid. The reason? Michael Jordan was a geography major.

The problem with these studies is they are looking at the average effect. I would suggest for any child of mine that if they have any interest in going to grad school they should take the better ranked route. Ranking in undergrad has a huge effect on ranking of grad school, and that really matters for both pay and who listens to you.

If the point of going to school is to maximize income, it makes sense to go to the highest ranked school you can get in to. The name recognition of the school, and access to vast alumni networks, seem worth it.

Of course, if you are going to school to maximize income, a lot of behavior follows that I think for most people isn't utility maximizing, particularly for the high-IQ.

What about academic success? I would think that going to a grad school at a prestigious university is highly correlated with good placement in academia.

"Fit" seems to me to make sense. When I chose what uni to go to (be it over the ocean in the little ol' UK) prestige only was one part of it - more important was whether the uni matched me - in location, in aims, in style and ambition. If you think you will be happy there then you will be, which in turn lends itself to much growing up and experiencing of life, and doing better academically than if you had been sad.

To add to Seinberg's point, it would be interesting to see a comparison of the amount of soft power obtained by graduates - say, political success, entrepreneurship ability, or record of technological innovation - to the prestige of their universities. This might allow for cleaner identification of the role of the social networks (with political success) or academic rigor (perhaps with technological innovation) of a prestigious school, as well as providing a more nuanced definition of "success." You'd still have to look for a way to control for family income and social status [i.e. degree of prior connection to useful social networks], though, which could possibly confound any identification of success with university prestige.

I was in that class of 1972 that the first paper analyzed. I was accepted but could not afford to attend an Ivy league school. I graduated from a state school w/ a science degree and went to work alongside a few interns around my age from MIT and other fancy schools. A few were frustrated that they couldn't keep up with our group and resolved to take the easier and more lucrative path into consulting/finance. I had never even heard of McKinsey&Co at my state school, but all the Ivy leaguers knew it intimately. There's your difference in income: maybe 30% of Ivy league grads eventually end up in finance/consulting, thus skewing the income comparisons significantly. Also, a large percentage of Ivy league students are legacies and/or from Rich&Connected families.

My brother and sister followed in my inglorious footsteps. She ended up on Wall St. anyway earning a LOT to run a department. He turned down a lucrative offer from McKinsey to be a post-doc at a big Ivy League med school. I ended up reading blogs after a PhD and a boring research career. The lesson for your kid: the path from Ivy league to finance/consulting is easy. The path to (science) grad school is not much easier. The undergrad education at Ivies is absolutely NOT any better than a good state school. The average student you meet at Ivies is much more ambitious and that will rub off on you. Stoners and frat boys killed my motivation in state school, but it was fun.

These studies need to stratify the population. If higher SAT => less money, yet students from prestigious high-SAT colleges don't earn less than students from less-selective colleges, then those prestigious colleges are offsetting the harmful effects of high IQ.

Another problem is that someone who goes to an ivy-league school has a good chance of getting an academic career. Someone who doesn't, doesn't, and is likely to go into industry, making their salary higher, but their job satisfaction lower.


"go into industry, making their salary higher, but their job satisfaction lower"

Speak for yourself Phil. I loved my time in Wall Street IT - really loved it - and trust me, I made beaucoup bucks. Almost all the academics I know struggle with the low pay, the onerous grading, the ridiculous and constant political nonsense, the student disrespect, and interdepartmental back-biting. Gimme the Champagne holiday parties and 6-figure bonuses any day over that, big guy.

How did I get that job? Sure, I had a nice web portfolio, but also what clinched it was my ability to speak easily in French about Aristotle's Ethics - at that time, it was a small French software company with a French founder and president, a guy who considered liberal arts and belles lettres to be Extremely Civilized. So you bet my crazy Johnnie education made the difference.

This paper contains a critique at footnote 27:

For instance, Dale and Krueger (1999) attempted to estimate the return to attending specific colleges in the College and Beyond data. They assigned individual students to a “cell” based on the colleges to which they are admitted. Within a cell, they compared those who attend a more selective college (the treatment group) to those who attended a less selective college (the control group). If this procedure had gone as planned, all students within a cell would have had the same menu of colleges and would have been arguably equal in aptitude. The procedure did not work in practice because the number of students who reported more than one college in their menu was very small. Moreover, among the students who reported more than one college, there was a very strong tendency to report the college they attended plus one less selective college. Thus, there was almost no variation within cells if the cells were based on actual colleges. Dale and Krueger were forced to merge colleges into crude "group colleges" to form the cells. However, the crude cells made it implausible that all students within a cell were equal in aptitude, and this implausibility eliminated the usefulness of their procedure. Because the procedure works best when students have large menus and most student do not have such menus, the procedure essentially throws away much of the data. A procedure is not good if it throws away much of the data and still does not deliver "treatment" and "control" groups that are plausibly equal in aptitude. Put another way, it is not useful to discard good variation in data without a more than commensurate reduction in the problematic variation in the data. In the end, Dale and Krueger predictably generate statistically insignificant results, which have been unfortunately misinterpreted by commentators who do not sufficient econometric knowledge to understand the study's methods.

All this is very interesting.

As for your son, I would say he should choose his college according to the types of students that go there, my thinking being the student culture at a college largely determines what students get out of it. I would guess a more prestigious school comes with good and bad: on the good side the students there are on average more interesting and intelligent and driven, on the bad side many of them may not be as interesting or intelligent or driven as they and others tend to think ;-).

You often see the "high quality of the average student" justification given for going to very selective schools. My issue is: so long as you are not the very smartest student, you are likely to find more very intelligent students, probably more than you could reasonably incorporate into your "friend group" anyway.

I can see value in not being among the top percentile of students at your school. At the same time, there are benefits: at selective schools, students make more demands of their professors' time. It is harder to differentiate yourself in their eyes when you can't casually stroll in during office hours.

I went to a large state school where I was likely a few standard deviations above the mean IQ. I quickly made friends with other bright students and built relationships with professors. My academic scholarship saved my parents a lot of money and the close proximity to home made the college experience a lot easier for me (some might say this is a bad thing, I disagree). I could have easily transferred to a much higher ranked school, but chose not to. I guess the point of all this is: there are so many factors involved in choosing an undergraduate institution, it seems crazy (to me) to make the choice based on possibly small percentage difference in lifetime expected income

One thing about the SAT score stuff is that I doubt it adjusts for what people major in. Smarter people probably tend to major in less lucrative majors, whereas the average and slightly above average people go for lucrative majors such as business. If this is true, then it creates a Simpson's paradox situation. A better way to look at the data would be to divvy it up by majors. Business majors in one category, physics majors in another, etc. Then, I believe, you would see that SAT scores correlate with higher earnings.

It is far easier for people from more elite schools to get into low paying professions like academia, non-profit work, politics, journalism, etc. I'm sure the people who choose these opportunity bring down the average income earned. What you need is the statistic, of the population whose goal is to make money, what schools do they make the most from?

We don't have that in the study, but we have a decent proxy. Low income students earned considerably more coming out of elite schools.

Evil Mutant,

Looking at occupational status shows a strong effect of IQ than on income.

A stronger effect rather.

Another variable worth considering is the ranking of the school for which the highest degree was awarded. For example, I went to a not-so-highly-rated public university, but then went to a prestigious school for a PhD. My guess is that put me in better position for future academic appointments than someone who did the reverse. I would guess similarly for MDs, JDs, etc.

Half Sigma might be right about this study, but I wouldn't expect him to be; he really likes jumping to conclusions.


Good thing I registered for Typepad at one point, or I would have had to resort to telling you what search terms to Google. Your Post button simply "cannot accept this data" when there's a link in there even when there's no HTML and I broke it up with spaces. Dumb.

Wait wait you mean Half Sigma, home of "Since mainland China is communist, all Chinese cooks in America must be communists as well, which is why they don't know how to cook beef" and "OH MY GOD THE MEXICANS ARE COMING" might have poorly thought out and ignorant ideas?

Surely a George Mason economist would know better than to let a racist, moronic blogger do his job of summarizing papers for him...

none, HS has said some stupid stuff but Noumenon did not provide any evidence that he & Robin are mistaken here. Your logic seems like a textbook example of ad hominem.

Some criticisms of this interpretation are made by "Reader" and "Adam" over in the Marginal Revolution comments:


I haven't read the paper, so I can't personally judge the issue, but I thought the link might be interesting.

TGGP, if Robin had actually read the paper in question instead of trusting the summary of someone who has already shown themselves to be pretty stupid, then he would have learned much earlier on that- no shock to anyone- the stupid person's summary of the paper is in fact wrong.

One thing I find very curious about this post, and all the ensuing comments, is that everyone seems to be treating "prestige" or "status" as having value only as a means to more concrete goods, such as money, education, or pleasant social and intellectual experiences. Then the temptation is to play a gotcha game, where we show that the prestige is over- or under-valued in terms of these other assets -- e.g., that an Ivy degree costs too much compared to a state degree, etc..

But who are we kidding? Surely prestige is valuable partly because people simply value prestige in and of itself -- even if that prestige is expensive in cash terms, and even if the prestige is not actually based on things it pretends to be based on (e.g., quality of education). Certainly an interest in prestige for the sake of prestige is not very high-minded, is maybe even a bit craven. But you can also see it as a virtue, a wholesome interest in what would be called "glory" by the ancients, or by the Adam Smith of *Moral Sentiments*. And anyway, a pure interest in prestige seems no worse than a pure interest in money.

To be happy we must act according to what we really want, rather than what we pretend to want. So I would advise your son to go to the most prestigious school he can afford. If he is normally constituted, he probably cares about what other people think. A lot. More than he realizes. And if he later discovers he truly does not care, then the high-prestige route still will not have cost him many other opportunities because of the secondary material value of the prestige.

Ironically, I would say the biggest downside of a high-prestige route is not the financial cost but the risk that it causes him to over-value prestige even further, corrupting his character and diminishing his soul. But most people who are corrupted in that way seem quite happy with themselves, so this concern is more a value judgment rather than a piece of advice about subjective welfare.

somebody else: Robin clearly stated he first read the paper, and later on found HS & Study Hacks through web searches. Nor would he have figured it our "earlier" rather than "at all", because he hasn't yet acknowledged being wrong in his interpretation. Noumenon or none could have done what rfriel did by writing or linking to some actual criticism of Hanson's view on the paper and the former was even willing to grant the possibility that HS had the right interpretation in this instance.

I suspect high lifetime earnings are well-correlated to having a father who can write blog entries doing a survey of research into the effect of college on lifetime earnings

For your consideration:

James, this is indeed a thoughtful relevant article.

Alexis, yes of course we value prestige itself; given my original reading of the paper I was going to post on it being nice clear evidence of this fact, but then messy facts intruded.

none, if I'm to refrain from citing anyone who has every said anything stupid, I'd just have to refrain from citing.

This is an interesting topic, but I find it somewhat disturbing that salary is the only thing mentioned in regard to the prestige of a college. This might be true for some degrees, and possibly all degrees. However, as a computer engineering undergrad in a non-prestigious college, I find that many companies in my industry do not bother do campus events at my college, despite being a large campus. These companies have obviously done their research, and there must be a reason that they consistently choose from specific universities before others. Additionally, a quick survey I performed recently on the founders of recent successful Internet startups showed that the founders tended to be graduates or drop outs of these prestigious universities. The longer I attend my college, the more I feel I am missing out on something better-- perhaps more opportunities or a better way of thinking.

The comments to this entry are closed.

Less Wrong (sister site)

May 2009

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30