September 17, 2009

College rankings

The neoliberal Washington Monthly magazine has gotten into the business of ranking colleges, but they do it based on their assessment of each university's "contribution to the public good in three broad categories: Social Mobility (recruiting and graduating low-income students), Research (producing cutting-edge scholarship and PhDs), and Service (encouraging students to give something back to their country)."

Interestingly, in their 2009 rankings of 258 national universities, the top three contributors to the public good are all prevented by law from practicing affirmative action: UC Berkeley, UC San Diego, and UCLA. (Of course, the UC administrations break the law, but Prop. 209 does keep them from doing it as flagrantly as they would wish.)

As a UCLA grad (MBA, 1982), I guess this should make my heart go pitter-patter with alma maternal pride, but, I dunno, I'm not sure I'm persuaded. Now, some of the high rankings of the UC universities stem from the Research measures being partly biased in favor of sheer size of school. For example, Berkeley, a huge place, is lauded for being first in the country in number of science and engineering Ph.D.'s awarded annually, while, say, Cal Tech is downgraded for being only 38th in this absolute measure.

But, what interests me is the credit the magazine gives for having a high graduation rate relative to average SAT scores and percent of students getting Pell Grants for being low income. For example, UCLA ranks 7th "best" in the country at graduation rates because a simple multiple regression model of its SAT/Pell percentages predicts that only 76% would graduate but 90% actually do.

I gather the idea is that UCLA is to be commended for inspiring its students with a love of learning or something, but, certainly, one simple way to boost your graduation rate is to make graduating easier. Consider a counter-example Cal Tech: A few years ago, I was walking across the Cal Tech campus and I stopped to listen to a speech given by a sophomore coed to a group of high school students and their parents considering Cal Tech. The young lady who had been chosen by Cal Tech to talk to potential students started talking about how big an adjustment freshmen year is, and how hard it is, and all the all-nighters you have to pull, and how brusque the professors are, and how emotionally wrenching it all is ... and then she broke into tears over her memories.

According to the Washington Monthly's (presumably linear) model, 104% of Cal Tech students should graduate, but only 89% do, so Cal Tech ranks 248th out of 258 in contributing to the public good by graduating people from college.

If I had a kid who could get into Cal Tech, I'd probably want him to go to Harvey Mudd instead. But, is Cal Tech detracting from the common good by being hard? Beats me.

By the way, Cal Tech is the only university out of the bottom 20 in graduation rates to be private. Colleges largely funded by dads writing checks tend to be more accommodating than at least some colleges funded largely by the taxpayers. The elite model these days is to have very high admissions standards and pretty low graduation standards. Is that better for the country? I don't know, and I doubt if the Washington Monthly knows either.

So, why are the graduation rates at the University of California schools so high? UCLA didn't use to be easy -- six of my high school friends went off to UCLA, rushed the same fraternity, drank a lot of beer, and flunked out together by the end of their freshmen year. Has it gotten easier? I don't know.

The affirmative action ban probably helps.

One obvious reason is that their student bodies are so heavily Asian, whose parents push them hard to graduate.

I'm also wondering whether the UC schools are jiggering the statistics without the Washington Monthly folks catching on. The UC schools have a little-publicized backdoor in that they accept a huge number of transfer students (UCLA takes in about 3,500 per year), typically from California's junior colleges. Are they counting transfer students who only spend half their college career on the UC campus?

My published articles are archived at -- Steve Sailer


birmingham said...

If I had a kid who could get into Cal Tech

Ouch. Does Sailer Junior read this blog? That's gotta hurt. You don't worry about his self-esteem if he hears his own father say things like that?

Regression to the mean, maybe?

Greg said...

Steve, no one uses "neoliberal" this way anymore.

"Neoliberal" means evil free market capitalism and globalization -- Friedman and Hayek -- in the current language fashion.

mutecypher said...

Those of us who went to Cal Tech also wonder if it makes sense to have such an oppressive level of difficulty. I'm told it's a pretty cushy place for grad students. So, send your kid to Harvey Mudd or MIT or Stanford and then they can go to Tech for the additional letters and skip the part where they're made to feel stupid. Unless your kid is a seriously arrogant little prick, then it might do them some good. For some of us, humiliation is a necessary part of growing up.

Will said...


Three questions I'm curious about: What schools were you accepted to as an undergraduate, and why did you choose to go to Rice?

What was your SAT score, and what was the verbal/math breakdown?

Why UCLA for b-school?

Anonymous said...

In the Liberal Arts list, they've ranked Morehouse 14th (above Grinnell and Wheaton?) Spelman 21st, Fisk 32nd, Dillard 40th, and Rust 57th. (All are historically black institutions)

I suspect there was some "special sauce" weighting involved here especially as Rust University's graduation rate is 29%!

For comparison's sake, the excellent Kenyon College is ranked #58 (one space lower than Rust), and yet it graduates 83%.

Mitch said...

I don't know about the others, but at Cal you can graduate with a liberal arts degree with just one math class, "pre-calc". This class is waived if you get over 600 on the SAT Math section, which is everyone except URMs.

Reg C├Žsar said...

The elite model these days is to have very high admissions standards and pretty low graduation standards. Is that better for the country?

If the country is Japan! That's been their "elite model" for generations.

UCLA didn't use to be easy...

It's so refreshing to see this phrase spelled correctly. I can't remember a time when one didn't use to see the wretched "didn't used" much more often.

Tom V said...

"Alma maternal pride" would have been a characteristically brilliant coinage, Steve, but it should refer to something that UCLA feels for you rather than the other way round.

Steve Sailer said...

Oh, well, no hope of that!

klaos said...

I guess once we're in full fledged Idiocracy they'll scrap the GDP in favor of a GDPPG (Gross Domestic Production of Public Good).

stari_momak said...

According to the Washington Monthly's (presumably linear) model, 104% of Cal Tech students should graduate

I would think that if they are going to do this sort of thing, they would at least consult with a social scientist aware of logistic regression.

Anonymous said...

What were you doing on the Cal Tech campus?

Were you running 'Game' on statistics nerd girls?

dearieme said...

"some of the high rankings of the UC universities stem from the Research measures being partly biased in favor of sheer size of school": are you using "neo-liberal" as a code for "stupid"?

Drawbacks said...

Maybe Caltech just don't care much about their students.

silly girl said...

There is an interesting discussion at the feministing blog of all places where one commenter, Alice, asserted that in fact college is an "investment" worthy of debt in only some cases. She further argued that college is actually "consumption" if the degree, college, etc. does not lead to a return on the outlay.

It really is a better paradigm. That is "investment grade" education vs. "personal enrichment" education.

It would be nice if a newspaper or magazine would do a review of colleges and degrees based on their value as an investment. This would take a fair degree of analysis because although there is info on the average salary for graduates in certain fields, say a CPA, some employers pay more and are more likely to interview and hire graduates of certain programs. The ratio of the cost of a specific degree at a given institution to the lifetime average income potential is its investment rating.

Just an idea.

silly girl said...

Comment by Alice at feministing blog

It has nothing to do with the extent to which different degrees are "worthwhile," but with the fact that a degree that can make you money is, besides whatever else it may be to you, an investment, whereas a degree which you pursue purely for its own sake is a form of consumption.
Going into huge amounts of debt can make sense if you expect the returns to be worth the wait, even if your real reasons for doing it don't involve money. The problem with unprofitable degrees is not that people get them, but with people treating such consumption, noble a form of consumption as it may be, as though it were an investment, and going into far more debt than they would normally be willing to for something that they did not expect to earn a return on.
So, saying that you should not go into a huge amount of debt for a degree that will not pay for itself is largely the same as saying that you should not go into a huge amount of debt for any other form of consumption.

Quant said...

Regarding CalTech's low graduation rate, it may not be a problem if its "dropouts" are simply students who find it too intense and transfer to another college. Anyone smart enough to get into CalTech is smart enough to graduate from a decent school.

How do transfers out of a school affect its reported graduation rate?

Anonymous said...

birmingham: That's not much of a self-esteem crusher. Cal Tech is damned hard to get into.

Alticor said...

I paid about ten thousand dollars, total to get a degree in Political Science. I went to a small town community college 2 years and the other two as an adult off campus resident local at a crummy teacher college.

I paid $24,800, upfront and in cash, to get an aircraft mechanic's license thru a private vo tech school. In both cases I cared little about the subject-I wanted a flying job and airlines wanted a degree, any degree. Aeronautical Engineering was no more attractive than music or literature. When the airline jobs collapsed, I wanted a corporate flying job and having an A&P is a big advantage. I had no intention of becoming an aircraft mechanic per se.

The quality of instruction at the vo-tech was vastly superior.

Truth said...

"The quality of instruction at the vo-tech was vastly superior."

I had a similar experience. There is nothing quite like the feeling of "graduating" from a Southwest directional school, then moving to LA and discovering that your freshman level classes at Los Angeles Valley College are more challenging than your Senior level classes back home were.

Anonymous Rice grad said...


I am also curious about Rice. I was there about 10 years after you. Two questions:

1. What did you think of it (educationally, socially, etc.)?

2. I remember being told that prior to the mid-80's or so, it had "the highest" or at least an on-par with MIT suicide rate due to workload and academic pressures. But then they had some sort of reform to mellow things out a bit by the time I got there.

My opinion of it was that it was challenging enough, probably harder than the Ivies but not at the level I would have expected at MIT/Cal Tech. What was it like when you were there?

Steve Sailer said...

"I am also curious about Rice."

My recollection of 1976-1980 is that Rice then was hard but not brutal; but I triple-majored in soft subjects (econ, history, management). It was clearly much harder for science and engineering majors. About 3/4ths of freshmen entered intending to major in S/E fields, but only about half graduated with an S/E degree.

I don't, however, recall many people other than architecture majors and computer science majors routinely pulling all-nighters. A roommate had to switch to working all night to graduate in CS because the mainframe was so slow during the day.

The serious dope smokers tended to fail to graduate, or still be at Rice after a half dozen years. Most people who did graduate seemed to study six nights per week and party one night. It was fairly monastic -- most guys didn't seem to have girlfriends, at least not until they got far enough ahead and could coast during their senior years.

I may be underestimating how hard it was, though. I had higher than average SAT scores, easier classes (especially in Econ, which didn't require any econometrics back then), and good, non-procrastinatory work habits (much better than I have today).

Horace Mann said...

It looks like some of the authors of this ranking share some of Stevil's ideas from an earlier post of his and may perhaps even be readers.

From Purpose:

we want to add even more, particularly with respect to the single most important thing colleges do: helping students learn. Remarkably, colleges report virtually no useful, comparable information about teaching practices or how much students learn between the time they arrive as freshmen and leave as seniors. It’s not that such data don’t exist...

There are, however, positive signs from the Obama administration, including a push for legislation that would require colleges to report data like graduation rates for Pell Grant recipients (another measure we called for four years ago), how much students earn after college, and whether they land a good job. This type of transparency will be key to understanding which colleges really serve the public interest—and giving others a reason to follow their lead.

Andrea Nyx Hemera said...

If they really want to be PC, why not include some city colleges or community colleges?