College Rankings Fiesta

I'm back from Costa Rica after 10 blissfully email / Internet / news-of-the-outside-world-free days, and while I'd still much rather be there than here, it's nice to be here.

Right to business: the annual U.S. News & World Report college rankings were released in my absence. The news cycle around this is funny, because there's never really any news -- as long as the way the rankings are calculated stays the same, the top schools never, ever change. Yet again, Princeton and Harvard are #1 and #2 -- yawn.

That forces the news outlets to search for different stories, and the Times went with the straightforward "dodgy practices designed to juice the rankings" angle. But in focusing on things colleges do to drive down acceptance rates (mostly by increasing applicants), the Times is looking in the wrong place. Acceptance rates make up only 1.5% of a college's rankings (see our paper on the subject here), so even you used every trick in the book to knock your rate down by five or 10 points, it wouldn't make much difference.

The biggest factor driving rankings that a college can realistically change (reputations being difficult to budge) is money. 30 percent of each college's rankings is based on either direct measures like spending per student or measures of things that cost money to buy, like higher faculty salaries and smaller class sizes. And that's one of the big reasons the rankings never change; our increasingly-winner-takes-all-society is reflected in the higher education sector, where a few elite colleges accumulate Scrooge McDuck-like mountains of cash in their endowments and pick among the brightest students, while everyone else scrambles for the leftovers. They more they have, the more they get, and so on.

InsiderHigherEd notes that U.S. News made a small change in the methodology, incorporating the percent of students on Pell grants for the first time into the "expected graduation" rate component, which rates colleges based on the difference between their actual and statistically predicted graduation rate. It's a nice gesture, but that's all it is. The expected grad rate measure is only 5% of the rankings, and based on some analyses I've done of graduation rates, the Pell percentage is probably only about 10% of that -- in other words, one half of one percent of the total. This won't make anyone redouble their efforts recruit more low-income students. They'd be better off recruiting rich students who pay full tuition and whose parents will donate to the endowment.

The Washington Monthly, by contrast, makes the Pell percentage a much more prominent part of their newly-released annual college rankings. That's the difference between rankings focused on what colleges do for society at large and the U.S. News rankings that are, in the end, mostly about what colleges do for themselves.