Monday, April 30, 2007

Who Cares If Harvard Doesn't Love You?

This weekend, my local paper reprinted a very thoughtful column from April 9, "Why doesn't Harvard love me?" by Meghan Daum of the Los Angeles Times. While I commend the whole thing, I wanted to focus on two parts and then add my own views. The first is the very low acceptance rates among the most selective colleges:

IN THE LAST few weeks, the anxiety of high school seniors awaiting news of their college fates seems to have spilled over into the general population. It's easy to see why. UCLA received more than 50,000 applications, more than any other university in the country, and accepted just 11,837 of them. Harvard turned down 91% of about 23,000 hopefuls, 1,100 of whom had perfect SAT math scores. Acceptance rates for Stanford, Yale and Columbia were 10.3%, 9.6%, and 8.9%, respectively. That means thousands of valedictorians and people with grade-point averages of 4.0 or higher were passed over in favor of whatever form of superhuman DNA now constitutes a worthy Ivy Leaguer.

The statistic we should care about is not the probability of a student getting into a particular college but of a student getting into at least one college of a particular caliber. Many of the applicants to any one of these schools also applies to several of the others. The probability of getting into at least one is higher than getting into a particular one. But this is a statistic that is calculated person-by-person, not school-by-school, so it is not readily available or calculated at all.

Second, Meghan reports a widely overlooked phenomenon; namely, that most schools don't have the luxury of being very selective at all:

Believe it or not, the National Assn. for College Admission Counseling's most recent Annual State of College Admission Report shows that the median acceptance rate for four-year colleges, private and public, is about 70%. Do less flashy schools provide a faulty education? Do they lack high-quality professors? Judging from the brilliant academics I know who would be grateful to get a job anywhere, I doubt it.

A median of 70 percent is very high. A look at any US News ranking will show that acceptance rates rise fairly quickly with rank. Here's a graph based on the 2004 acceptance rates of national universities, from the August 2005 rankings:

I've taken a 5-school moving average by rank here to make the graph more readable. By rank 11, we're at an acceptance rate of about 20%. By rank 21, we're above 30%. In the low- to mid-30s, we cross 40% and then 50%. There are plenty of well regarded schools in those ranks.

Daum's larger point--with which I agree completely--is that it is a huge mistake to elevate these rankings in importance. It is true that on average, going to a school with a better ranking opens up more opportunities later in life. But I doubt that the difference between schools accounts for much of the variation in lifetime opportunity.

More importantly, I would hate to think that our colleges and secondary schools are conditioning students, particularly our best and brightest, to rely so heavily on factors like "which school you went to" in their thinking about their futures and how to succeed.


Anonymous said...

Didn't Alan Krueger publish a paper where he found that, controlling somehow for student ability, going to Penn or Penn State (say) made no difference in the long run?

Tom said...

There is a published study by a professor at Indiana (I think) that showed people who are accepted at Ivies but DON'T go are just as successful as those who do.

Unsurprisingly, this leads to the observation that what the Ivies are best at is accepting people with the highest probability of success. (No jealousy here...I went to MIT & Dart, my son is at Dart now).

I don't know how well correlated the different admissions departments are. My son's experience (wait list Yale, Cornell; accept at Amherst, Middlebury, Dartmouth; FULL RIDE at McGill -- oh, what is money?) is pretty typical of his classmates; if the process is random, applying to six schools at 20% acceptance gets you into at least one 3/4 of the time.

If you throw out 50% of the applicants as being obviously not qualified, and factor in twice the rate of acceptance for early action, you're up to 95%. Parents and kids should calm down.

Anonymous said...

I thought the studies showed that those who were accepted to Ivies, but declined were as (financially) successful as Ivy League attendees only after factoring in the investment returns on the money saved by going to a cheaper school. I.e. not attending the Ivy school reduced earned income, but the lower tuition made up for it. So in lifetime wealth sense, they were as successful. In the earning potential sense, they were less successful.

Also, I would imagine that the gap is much more stark when measuring success in academia. The prestige of the schools you attend strongly affects where you can hope to get tenure-track positions.

ZH said...

I don't know about the studies tom mentioned above, but I do know that I have seen studies show that your graduate school matters a lot more than your undergraduate school. And, although I have no empirical evidence for this, I seem to notice that top grad schools care less about where you went for undergrad, so long as it not a garbage school (although I would assume there is a bias toward large research universities, whether it is Harvard or "merely" some large State U), than they do your grades, courses, level of difficulty of courses and GRE, GMAT, LSAT, scores.

Independent Accountant said...

That's exactly what they do. Decades ago I suggested the Ivy League schools sell "certificates of admission", which would enable a holder to claim "alumni" status. How much to charge, auction them off. I bet Harvard could get $100,000 each. If say, Harvard claims 80% of its applicants could do Harvard work, offer each such rejected applicant the opportunity to buy s certificate. Why not?