Monday, January 11, 2010


In today's Boston Globe, Neal Gabler takes on the issue of "meritocracy" in college admissions. His thesis — one that's been made before — is that the admissions process at elite schools isn't meritocratic at all, but based more on wealth and privilege. It's not exactly a courageous, or new, position to take — tearing down our great academic institutions is a great pastime for populists on the right, who resent East Coast intellectuals, and the left, who make their living telling people that the system is designed by the rich to keep the poor in their place.

Gabler says the admissions process isn't meritocratic. He keeps using that word, but I do not think it means what he thinks it means. Let's go with Merriam-Webster on this one:
Main Entry: mer·i·toc·ra·cy
Pronunciation: \ˌmer-ə-ˈtä-krə-sē\
Function: noun
Inflected Form(s): plural mer·i·toc·ra·cies
Etymology: 1merit + -o- + -cracy
Date: 1958
1 : a system in which the talented are chosen and moved ahead on the basis of their achievement
2 : leadership selected on the basis of intellectual criteria
So let's look at Gabler's analysis:

They [high school seniors] know that America, for all its professions of meritocracy, is a virtual oligarchy where the graduates of the Ivies and the other best schools enjoy tremendous advantages in the job market.

According to the above definition, in a meritocracy people move ahead based on their accomplishments. This opening salvo is a poor attempt at proof by assumption. Moving people ahead based on their academic pedigree is only un-meritocratic if you believe earning a degree from MIT or Stanford is no accomplishment at all. Gabler himself doesn't even appear to accept that premise, so moving people ahead based on graduating from the best schools is actually evidence in support of a meritocratic society in the post-college world.

So let's look at the admissions process itself.

Daniel Golden demonstrated in a Pulitzer Prize-winning series for the Wall Street Journal and then in his book, The Price of Admission,’’ the so-called “best’’ schools give heavy preferences to the wealthy; as many as one-third of admissions, he writes, are flagged for special treatment at the elite universities, one-half at the elite liberal arts colleges, and the number of open spaces for the non-privileged is reduced accordingly. As Golden puts it, the privileged take so many spots that the “admissions odds against middle-class and working-class students with outstanding records are even longer than the colleges acknowledge.’’

Some of this is undeniably true, and obviously un-meritocratic. But the development process at Duke analyzed by Goldman represented 3% to 5% of the enrolled student body. It's possible that Goldman selected a year and school that helped make his point, so that the figures might be even less at most schools. Legacies, however, make up a far bigger portion of the student body at the top schools. From Golden:

Sons and daughters of graduates make up 10% to 15% of students at most Ivy League schools and enjoy sharply higher rates of acceptance. Harvard accepts 40% of legacy applicants, compared with an 11% overall acceptance rate. Princeton took 35% of alumni children who applied last year, and 11% of overall applicants. The University of Pennsylvania accepts 41% of legacy applicants, compared with 21% overall.
While I'd personally like to see legacy criteria dropped from my alma mater's admission process (sorry, Junior), it's unclear how much of an advantage it really is. The naive analysis of the above numbers suggests it's huge, but without any explicit advantage given to legacies, we'd expect them to have higher admission rates anyway. Why? Well, lots of reasons. If Mommy was smart (or had some other quality that got her admitted), some of those genes might have passed on to Junior; Harvard grad Dad probably corrected Junior's English and helped him with his homework; rich parents got Junior tutors, paid for piano lessons, and put him in private school; finally, when Joe the Plumber's son came home with Bs he got a pat on the back, but Yale Junior got sent to his room without supper.

So it's hard to tease out the explicit advantage a legacy gets over a similarly qualified non-legacy from the numbers above. But we have this, again from Golden, about the University of Michigan's admissions process:

The University of Michigan has a 150-point "Selection Index" for undergraduates, with 100 points usually enough to get in. The university awards a four-point bonus to children and stepchildren of alumni, or one point to grandchildren, spouses or siblings of alumni.

Michigan also gives an automatic 20-point bonus to blacks, Hispanics and native Americans (though not to Asian-Americans).

So yes, there's an explicit legacy advantage, but it's one-fifth the advantage given to under-represented racial minorities. While I'd personally like to see it go away altogether, it's a very small advantage compared with other selection criteria. Certainly nowhere near enough to claim the admissions process is devoid of meritocracy. As far as privilege itself, Gabler points out:
They are aiming instead for a diverse student body: an exceptional athlete, an exceptional musician, an exceptional scientist, an exceptional poet. Except that exceptionality, as most parents can attest, doesn’t come cheap. Athletes require coaching and often traveling teams; musicians require lessons and instruments; scientists require labs and internships; poets require classes and opportunities for publication.
Meritocracy means that the selection should be made based on "achievement." It does not say that said achievement can't be aided by wealth, genes, or other opportunity and advantages. That is, to be meritocratic the schools should accept the best violinist. Period. Not the richest , not the poorest, but the best — however she got to be the best. Those tutors and instructors, as noted above, might have been awarded to young, richer children based on wealth — but the decision to admit based on accomplishment conforms to the very ideal of meritocracy.

One more note from Gabler about poorer applicants.
The prognosis is equally poor for economically disadvantaged students, whether black or white. According to Golden, economic diversity counts the least in admission considerations, and only 3-to-11 percent of admittees come from the lowest economic quartile.
See the above "legacy" analysis to see why this might happen naturally, without any explicit preferences. You wouldn't expect 25% in a meritocratic admissions process — or even in an entirely meritocratic society. Genes, environment, expectations, and other opportunities would keep this number well below even if the college admissions process were 100% "meritocratic." In fact, 11% from the lowest quartile suggests a pretty healthy opportunity for social mobility and refutes Gabler's following statement:
Indeed, the system exists not to provide social mobility but to prevent it and to perpetuate the prevailing social order.
Gabler also laments that SAT scores correlate with wealth, but the same argument applies for the most part. Studies suggest that paying for coaching helps a little bit (and I'd personally like to see that entire industry go away), but for the most part you'd expect richer children to be better at math and have stronger vocabularies. You want to make the process more meritocratic by ignoring verbal and math skills in the admissions process? That seems silly. How exactly would removing the part of the application process most associated with intellect promote "leadership selected on the basis of intellectual criteria?"
A good compromise might be to recognize that a poor person who got a 1200 despite their background has achieved something more impressive than a rich kid who got the same. Some colleges claim they do this, but its hard to say for sure how much this is really done.

[Interesting side note, in his commentary in the Economist, Lexington notes that more emphasis should be placed on test scores to make the process more meritocratic.]

Yes, we'd have an even more meritocratic society if tutors, trainers, educated parents, and private schooling were allocated to the most promising students instead of the richest. That seems true enough, but would require an awful lot of socialism and government interference in raising our children. But the lack of meritocracy is in the years leading up to college admissions — the evidence is lacking that the admissions process itself is significantly un-meritocratic. Could it — should it — be more meritocatic? Yes, I think. But the numbers actually suggest that our top academic institutions are doing a good job of accepting the best and allowing for advancement.

Final note: while tangential to the rest of this post, as often occurs when we try to discuss race in America, Gabler doesn't seem to understand what "race" actually means.
Nor does diversity extend to racial composition. Of course every college boasts about its efforts to enroll a more racially diverse student body. But here are the facts: A New York Times article in 2004 revealed that Harvard’s incoming freshman class was 9 percent black, but between one-half and two-thirds of those black students were actually West Indian or African immigrants or the children of immigrants, and many others were biracial. In short, they weren’t African-American.
Hmm, so Harvard's class had about the same mix of blacks as the US population as a whole, but this doesn't count as racial diversity because many of those blacks are from other countries or are the children of immigrants? Racial diversity is about race, not country of residence. If more of those blacks were from the US, the student body wouldn't be any more racially diverse.


Vercingetorix said...

I think you're right that this has something to do with confusion about the word meritocracy. Most definitions are similar to Merriam-Webster's and use terms like ability, achievement, talent, etc. This is, I believe, what it means in standard usage.

Unfortunately, the waters get muddied when people relate it to the word "merit" since that word is fairly vague and isn't exclusively tied to achievement and ability. Rather it often carries connotations of deserts – someone with merit is deserving.

This is much more appealing to some than the harder, outcomes-based view. In this mindset, meritocracy should allocate positions of leadership and status based on who works hardest, who improves the most, who overcame the most obstacles – not who plays the best, knows the most, or performs at the highest level.

Until you figure out which side of this is/ought divide the person you're talking to is on, it can be dangerous to bring up the idea of meritocracy at all.

Mithridates said...

"Who works hardest, who improves the most, who overcame the most obstacles" might get admitted on a meritocratic basis if you consider those things to be achievements (which I personally do).

I would say it's a greater "achievement" to get a 3.8 GPA while working 30 hours a week after school to help the family pay the mortgage than it is to get a 3.9 without the extra burden. I would say it's a greater "achievement" to get a 780 on the Verbal SAT if you're the child of uneducated parents and went to a bad public school than to get an 790 as the son of Harvard alums.

I would also say that there's evidence that the former student in each case has more "talent". I'm not sure exactly how much those are worth in terms of GPA and SAT points, but they're worth something in my book.

I think both definitions of "meritocracy" would reward the former students. So I'd caveat my above point about the violinist. You might take a slightly less heralded, but self-taught violinist, for example, and still call yourself a "meritocrat", on the basis that this person achieved more and probably has more talent. But you certainly wouldn't ignore violin-playing itself as inherently un-meritocratic.

Vercingetorix said...

Oh, I agree completely. That's why I think socioeconomic preferences are often okay – equivalent outcomes with a larger handicap (whether it's poverty, a language barrier, having a single parent, or something else) is definitely an indicator of having more talent or ability.

I was simply saying that some people seem to think that merit is the same as deserts. And that "undeserved talent" (often conflated with anything heritable) shouldn't factor in.

Phutatorius said...

I think one useful way to look at this — meritocracy aside — is to consider the two very different qualifications of aptitude and achievement. Now, of course the two are related: aptitude enables achievement, and achievement (often) reflects aptitude. The way I see it, the availability of educational benefits ought to be keyed more to aptitude early on, but as students advance through the system, we ought to look more to achievement. That is, there ought to be a sliding scale.

The problem isn't so much on the back end — colleges are right to emphasize achievement, while only occasionally making allowances for/taking chances on applicants with demonstrated aptitude, but little opportunity to build on that aptitude along the way. The problem is on the front end: identifying people in every social stratum with extraordinary talents and extending resources appropriate to their needs.

Mithridates said...

And then the NYT publishes this unhelpful bit of nonsense.

Post a Comment