Colleges have been under pressure to admit needier kids. It’s backfiring.

Applying to College
41

MIT is just as holisic as Hrvard or others. People seem to think it’s more stats based (and throw in some accompished research,) but look deeper.

Holistic is not an excuse. It’s a challenge. Despite top high school performance (c’mon, between 14 and 17?,) many kids can’t fill out the app in a compelling and meaningful way. They don’t understand what matters more and what less. I say, that’s not because the colleges are hiding something. It’s more that hs kids are used to the hs scheme, where being a BMOC makes one ‘desirable’ and oodles of respect go to the top %. Look at CC. They talk mostly of rank, weighted gpa, titles in clubs, how much money that fundraiser party raised. How they started a blog or proclaim passions unreached for. Far fewer get the “more.”

And then, adult posters who tell kid posters just to do what they like, that some small number of ECs sounds like plenty, that unilateral is great…and that planning more broadly is padding.

And then point fingers at something like legacy. Ime, most kids can’t even do a decent "Why Us, even when the main bullets of a college are obvious. By most, I personally mean more than 80%.

But so many legacies have the advantage of knowing more about a college than USNews, “You have my major,” or what jobs it may lead to. They have that benefit.

It’s nuts to say legacy is some conspiracy. They get read same as every other app and when they fail, they fail. The real issue, to me, is that with all the assumptions swirling around, plus the palpable distrust so many adult posters have of top college processes, these top kids don’t get as far as properly matching themselves. And filling in the right blanks in the right ways. Then, they’re shocked by a reject. “But I had higher grades than she did!!!” So missing the point.

Nor is every poor kid a compelling candidate. But in some regions, the mentoring of the best of these is great. The ways they get out and get involved can be more meaningful. And more.

But then CC goes right back into the rabbit hole of assuming this is a minority tip. Assuming that poor kid can’t possibly be doing anythig more meaningful than your kid.

And it swirls.

42

Exactly my point. What good is it to get accepted when they know you can’t afford it. We all know the two groups are working together when they want to. Just not for every kid, just the chosen ones to make them look good.

43

For an alternate, thoughtful discussion of the limitations of holistic admissions, read Harvard Professor Steven Pinker’s well known article from 5 years or so ago: https://newrepublic.com/article/119321/harvard-ivy-league-should-judge-students-standardized-tests

Concerns about ethics and moral justifications with respect to legacies and development admits are misplaced. @DeepBlue86 basically nails it in post #36 above, but with two corrections. First, Harvard chooses “academic superstars” for far fewer than 20% of its class. Pinker in that article talks of 10% at most and likely substantially lower.

Second, there is a more fundamental reason that MIT does not favor legacy than its reliance on research funding rather than donations. The nature of the curriculum means that relative academic weakness is brutally apparent. So the institution cannot afford to lower its standards for a substantial portion of its admits. It wouldn’t be good for fostering the “giving spirit” anyway. No one wants to feel outclassed, at least not by more than 5, 10 or even 20% at the very most. MIT is therefore not losing much by ignoring legacy status.

44

5 years ago, when Pinker came up, I seem to remember he has no admissions experience.

See, just finding any article that agrees something is wrong is not the key. And to the best of my knowledge, H is taking about 300 superstars, so yes, less than 20%. And those are still subject to holistic scrutiny. They don’t get in “just” because of stats and some research.

“So the institution cannot afford to lower its standards…” See the assumption in there, that legacies are less worthy, as a whole? I have no idea what people base that thought on. MIT isn’t losing out because they are still applying the same high standards, as do other top colleges. Including holistic. It’s not because there’s something inherently wrong with legacies.

Other than some concerns about athlete hooks, there’s little reason to fear someone else got an undeserved tip your kid won’t, IF your kid properly matches, both in the first place and by doing the required due diligence about what the target colleges want.

But now, you can’t just swing back to, “But they don’t tell me.” Certainly kids qualified can look a little deeper than how their high school operates.

45

If they are fully competitive on their own, then why does the college need to have a preference for them?

Especially when they already have additional advantages like this. If the college prefers legacy, then it is like adding privilege to existing advantage.

46

Legacies as a group will be less qualified simply because of the mathematics of the normal distribution and the dramatic ratio on non legacy applicants to legacy applicants. I believe at Harvard it is on the order of greater than 30 to 1, although I may be misremembering that.

While there are reasons to suspect there is some self selection in the legacy applicant pool, there is also reason to suspect that some weaker legacies apply disproportionately because of the knowledge of legacy bump. We see this convincingly in the URM applicant pool: knowledge of the bump leads to a dramatically weaker applicant pool, with less self selection than would be expected. It is not a stretch to imagine similar dynamics at work with regard to legacy.

There is no evidence in the Harvard documents that the legacy applicant pool is stronger than the non legacy applicant pool and actually some evidence that it is weaker on GPA, AP tests, and SAT subject tests. SAT/ACT scores, though, appear to be about the same.

47

Thanks for posting this. Excellent article

48

I agree that MIT doesn’t favor legacies or development cases primarily because of the nature of its curriculum and its mission. Majority of Harvard’s students won’t flourish, or even survive, at MIT. Unlike Harvard, they have fewer alternatives to fall back on if they don’t succeed in a harder major at MIT (Caltech, perhaps an even more extreme example, is another school that doesn’t consider legacy status or development cases for the same reason). However, I don’t agree that schools like Harvard need to favor legacies or development cases in order to attract donations. There’re plenty of donors who want to be associated with Harvard for reasons other than their kids’ college admission.

49

At Yale the ratio seems likely to be of the order of 35:1, on the following basis. When legacies apply, Yale informs the alum that legacy applicants are admitted at a rate of about 20%. Legacies comprise 11% of the class of 2022, as disclosed by Yale (see below link). 1,578 students enrolled, so there would have been 174 enrolled legacies. Assuming that admitted legacies have a 90% yield rate, that implies that 193 were admitted, so (assuming an admit rate of 20%) 964 applied. There were 35,308 applicants, so the ratio is (35,308 - 964) / 964 = 35.6x.

That said, if I’m understanding correctly, I’m not sure it’s prima facie obvious that “legacies as a group will be less qualified simply because of the normal distribution” coupled with the above ratio. I suppose you might argue that if the tail to the left of two standard deviations in a normal distribution contains 2.28% of the population, and the “qualifications” of the legacies and non-legacies were distributed normally, then the number of individuals in the left tail of the non-legacy group is equal to about 80% of the total members of the legacy group, implying that (if they have a 20% admit rate), legacies are being admitted at 4x the expected rate for their level of qualifications.

I think this is questionable for at least two reasons, though. First, we don’t know that the legacy and non-legacy distributions are normal. I’d guess that the legacy distribution is leptokurtic (has fat tails), reflecting a large number of both really bright legacies and failsons. It might also be positively skewed - we don’t know. I have no idea what kind of distribution you’d find for the non-legacy pool, but I would guess it’s closer to a normal one.

That said, though (and this is my second objection), what do “qualifications” mean in this context? We know from other threads that Harvard sees maybe 100-200 “Academic 1s” per year. Those are the true right tail, and I’m certainly prepared to believe that numerically there are many more of them in the non-legacy than the legacy pool. They can only at best comprise a small fraction of the class, though, and after that, the number of individuals in the next level is dramatically larger, enough to fill the class easily. More to the point, the individuals in the next level are much more interchangeable from the point of view of intellectual firepower. Given that, there’s scope for other measures of quality to be taken into account and become more important.

Now, I’m not claiming that every admitted legacy is at that intellectual level or better (and anecdotal evidence certainly suggests otherwise). A good number of them are, though. And, to touch on a point in your other post, I’m going to posit that if MIT were to give a slight legacy tip to candidates at that level, it wouldn’t affect the quality of the class noticeably.

https://admissions.yale.edu/sites/default/files/class_profile_2022.pdf

50

Actually, Caltech and Harvey Mudd (two schools where the minimum curricular rigor in math and science is probably significantly harder than at MIT) do list “relation with alumnus” as “considered”:
https://www.collegedata.com/cs/data/college/college_pg02_tmpl.jhtml?schoolId=706
https://www.collegedata.com/cs/data/college/college_pg02_tmpl.jhtml?schoolId=850

51

That’s incorrect with respect to Caltech. Directly from Caltech President Rosenbaum:

http://www.caltech.edu/content/2018-19-academic-year-welcome

52

We will never prove that a legacy is less or more qualified than a non legacy applicant. But we can prove that with all things being equal the legacy kids get accepted. That is tough for the other kids that have worked extremely hard and it comes down to the parents’ schooling. Funny these same admissions people and these big liberal schools vote to keep the inheritance tax and want to even raise it but they agree that legacy is moral. Interesting.

53

Re: #50

Looks like Caltech’s president needs to inform whoever produces Caltech’s Common Data Set (from which the information in collegedata.com seems to be derived): http://finance.caltech.edu/Resources/cds

54

Thanks, @DeepBlue86 for keeping me on my toes. Your information about Yale appears to be broadly consistent with the Harvard data, but I have only looked into the latter.

There is a lot to unpack in your Post #48 above, and much that I could agree with, so I’ll confine myself to a few points, hopefully avoiding the type of response that could be characterized as “debate.”

Harvard’s non-legacy applicant pool is approximately 34x the legacy applicant pool. This can be verified through Table B.3.2R here (p. 145 in the pdf):
http://samv91khoyt2i553a2t1s05i-wpengine.netdna-ssl.com/wp-content/uploads/2018/06/Doc-415-2-Arcidiacono-Rebuttal-Report.pdf

On academic measures, Harvard roughly bifurcates its applicant pool. Academic 1s, which as you note are very few, and 2s are the top half, with 3s and 4s forming the bottom half.

Harvard reported its legacy admit rate for Academic 1s and 2s as 55%, ~3.7x the admit rate for non legacies with those credentials, which can be seen in Exhibit 4 to this document: http://samv91khoyt2i553a2t1s05i-wpengine.netdna-ssl.com/wp-content/uploads/2018/06/Doc-421-112-May-1-2013-Memorandum.pdf. No tail is fat enough to account for that disparity if academics were the only factor.

However, we know that Harvard cares about much more than academics - which goes to your question of what “qualifications” we (and Harvard) should be looking at. Rather than getting lost in the weeds on the math (apologies for my unnecessary reference to the normal distribution), lets just use common sense. There are >34x more nonlegacy than legacy overall. Granting that legacy applicants are on average stronger than the overall applicant pool, the multiple of nonlegacy applicants to legacy in the top half only of the pool is going to be lower than 34, but still a very large multiple. Legacies constitute less than 3% of applicants, and the 55% admit rate tells us that legacies are not all in the top half academically as the overall legacy admit rate is ~33%.

It is not plausible that on the “winning combinations” of other metrics that Harvard cares about, legacies will just so happen to be overrepresented by a factor of 3.7x. This intuition is bolstered by the fact that legacies by definition cannot enjoy first generation preference, and are underrepresented on other attributes known to be favored by Harvard (primarily economically disadvantaged and URM).

Ultimately, I go with what both Harvard’s and the plaintiff’s experts found: an increase of 11.0x (Harvard analysis) or 10.5x (plaintiff analysis) in the odds of admission over similarly “qualified” nonlegacy applicants. If there are some special qualifications that legacies are bringing to the table - other than their legacy status of course - neither Harvard nor plaintiff’s expert could find them using dozens of controls and plausible explanatory variables.

I understand the reasons for legacy preference, not only for donations, but also with regard to reinforcing institutional continuity and culture and cementing influence. Like you, I do not think this is a moral or ethical issue. However, as for being just as “qualified” as unhooked admits on the other measures Harvard (and by extension Yale too) care about, not a chance.

55

@privatebanker I find it curious why people think you cannot study by yourself and improve your SAT/ACT scores. IMO people who keep on saying that it’s no big deal that kids from “rich” families get higher test scores because they can afford to go to test preps are ignoring several things:

  1. Many kids from rich families don't go to paid test preps. It's not as if studying for SAT/ACT is such a difficult endeavor. You buy a book, do practice tests and figure out your weak areas and practice more.
  2. You do not have to go to paid test preps to get good test scores; you can study by yourself.
  3. Even going to test preps take effort and time.
  4. Some people are using this as an excuse for their not trying hard to increase their test scores.

I think the attitudes of the parents toward education and their support and/or involvement are the biggest determining factors in their kids doing well academically. But I do agree that the richer parents are generally going to be more “involved” and supportive of their kids’ education. To me, that might be the difference, generally speaking.

56

Generally speaking, it’s the birth lottery. Sure, any kid can buy a book and study. The rich kid goes to a great pre school , goes to private school or a top tier public education. This matters. Richer parents tend to be more “involved” because they have the means to be. They’re not worrying about the electric bill or paying rent or working two jobs. Their kids aren’t going to schools where being smart is grounds for bullying or worse. Obviously, I’m generalizing, but it isn’t like these kids are all on the same level when it’s time to start taking these tests.

57

^ not to mention the rich kid isn’t growing up with lead paint in the house, in the school fountain water, dealing with understaffed and underfunded schools but rather is likely to have parents who read to them rather than work a second shift, make sure they eat not only enough but nutritiously…the score/stats discrepancy starts way before test prep time.

58

The solution to the “problem” of poor kids not being prepared for standardized tests is to make the tests more sensitive to innate ability. So much easier than going on about unequal family and educational conditions that will never be equalized anyway. Brains is the one thing money can’t buy.

59

I agree with @privatebanker. I scored terrible on the ACT because I didn’t pay attention in class and never did homework. Basically I was a lazy student with no interests in education. Being middle class didn’t slant the test toward my limited skills. My children love school. My son got the book. He studied it. He took the SAT once and scored 1550. He thought it was easy and was surprised he missed any. The “uber” wealthy are getting into elite schools. What is so valuable about being first generation? We arrived here in 1615. It took us 400 years to have a qualified applicant. What is wrong with the first generation going to their local JC and then furthering their education at the nearest state school? Maybe their children or their children’s children will make it to an elite school. My wife and I both work. We do not receive a free lunch. Our children do not apply to all the schools in the country for free. We could afford three. My son had little chance with 1 in 20 odds. Of course if we weren’t middle class and we were more needy he would have been considered. Its fascinating to me that with stats in the top 1-5% of those accepted, he wouldn’t get accepted at all. Middle class kids cant afford Harvard either. Of course I wasn’t worried as I raised a NMS who was going to a great school on a Full Ride. IMHO, there is nothing valiant about accepting lesser qualified students because they are needy and then turning down excellent middle class candidates who may have earned the right academically to be there. The wealthy kids are going to Harvard. The 1st Generation URM from the inner city is going to Harvard. The brilliant middle class kids…?

60

I wholeheartedly agree with post #55 and 56.

And, the legacy kid likely has been to that campus multiple times, may know professors, and certainly understands the culture of the school in a way that the average applicant may not. I’ve met with many legacy students and the nuance of “why us” is much more refined than most of the non legacy applicants.

It is also almost guaranteed to be a boost to yield for a university to admit a legacy. Especially since most schools give the boost only in ED.

That said, the legacy students that I’ve met over the years are generally extremely well qualified. Those that aren’t don’t tend to be admitted. But even the well qualified students are often denied. We know plenty of friends and fellow volunteering interviewers whose kids were denied but accepted at Princeton, Duke, Yale, etc… Obviously well qualified students but the university didn’t think they were as good of a fit.