Comment by virtuous_signal | Hacker News Reader

Comment by virtuous_signal | Hacker News Reader

virtuous_signal

>The year I arrived, Hopkins had more legacy students in its freshman class (12.5 percent) than students who were eligible for Pell Grants (9 percent). Now those numbers are reversed—3.5 percent of students in this year’s freshman class have a legacy connection to the university, and 19.1 percent are Pell-eligible

This change happened in five years. Such a policy change would be extremely popular. For proponents of race-based affirmative action, this would serve as somewhat of a proxy due to the correlation between race and class. For opponents, removing legacy preferences makes admissions more meritocratic.

The only groups with something to lose are (i) mediocre children of alumni, and (ii) the university's endowments if alumni decrease their donations. I don't believe (i) is a strong factor in a university's calculations, which leaves (ii). I believe universities are happy to see the public fighting about race-based affirmative action: indeed, dividing and pitting the low- and middle-income groups against each other over their limited subset of the pie, means they will not be working together to increase their overall share of the pie.

(I think this issue could and should be considered orthogonal to race-based affirmative action. Race-based AA was never put on a stable enough legal footing and will continue to bounce back and forth in the courts until it's resolved, while this could be resolved quickly and uncontroversially.)

Another group that stands to lose are those students who were admitted on merit and no longer have access to “mediocre” students with good connections.

That's true, and in addition to the claim that legacy admits make the entire college experience more affordable for everyone else, is probably the strongest objection.

I would venture the following: if we agree on the norm that "power should stop being transferred dynastically or through nepotism in a modern civilized society", and abolishing legacy preferences bring us closer to this norm, then the negative effects will be transient, and worth it.

That's never going to happen unless you truly change human nature. We are tribal and protect our families. Nepotism is a family member making life better for other members of the family. I'm not sure we will still be a society if we eliminate family bonds.

> I'm not sure we will still be a society if we eliminate family bonds.

There's a significant difference between not having nepotism and eliminating family bonds.

> We are tribal and protect our families.

Many of virtuous_signal's arguments boiled down to the observation that protecting your family via nepotism or dynastic preference isn't even necessarily good for the effected family member in the long run. I think that poster makes a good argument that dynastic preference in university admissions is a net negative for its beneficiaries.

The negatives of nepotism also show up in business, where, with rare exception, nepotism tends to genereate a huge drain on both productivity and external respect.

I always hesitate with these sorts of effects calculations. Depending on how long or how transient these negative effects are, a lot of folks could get screwed through no fault of their own.

This relies on a correlation between “legacy” and “good connections”. But if it’s about good connections, why not just have a policy like: “if you’re sufficiently good you get in and if you don’t have a ton of money we can provide scholarships to make things more affordable or free; and if you’re ok but rich, you can pay 2x the usual rate for admission up to some quota of these admissions”

You could even replace “rich” with “rich and well-connected” though it may be hard to determine the second part.

This seems like a more honest system and it doesn’t strike me as obviously unfair or unreasonable to have. Someone paying double fees isn’t taking someone else’s place so much as providing for someone else’s place.

That's the current system. The rich people pay the MRSP (about 70K/year), and the poor pays half as much on average due to various scholarships / grants / aid.

Your way is just that in terms of outcome, except it's more antagonizing to the rich.

They pay more than that. Legacy enrolment is one of the key ways that colleges grow their endowments, which are (amongst other things) used for scholarships.

> You could even replace “rich” with “rich and well-connected” though it may be hard to determine the second part.

Connections are traditionally determined through the very old method of asking for letters of recommendation. US military academies still work this way; you can't be admitted without a recommendation from one of the three (3) legislators representing you in the federal government, or the vice president. The method is cheap and accurate; you test someone's connections by having them exercise those connections.

For some reason universities still ask for letters of recommendation, but they've shifted to asking for letters from random teachers. This completely defeats the purpose; a recommendation from someone who is unknown to you [where "you" are the admissions officer] and who holds no power of any kind is obviously worthless.

isnt this what happens already?

Maybe this is an added benefit since those well connected kids will go somewhere else and increase the access to those networks.

The good connections' access to outstanding smart college kids would also be threatened. Which means these connections will have to find other, hopefully less dynasty-reinforcing ways to recruit.

Love this point. Not sure if it can be called cynical or not, but a really smart kid would definitely miss out on a networking opportunity with the parents of a legacy friend. Think Bighead from Silicon Valley.

I once heard the dean of admissions at Columbia claim that they could get rid of the entire incoming class and replace them from kids in the applicant pool without suffering a decline in quality. I'm guessing this is mostly true, save for a few truly impressive students at the tippy top. Given the low number of spots compared to applicants, it's likely that you can fill an entire class with children of alumni without much of a dip in quality. But it sure is unfair to people who don't have parents who went to elite schools.

I think it's fair to assume that the "quality" of an applicant is some combination of innate ability (normally distributed) and the amount of resources that have already been invested in that student (exponentially distributed). this implies that the differences between students increase at the very high percentiles.

if the dean's claim is true, they are basically admitting that their admissions criteria sucks.

I think the dean would freely admit this. If elite colleges had a perfect admissions process, they would never fail to admit people who go on to accomplish great things. But they do. All the time.

The issue is that they have a limited set of things to go on. And high school performance, test scores and extracurriculars just aren't a great predictor of success, but they certainly better than just picking students randomly.

Another way to look at it: The number of legacy students has fallen almost to zero, so now Johns Hopkins announces they are getting rid of legacy students, not before.

Yes this seems more like an attempt to capitalize on a negative trend (legacies no longer applying and matriculating) to achieve positive brand association (Hopkins is now leading the way towards a meritocratic future). I'd be curious what led to the fall in legacy students in the first place.

This totally might be the case. However it's still a good change and might lead to other colleges following suit. I'm reminded of a recent trend among the Ivies where Princeton was the first to eliminate loans from its financial aid packages[1] and instead offer only grants. Of course they have the largest endowment per student in the country, so the amount they collected from low-income students via loans was probably trivial. However it did seem to kick off a trend to the extent that all of the ivies[2] (dartmouth and cornell being more stringent) have replaced loans with grants for low-income students.

[1] https://paw.princeton.edu/article/no-loan-pledge-decade-late... [2] https://blog.collegevine.com/does-the-ivy-league-offer-schol...

> (ii) the university's endowments if alumni decrease their donations.

The tax law says you aren't supposed to be able to deduct contributions to charitable organizations if you get something in return.

Edit: Point being that this system is illegitimate in the first place, not that it doesn't happen.

I'm not sure that's true. There are rules about explicit quid-pro-quos, and although they're not entirely banned (if you donate $250 for a concert ticket worth $50 you can deduct the $200) there are rules that clearly aren't being followed in this case. But tax law doesn't (and probably couldn't) attempt to restrict vague returns like "the recipient sees you and your family in a more favorable light".

Explicit quid pro quos are banned, your concert ticket example showing as such. And the fact that we can’t define or prove every situation where there is quid pro quo is a weakness that is exploited by many rich people, also shown by your example (is that ticket worth $50?). Doesn’t make it the right thing to do, and is not optimal for society in the long term.

You are supposed to reduce the value of the deduction claimed by the fair market value of anything in return. So if they send you a $10 sticker or invite you to a banquet with $500 in food, you can't deduct that part. The question is, how exactly can one determine the FMV of a boost in admission consideration (not even just the admission itself, but an un-quantifiable additional favorable factor in review)?

Question would be moot if we simple got rid of deductions. Only reason to have them is to enable this type of gaming of the system.

Legacies are considered more favorably for admissions regardless of how much (or even whether) they donate, right? Or have I misunderstood the system?

That is, the only thing that donations get you is the vague hope that your university will never make the decision Johns Hopkins just made, and it's a vague prisoner's dilemma - as long as enough alumni donate, legacy admissions stay around for all alumni.

No -- I'm going off news that came out in light of the lawsuit against Harvard recently, but I have no reason to believe other elite universities handle it differently. These things are shrouded in secrecy for the precise reason that it is massively unpopular, but at least in this case we got an inside look because of discovery for the lawsuit.

Harvard creates a special "Dean's Interest List" (https://www.thecrimson.com/article/2018/10/18/day-three-harv...) which includes celebrities, "top donors", and certain athletes. The admissions rate restricted to this list was 42.2% in 6 recent years.

my guess is that legacies are just a good proxy for donations in general. if someone encourages their child to apply to the same college they attended, they probably feel more connected to the institution than the average graduate (and have donated more accordingly). legacy admissions policies may also encourage people to donate more because they know their children will have a better chance of benefiting from the investment. it's likely a decent proxy for wealth, too; someone who graduated from an elite institution is probably well above the median income.

so basically it correlates with a lot of things that the institution wants, but doesn't want to state as an explicit policy.

the university's endowments if alumni decrease their donations.

Most universities where legacy admissions are a thing have self-perpetuating endowments by now. Harvard could easily fund itself forever without any donations or even tuition fees at this point, just off the interest and trading profits.

It would be interesting to study the effect of this on donations and the endowment. There should be some measurable effect. Additionally, if there is an effect on donations and the endowment, it would be interesting to study the effect that that change has on endowment payouts. Does lower donations, due to lower legacy admissions result in less funds for University operations?

I haven't heard of studies on this but I would say the effect is potentially huge. For instance, just going by numbers of legacy admits at Harvard: www.cnbc.com/2019/04/07/harvards-freshman-class-is-more-than-one-third-legacy.html

And any natural experiment that could have been done due to JHU's change was probably messed up by the largest donation to a university, ever: www.nytimes.com/2018/11/18/us/michael-bloomberg-johns-hopkins-donation.html

And given the size of Bloomberg's donations during his lifetime, I suspect the university also has assurances for more gifts in his legacy allowing to close the gap from former alumni no longer donating.

My single datum indicates it results in a 100% drop in donations from families with a rejected student.

There is a third group: (iii) individuals who hope that some sort of chance nepotism will benefit them someday and don’t want society to move away from it before it does.

If someone has to decide openly between allowing or denying an instance nepotism, based on a secretly-held belief that it’s acceptable because it could benefit them, then they are able to take the stance that “nepotism is bad, but anti-nepotism has gone too far” in order to extend the life of the benefit (and their chances of benefiting from it).

"mediocre children of alumni" This is an unfair description of the advantage of legacy at an elite school. It's typically an edge, like playing a sport or belonging to an under-represented demographic.

I'm aware that it's typically an edge; what I mean by "mediocre" is not a personal description, but rather the state of being quantitatively less meritorious than those students who gained admission without the legacy edge.

The non-mediocre children of alumni can still get in. "Mediocre" is being used here in a relative sense.

If the impact of legacy is small - equivalent to gaining 0.01 points of GPA - then if they need legacy to be admitted, they're only 0.01 points above being rejected - one of the least qualified students at the school.

And if the impact is large - equivalent to 1 point of GPA - then there are legacy admissions a full GPA point worse than the least capable of their non-legacy peers, which is even worse.

I agree mediocre is inaccurate, because mediocre means half way to the top, and students who need legacy admissions are at the very bottom of the class.

Reply on news.ycombinator.com