Amherst to stop offering legacy preference in admissions

Maybe Stanford is just being honest that many of the non-admit EA applicants have basically no chance of admission, so it delivers the rejections early, while deferring only the few non-admits who are in the borderline zone. Other colleges may defer even no-chance EA applicants out of politeness or whatever, even though they are not going to be admitted.

2 Likes

Applicants or admits? At a college with significant legacy preference, legacy applicants may be stronger than unhooked applicants, but legacy admits may be weaker than unhooked applicants.

The vast majority of legacies are White. A comparison of the AI stat distribution for White admits is below. The decile refers to the full applicant pool. For example, 27% of White non-ALDC admits were in the top decile of the full applicant pool compared to 14% of White LDC admits . The non-ALDC admits are clearly skewed towards having a very high Academic Index. In contrast LDC admits have a more varied Academic index.

Top Decile ā€“ 27% of non-ALDC admits, 14% of LDC admits
2nd Decile ā€“ 22% of non-ALDC admits, 18.5% of LDC admits
3rd Decile ā€“ 17% of non-ALDC admits, 16% of LDC admits
4th Decile ā€“ 12% of non-ALDC admits, 13% of LDC admits
5th Decile ā€“ 9% of non-ALDC admits, 12% of LDC admits
6th Decile ā€“ 7% of non-ALDC admits, 10% of LDC admits
7th Decile ā€“ 4% of non-ALDC admits, 8% of LDC admits
8th Decile ā€“ 1% of non-ALDC admits, 5% of LDC admits
9th Decile ā€“ ~0% of non-ALDC admits, 3% of LDC admits
Bottom Decile ā€“ 0% of non-ALDC admits, 1% of LDC admits

As BKSquared noted, Hispanic admits have a different pattern. This largely relates to there being a far larger difference between Hispanic non-ALDC applicants and LDC applicants in academic ratings than occurs for other races. So instead of a weaker academic rating, Hispanic LDC admits have weaker EC, personal, and teacher LOR on average compared to Hispanic non-ALDC admits. The blunting of multiple hooks also has an effect. The analysis suggests the combination of URM + legacy is stronger than either alone, but not as strong as average URM boost + average legacy boost in isolation. Rather than make URM + legacy + REA + ā€¦ .a super hook , the effect gets blunted.

White LDC admits also tend to have weaker EC, personal, teacher LOR, ā€¦compared to non-ALDC admits. Scores are often a strong point of the application for legacies, which compose the bulk of the AI calculation, so the legacy boost often helps more with non-score portions of application for which a detailed decile distribution is not available.

1 Like

Essentially, if you are unhooked at Harvard you have virtually no chance of admittance unless you happen to fall into the highest academic rating category. Not sure this is a surprise and Iā€™d be shocked if it is much different at other Ivy League schools.

3 Likes

I was referring to admitted students. For legacy applicants, the amount of skew to the right may not be much different. The impact of legacy boost is much greater on those legacy applicants who are academically relatively weak (i.e. those in the left tail of the distribution), as @Data10 has shown. Therefore, I expect the fatness of the left tail in the distribution of legacy admits to be more pronounced than that in the distribution of legacy applicants, relative to the respective distributions of their unhooked pools.

I would add that even if an un hooked student falls into the highest academic rating category, their chances of admission are still extremely small. The are a lot of applications for very, very few spots at Ivy and similar schools, so even a stellar record is not really enough.

1 Like

Few applicants or admits have the highest academic rating. That category usually involves faculty review of academic works. It may also include persons who have national/international level achievements in certain academic contests. The bulk of unhooked admits instead have a 2 academic rating. The admit rate for 2 academic rating kids is relatively low (10% at time of lawsuit sample) because a 2 academic rating alone is insufficient. Unhooked admits also generally do well in the non-academic criteria of the application, such as ECs, personal, LORs, etc. Some specific numbers are below. The admit rate jumped to 43% for applicants that had three 2 ratings such as 2 academic, 2 EC, and 2 personal. I expect this is the most common rating combination for unhooked admits.

Ratings for Academic, EC, Personal, and Athletic
1 EC Rating ā€“ 48% admit rate, 220 admits
1 Academic Rating ā€“ 68% admit rate, 450 admits
1 Athletic Rating (recruited athlete) ā€“ 88% admit rate, 1200 admits
Three 2 ratings and one 3+ rating ā€“ 43% admit rate, 4000 admits

Not only Lā€™s far outnumber DCā€™s, but the effects of Dā€™s and Cā€™s tend to offset each other, as Dā€™s are likely weaker and Cā€™s stronger than Lā€™s academically.

Actually, 66% of non ALDC admits (of all non-ALDC admits) were in the first 3 deciles, leaving a third of non-ALDC admits in the remaining 7 deciles. 21% of non ALDC admits were in the 5th decile or lower if we look at Dataā€™s table at #243. That table however does not account for the URM lift as non ALDCs and LDCā€™s are taken as a whole. Note that these deciles relate to the admittees Academic Index which is derived from a formula comprised of GPA and test scores that is used for purposes of athletic recruiting requirements. While correlated to the Academic ratings, more goes into the ratings and the number of students rated in each group is not equal as Data also points out.

Trying to triangulate this from another angle using the table I looked at which does distinguish between races so that we look at qualifications of ORMs (Whites and Asians), we do see 11% of White non-ALDCā€™s in the admitted group with a 3 rating of all admitted White non-ALDCs , and 5.6% of Asian non-ALDCs. While these are unfavorable numbers for Academic 3ā€™s of ORMā€™s, itā€™s not as if they have 0 chance.

To close the loop where we might speculate that only a small number of non ALDCā€™s are admitted, I looked at Table 4 of the Arcidiacono report which provides gross numbers of admits by race and ALDC status. The LDC % and non ALDC data are my calculations using the Total Admits, LDC Admits an Athletic Admits in Table 4:

White African American Hispanic Asian American
Total Admits 4,993 1,392 1,283 2,443
LDC 1,362 81 122 270
LDC% 27% 6% 10% 11%
Athletes 817 124 54 101
NonALDC 2,814 1,187 1,107 2,072
Non ALDC % 56.4% 85.3% 86.3% 84.8%

So from here, we can see that LDCā€™s still are a minority of admits, with Whites having a much higher percentage of LDCā€™s relative to other groups.

Most people assume ALDC had it easy. By my observation that is far from the truth.

A, If you can make it thru 1%-3% admission to be a recruited athlete the 89% success rate doesnā€™t seem so high. Just being an outstanding athlete is no guaranteeā€”a case in point Nathan Chen was rejected at H because he wasnā€™t among 1-3% recruited.

L, can you imagine being told since you were six that you were favored to go to Harvard? How many times we all feel sorry for some kid who asked something like ā€œI am an eighth graderā€¦ what should I do to get into Harvard?..ā€

D, if you have $12m to spare you can find an intermediary to get a pre-read on your kidā€™s file. If they think the kid had the chops to survive then congratulations, you have the privilege to lose that $12m.

C, many schools have the practice of giving the children of faculty a boost. What if the school uses it as part of the faculty recruiting, which has <1% acceptance too.

2 Likes

Just to share a story to back this up. Iā€™m a Harvard grad and have given very little to Harvard over the years. Most years I gave nothing, and some years I gave $25-100. I did give $500 the year of my 25th year reunion, a big deal for Harvard. (For the uninitiated, $500 is chicken feed to Harvard). This was also the year that my son applied to Harvard.

My son had a very strong application and applied early action. He had straight As at a very challenging magnet science school, he took very advanced math and science from a young age, he is an accomplished pianist, he had several leadership roles, plenty of evidence of being a good person, high test scores, and at least one physics teacher who wrote a letter saying that he was In the top 99.9th percentile among students this teacher had seen. He was deferred from Harvard. He applied regular decision to many other great schools, and was accepted at Princeton, Yale, Duke, UPenn, Brown, UCLA, Berkeley. He was rejected from Stanford. At RD, Harvard waitlisted him.

I donā€™t pretend that anyone ā€œdeservesā€ to be admitted anywhere, so only an evaluation here. His acceptances tell me he was the quality of student Harvard seeks (as Iā€™m sure many applicants are). If alumni status by itself were valuable to Harvard, he would have been admitted. If an alumni boost were in the making (independent of lack of donations), it would have gotten him in. It is clearly pretty unimportant to Harvard in our case.

My son was not interested in trying to convince Harvard, so who knows if he would have gotten off the waitlist had he tried.

3 Likes

The year my youngest kid applied to Princeton, my spouse and I increased our annual donation to the university from $50-75 to $100. Our kid was accepted, and we liked to joke that it was the extra $25 that did it.

5 Likes

This matches the data indicating that the one who mostly benefit from their legacy status are the wealthiest alumni families.

Harvard publishes their data on their incoming class, including the income classes of their admitted legacies. Until their Class of 2023, when they added ā€œprefer not tellā€ to the choices for incoming legacies, almost 70% of the legacies were from families making more than $250,000 a year, or the top 5% (in 2019), while 46.4% come from families making more than $500,00 a year, or the top 1%. Only 12% were from families that make less than $125,000, so maybe 10% from the bottom 80% by income. Looking at the Chetty analysis, about 42% of Harvard graduates end up in the top 5%, 21% end up in the top 1%, and 35% end up in the bottom 80%.

So the wealthiest legacies are about 6X as likely to attend Harvard as the middle class legacies, and, even assuming that at least half of the poorest alumni improved their SES, their kids are being accepted at a small fraction of the rate f the wealthiest legacies

So yes, it definitely seems that legacy status at Harvard is only really meaningful if the family is very wealthy, and almost certainly, are serious donors.

As an aside, it also demonstrates that claims that Harvard prefers legacies because these kids are somehow more prepared for Harvard.

As a second aside - why are Harvard alumni who arenā€™t super wealthy donating their money top Harvard? I mean, they have these vast resources of wealth, and hundreds of super wealthy donors who want their names on buildings and their kids at Harvard. While that $25 may be very little to a reasonably well off family, Harvard needs that money a LOT less than you do. Itā€™s as though you are giving Bill Gates $25 to help him buy coffee.

Harvard accepting that donation is as though Gates takes your $25 and says ā€œthank you, every little bit helpsā€.

1 Like

You werenā€™t addressing me, but Iā€™ll answer your question about why non-wealthy alumni would contribute to their uber-wealthy alma mater: I contribute to Princeton because they gave me a ton of financial aid when I was a student there. I am grateful for the opportunity to have attended, and my small contribution each year feels like repaying a debt, even if the practical benefits to the university are negligible.

Of course, one could argue that my meager donations would be better contributed to a local food bank, or to a nonprofit that makes microloans to female entrepreneurs in India, or to a medical NGO that repairs fistulas for people in low-income countries. And that is a rational argument that I completely agree with. But for me, donating to my alma mater is an emotionally satisfying act, and an opportunity to (forgive my use of this cliche) pay it forward.

5 Likes

ā€œNever heard of a legacy getting the same pre-read as an athleteā€. My son is at Amherst (not a legacy) and for three years he had roommates that were legacies. When it came time for the younger sibling of one of the roommates to apply he was given a pre-read by admissions that would come back as ā€œlikelyā€ or ā€œunlikelyā€. The pre-read came back as ā€œunlikelyā€ and he is now attending another NESCAC. Not sure if this is offered to all legacies or if this was an isolated case.
P.S. I am always surprised by the number of posters on this site who believe that any high stat student who EDā€™s to a ā€œless selectiveā€ school is doing it as a ā€œstrategyā€. What about they just preferred the school??

That makes sense, though, as you said, it could probably be used better elsewhere, but, as you write:

Feelings are not logical, which is absolutely OK.

Personally, I would contribute to a program sponsored by a college, or associated with the college, especially if it was receiving less money in donations (I actually do, for my kidā€™s college). But that is based on MY feelings, not on any better logic than yours.

1 Like

I didnā€™t go to Harvard, but I donate to a specific program at Brown because I participated in it as an undergrad and it meaningfully changed my life. I am not a big donor; I try to donate as much as I can to the other, more ā€œneedyā€ type causes that Tiger in Winter discussed; I would not continue to support Brown if it meant that I couldnā€™t donate to these other causes at all. But since Iā€™m lucky enough to do it-- I do. I have friends who do not support their own alma mater (ā€œthey donā€™t need my moneyā€ or ā€œI hated every minute I spent there, why would I support them?ā€ or ā€œIt took me ten years to pay off my loans- they should be donating to MEā€.) and I am grateful that I get to feel differently about my college years.

I like reading Nick Kristoffā€™s analysis of how to donate to do the most good in the world; I follow various other columnists and scholars on how to give and I follow some of their advice but not all. (I am reluctant to donate to some fantastic humanitarian causes overseas because the level of corruption and kleptocracy in some of these countries means that 90 cents of every dollar you give is getting siphoned off by some oligarch who lives on the Champs Elysee and hasnā€™t been ā€œhomeā€ in ten years). But giving to a specific program which set me off on my lifeā€™s course ā€œfeelsā€ like the right thing to do, whether or not itā€™s a logical way to look at money- which after all, is fungible.

2 Likes

I forgot to ask in my previous post - can you earmark your donation to Princeton so that it goes to financial aid?

1 Like

I donā€™t know, but thatā€™s an excellent question and one that I will investigate.

1 Like

A friend of mine once explained that she gave a modest yearly donation to Harvard because she thought if she failed to do so, it might harm her child because she would stand out negatively. So the yearly $100 donation was a way to stay off some imaginary ā€œbad alumniā€ list.