What would be the point of RD, then? Wouldn’t you just likely be moving 75-80%, or more, of admissions decisions to EA? For people to act on in the space of six weeks, vs. three months?
Why not just tell everyone to do rolling admissions? Much easier. You start reading applications, and when you feel comfortable that you are going to accept this person, you tell him or her, and if you feel certain about rejection, you will tell that, too. That could happen 15 minutes after the file is complete, or months later after lots of talk and comparison. Whenever for any reason you are still on the fence about somebody, that person gets held until you’re no longer on the fence, but everyone has to get a decision by April 1.
“What would be the point of RD, then? Wouldn’t you just likely be moving 75-80%, or more, of admissions decisions to EA? For people to act on in the space of six weeks, vs. three months?”
Well we are already on the way there if 50% of seats are being filled ED at many places. And what’s the point of the current crazy RD phase with huge numbers of apps and tiny admit rates?
Looks like MIT, ND and Gtown are all steady with about one-third of their total apps coming at the EA stage. That seems like a reasonable split to me. But having some kind of limit on the number of EA apps a kid could do might be needed to keep things reasonable.
On the ED side, NW and Duke run at about 10/90 between ED apps and RD apps. Reasonable front end, but then a crazy back end.
ED is a mechanism by which colleges can lock in candidates who fill specific institutional priorities early - it draws a disproportionate number of full pay or almost-full pay students and allows athletic coaches to shape their teams, etc.
If ED was abolished as @northwesty suggested, then ED would no longer serve that locking-in agenda, and would no longer be used to fill 50% of seats. To start with, even if the same number of students were admitted EA, there would no longer be the ~95% yield that ED produces – so more students would have to be admitted RD to make up for the yield shortfall of a nonbinding early admissions system.
Sorry to disappoint, but there is not going to be a Hail Mary solution here. There is no such thing as “some amount of collaboration that would be permissible legally” among the colleges. And we have already discussed why; namely, the admission mechanics are intertwined with pricing. What we have here is no different than the situation in the airline industry. The airlines are not allowed to sit down around a table and discuss the mechanics they might hypothetically use to fill the seats on the planes. I guarantee you that the Antitrust Division at the DOJ will not allow the colleges to do any such thing. Have you met the economists who work there? I have. They have PhDs in economics from HYPSM in things like microeconomics, game theory, and pricing…
I like @JHS’ idea. Rolling admission everywhere, and any offer expires on the earlier of (i) 60 days after you get it or (ii) May 1. So you get an offer from Penn and you have 60 days to wait and see if you get into Harvard, or you get into Duke and wait and see if you get a Northwestern offer with a better aid package in the timeframe. Once you accept an offer, every other school to which you’ve applied is automatically notified that you’ve committed and are out of the game.
In some ways, though, this would complicate life for everyone, because applicants and their GCs would be actively negotiating with multiple schools simultaneously and you’d have to handle it like athletic recruiting (e.g., imagine this kind of call from a GC: "Hello, Penn, our student Johnny has an offer from Columbia that expires in two weeks - if you give him a no-loan package in the next week and fly him out to see the campus, I’m sure he’ll commit before the end of the month).
@DeepBlue86 I think your proposal would be a nightmare. Since 60 days is not enough time to send additional apps and get an answer after the first response, you are forcing kids to apply to all their colleges at once: the desired, the next down, and the safeties. You have now increased the number of apps for students (who otherwise would have had an early accept) and for colleges, and removed the ability for colleges to determine if there is high interest. It’s just as if you eliminated all Early and everyone must go RD.
It seems to me that everyone keeps proposing solutions to a problem that the schools don’t seem to mind. The increase in applications means that each school is getting much more money in application fees and the admit rates are decreasing which improves the perceived selectivity of the schools. The ED rounds, in general, lowers the bar a bit to accept more students that may have been borderline in RD but guarantees that a significant portion of the incoming class is pretty much locked in at full pay. The increase in applications also means that the schools have more choices in selecting students and can increase the stats of the “average” accepted student. Nothing here will change since the schools are the ones calling the shots and all the benefits of this “problem” go to the schools.
@DeepBlue86 Going back to your discussion of legacy in post #399, I would be interested to hear what you think of my analysis. My sources of data are the admission statistics for the class of 2020 from the Princeton website, the news article posted on the Princeton website that gives some statistics for the SCEA round for the class of 2021, and a letter they sent us saying that they have accepted about 30 percent of legacies in recent years.
According to their website, the class of 2020 had 1312 enrolled and of those, 14.5% were legacies, so that means about 190 students in the class were legacies.
If we assume that although the number of non-legacy candidates can grow tremendously over time, the number of legacies in the applicant pool is about the same from year to year because there are a fixed number of alumni generated each year to have kids later, we can do some calculations. I know this is big assumption, and there are likely to be fluctuations.
If the number of legacies enrolled for 2020 was 190, and they are admitted at 30 percent, there would have been at least 633 legacies who applied. Probably a few more because maybe not all legacies who were admitted enrolled.
So going forward to the class of 2021, for the SCEA round there were 770 students admitted, and 16 percent of those were legacies. That is 123 legacies admitted so far, with 190 potential slots, meaning there are still 67 legacy slots available for the RD round.
If there are a similar number of legacy applicants in total as the year before (633) and 123 of them are already admitted, there are 67 legacy slots still available with (633-123) 510 students left to fill them, the percentage of legacy applicants admitted in the RD round may be 13%, which is still better than non-legacy chances.
Your speculation that many of those would have applied SCEA to show interest gives a little hope to those who were legacies but deferred.
@TrudiRexar I obviously have no knowledge beyond what’s been disclosed publicly and some anecdotal evidence, but I’m pleasantly surprised to hear that Princeton discloses how many legacies were admitted SCEA - very transparent of them. I’d note that, if in fact 123 legacies were admitted so far, and on average 190 are enrolled in a class, this would square quite neatly with @al2simon’s intel that at one of the HYPS schools approximately 2/3 of legacies are admitted SCEA and the rest RD. I would assume ~90% of Princeton legacies who are admitted enroll, so I’ll bet the university admits a little over 200 in all, year in and year out (as elsewhere, though, I imagine this proportion may be slowly declining as other priorities, e.g., increasing the number of first-gens, assume greater importance).
I’m far less experienced than others here (I’m looking at you, @JHS), but it seems to me that the balancing act at places like Princeton is to make alumni feel that legacy counts for a lot while encouraging the rest of the world to believe that it barely matters. Having 10-15% legacies in the class, with a 20-30% admit rate, seems to be the happy medium.
Beyond that, I’ll add that I know multiple HYP legacies who were deferred and then admitted (and one Princeton case in particular within the last two years). They were excellent candidates, but not quite the slam dunks that got in SCEA, and the schools clearly wanted to compare them to the rest of the pool and see who else showed up. I also suspect that if the university sizes you up in the early round and concludes that you’re good enough and they’re very likely to take you because of the legacy tip combined with your other attributes, but they’re very confident you’ll accept their offer and they’re not worried if you go elsewhere, this is how they deal with you. I don’t know what makes the difference between defer-admit and defer-deny, but being a connected, generous alumnus seems not to hurt. Also, a couple of updates from the applicant over the next few months to the admissions office reiterating interest and providing appropriate supplemental information (@gibby is an expert on this subject) may be helpful.
That said, if it doesn’t work out, the sun will still come up the next morning (easy for me to say, I know).
or… if one of the SMHYP schools go ED… watch they will all go ED… although it is more true of the HYP cohort … which are outliers with SCEA admission rates 4x higher over RD… and all admit about half their class SCEA… .no doubt these schools are closely watching each other… so if one went ED the others would be forced to go ED to keep yield up.
This is a classic prisoner’s dilemma example. here is a link to a description of this game theory concept.
I came across a paper that answers many of the questions that are being debated in this thread. It’s a study of early admissions programs by Avery and Levin. The data they analyze is old (from 15 years ago), but it yields some interesting answers.
They study 14 elite schools with EA/ED programs (Ivy League plus MIT, Stanford, Duke, Chicago, Georgetown, Boston College). 6 schools had EA and 8 schools had ED back then.
They then put together some models that control for SAT scores,dummy variables for race (Asian, Asian American, Black, Hispanic, Other), family connections (alumni child, sibling attended given college), and indicators for athletic recruit, male, private school student, financial aid applicant, applying from high school in the same state as the college, and having at least one parent attend graduate school.
They also had a college admissions officer rate the quality of the applicant’s high school and their extracurricular activities. The models then control for these variables too.
Here’s their conclusion -
So, even though the EA pools are stronger than the RD pools, EA applicants still get an admissions preference after controlling for their stronger qualifications using all the above variables. The ED pools are weaker, but applicants get an even greater preference.
This study agrees with my own opinions, so it must be right
This makes sense because he tippy top stats kids are only in the early pools at HYPSM but they’re back in the RD pools at the “lesser but still great” schools with ED.
Timing makes a difference. Right now, admissions departments have 4-5 weeks to evaluate early applications, including in many cases arrange interviews, and those weeks coincide with a period when they are also still travelling around visiting high schools and communities and beating the bushes to attract more applicants. (I am assuming that there’s a week of logistics at the end after decisions have been made.) For regular round applications, they have 12-13 weeks, and it’s really all they do during that period.
I think if you triple early applications, it would put a lot of stress on the system. The system works now in large part, I suspect, because there is a “defer” option that avoids having to make decisions that would require more time to make well, so if you dumped a lot more early applications into the hopper most of them would come out the other end undigested, so to speak, as regular applications, which would be frustrating to the applicants. As it is now, but I think more upsetting to more people.
Back when the elites first started with early rounds, I believe I recall that Harvard began with ED then went to unrestricted EA, and the next year instituted SCEA, because they were horrified by the huge number of applications unrestricted EA produced. (It was about 4,500, I think, roughly triple the previous year’s ED applications. A simpler age.)
Why have an early round at all? I’m not certain, but I think it does serve several non-nefarious purposes. It somewhat spreads out the admissions workload. It allows kids to identify a first-choice or high-choice school. For kids who are accepted, it is nice not to have to go through another 3-1/2 months of tension, and it reduces competition among high school classmates. And, very importantly, it provides a signal to kids if their expectations are out of line. I saw a number of kids apply early to a college they liked, and that they (and their advisors) thought was sure to accept them. Often, that worked out as planned, but sometimes it didn’t, and those times were a big flashing red signal to the kids and their advisors that their thinking was off. It happened at a time when they still had a little opportunity to recover and to shift direction. It wasn’t optimal in terms of timing, but it was sure better than having that happen in late March.
Maybe the reason why UChicago goes EA/ED1/ED2/RD is to get as many top students as possible, peeling them away from schools ranked higher.
If all schools use RD only, then better schools get better students (who shop around) and that benefits only the top one school. But this would be a better choice for applicants.
The vast majority of applicants who aren’t interested in highly selective schools won’t care at all.
JHS – strongly agree that there are very big benefits to having an early stage. Getting some feedback (positive or negative) by 12/15 rather than 4/1 is good for everyone.
But all the concerns being discussed here are basically variations on the same theme – how best to keep the early round within bounds. Because when the early round gets too big or too important, it creates many other problems, distortions and gamesmanship in the rest of the system.
IMO, the more level the field between early and regular the better. I doubt all the top schools would ever adopt one model. But each level could be dialed back a bit.
Unrestricted EA as done by Gtown and MIT is best. SCEA could be improved by evolving into double/triple/quadruple EA rather than single choice. Binding ED could be dialed back by limiting the number of seats that get filled via ED or by going to SCEA instead.
Sure they could, but why? The Trustees that run the colleges have a fiduciary responsibility to manage their college in the best interests of their colleges. And if that means a large ED acceptance, so be it.
That makes no statistical sense to me. A student can only attend one college. And there aren’t hundreds of thousands of “better students” (however defined) that would even make the maybe pile of tippy top schools.
If they wanted to spread the workload, they could easily move up the RD date to Dec 1 or November 30. (Working on last minute essays over Thanksgiving weekend is a rite of passage for every California senior since UC apps have been due the last day of November since the dark ages.)
“If all schools use RD only, then better schools get better students (who shop around) and that benefits only the top one school. But this would be a better choice for applicants.”
But this will happen under any system (especially since it is a market and all the schools adjust their practices in response to each other). Even with all the varying practices in the current system, Harvard is still Harvard. Harvard is going to get the biggest share of the tippy top kids.
That same sorting happens now. The overwhelming majority of the kids who opt out of applying to Harvard SCEA and who ED to Penn or Duke would not get into Harvard. And the close misses and defers from Harvard SCEA will wind up being very strong applicants in the RD pool at Penn and Duke or the ED2 pool at Chicago or Vandy.
At the end of the day, the ED games really aren’t about getting the best kids or getting the kids who love your school the most. It is more about hitting budget for full payors (just as legacy admissions largely are). But that can be done a lot of other ways too.
Under a RD only system, a student may get accepted at Harvard, UChicago and USC. H gets her. But under an EA/ED1/ED2/RD system, she may end up at one of the other two schools. Why would UC or USC agree to a RD only system, let alone other reasons?
