The Plague of ‘Early Decision’

Parent Cafe
421

That’s not really responsive to my point, @sbballer. I’m suggesting that the reason that the SCEA admissions rate at HYP is ~4x the RD rate is that the SCEA pool at these schools is very strong (much stronger than the RD pool, on average), at least half of the applicants HYP consider to be the most desirable (and available to them) apply SCEA, and HYP admit them then, without much fear of losing them. HYP don’t need to give anyone a boost to apply SCEA, because they aren’t nearly as worried about yield protection or fin aid budgets as the ED schools. Accordingly, I don’t think they do.

The principal consideration for HYP, if I’m right, is how much room they want to leave for people in the RD round (in which many of the best applicants participate only to a limited extent, if at all, because they already got into their preferred school early), while being aware that they’re somewhat less likely to land the ones they admit because they haven’t indicated a preference and may have other options within this peer group of schools. It’s more about class planning than yield or fin aid management.

Regarding recent history, H &P dropped early admissions for a few years, then reinstated it because Y (and S, I believe) didn’t play and got too many of H & P’s coveted admits early. As long as all these schools are behaving the same way, the equilibrium holds.

In the abstract, the math is that an average applicant is ~4x as likely to get in if they apply SCEA, but I don’t believe applying SCEA helps any individual applicant to anything like that extent. If you apply early and H wants you, they’ll take you. If they wouldn’t want you in the RD round, they’re very unlikely to want you SCEA. This is in contrast to the ED schools, which give some preference to ED applicants to induce them to commit early and provide these schools with the yield protection, selectivity benefit and fin aid management they need.

By the way, I would guess that there’s also very little difference in the likelihood of admission SCEA or RD to S, for the same reason. I think if you’re going to get in, leaving aside special cases like legacies who need to demonstrate interest by applying early, you’ll get in, in whichever round you apply.

422

“MIT has stated in the past that they don’t want to make EA the norm, so were committed to not accepting a large percentage early. However, at least when my son applied, the admissions rate for students who were deferred to the regular round was twice the overall admissions rate - leading me to believe that their early pools were as strong as Harvard’s or Princeton’s - academically at least.”

Gtown is similar to MIT – they intentionally keep the early admit rate the same as the RD rate.

Which in practice can mean that EA standards are actually tougher than RD. When you see how the EA defers do in RD, you see that the EA pool was much stronger and (assuming level standards) should have had a higher admit rate based on pool strength.

ND also to a lesser extent. Despite a 2.2/1 early/regular ratio, they use slightly higher standards at the early stage.

Harvard’s 4.2/1 ratio is harder to believe, but I tend to believe Harvard when they say there’s no early advantage.

423

Georgetown’s strategy makes sense because it counts on a lot of rejects and deferrals from HYPS SCEA in the regular pool. It is highly likely in Georgetown’s case that it has more qualified candidates in the regular round than in the early round.

424

If you think about it, I believe you have to conclude that everyone has more qualified candidates in the regular round than in the early round. They may not have as high a percentage of qualified candidates in the regular round. There may be – probably are – more unqualified candidates there, too, comparatively. But in absolute numbers, there have to be more qualified candidates in the regular pool.

425

I’m sure that’s right, @JHS. I would argue, though, that not only is the proportion of qualified applicants higher in the early round, but the average quality of the qualified applicants is higher as well (this was my point to @al2simon). Also, all of those applicants have expressed a preference for the school in question by applying early. For these reasons, I’m not surprised or bothered by SCEA schools filling half their spots in the early round, and this is why I don’t think applicants get a substantial advantage by applying early to them, because a lot of the best kids do so, and therefore the competition is even tougher.

Also, a lot of the candidates admitted SCEA quit the game or apply to only a small number of other schools after getting an offer from their first choice, so it could be that even though there are greater numbers of qualified candidates in the RD round, the average quality of the qualified candidates might be lower.

In summary, in the early round these schools are offering half their spots to some of the most qualified applicants, who’ve indicated to the schools that they’re very likely to accept their offers, so it’s only logical that a substantial number of the available spots would go to them. The schools know with a high degree of confidence what they’re getting, and that it’s excellent, so it’s not surprising that they take a lot of it.

426

Colleges are in a game theory environment and they are all trying to increase yield… So students considering top schools should use this to their advantage.

So basically you have major statistical outlers with Harvard, Princeton, Northwestern who have early admission, (ED for Northwestern) rates over 4x the RD rate. In the case of Harvard and Princeton they are admitting half their incoming at a rate 4X higher than the RD… This is a major statistical advantage… no way around it. Harvard drops EA and has to reinstate it because their yield drops is a classic prisoner’'s dilemma case. textbook actually.

MIT is the school where I would say definitely do not apply early… basically its admission rate is the same for EA and RD… and when you factor in the small advantage for sports (I say that in jest… because MIT along with Caltech and possibly Chicago are the schools that place the least emphasis on college athletics), legacy etc. it may actually be a disadvantage… in the case of MIT the fact that these other schools are single choice or restrictive… essentially makes MIT single choice if a student is considering these schools.

Stanford’s ratio is 2.6 admits about 35% of their class EA so there is also an advantage to applying early but considering athletic recruits admitted early it’s probably significantly lower.

Harvard and Princeton admitting half the class at ratio of 4.2x higher over RD. No brainer.

the stats tell the story. rationalizations result in heuristic errors (very common) IMO.

and if one of the more selective schools goes to defcon 1 and does ED… .all the schools will follow…

classic prisoner’s dilemma

427

Not withstanding your insistence, @sbballer, I agree with @DeepBlue 86 that despite the 4.2 early admit ratio, applying SCEA to Harvard does not provide an advantage over applying RD because Harvard attracts only the very best in that round. But we can agree to disagree without being disagreeable.

The funny thing about MIT is that they actually do place a great deal of emphasis on college athletics, but it is not clear that it provides an advantage in admissions. They have 31 varsity sports, the most of any D3 institution. Perhaps this is a little-know fact but it is true. They have to fill all those spots, so a pretty good proportion of the student body is an athlete…very smart athletes!

428

JHS – that’s true, but not all that helpful.

The really helpful data is Mathmom’s observation in #419 on MIT. If MIT’s EA defers fare better in RD, that tells you that (i) MIT’s EA pool is stronger than RD and (ii) MIT is applying higher standards in EA than in RD. So the exact opposite of the ED deal as practiced by say Duke. My anecdotal experience on MIT is the same – getting deferred EA is an accomplishment and is a strong plus factor for your chances in RD.

You really can’t see that same info at Gtown. Since another quirk of Gtown’s model is that everyone who is not accepted EA is automatically rolled over into RD. There are no rejections of weaker apps and no identification of competitive apps via deferrals.

I’d agree that Gtown would pick up strong apps which are misses from HYP SCEA, but they also pick up lots of other apps too (weak, strong and in between). Overall, the pools are probably equal in strength (as evidenced by a similar admit rate in the two rounds).

429

MIT does not consider legacy in any round.

430

you can make the case for every school I mentioned… all top schools… all get top candidates applying early… yet the higher EA rates and higher proportion of students admitted early give students a major statistical advantage to apply early for schools like Harvard and Princeton compared to others… the same holds true for Northwestern when compared to other ED schools.

MIT gives you no benefit to applying early by this analysis.

431

@sbballer - I believe @al2simon’s numerical analysis in post #395 contradicts your assertion. If you just looked at the headline numbers, you’d conclude that with a 17.1% SCEA admit rate and a 3.8% RD rate last year, you were 4.5x as likely to get into Yale if you applied SCEA last year. Correcting for the number of athletic, legacy and Questbridge admits as @al2simon has done, though, and making an educated guess as to what proportion of the two pools consisted of “serious” applications, reduces the advantage to 1.8x (13.6%/7.6%) for unhooked applicants. This doesn’t take into account other hooked SCEA applicants, which would narrow the gap further. Most importantly, though, if the quality of the “serious” applicants in the SCEA pool were significantly greater than that of those in the RD pool (as I believe), and you adjusted the number of “serious” applicants in each pool to “tippy-top” applicants to account for this, the difference would narrow substantially or possibly even disappear. An unhooked applicant who wasn’t a tippy-top candidate might have no advantage at all applying early to Yale - either way, they wouldn’t get in. Similarly, a tippy-top applicant would get in no matter which round they applied in.

Harvard’s numbers are in the same ballpark as Yale’s, and I think the conclusion would be similar if you went through the math. The early pool contains many more applicants that satisfy these schools’ institutional needs, so they take more of them because it provides greater certainty as to whom they’ll end up with. Of course, if one of these schools were to change the rules of the game and move to ED, it would disturb the equilibrium and they’d all follow, because the schools are of similar quality and the outlier school would have an advantage. I’m not sure what that has to do with the question at hand.

Also, the fact that Stanford admits fewer students early proves nothing except either (i) Stanford believes they have particular competitive advantages (e.g., much less regional competition, big-time sports, Silicon Valley proximity and better weather), so aren’t really comparable to HYP and they’ll get whichever kids they want anyway; or (ii) Stanford really doesn’t care how many kids they lose to Harvard (and Parchment would suggest Stanford is a net loser to Harvard in cross-admits).

432

SBB – the raw 4/1 ratios between SCEA and RD at HYP certainly do jump out. And they look similar to the ED/RD ratios of NWU and Penn, who are admitted ED power players.

But I still think they are different animals on whether the early round equals easier admission standards (after backing out all the legacies and athletes).

First, there’s what HYP actually say – no strategic advantage, no boost. I really doubt they would be so clear in saying that if it was false.

Second, HYP really aren’t trying to IMPROVE their yields like the big ED schools do. Seems like they are more trying to maintain/defend their yields from erosion by the ED schools. So long as the field is sort of level, Harvard is always going to have the highest yield.

When HP unilaterally disarmed on ED, the resulting playing field was not level. So they had to retrench with SCEA.

433

sure there are mitigating factors… questbridge legacy… for every school… while the EAadmission rate is 4X higher for schools like Harvard and Princeton, Yale is close too… .it doesn’t mean it’s 4X easier to get in… it may be 2X easier to get in all other things being equal… if you’re considering these schools… apply early… no brainer.

and if you’re also considering MIT… apply early to these other schools… and MIT in RD. Students should use this to their advantage.

434

Except I don’t think it’s 2x easier to get in SCEA, @sbballer - I think you have negligible advantage if you’re unhooked, because if you’re not going to get in RD, SCEA is very unlikely to get you over the line. Better, by far, to assess your chances realistically and, if appropriate, apply ED to one of Penn, Columbia, Duke, Northwestern or most of the top LACs, because doing so could make all the difference.

435

It seems that schools like Duke, Vanderbilt, and Northwestern would have a definite ED advantage considering that the “first choice” of highly qualified kids is split among each other, the Ivys, Stanford ect.
After Early apps are over, most-other than ED’s- are back in the RD pool…even the SCEA admits!

436

@DeepBlue86 - Thanks for your response. I had a chance to run this by someone who is “very familiar” with the data at one of these schools. They wouldn’t share anything confidential, but they did give me some interesting feedback.

  1. Regarding legacy admits. The actual data at this school is that about 2/3’s of admitted legacies are admitted under SCEA, with about 1/3 admitted RD. That’s a little less than the 3/4’s SCEA that I had estimated. One factor that I had forgotten about is that there are a few applicants who come from “mixed marriages” (e.g. a Princeton grad married to a Yale grad) or who have a parent who’s an undergraduate alum of one of the HYPS colleges and a graduate alum of another (for schools like Yale that count children of graduate alumni as legacies).

  2. Some Questbridge etc. students apply RD. I had assumed they all get admitted SCEA, but my contact said that there are a number who get accepted outside of the official “Questbridge match” process.

  3. There are a few other hooks that are about the same as being a legacy. The most important one is being a faculty brat. The SCEA pools are heavy with these applicants, and these students need to be accounted for too.

  4. The big hook we haven’t mentioned is for URMs. However, the SCEA acceptances aren’t tilted towards URM admits. One of the reasons why Harvard and Princeton suspended their early admissions programs a few years ago was because they felt early admissions worked against low-income and URM students. My contact says that the situation has improved quite a bit over the last 10 years, but I gather that URM admits are still slightly underrepresented in the SCEA round. Because of this, omitting URM students from the calculations moves the results in the opposite direction from what you suspected.

If I update my numbers, the SCEA / RD numbers change from 13.6 vs 7.6 to 12.6 vs 7.2. Still means that SCEA is a bit easier than RD, but not dramatically so.

I agree with the point you made about “tippy top” applicants. I would guess that there are 200 or 300 unhooked applicants in the US each year who are in the very enviable position that HYPSM are match schools for them. In my area there are often 1 or 2 students a year who fit into this category. From what I can tell, they generally decide which school they like best and just apply SCEA/EA to it. After they get accepted they may send in an RD application or two, but often they don’t even bother. As you said, if this is the typical case then this would reduce the spread between the SCEA and RD admit rates even more for unhooked applicants who aren’t in the “tippy-top”.

I still believe that even after accounting for all this there is still an advantage to applying SCEA. It’s certainly what I told my own children to do.

437

I find the discussion between DeepBlue and SBB very interesting regarding whether there is an advantage to applying SCEA to HYP. I have a junior unhooked D who is contemplating the merits of SCEA so the discussion isn’t just theoretical to me although I am intrigued by the nature of the discussion.

My thinking is that if the unhooked student feels like she has a realistic shot (after due diligence) of getting admitted to HY or P, then SCEA seems to offer better odds than RD. I’ve never crunched any numbers but it just seems like the competitive HYP applicant is competing against a smaller pool of competitive students in SCEA than in RD. Of course I could be wrong about this but it just seems intuitive.

438

I think you are looking at the question wrong. We all agree that SCEA can’t hurt - so the question is, what’s the downside.

The downside isn’t a matter of chances – it is that the student who uses SCEA is precluded for applying EA elsewhere. EA usually doesn’t confer a significant admissions advantage, but at some schools it may be required for consideration for some merit scholarships.

So the real question isn’t what’s the advantage --but rather, for any given student, what is the downside? If admission to HYP is something of a long shot, the student may want to think carefully about what possible benefits are forfeited by giving up the ability to apply EA to other colleges. It makes sense to develop the basic college list (safety-match-reach) early on and get familiar with any specific issues related to EA when offered at other schools.

439

Applicants to SCEA can apply EA to any public university or to any private university that has a scholarship deadline.

440

While there is a clear statistical advantage to applying early to schools with High EA vs RD rates (Harvard Princeton). this ratio admission advantage has been going down the past several years…as students have been figuring this out and applying early in greater numbers to these outlier schools.

MIT is an interesting case… the EA/ED rate is 1.1 to 1. the lowest of any selective school.

MIT saw a dip in early applicants for the Class of 2019, with 6,519 students applying early and the admit rate went up to 9.6%, compared to 6,820 for the Class of 2018 where the admission rate was lower at 8.9%

this past year MIT allowed international students to apply for the first time resulting in a record 7,767 early action applicants this year with the admission rate going down to 8.4%

EA Rate
2018 8.9% Ratio 1.1/1
2019 9.6% Ratio 1.1/1 applicants went down
2020 8.4% (international students allowed to apply EA for the first time) because of early international applicants the EA/RD Ratio may go below 1.1 when final admissions numbers come out in the spring

Applying early to MIT is a bad bet… in fact you can make the argument that students applying early to MIT are actually at a disadvantage compared to RD… especially when you factor in that MIT’s EA pool is probably stronger than RD. In fact this is probably one of the reasons MIT saw a drop in EA applications in 2019 while other schools recorded record apps.

This year without international students EA applications would have been around 6700 still lower than 2018 numbers which is probably the ceiling put in by the “market” because applying early is not a good bet.

If a student is considering other schools that are single choice than MIT effectively becomes a single choice school for them. this would put the US EA MIT applicant at a further disadvantaged compared to international students (where MIT may be the only school they are applying to in the US).

my prediction is that MIT will see a drop in EA applicants from the US next year as the “market” adjusts to the bolus of international students.