Early Decision: Policy and Data

Parents Forum
1

<p>I'm interested in information that members of the CC board may have on ED across the country this year (and for the coming year). Has anyone learned of changes in ED policy at the LACs or otherwise? Williams, for example, has stated that its ED acceptance rate will stand at no more than 30% from the class of 2007 and afterwards. Is there a shift underway (apart from the move to single-choice early action) among the schools that will stick with the traditional ED? </p>

<p>In the absence of any statements of policy, one of the ways to measure the effect that early decision has at any given school would be to look at the difference between acceptance rates for RD and ED and, thereafter, the change in differential over the years. For example, take the numbers that I have for Haverford from the Common Data set for the incoming classes of 2003 and 2004. In 2003, the ED acceptance rate was 51% and the non-ED rate was 28%. Setting aside the populations of both pools for the moment, and the self-selection bias at work for most of these schools, the differential (i.e., 23%) for 2003 at Haverford suggests that ED helps. However, the ED acceptance rate dropped to 47% there this year. There were similar drops at other schools, although the data can be hard to come by for this year at least. </p>

<p>I have read The Early Admissions Game (Avery, Fairbanks and Zeckhauser), so I understand the data for particular schools and the general conclusions offered by those authors ("[E]arly applicants were much more likely to be admitted than regular applicants with the same qualifications." (150)).</p>

<p>Can anyone offer more recent data, policy statements, or anecdotal evidence about what's going on with ED these days? Thanks much.</p>

2

<p>Early decision programs have served elite liberal arts colleges well since the 1960s. I don't look for these schools to make any significant changes in the programs.</p>

<p>HYPSM are taking the "heat" in the popular press for ED/EA. Their changes will suffice to reduce the "outcry" and deflect the attention, leaving the elite LACs to continue on with business as usual. Early Decision programs are a win-win situation for the LACs. Both the school and its most eager customers benefit enormously from ED.</p>

3

<p>Williams ED to final class total seems to have been fairly consistent over the past three years: 36% for 2007, 40% for 2008, projected at 35-38% for 2009. What's different is that they received fewer applications for 2009 than for 2008 which will presumably increase the chance of getting in via ED.</p>

4

<br>


<br>

<p>Here's the flaw. The only empirical measure of "qualifications" that provides sufficient resolution to distinguish between ED and RD is SAT scores. You can probably "prove" some slight differences. For example, I seem to recall that Swarthmore's ED acceptances have marginally lower SATs (by five or ten points) than the overall accepted class. However, they may well be marginally higher (again by a couple of points) than the overall enrolled class.</p>

<p>But, while the numeric precision may be there, the schools themselves don't put much meaning in small SAT differences. Once you are in the range, other qualifications make for more difference.</p>

<p>For example, here are qualifications of two students -- Student A and Student B:</p>

<p>Student A says that Podunk U. is her absolute top choice, no matter where else she is accepted. Student B says that he'll wait around and find out where else he gets accepted before making a decision.</p>

<p>Student A feels so strongly about Podunk that she will accept whatever financial aid package can be worked out. Student B says he's going to shop around for the best offer.</p>

<p>From the college's perspective, which student has "better qualifications"?</p>

5

<p>I think you will find, overwhelmingly, that the students who both apply and are accepted early decision are richer, and much less likely to need financial aid. It is a way for colleges to assure that a large percentage of the student body will be paying their own way - and more - as legacies and development admits are much more likely in the ED round. UPenn and Princeton say as much. Admits may have the same or even better "qualifications" - but wealth more than accounts for the "equality". You'll never see them publish the numbers....</p>

<p>If you eliminate the legacies (much higher percentage ED), the developmental admits (much higher percentage ED), the recruited athletes (much higher percentage ED), and the recruited URMs (not necessarily higher ED), you are likely to quickly find that, at most top schools there is little or no advantage to ED at all. In fact, one college - Pomona - has actually demonstrated this. As they attract few legacies, few developmental admits, don't recruit athletes or URMs, their ED admit rate is 18%. Their RD admit rate? 18%. (I wish they'd publish the wealth statistics on who they admitted each round....)</p>

<p>"From the college's perspective, which student has "better qualifications"?"</p>

<p>The one that doesn't need any financial aid at all. Just another way to be "need-blind" without being "need-blind".</p>

6

<p>I have noticed a very slight - and only ancedotal - trend over the past three years of some schools adding an non-binding Early Admissions option in the past three years.Some have kept the ED option in addition to EA, others have switched just to EA. Most of these schools are in the second half of the top 100 LAC's. Most tout either overtly or subtly that applying EA increases your chances of acceptance somewhat over RD. Some tie eligibility for merit scholarships to EA application. The schools that have kept ED too still seem to imply that that is your best bet at acceptance. Just something I've noticed, not sure if it's a trend.</p>

7

<p>I think what ID is saying in summary is you still can BS a college with an ED application vs RD. The logic is since I prefer Podunk U, I am better qualified even though my SATs and other statistics are somewhat weaker. Win Win...I don't know, certainly not a meritocracy. The morale of the story is if you show you are interested then you get the Early Admission Game advantage (statistically shown to be 100 SAT points) improvement in your odds of admission. The only real quality statistics I have seen are the Early Admission Game by Avery et al and I don't see any verification of Mini's hypothesis there.</p>

<p>The advice to kids is apply ED if you really want Podunk. It is the easiest way in. Whether it is right, good or fair is entirely another question.</p>

8

<br>


<br>

<p>No. I'm not saying that at all. Quite the contrary, actually. </p>

<p>At the schools I am most interested in, and familiar with, the measured "stats" for the Early Decision round (application, acceptancees, and enrollees) are essentially identical to those in the Regular Decison round. When you are talking a median SAT of 1450, a few points up or down is insignificant, especially with small sample sizes. For example, my daughter's school only receives roughly 300 ED applications.</p>

<p>I believe that the empirical standards used by the admissions commitee (GPA, class rank, SATs, etc.) is exactly the same for the ED and RD round. If you "make the cut" in one round, you would "make the cut" in the other round. I don't believe you'll find one example of a student who would be rejected because of "stats" in the regular round and accepted in the ED round.</p>

<p>Once you've "made the cut", then they start looking at the intangibles. At any school, being willing and able to pay the tuition bill is certainly a valuable "intangible". Without enough full-fare customers, there isn't any money for need-based financial aid.</p>

<p>At small liberal arts colleges, with tight-knit campus communities, enthusiasm for the school is a HUGE intangible. At UMich, it may not matter if a couple thousand undergrads hate their school. But, when you only have 1450 students, everyone benefits from having a significant percentage of students who absolutely love the place. By and large, you have to love the place to apply with a binding ED commitment.</p>

<p>Having said that, ED does make overall admissions to a college more difficult -- entirely because of yield. If a college has a 50% yield, it has to accept twice as many applicants as there are slots in the freshman class. With a yield of 25%, it would have to accept four times as many students. Obviously, as you accept more students, you have to go deeper into the applicant pool. Since ED boosts yield, it reduces the number of acceptances. Whether that is a good thing or not depends on your point of view. On an individual applicant basis, a low yield means increased admissions odds. However, I think the system as a whole would work better if there were more initial self-selection and higher yields.</p>

9

<br>


<br>

<p>Not according to the data Pomona supplied to the Journal of Blacks in Higher Education:</p>

<p><a href="http://www.jbhe.com/features/43_early_admission.html%5B/url%5D"&gt;http://www.jbhe.com/features/43_early_admission.html&lt;/a&gt;&lt;/p>

<p>Pomona's ED acceptance rate for the fall 2004 entering class was 30.8%. Their African-American ED acceptance rate was 50%. BTW, this article has some interesting and very hard to find data -- including the overall acceptance rates for black applicants at each of 30 elite colleges and universities.</p>

10

<p>We're talking about two separate percentages here: first the percentage of ED acceptances to the total of the final class size. So if the class size is targeted to be 500, and the school's guideline is to accept 30% of the final class then they are going to offer 150 places during ED, no matter how many applications they get. The second is the chance of acceptance. If they get 300 applications those 150 acceptances reflect a 50% acceptance rate. If they get 400, 37%. Either way the number of acceptances stays the same. </p>

<p>I see the percentage of ED to final class to be fairly static, but the chance of acceptance to be the more variable figure. In the latter, it appears that you have a higher percentage (chance) of being accepted ED than you would have RD, but it may be that ED applicants self select.</p>

<p>Using Williams as an example, if they accept the exactly same number of ED applicants this year as they did last year, the ED acceptance rate would increase from 40% to 44% just because there are fewer applicants. Would this mean that Williams would have dropped its admissions standards? No, I don't think so. I think it just means that total number of applicants, either ED or RD, is a variable that is very difficult to predict and that doesn't follow a predictably logical course. It will be interesting to compare ED figures at other schools. </p>

<p>I don't know if it's true that ED applicants get short changed financially. I know several kids who got brilliant ED packages. Maybe some other ED parents could offer examples. On the other hand, since the conventional wisdom is that you get more if you wait and comparison shop, it could be that the kids who need the most aid just are not applying ED. We didn't qualify or apply for financial aid and thus fuel the gripe that ED is a way for colleges to bankroll their tuition income. This rankles me because it implies that kids paying full fare are somehow are less deserving.</p>

<p>ED to me IS a win-win situation for students, their families and the colleges. Everyone ends up where they most want to be. I don't see anything unfair about that.</p>

11

<p>"Once you've "made the cut", then they start looking at the intangibles. At any school, being willing and able to pay the tuition bill is certainly a valuable "intangible". Without enough full-fare customers, there isn't any money for need-based financial aid."</p>

<p>Ah, but that's only half the story. You receive an admissions boost because of income, regardless of the school's need to dole out financial aid. No one would deny (because the data are available) the link between income and SAT scores; and, certainly, the link between income and certain types of ECs are obvious to me, as are links between GCs and certain schools (attended predominantly by high income folks.) Income figures into the cut BEFORE it becomes an "intangible' (but money is very "tangible", don't you think?) "Equal" qualifications do not indicate the factors that go into obtaining them. A 1450 for one student is NOT the same as a 1450 for another (and every admissions officer knows it.)</p>

<p>"The only real quality statistics I have seen are the Early Admission Game by Avery et al and I don't see any verification of Mini's hypothesis there."</p>

<p>You actually do, but you didn't notice them, because the authors did not examine how candidates came to have those qualifications. Nor did they reveal the data on percentage of financial aid admits in the early round relative to the later one. The packages could be wonderful in the ED round, or they could be ho-hum, but the likelihood is that there are fewer of them. Williams, for example (it being my alma mater), is well known for raising its packages in April in keeping with the "competition" - students all of a sudden are found to be poorer than they would have been in December, without any change in family circumstances. (It is a great wonder, isn't it?)</p>

<p>Can't say I'd do it any differently, though perhaps substantially raising the tuition for full-fare customers who apply ED would be a good idea, so that they could spend my alumni contributions on needier ones.</p>

12

<br>


<br>

<p>This is true. However, I have yet to see an account of the admissions process that does not factor background into the evaluation of the SAT scores. The adcom quote that stands out in "The Gatekeepers" comes when the adcom picks up a silver spoon application (prep school, parents' info, etc.) and says before looking at the students' stats, "This one better be good...."</p>

<p>IMO, there is a strong bias IN FAVOR of low-income applicants at elite colleges. Of course, the fact that they have a stable and predictably high percentage of high income customers is what gives them the freedom to be so "enlightened".</p>

<p>The fact that few low income students apply to elite colleges is a separate issue.</p>

13

<p>"I believe that the empirical standards used by the admissions commitee (GPA, class rank, SATs, etc.) is exactly the same for the ED and RD round. If you "make the cut" in one round, you would "make the cut" in the other round. I don't believe you'll find one example of a student who would be rejected because of "stats" in the regular round and accepted in the ED round."</p>

<p>Ahh...but the devil is in the details. For a minute look at Amherst which is one of a few schools that discloses its acceptance rate by SAT bracket and I think is representative in admission rate to the rest of the elite LACs. Amherst's admit rate for verbals of 750 to 800 is 50% and for 700 to 749 is 22%. Likewise for math scores the acceptance rate for 750 to 800 and 700 to 749 is 45% and 27%, respectively. My point is the Early Admission Game states (based on exhaustive statistical analysis of data not provided to the public) that elite schools on average give the same probability of acceptance to an applicant with a 100 less SAT points in the ED round vs the RD round. Hypothetically, a 1400 SAT scorer at an elite LAC would have, based on the Amherst data, about a 20 to 25% imporvement by applying early versus applying RD. Clearly, the LAC is not taking unqualified students in the early round. But, what they are doing is giving the 1400 scorer (in my example) the ability to be statistically treated like a 1500. Not guaranteed admission but significantly aided. I am not saying they would take unqualified kids through ED. The devil in the details...which bright well rounded kid gets in and which don't get in. If the kid's heart is set on Podunk, apply early and get maybe a 20 or 25% advantage. Or alternatively, get a higher test score and apply any old time! I know test scores are only one factor of many but it makes the illustration more clearly.</p>

14

<br>


<br>

<p><a href="http://www.jbhe.com/features/43_early_admission.html%5B/url%5D"&gt;http://www.jbhe.com/features/43_early_admission.html&lt;/a&gt;&lt;/p>

<p>Pomona's ED acceptance rate for the fall 2004 entering class was 30.8%. Their African-American ED acceptance rate was 50%. BTW, this article has some interesting and very hard to find data -- including the overall acceptance rates for black applicants at each of 30 elite colleges and universities. <<</p>

<p>Not only are the blacks who apply early admitted at higher rates, but their stats (SAT I & II scores, GPAs, class rank) are much lower than the rest of the early applicant pool, and the blacks admitted early also have lower stats than the rest of the applicants admitted early. This will give the black early applicant a bigger and huge advantage over the rest of the early applicants (ED/EA) , and of course, the regular applicants (RD). Again, it is all relative.</p>

<p>The admissions process is not transparent, and these schools will never publish this data for admission rates and stats (SAT scores, GPAs, etc.), especially when disaggregated by race and ethnicity. This will be damning because of the unfairness of the admissions process, and shows that the race preference for blacks in admissions in the adcom's rush to admit warm bodies, renders their lower stats almost meaningless in many cases.</p>

15

<br>


<br>

<p>The College Board addressed your point above, several years ago, by wanting to give ALL blacks a 200 point spot, and automatically add these extra 200 points to their SAT I scores. The CB called them "Striver's Points" for blacks, regardless of their economic class. This would include rich and affluent low performing and underachieving blacks. Of course this was not done, but the competive colleges which use race preferences for blacks have been doing this for years by giving black applicants a lower SAT I standard for admissions (a double standard), than for than rest of the admitted class. In some cases, this advantage or "spot" for blacks amounted to blacks being admitted with as much as 500 points lower or more (2 1/2 standard deviations lower) on the SAT I, than the median SAT I scores of the class. This was true at UC Berkeley, before Prop 209, banning race based AA.</p>

<p>"1,877 African American students nationwide scored higher than 1300 out of a possible 1600 on the SAT last year, compared with nearly 150,000 students overall who achieved that score. Minority students with higher SAT scores have become the target of frenzied competition between state and private colleges."</p>

<p>1.877 out of 150,000 black SAT I test takers in 2003 scored above 1300. Only 192 of these black test takers scored above 1450 and only 70 scored above 1500.</p>

16

<p>I'm puzzled as to how much of a boost ED gives a less-than-top student at a school like Williams. Williams admits such a tiny number ED. I can't imagine that they have to "settle." I am doubtful that a not-superb student would get in simply by applying ED. Seems to me that what does happen is that a really excellent student can feel more confident that things are more straightforward and that she or he will be admitted ED, whereas during RD there's a major crapshoot factor that leaves even the most outstanding students unsure of admission at the most highly selective schools. I have read The Early Admissions Game and found its data complex; I would also note that its data is, at this point, somewhat out of date.</p>

17

<p>"IMO, there is a strong bias IN FAVOR of low-income applicants at elite colleges. Of course, the fact that they have a stable and predictably high percentage of high income customers is what gives them the freedom to be so "enlightened".:</p>

<p>There is a bias in favor 1) of high-income students (top 5%), and 2) low-income (bottom 35%) students. The heavy bias for high-income students comes from legacies, developmental admits, needed ECs, contacts with GCs, prep schools, foreign languages, letters from famous people, well-spoken interviews, and, finally, the need not to provide financial assistance - in other words, entitlement. (SATs are really just the tip of the iceberg, and I know of NO adcom who takes account of the fact that the low-income student doesn't play the English horn because of income/school when they need English horn players. Nor do I know of any adcom that considers the lack of wealth of a competing applicant when considering a "development admit". LOL!) The bias for low-income students is for "diversity" or "institutional mission". The data on this bifurcation ARE in fact available. Williams (again favorite example - the numbers are the same at Yale) has 60% of its student from the top 5% of the population, and roughly 9% from the bottom 35%. The remainder goes to the broad middle, precisely where the overwhelming number of potential candidates are. This bifurcation is even more extreme in the ED. If Yale admits 1 out of 3 or 4 among high-income candidates overall, and 1 out of 7 low-income ones, and 1 out 20 in the middle income range, just exaggerate the spread in the ED round and one won't be far off.</p>

<p>So you missed the point: ED is not biased against low-income candidates particularly because, as you note, not many apply. It is heavily biased against middle income ones, and for very good reasons. (Let me be clear: I'd do the same thing.)</p>

18

<br>


<br>

<p>That's why I think the "equal to 100 points on the SAT" is so misleading. The implication is that adcoms sort the applicants by SAT and go down the ranking list, in order, until they have filled the class. That's not the way it works.</p>

<p>The other misleading aspect of data like "equal to 100 points" is that it falsely tries to isolate one factor in admissions. The reality is that there are many things that would be "equal to 100 points" using the same simple statistical analysis. For example, being black would be "equal to 100 points". Starting a community service project would be "equal to 100 points". And so on and so forth.</p>

<p>Here's the way I have come to see admissions: </p>

<p>a) One group at the top of the applicant pool is so strong academically that they are "auto-admits". Various adcoms have called them "academic 8s and 9s", "walk on water kids", etc. </p>

<p>b) At the other end of the scale are applicants that have simply applied to the wrong college academically. For example, a 1200 SAT, a B average, ranked in the third decile at a public high school is NOT going to get you into Williams, Amherst, Swat, etc. These are the auto-rejects.</p>

<p>c) A third category includes the "hooks". Kids with special attributes, be it the color of their skin, their ability to play quarterback, or their grandfather's name over the door of the school library. There is no question that these special attributes can overcome lower "stats".</p>

<p>d) The biggest group of kids are what I call "standard, well-qualified" applicants. They have SAT scores (and other stats) comfortably in the middle 50% of the school's range. As far as academic qualifications, they would all get accepted. Unfortunately, there are at least twice as many of them as there are admissions slots. So, very small differences result in one being accepted, the next rejected or deferred.</p>

<p>In my opinion, it is this LAST group that benefits the most from an ED application. It has nothing to do with an "extra 100 points" because their SATs were comfortable enough to begin with. It has to do with the fact that being enthusiastic about the school or being able to attend with whatever aid package is offered are both things that have a tangible benefit to the college. Thus, a "standard, well-qualified" applicant becomes a value-added applicant.</p>

<p>The flaw of "equal to 100 points" data is that a standard applicant with 1350 SATS is not suddenly going to get accepted at Williams by applying Early Decision instead of regular decision. It is the student with 1450 SATs who gains an admissions edge by applying ED rather than RD. The "equal to 100 points" may indeed be valid, but only when the student is already in a "stat-range" that is good enough to make him a solid applicant for the school to begin with.</p>

19

<p>"That's why I think the "equal to 100 points on the SAT" is so misleading. The implication is that adcoms sort the applicants by SAT and go down the ranking list, in order, until they have filled the class. That's not the way it works."</p>

<p>At risk of belaboring the opposite, a full 25% of students admitted and attending Harvard have SATs below 1370. (that's what the 25%-75% spread means.) And, as I remember, they rejected half of their 1600s. I think that for every other college in the U.S., the number would be even higher.</p>

<p>But I doubt they rejected ANY candidates whose dads gave $20 mil.</p>

20

<br>


<br>

<p>Actually, I bet that Harvard has rejected applicants whose Dad gave $20 million. The scorn of "development applicants" among adcoms is very strong. There is a built-in tension between the admissions office and the development office that can reach the point of "standing on principle".</p>

<p>Obviously development candidates have a huge leg up in the admissions process. But, there are thresholds below which the admissions office would take great glee in rejecting a kid and standing their ground right on up the chain of command. Especially at Harvard, a school that really doesn't need ANYONE's money. You might be shocked at how much money it takes to achieve "player" status in the Harvard development office. $20 million might do it. A few $ million probably would not.</p>

<p>Also, keep in mind that many kids who grow up in families capable of giving $20 million are plenty well qualified academically to get into Harvard.</p>