The advantage of being from a high SES family often comes long before the point of making the application to a college. This is the advantage in that parental money can purchase increased opportunities for and remove barriers against earning stronger qualifications. An obvious example is that high SES parents are much more likely to be able to avoid having to enroll their kid in a low quality public school – they can either afford to live in a place with a better public school, or pay for a better private school.
Colby’s income distribution from the study’s sample is at https://www.nytimes.com/interactive/projects/college-mobility/colby-college and summarized below. At the time of the sample, Colby had the 7th wealthiest undergrad student body in the United States, as measured by median income… a much wealthier student body than any Ivy+ college.
The high income top quintile was dramatically overrepresnted at Colby, encompassing >3/4 of students. The overrepresentation became even more extreme as income increased further within the top quntile. 20% of families were top 1% income, for a 20x overrepresentation. The middle income families were underrepresented at ~4x less than expected in a random distribution. The low income families from the bottom quintile were even more underrepresented, occurring ~10x less than expected.
Colby Income Distribution
Top Income Quntile – 76% of students
2nd Highest Quntile – ~6% of students
Middle + 2nd Lowest – ~10% of students
Bottom Quntile – 2% of students
It suggests that wealthy families were dramatically overrepresented at Colby, but does not suggest middle income families have a donut hole type region. More recently Colby has made some key changes that influence portion of lower and middle income applicants such as no longer charging an application fee, no longer requiring a supplemental essay, having $0 expected cost to parents for families with below ~median US income, and having $15k expected cost for $150k income families near/above the top of “middle income”. I expect these changes to have increased the degree of representation of lower and middle income families since the time of the sample.
Colby FA percentages for the most recent IPEDS year are below. Note that Pell grants were only 7% two years before this, so there may be a good amount of year-to-year variation, due to the small class size.
Colby FA percentages
52% received no FA
34% received institutional FA grant, without Pell grant
14% received Pell grant
There was a type in the income distribution from by post above. The corrected percentages are:
Colby Income Distribution
Top 1% Income – 20% of students (20x overrepresented)
Top 1-5% Income – 31% of students (8x overrepresented)
Top 5-10% Income – 12% of students (2x overrepresented)
Top 10-20% Income -- 13% of students (balanced)
Total Top Income Quintile – 76% of students
2nd Highest Quintile – ~11% of students (2x underrrepresented)
Middle + 2nd Lowest – ~11% of students (3-4x underrrepreesnted)
Bottom Quintile – 2% of students (10x underrrepresented)
Quote: “However, I would not assume that an unhooked higher income kid has a big advantage over unhooked kids at other income levels, who have similar qualifications.”
Interesting post @Data10. I agree with some of your analysis, but this point deserves to be challenged. Yes, I was referring to the link I provided rather than the original post link. Good point.
The data for Colby’s self reported class of 2024 ED would appear to contradict the quote above if you are generalizing this to elite colleges. If you are limiting your comment to HYP, I don’t know if this is true or not true. But among elite colleges below HYP it is certainly not accurate. Very few elite colleges claim to be needs blind.
Many are needs aware for all or part of their applicant pool. Even among those claiming to be needs blind, some experts have written that it’s not always true. The fact that many elite colleges themselves say they are needs aware is proof that, among equally qualified candidates, income levels matter.
A look at common data sets show that Colby is not alone in favoring wealthy full pay applicants over less affluent applicants with similar qualifications. Books have been written on this by experts including Daniel Golden and Mitchell Stevens.
Unhooked applicants from the lower and middle class families, above the Pell grant level still needing significant financial aid, have the hardest path to acceptance at elite colleges, I don’t see that as disputable if you have familiarity with the research.
The title of this thread includes the words “Ivy Plus colleges”. All Ivy Plus colleges claim to be need blind, as do the vast majority of other colleges that are called “elite” on this forum.
How are you determining whether a college favors wealthy applicants by the CDS? Having a significant portion of entering students being full pay does not mean that the college directly favors wealthy students over less affluent students with similar qualifications (including hooks). As discussed in the study from the OP, many admission criteria are well correlated with wealth, such as test scores. One of the biggest factors is also who chooses to apply. Wealthy students tend to be dramatically overrepresented in the applicant pool at “elite” colleges. The quotes I’ve seen from Golden’s book address different subjects from my interpretation of your claim, such as favoring hooks that are more common among wealthy students.
The Pell grant threshold chart was for 2 cherry picked colleges that had a huge increase in Pell grants over a short period. They are not representative of all colleges. The previously linked Harvard internal study, showed that the only income levels that had a significantly lower admit rate that expected based on qualifications (including hooks) were wealthy families, with incomes of $160k to $200k+ ($120k to $160k had negligible difference), and unknown income families who did not fill out FA. Middle income families with similar qualifications (including hooks) appeared to have a slightly higher chance of admission that wealthy families. There was no evidence of a Pell grant threshold. I’d expect Harvard to be more representative of “elite” colleges than the 2 cherry picked colleges from the article, which I expect were non-elite publics since the colleges from the Hoxby-Turner study were all non-elite publics.
However, if by “path to acceptance”, you include path from birth and throughout primary education and chance of becoming academically qualified, then I’d expect the lowest income families to have the most challenging path. This group tends to attend lower quality primary schools, tends to be surrounded by community who does not encourage applying to “elite” colleges, etc. This relates to why a smaller portion of lower income kids attend “elite” colleges than portion of kids from any other income level.
For example, one of my relatives was the first kid from the history of her low income high school to ever apply to a selective college. The high school did not offer any AP or college level classes and offered few classes with multiple levels (for example honors vs normal). The classes were generally at a low level that did not adequately academically prepare kids for college or standardized achievement type tests. Kids from the high school usually worked on the family farm after HS and did not apply to college. The few who attended college chose non-selective ones that were usually local. Kids who grow up in this environment have an especially challenging path to elite colleges, even if the college does give a small boost for being low income.
“…at the most selective schools (Ivy-plus), the middle class is heavily under-represented relative to others with the same test score”
https://opportunityinsights.org/paper/undermatching/
It isn’t just low and middle income students not being academically competitive.
Some of it is due to self selection. Some of it is due to elite colleges favoring certain demographics. Some could be the holistic admissions that favors families that can put together an application utilizing experts. “Inside the Pricey, Totally Legal World of College Consultants”
https://www.nytimes.com/2019/03/13/us/admissions-cheating-scandal-consultants.html
Examining all the evidence hasn’t convinced me that need blind colleges are fully honest when they claim the need blind mantle.
Many different factors are involved. All these factors disadvantage low and middle income families.
They are truthfully need blind for individual applicants in a narrow sense of not having admission readers know the FA status of applicants.
But they are need aware for the entire class when setting policy of admission process and criteria for the upcoming cycle. For example, if the college wants to raise the SES level and decrease FA need, it can increase ED admission and instruct admission readers to increase the weight of legacy and other high SES correlates, without having to be need aware for individual applicants.
Nobody has claimed otherwise, but this does not mean middle income families are at an admissions disadvantage compared similarly qualified (including hooks) high income families at Ivy+ colleges. As you noted, there are many possible other contributing factors. Some important ones include high income kids with a particular SAT score are both far more likely to apply to Ivy+ colleges than the general population and are far more likely to have ALDC hooks. These 2 factors are too influential on the portion of high income kids in class to be ignored.
If you look at CDS information without considering test scores, then it is also important to consider rate of applicants being academically competitive. For example, if you choose a 1400 SAT minimum to be academically competitive at Ivy+ colleges, the study linked in the OP found that 1400+ SAT (or ACT equivalent) kids have the following income distribution. The study found that most 1400+ SAT/ACT kids were high income, and extremely few were low income. This income distribution of 1400+ SAT/ACT kids is not that far from the actual distribution of entering students at Ivy+ colleges.
Portion of 1400+ SAT/ACT Kids in USA with Different Income Levels
Highest Quintile Income – 59%
2nd Highest Quintile – 20%
Middle Quintile Income – 12%
2nd Lowest Quintile – 6%
Lowest Quintile Income – 3%
What would be more meaningful is chance of acceptance for similarly qualified applicants with similar hook status at different income levels. The Harvard study does something similar to this comparison and found that middle income kids had a slight advantage over high income kids with similar qualifications + hook status. Unfortunately few, if any, other Ivy+ colleges provide this information.
As I stated in my earlier post, if you want to consider all pre-application factors rather than just qualifications at time of application, then I expect low income kids have the most challenging path on average, not middle income kids.
“For roughly 75-80% of the population, HYPSM would be essentially free.”
People make this comment a lot, it could be free sure, but you have to get in, right? That’s the little part people overlook and why 99% of students go to public universities, because they got in there and not HYPSM. Middle class students with no hooks don’t apply to Harvard, not because of affordability, they do think that Harvard will give them a good FA package, but because their chances of getting in are zero, and guess what, they’re right.
Gaining admission to a HYPSM is difficult, no doubt about it. But, I would never tell a high-stats kid they had zero chance (hooked or not).
IMO this assumes facts that are not necessarily in evidence.
As we all know, there are many need blind schools trying to increase the number of Pell grant recipients in every class.
Does anyone think these schools are leaving that to chance?
A blanket policy is not going to get at that, someone (or a computer model) must look at an individual applicant’s details to make the determination if they qualify for Pell so that enough of these applicants are admitted to reach the school’s goals.
Whether it’s using zip codes or other correlates…
or being need blind thru the first iteration of admission decisions, but not for the subsequent rounds where many ‘trades’ happen…
or the FA office is talking to the 3rd party enrollment management company who in turn directs admissions…
or any number of other strategies…
IMO that is need aware at the applicant level.
Estimating portion low income can still occur while being need blind. For example, the Harvard lawsuit docs state the following. Harvard readers are asked to identify low income students without having access to information about family income levels. The depositions from Harvard employees suggests that in addition to HS and neighborhood characteristics, parents’ occupations are also a key factor in how admission readers decide whether an applicant is likely to be SES “disadvantaged” without knowing family income. Getting this “disadvantaged” flag was associated with a significantly increased chance of admission (Black applicants were an exception and did not appear to have an increased chance of admission for getting the flag).
The more recent Harvard reading procedures since the time of the lawsuit state that this reader assigned SES “disadvantaged” flag was accurate ~78% of the time, as quoted below. That’s a lot better than random chance, but there are also a good number of errors. With this degree of error, I think it is safe to say that they are not using parents reported income on FA to make the estimate.
The reading procedures also mention applicants have a low income predictor metric, which sounds like it is calculated by a computer model.
Yes, there are many ways to gauge SES status that don’t require knowledge of ‘family income levels’…which is not necessarily the definition of need blind.
Edited to add: we can construct many situations where an AO wouldn’t know ‘family income level’, yet they could know if the applicant is pell grant eligible or even know the bottom line aid number. Neither of those scenarios would divulge ‘family Income level’ to the AOs.
IMO this type of profiling is not need blind…even if many might agree the end goal is worthy.
I do not believe that the “need blind” schools do not know the individual finances and/or need of a prospective admit prior to extending an offer of admission. I simply do not believe it.
It doesn’t matter whether someone believes it or not. What matters is that individual colleges are fulfilling their own strategic objective of enrolling first gen or Pell eligible or Native American or Af-Am or whatever their own goals happen to be. And do so without breaking their financial aid budgets.
And gosh- in real life, do none of you recognize the address of your local housing project/subsidized housing apartment when you see it? You don’t need to be Sherlock Holmes here…
Agreed, @blossom. It is not difficult to guess which applicants are Pell kids, so I am surprised the readers are only 78% accurate. Moreover, an applicant’s ability to overcome such circumstances is often highlighted in the application or letters.
I used the Pell Grant example because enrolling more of these students has become an important initiative. But the same strategies apply to all applicants, regardless of family income. It’s no coincidence that every year at some highly selective unis and a number of the elite LACs there is roughly a 50/50 split between matriculants who are full pay and those who have financial need.
In regards to the 78%, note that Harvard’s definition of “economically needy” seems to correlate well with parents having less than ~median US income. They are flagging kids as economically disadvantaged who have below ~median US income, not just Pell grant kids. The Plantiff was critical of the practice in the lawsuit saying, “Harvard’s current admissions system thus asks admissions officers to treat the bottom half of the U.S. income distribution as an undifferentiated mass.”
I expect that admission readers are rarely are grossly errored in their SES estimate, such as incorrectly estimating that a $300k income kid is economically disadvantaged. Instead I’d expect it more to be guessing that a particular $90k income kid is disadvantaged or a $50k income kid is not ($50k is expected to get SES “disadvantaged” flag under Harvard’s system). You can see this effect in the previously linked Harvard internal study. The $80k+ income group is not expected to get the flag, yet a small portion of kids in the above average $80k to $120k income group appear to have a low income advantage, with a slightly higher admit rate than expected. However, few above but few above $120k appear to get the low income advantage, as measured by difference between expected an actual admit rate. Readers’ estimates of SES “disadvantaged” appear to be very accurate for >$120k income kids.
I wouldn’t be surprised if over 50% of the qualified (by some minimum score) applicants to the Ivy plus schools were full pay. No do I think it matters to Harvard at all if they enroll 100 more or fewer full pay, so it quite likely does work out.
Among Ivy+ colleges, I think MIT has the smallest portion of full pay students. They have some different practices from Ivies that result in a reduced enrollment of high income students, including giving less weight to ALDC hooks. In the most recent CDS class, 90% of MIT students received FA, and 63% of students received need-based grant aid from MIT. The increased FA need among MIT students likely contributes to MIT not being as generous as HYPS… with FA for individual students. For example, in another thread I compared the NPC cost to parents at MIT and Harvard, as summarized below. There were large differences in cost for middle income families, with MIT sometimes as much as $20k/yr more than Harvard.
$50k Income: Harvard is $0k, MIT is $3k
$65k Income: Harvard is $0k, MIT is $8k
$100k Income: Harvard is $5k, MIT is $19k
$150k Income: Harvard is $15k, MIT is $39k
I expect that colleges with as large a class size as Harvard and MIT have a good idea about what portion of class will be near $0 cost and what portion would be full pay, However, this does not mean they are selecting their class according to meeting a quota of full pay or full need kids. I’d expect it’s more they have a good idea from what happened in their previous admission cycles, then adjust as desired based on future goals. If they want to increase portion low income students, then they might have an improved FA incentive or adjust recruiting to better target low income kids. They might increase sticker price more than inflation for full pay kids to compensate for increased near $0 pay kids. Ivy+ colleges tend to get the bulk of their revenue from non-tuition sources, particularly endowment returns. This gives them a good amount of flexibility with year-to-year changes in tuition revenue.