That’s a bet I’d take, zoom. There are an awfully lot of idealistic AO’s ready to blow the whistle if a thing like that would happen on their watch. It would not only be a matter of moral obtuseness but a flagrant lie - many lies - purveyed in all the materials and statements of policy emanating from this University over many, many years. That would be a pretty big risk to run. In the end the conduct of an institution of higher learning is not directed to maximizing the bottom line, hard as that might be to believe. John D. Rockefeller complained bitterly to President Harper about this, but he kept on giving.
Marlowe. what do you mean “blowing the whistle”? As I said, the annual FA budget is determined by the Board/President/Provost and the Admissions office is charged with meeting that budget fully by admitting enough kids who need FA up to but not exceeding their pre-specified budget. Unlike many things, the FA budget is quite transparent and everyone in the AO knows what it is.
I am just saying that once the FA budget has been met, then any additional kids who are full pay will be highly favored vs admitting an additional kid who will “bust” the FA budget. Are you trying to say that colleges don’t have annual FA budget that is fixed and determined by the President/Provost and approved by the Board just like every other aspect of the university’s annual budget?
You keep on insisting that admissions offices seemingly just admit kids without any regard to their finances and that any consideration of finances is therefore a lie. Yes, the admissions office does admit enough kids who need aid to meet its FA budget and this is what colleges really mean when they say they are need blind. Need blind doesn’t mean that colleges will gladly exceed their Board approved FA budget and not take the kid’s finances into consideration once the FA budget has been met.
I don’t know what you do for a living but if you ever worked in a corporate job I assume you understand how budgets are set each year both overall as well as departmentally and by line item. And that you also understand how difficult it is to ask for more than your departmental budget unless you have a very compelling reason and how even these pre-set annual budgets can be adjusted downward if there is an unexpected downturn in the business.
Maybe we are just arguing past each other because we don’t share the same understanding that a college’s FA budget is fixed and set each year by the Board and admin. But once you accept that fact, then then there is no need to argue that anyone is lying because the AO IS need blind up to the limit of their FA budget but not $1 beyond that limit.
I’m with zoom on this one. Schools have a net revenue target (revenues less FA), and they can’t leave it to chance that the institution meets that goal. Schools are using predictive analytic models to maximize yield and/or achieve net revenue goals, and strategically disburse FA.
While I believe (most) need blind schools are need blind in the sense the AOs don’t know whether an applicant applied for FA and/or their level of need if they did apply, there are many ways AOs infer income levels from an application. Zip code, HS, parents’ education levels, parents’ job titles, URM status, first gen status, and ECs, to name a few…and these variables are absolutely a part of the admission decisions when the team is building its class.
There are also direct ways schools know whether an applicant has financial need or not. For example, high school counselors (especially those at boarding schools) may tell College AOs which applicants have need and which don’t (FA applicants from boarding schools are highly attractive).
Also, during the DIII athletic recruiting process, coaches at some schools (don’t know about U Chicago) ask recruits whether they will be applying for aid. Then, they pass this info on to the AOs with the pre-read materials.
Chicago is need aware for internationals, so for those applicants (13% of the enrolled class) they do know FA status and level of need. Many need-blind schools are need aware when going to the waitlist as well (again, don’t know about U Chicago)
In summary, schools like Chicago are not leaving important institutional priorities and goals like yield and net revenue to chance.
For an example of what I mean by blowing the whistle, zoom, here is a good example from a thread from March 17, 2019:
By comparison with what you are suggesting this was a tempest in a teapot: A staffer who felt wealthy kids were being favored for summer internship opportunities was so incensed by this that she took what she thought were the incriminating emails in the affair to the Maroon, where it became a front page story. The event complained of was far from what she made it out to be, but for my purpose it is Exhibit A in proof of the high level of sensitivity on this campus to any conduct that would show the sort of favoritism you advocate and believe to be actually in progress - that of a policy of favoring wealthy kids over poor ones or at least less wealthy ones in the admission process. How such a thing doesn’t stink in your nostrils is hard for me to fathom, but I can assure you that the nostrils of these young AO’s would be at least as sensitive as those of the staffer in the situation above, both as to the injustice and “the overpowering odor of hypocrisy”.
As for budgetting, the University ought to have a pretty good idea in advance of the admission of any class how the breakdown among demographics in the class will fall out and thus the projected level of FA. You can call that figure a target if you wish, but it would be surprising to me if the figure couldn’t be predicted with a very high level of accuracy well before the admission of a single applicant. But what you are suggesting is much more than this - that decisions on individual admissions all down the line with respect to all the 2,500 or so admittees will be made with a view to achieving this target, as if that was what the purpose of the enterprise. Forgive me if I say that this is simply ridiculous even if it were possible and even if the University had any chance of getting away with such conduct without covering itself with shame and embarrassment. That’s no way to run an institution devoted to learning.
Mwfan, you are absolutely correct about how schools can infer a kid’s family wealth but we don’t even need to make this more complicated than it needs to be. There is a very simple and direct way that schools know if a kid will be asking for FA because this very question is asked on the Common App! The answer choices are Yes or No!
Now an individual AO is not supposed to take this into account when he/she reads the file and makes a recommendation, but I guarantee that the responses to the Yes/No FA question are entered into a predictive analytics model along with a ton of other demographic data which then informs the AO what their preliminary wish class will look like on all sorts of dimensions including whether the amount of aid needed is within its FA budget. And if the model indicates that the FA budget will be exceeded, then the AO will revise its preliminary admit list by taking some kids off and adding others until the model confirms that the FA budget will not be exceeded.
And when it comes to the wait list, unless the FA budget has not yet been met based on which kids have accepted or denied their offers, you should bet that kids admitted off the waitlist will be full pay or close to it, and that is exactly the rational and financially correct thing to do from the point of view of the college.
Marlowe, again, you seem to be suggesting that I am arguing AOs review individual files with an eye towards their finances. That is not at all what I am saying. Rather this is how I think things work:
- Each applicant file is initially read by 1 or more AOs WITHOUT regard to his/her finances (need blind) and a preliminary recommendation is made to Admit, Deny, or Waitlist
- Then all of these preliminary decisions are assembled into a preliminary class of admits (let;'s say this list of admits is 2000)
- Then this preliminary class list is run thru predictive analytics model that informs the committee what the class will likely look like across a whole spectrum of demographics. For example, if the model says that the class will be 70% men and 30% women, then the committee knows it needs to make some changes.
- Then adjustments are made by taking some kids off the list and adding others and the list is rerun multiple times iteratively thru the model until the class reaches the desired proportions across a wide range of demographics and institutional needs.
- Everything up until now is just preliminary. Only when the model confirms that the class will represent the desired demographics are the formal admissions decisions finalized and released. This process ensures that if the FA budget is $200 million that the estimated FA need for the admitted and yielded class is also very close to $200 million.
Note that the AO is experienced enough such that its 1st preliminary class list will likely come reasonably close to what is desired so the number of kids who are moved from Admit to Deny/Waitlist and vice versa is probably just a few hundred rather than a complete makeover of the list.
Based on this process, the colleges can claim that they are need blind because they are indeed need blind when doing the initial review and recommendation of each file. But make no mistake: while the preliminary draft class list has been composed in need blind fashion, there will be adjustments made to this list before official decisions are released based on predictive analytics modeling to ensure that the FA budget is not exceeded and this goes for all other demographics deemed to be critical, not such FA.
Need blind schools suppress the ‘did you apply for FA question’, so the answer to that question is not visible to AOs, nor in the analytic model…not sure how you can ‘guarantee’ something is so for all colleges.
For need aware schools, the yes or no question may not be a variable in a model, because they can use the actual need number which is far more valuable as far as information goes.
Agree on the waitlist situation, although this year, applicants are being pulled off waitlists that do have need…but not surprising as this year’s admissions cycle is highly atypical.
Some colleges do look at need after the first read, but I don’t know any colleges that say they are need-blind that do this…do you?
The process you describe is how some schools do it, like Trinity college in the NYT article linked above…but that is a need aware, or need sensitive, school.
sgopal2 wries: In my son’s graduating class at Lawrenceville, there were about 10 students who were accepted to UChicago. I think 7-8 decided to matriculate. Of that I am aware of only one student who needed financial aid. This is obviously a small sample to extrapolate from. But I highly doubt that half of the matriculants from Andover/Exeter would need FA at UChicago.
Well, that’s the difference between Lawrenceville–which, like the StGrottlesex schools–was founded for the rich and the older New England academies, like Exeter and Andover, which are, as I said, 50% FA. In my son’s graduating class at Exeter, my impression is that more than half of than half of his classmates going on to UChicago are on FA there.
Exeter: 48 % on Financial Aid https://www.exeter.edu/admissions-and-financial-aid
Andover: 47% on Financial Aid https://www.andover.edu/admission/tuition-and-financial-aid
@zoom10 , all these Rube-Goldberg-like levers, trapdoors, and pulleys that you envision in your #45 are inventive and detailed, but implausible, though I credit you with creativity. The high-technocratic-phrase “predictive analytics model” is doing a lot of heavy lifting, but here’s a word I prefer - “bafflegab”. Can you cite a source for anything like the detailed process you are asserting so confidently? Or is this really just speculation on your part? Perhaps such a dystopian nightmare of a scheme exists at some lesser institution, but could anything like that be in actual operation at a school anything like the U of C? You still haven’t dealt with the plain vanilla envelope under the door that would surely surface if that was the case.
That NYT article was quite interesting.
One key difference is that Trinity is clearly not need-blind. That’s very obvious when you read the article, and it’s confirmed via Wiki’s list of need blind vs. need-aware institutions. Trinity is on the “need-aware” list (for domestic applicants). Chicago is need-blind for domestic and need-aware for international. Both institutions apparently meet full demonstrated need, although in Trinity’s case that’s of course conditioned upon the need-aware admissions process.
Chicago’s algorithm is a bit more complicated for a couple of reasons:
-
Some will check that Common Ap. box even though once admitted they are not given FA. Among other things, UChicago will waive the application fee if you apply for need-based, regardless of the outcome. Some will be “on the edge” financially and will figure it’s worth a shot, others will check the box just see what happens.
-
Some will NOT check the box because they don’t trust that any school’s admissions office is truly “need-blind.” However, they do qualify for FA and, once admitted, will apply - or will apply in Years 2-4.
Note: as long as the student hasn’t missed a deadline (and sometimes even if the student has) at UChicago they will be given aid meeting 100% of demonstrated need.
There’s a reason why only a handful of universities offer both “need blind” and “full demonstrated need.” It take a very healthy-sized FA budget to be able to do this. UChicago has years and years of data at its disposal to create a reasonable “worst case” drawdown- which of course won’t be as extreme as offering free tuition and R&B to everyone in the class. As long as they have the funds to meet that scenario, then they are NOT choosing their admits a la Trinity College.
The question about the FA box on the CA has been brought up on the UC thread in the past and I see that Mwfan already answered it. It’s pretty obvious that the CA FA question would be a violation of “need-blind” policies if that information gets fed to their predictive models. And, heck, in that case what’s to keep them from entering complete financial information from an older sib’s application? Does any of us believe that really happens? Major whistle-blower moment, as Marlowe has pointed out.
Nope. They follow “need-blind” procedures. Now, clearly there are carve outs. Questbridge and other college match services for low-income candidates or specific development cases are examples of where the decisions technically happen outside of the normal “need-blind” process. Whether they are determined outside of UChicago’s own predictive analytical process might be another matter. A lot depends on what UChicago is predicting with those models: success in the College, or “ability to pay.”
@Zoom10 in #45: I tend to agree with the process you’ve laid out but you were a bit vague in parts 4 and 5 w/r/t the financial aid budget. How exactly do they ensure that those reiterations will result in a class that matches the budget w/r/t ability to pay? And how do they do so w/o violating “need-blind”?
Note: like you, I believe that these prediction models potentially allow them to get around the question of “financial consideration,” but I’m not sure how. Is it a simple matter of parental education, what schools you and your sibs all attend, your EC’s etc? Are the prediction models that good?
The prediction models can be highly accurate without even inputting parent’s alma later or job title or ECs pursued or Yes/No box for FA.
All you basically need is the zip code/home address of the family’s residence and the name of the high school. The zip code and home address instantly gives the model the median household income and the model can also use the market value of the primary residence from a Zillow search as an input. The name of the high school provides info whether the high school is public or private and what the tuition is at the private high school which is a strong predictor of ability to pay. Also, the college can use historical data of kids’ FA status from that high school.
Again, the model doesn’t have to use any personally identifying information for any individual student in its calculations, it just uses anonymous data inputs to spit out a estimated class composition across all dimensions.
If people still have a hard time understanding how this works, or accepting this reality, let’s put aside the topic of FA and need blind admissions for a second. How do you think colleges magically mange to produce all the other demographic proportions that are remarkably stable within +/- 2% points each and every year. By these stable proportions, I am referring to gender balance, racial/ethnic proportions, Ist gen proportion,proportion of athletic recruits, etc. The Harvard lawsuit revealed and asked the question: how is it that Harvard’s class has exactly 14% blacks, X% whites, Y% asians, Z% hispanics, etc. that is essentially constant within +/- 1-2% each and every year when the proportions of these groups that are applying each and every year is not constant. The politically incorrect but honest answer is that these results are intentionally and purposefully manufactured… No college’s AO wants to reveal how the sausage is really made, which is why it takes a lawsuit for these uncomfortable details to come out.
So yes, I stand my assertion that all colleges will ensure they don’t exceed their FA budget and this statement is not contrary to the fact that they do practice need blind decisions to generate a preliminary admit list; after this list is generated, adjustments will be made as needed to ensure all institutional needs are met, including ensuring their net revenue target as described earlier.
Completely agree with @zoom10 post #45. This lays out clearly how things are hashed out. There is no way this could be done without knowing about financial need – not an individual student’s need, but the need of the entering class as a whole.
How else would the financial aid budget be so remarkably similar from year to year?
- This might be so. Application data is fair game, since it's on the CA. Parental education and profession, zip, high school, where your sibs go to school (grammar, secondary, post-secondary) are all predictors. Every high school will also have a profile as well that is submitted. This data can be analyzed completely without reference to the individual applicant.
- To answer an earlier question that Marlow and Zoom were pondering, I'm not sure that UChicago has to worry much about having to choose between a full-pay kid and an equally deserving FA kid. Obviously the data analytics can anonymously predict where a proposed class falls amidst the income distribution and I don't believe that violates "need-blind." I think this means that you compete against others in your (predicted) income segment for a spot in the admitted class but not necessarily against other segments.
Here is how it might work: very low and low SES, ie the kids who have access to Odyssey or Empower funding, aren’t able to attend without that funding. That funding is also reserved specifically for them and not for wealthier admits. Finally, it’s limited; UC can only admit so many Odyssey and Empower kids before the funding is used up. So Admissions already knows it will be admitting up to a certain number of low/very low SES who are covered by dedicated funds. Per Zoom and others, the predictive analytics models can identify pretty well who in the admitted pool (first iterations) fall in that group. They will be tallied up and checked against the available funding and if numbers outstrip funding, some will be eliminated and the process will begin again. Eventually, the model will optimize over this characteristic and others as well to arrive at the well-distributed class.But here is the bottom line: low income kids are basically competing against each other. They are not competing against wealthier kids. So the models can be calibrated over a few segments: full pay, partial-pay, free tuition, free tuition + R/B. Within those segments, some students will be added and some eliminated until the model generates the “optimal” class. I think what this means is that you compete against others in your predicted SES segment. You really don’t compete against other segments.
I have to wonder whether low SES applicants in need of very full FA are ever likely to exceed a modest percentage of the class. This is due to limitations in their academic preparedness, not the threat that they might bankrupt the University through their large numbers in a head-to-head competition with more privileged fellow applicants. I can also readily agree that the attempts of the University to expand the representation of this demographic through preferences and FA will have to reckon with limitations on the University’s financial resources.
I take zoom as saying something more than this - that in a true competition of academic equals the wealthy kid gets the preference. Wealth would in itself be a hook, papered over on the basis that the first sorting out would be need-blind and that the zip codes would come into play only in making the ultimate determinations, thus allowing the University to continue to claim that admissions is need-blind (with an unacknowledged asterisk). Have I got that right?
I am still awaiting anything like proof of this at the U of C. Given that it would be such an egregious breach of faith and simple honesty on the part of the University, a bare assertion is not enough to convince me. Show me the plain vanilla envelope!
That NY Times article is fascinating - thanks for sharing. It nicely summarizes the nigh-impossible tight rope AOs have to walk, with so many competing interests (have diversity! Lots of low SES kids! We want lots of tuition revenue!)
@marlowe1 - you may be looking for a controversy when none exists, because…
@Zoom10 - is it fair to say that for any particular (domestic) applicant, Chicago is need-blind, but FOR THE CLASS, Chicago (like any other college) is hyper-aware of the class’ needs, tuition revenue, etc.?
I’ll point to the classic article on Chicago’s admissions from the 90s:
https://www.newsweek.com/inside-admissions-game-164802
Even then - 20 years ago - Chicago put all the admits “into a computer program” that gave them all sorts of data/breakdowns about the admitted group. It seems likely that - zoom forward 20 yrs - predictive modeling would gain traction.
The biggest takeaway: while the outcome of any particular applicant is NOT preordained, the composition/makeup of the entire CLASS is. This is Zoom’s point about every class having such strikingly similar makeups (e.g., within 1-2% points of each other, year after year).
This is what it means to “shape a class” - to have a certain number of these, and a certain number of those, every year. In the past, Chicago did not “shape” their classes - and it led to more wonky numbers, one year to the next (e.g., more lopsided gender breakdowns, less diversity, etc.).
More recently, it looks like Chicago “shapes” their class more and more.
Ultimately, this shows you that elite college classes don’t always reflect pure “talent.” Rather, they reflect the institutional priorities of the school.
As Zoom says, and I agree, at any elite college, “the [overall admissions] results are intentionally and purposefully manufactured.”
It’s a pretty twisted game, once you pull back the curtain.
Do we really think that UChicago believes it is “sacrificing talent?” For instance, if the first iteration skewed 70/30 females, does re-running the iteration in order to achieve a better gender balance mean giving up talent or perhaps compensating for human error? The school has enough historical data to know which it is. There might be significant costs associated with being a lopsided class.
Also, @Cue7 my guess is that they tried to shape the class 20+ years ago as well, but the high admit rate and low yield limited their ability to do so as successfully.
Bingo! JBS and Cue, you both stated exactly the points I was trying to make to marlowe but in a much clearer way that I was able to.
Under the guise of “holistic” admissions, “shaping the class” to achieve intentionally manufactured outcomes is exactly what the Ivy’s have been doing for decades. 50 years ago, this shaping of demographics and institutional needs was done manually and with crude spreadsheets; although inefficient and laborsome, the number of applications was also much smaller then . Later, this sorting of variables was done on an Excel spreadsheet, and today, the use of sophisticated algorithms and computer models has made the shaping of classes exquisitely precise.
And yes, JBS is correct that kids in different major socioeconomic and racial groups compete against each other within those groups and not across those groups, all with an intent on achieving the desired (and stable) proportions that you see displayed on the school profile for each class.
Now for some nuances.
First, no kid is monolithic, meaning each kid has the potential of knocking off more than one institutional need which is great for the AO because they can now doublecount (or triplecount or quadruplecount). For example, Joe Smith is an URM from Nevada who is recruited for football and he is the first to attend college in his family . Bam!: Joe simultaneously meets 5 institutional needs: URM, 1st gen, will require financial aid (Odyssey), comes from an underrepresented state, and recruited athlete! Next is Mary Jones who lives in Silicon Valley, attends an elite prep school whose tuition is $50K, her dad is a partner at a VC firm, she is an accomplished dancer with 35 ACT, and she states she wants to major in Art History. Wham! Mary is also an auto admit because she also knocks off 5 institutional needs: full pay, family is likely future donor beyond just the full pay, her ACT score boosts the overall class average, she will continue performance dancing in college via RSO or other group, and she claims to want to major in an unpopular department which makes the Art History Chair happy (until she actually gets there and switches her major to Econ LOL)…
Second, while I said that these admitted class percentages to meet institutional needs are very stable over time, occasionally the college needs to adjust the weighting of its percentages and institutional needs to remain competitive with its peers or because it really needs to enhance one institutional need over everything else like having more money. When US News changed its rankings formula to eliminate admit and yield rates and instead include a “social mobility” score, what did UChicago and others immediately do? Right, these schools all tried to increase the percentages of low SES kids as measured by Pell Grants. This caused Chicago’s percent of Pell Grant recipients to rise to about 15%, and it will stay there until other peers increase their % to 17%. And so on.
Third, this year’s admissions cycle has been thrown a monkey wrench because of Covid, so the predictive models are going to be off more so than prior years… For example, Chicago wants about 13% of class to be International because it adds diversity and full pay. However, the actual % of International students might be around 8% instead. So what will Chicago and its peers do? They will replace this shortage of yielded Internationals with full pay kids off the waitlist. But these kids aren’t just randomly selected among the pool of all waitlisted full pays. Rather, the model is asked to spit out 100 full pay kids from the waitlist who meet multiple institutional needs that still exist and then a human AO carefully sorts thru this list of 100 full pay kids and selects and ranks them to fine tune the final overall class. This year’s class will likely have a lower % of International students than ideally desired but it will generally meet all other institutional needs because none of this was done by accident or left to chance.