"Race" in College Applications FAQ & Discussion 12

Applying to College
3141

@roethlisburger - “32% of undergraduates receive Pell Grants. If you want to brag your college has SES diversity and increased exposure to low SES students, your institution should exceed the national average.”

I have ZERO interest in bragging about anything, so please let’s not go there. If Princeton and its peers CAN increase the low SES kids to the point where they can replace the entire aristocratic SES kids, I’m all for it. However, these institutions are bending backwards to get those academically qualified low SES and URM kids to their campuses, yet the same old hackneyed criticism gets hurled at them for…lack of SES diversity! One can only deduce from this interesting phenomenon are questionable psychological motives of those critics rather than anything grounded on intellectual honesty.

3142

@OHMomof2

I’m not going to argue with you over $4,000 in student summer work and work study grant or whether that technically constitutes a full ride or not. All I know is that I’m sure there are many poor families that would be more than happy to attend the likes of HYP for $4,000 a year in student work money. Wouldn’t you? I would.

3143

It’s more complex than that. Harvard takes risk with their endowment, which leads to a highly volatile return. One year might be a 30% gain and another a 25% loss. They can’t just count on a 10% return. For example, Harvard’s inflation adjusted endowment was higher 10 years ago in 2008 than it is today. They can’t even count on beating inflation over a 10 year period. No high endowment colleges have target endowment payouts that average 5+3 = 8%, for good reason. The endowment is also not just a bank account that Harvard can use how ever it chooses. The endowment represents all of Harvard’s schools – the business school, law school, medical school, … not just to serve undergraduates at Harvard College. The vast majority of the endowment related spending is required to be used for specific purposes and/or specific schools, not unrestricted funds.

There are also less direct financial consequences to admitting an extra 30% of the class of no parental contribution students besides just the cost of additional financial aid. You also need to consider how much smaller the portion of revenue from tuition becomes, as well as less direct consequences from removing 30% of the class who are mostly wealthy. If the class distribution changed like predicted Harvard’s expert, then the portion of legacies in the class would decrease by a factor of ~70%. Children of big donors, wealthy/connected, and so on would have similar if not larger decreases. This impacts donations/gifts, which is another key source of operating revenue.

There are good reasons why elite colleges with a larger portion of lower income students than HYP… do not have the degree of financial aid that HYP do. For example, MIT was mentioned above with 42% of students from bottom 80% income in the NYT study, which was notably higher than other HYP… type privates. With the larger portion of students from not wealthy families, MIT doesn’t match the FA of HYPS and similar peers. A comparison of parental contribution for Havrard and MIT at different income levels, according to their NPCs is below. If Harvard were to admit a larger portion of lower income students, I expect they’d need to do something similar to MIT – not being as generous with FA for middle class families, and requiring a greater parental contribution from lower income families. The larger the change in portion of “disadvantaged” families, the larger the change in cost to parents off lower and middle class families would need to be.

$65k Income: Harvard is $0k, MIT is $8k
$100k Income: Harvard is $5k, MIT is $19k
$150k Income: Harvard is $15k, MIT is $39k

3144

Some replies randomly listed. (Sorry, too tired to be diligent about quoting right now.)

Someone mentioned summers, and EFC (if not those exact letters).
At D’s Ivy, our expected family contribution was 500/yr.
She had to contribute by working a part-time campus job – a limited number of hours per week, which she did gladly for the fine education. Different positions as an underclassman, a position in her field as an upperclassman. That position worked handily into her eventual after-graduation job because she could claim experience. And the latter in turn gave her a significant edge getting into another Ivy for grad school. Given such a pathway (again), she was not complaining about having to work, nor was I. I also believe that, while she was always a good time manager, having small part-time jobs at a demanding academic institution, while still keeping up her extracurriculars, was really good life experience for her and also gave her confidence in nailing her first job after graduation, which she was offered 2 months after she finished her senior year.

Also, her school had a routine for FA students where they would sit them down at the beginning of each academic year and “adjust” their package to meet new needs, and they would also make adjustments during the year. That is, if she ran out of money for whatever reason, they simply gave her more. I’m not making that up.

The reality with low-SES kids is the travel (and yes, the books – someone mentioned). The time factor of travel, too, by the way, if it cuts into the part-time job or consumes a lot of time, such as at Thanksgiving if it’s coast to coast. There were several times when she could not afford to come home. She was fortunate that she was invited by classmates to their homes in the East. One summer I could not afford a ticket for her, nor could she. It was very difficult for the family. Ditto for some Thanksgivings – an expensive time to travel.

On the question of “majority” students “wanting diversity.” Yes, but it depends on what you mean by the word. For my D, it meant that she didn’t want a duplication of the imbalance she would have experienced at her state flagship, but it also meant that she wanted to meet * international* undergraduates, which she did at an Ivy and would not at her public. Almost by definition (not quite, but close), public universities are less nationally diverse and certainly less globally diverse on the undergraduate level. She had plenty of exposure to both URM’s and low-SES throughout her childhood, but broader exposure she did not have.

Finally, regarding any claim that if a college just becomes “diverse enough” (quantitatively), social integration among races will just “happen” due to greater visibility, that is quite naive. I am well acquainted with institutions that are 50-50, yet self-segregation persists in the 21st century – whether that’s whites + "ORM’s’ or URM’s + non-URM’s.

It is not about “numbers.”

3145

I agree with the gist of @TiggerDad comments when it comes to costs. If a low SES student can find a better deal than the Ivy’s (from an academic and cost perspective) I would like to know about it. Hard to complain over $4,000 a year for that education, no matter how poor you are.

@epiphany You are correct that the self-segregation of races on college campuses across America is real. There is also many examples of intergration between races on college campuses. My own thoughts on the subject is that URM students look for that “critical mass” of students who look like them, which actually makes some of the self-segregation possible on college campuses. But the social integration of races can not happen if there is no diversity on a college campus in the 1st place (like your example of internationals at your public flagship example points out).

3146

Harvard endowment averaged more than 10% over the last 20+ years. It actually underperformed many of its rivals (especially Yale recently).

That 30% number you used is bogus (see post #3135). I entertained it in a prior post to show even if that were the case (which will never be), what the economics would be like. Harvard won’t, and can’t (as some of you have argued), fill its class with additional 30% low SES students (let’s define low-SES to be household income below US median) to reach the 50% goal (quota?) in order to represent the socioeconomic distribution of the entire US population. No one suggests that’s feasible, or even desirable.

That statement proves HYP CAN increase the proportion of lower SES students, doesn’t it? They just have to realign their priorities.

3147

Note that the “critical mass” notion may also apply to white students, if one believes a comment by @Hanna that white students prefer white majority colleges.

In terms of self-segregation, sometimes that is most obvious in fraternity and (especially) sorority systems, where most chapters at some colleges are highly segregated, and even the umbrella organizations (IFC, NPC, NPHC, MCGC) are separated mostly by race/ethnicity.

3148

"You think more than half the accepted class at Harvard got into MIT and CT?? Please. I doubt half the accepted H class even thought about applying to MIT and CT, they’r eon opposite coasts for one thing, and have very different focus and reputation.

I’m thinking the number of kids choosing from those 3 is more like 10."

I never said that, the size of Cal Tech is too small to have that many overlaps with Harvard or Stanford. I said that 200 kids up to a 1000 could be in the pool. Taking just the Intel finalists, there are 40 of them each year and according to the Crimson, 25% of Intel finalists are at Harvard, so 10 enrolled, figure 12 admitted. Most if not all the 40 finalists will get in to MIT or Cal Tech. There are 250 semifinalists, I suspect the semifinalists admit rate is 10% so another 25 kids, so you’re at 40 with the science kids. Then you have to add in math olympiad winners, may be another 25 or so getting admission to these four (HSMC) out of about 800 finalists.

And there are 750 students at Harvard majoring in CS and Applied Math, why wouldn’t at least half if not more be cross admits to MIT, Cal Tech and Stanford? The numbers are there.

3149

@ucbalumnus I totally agree with @Hanna . I have seen a similar phenomenon with certain school districts where once the district is no longer majority white, the number of white students drop precipitously until very few are remaining. This “white flight” is partly due to the same dynamic mentioned with minorities and a having a “critical mass” and some institutions may try and avoid losing those white students (hence the hooks that benefit white students most to keep the numbers above a certain threshold).

3150

The return is highly volatile. Over the past 10 years, Harvard had an 8% inflation adjusted loss in endowment value. Had they increased payout by 3% and kept everything else the same, the inflation adjustment loss would have increased to 32% of endowment value. Note that Harvard’s fiscal year ends in June, so this 32% loss does not include the recent market downturn. It would take many decades for Harvard to dig itself out of this hole while maintaining an 8% payout, with a constant 10% market return. The longest bull market in history will not continue forever, and the time would extend to even more decades, if there is another stock market crash. Yes, Harvard made some poor investments and did worse than most rivals, but Harvard cannot count on making good choices in investments every year. In some decades they do well, and some decades they do not. The returns are even more volatile on a year to year basis, yet expenditures need to remain relatively constant from one year to the next. None of Harvard’s rivals attempt a payout anywhere near that high for good reason.

Saying “For an endowment that generates 10% return annually, it’s far from financially non-viable” certainly does not sound like it’s not feasible. In any case, the numbers I quoted with the extra 30+% were from the Plantiff’s proposed race neutral admissions policy. It’s not just an extreme hypothetical discussion. It’s what the Plantiff argues Harvard should do.

No one said Harvard cannot increase the proportion of lower SES students. However, there are consequences for doing so. For example, MIT has made the admirable decision to not give preference to legacies. Harvard takes a different approach and gives a strong preference to legacies, on par with preference for URMs. The legacies that Harvard admits tend to be extremely wealthy, with ~half reporting an income of more than $500k/yr in the freshman survey. The lawsuit indicates that the legacies Harvard admits tend to be extremely White, with all over races underrepresented. If Harvard switched to MIT’s approach of not favoring legacies, they’d increase both SES and racial diversity, while also improving the quality of the class as measured by admission stats such as scores, and academic/EC/personality admission rating. However, there would be financial consequences for doing so, which would likely result in less financial aid being available for middle class families and some lower income families, moving in the direction of MIT’s financial aid.

3151

Harvard is raising future leaders of America, MIT the future engineers and scientists, i.e. future workers of America. As most of us know, leaders don’t really need to learn or know anything, while workers do. :))

3152

That seems to be incorrect. That 50% number (or an increase of 30%) came from deposition or testimony (not clear which) by Sally Donahue, director of financial aid at Harvard, according to this (2nd paragraph, p30): https://samv91khoyt2i553a2t1s05i-wpengine.netdna-ssl.com/wp-content/uploads/2018/06/Doc-416-1-Kahlenberg-Expert-Report.pdf

That same report also shows what Plaintiff’s modeling results (for Harvard’s admitted class of 2019) are (with race-based approach replaced by SES-based approach):

Race-Based Admissions…Race-Neutral Admissions
White 40.4%…39.6%
African American 13.6%…10.1%
Hispanic 12.9%…13.5%
Asian American 23.7%…27.6%
Other minority 9.3%…9.2%
Disadvantaged 17.4%…54.3%
Advantaged 82.6%…45.7%
Academic Index 227.8…225.9

Advantaged/Disadvantaged (students) are based on Harvard’s own definitions, which takes into account family income, 1st generation status, schools/neighborhood, etc.

3153

The number I quoted in this thread was “No Cost To Parents: 18% → 48%” or a 30% increase. This 30% increase was computed by Harvard’s expert, Professor Card, in model 1 at https://projects.iq.harvard.edu/files/diverse-education/files/legal_-_card_report_revised_filing.pdf .

The numbers you are quoting were computed by Professor Arcidiacono – the Plantiff’s expert using model 4. The main differences between model 1 and model 4 are model 4 doesn’t remove athletic preference because the Plantiff thought that might be considered too controversial, and adds a new requirement about spreading out admits among a variety of neighborhood clusters, which leads to greater racial diversity and too a lesser extent a larger portion of admits “disadvantaged”. This difference in models is the primary reason for the difference in percentages from my post. However, there are also some minor additional differences in Arcidiacono and Card’s respective models that lead to smaller differences in percentages, as well different conclusions about whether there is bias against Asian applicants.

3154

So which one is correct?

3155

The Plantiff argues that Harvard can and should achieve racial diversity through race neutral admission models that emphasize a boost for low SES instead of race. The Plantiff proposed several admissions policies that boost low SES and remove URM preference, such as the model 1 and model 4 that were discussed. Both the Plantiff’s expert and Harvard’s expert simulated the Platiff’s proposed race neutral admissions policies and estimated that the class would change in similar ways under those policies.

As to who is correct, I side more with Harvard’s expert on this issue for the reasons I’ve discussed in this thread. Harvard’s expert claimed that none of the proposed changes reach the same level of URM diversity as Harvard’s current class, which is true, although the importance of having a decreased percentage of Black students in the class is debatable. And Harvard’s expert claims here would be a variety of complications from making ~half or more of the class financially “disadvantaged” and removing all hooks except for athletes including, but not limited to financial issues, which I also agree with.

3156

By “So which one is correct?”, I meant which one of YOUR statements is correct. They seem to conflict with each other.

3157

Really, what CC has less SES diversity, defined as and measured by the percentage of students receiving Pell Grants and the percentage of students from the bottom 20% of the income distribution, than Harvard? It must be an unusual CC to skew that wealthy.

3158

I thought my previous post was clear. The statements do not conflict with each other.

**Statement 1: the numbers I quoted with the extra 30+% were from the Plantiff’s proposed race neutral admissions policy. It’s not just an extreme hypothetical discussion. It’s what the Plantiff argues Harvard should do.

Statement 2: The number I quoted in this thread was “No Cost To Parents: 18% → 48%” or a 30% increase. This 30% increase was computed by Harvard’s expert, Professor Card**

Both the Plantiff’s expert and Harvard’s expert, Professor Card, simulated the Plantiff’s proposed race neutral admissions policies of what the Plantiff argues Harvard can and should do. The referenced simulation by Harvard’s expert, Professor Card, found a 30% increase from 18% to 48% under the Plantiff’s proposed race neutral admission policy.

3159

Where did you get the information that the Plaintiff proposed the extra 30+% increase? The exhibit I linked to earlier contradicts your statement 1.

Additionally, the plaintiff didn’t propose for disadvantage admits there should be “No Cost to Parents”. That’s entirely Harvard’s assumption in its model.

3160

The quote stated “the extra 30+% were from the Plantiff’s proposed race neutral admissions policy.” The Plantiff proposed a race neutral policy, and both Harvard’s expert and the Plantiff agree that the proposed policy would increase by ~30+ percentage points, with specific numbers depending on model. I referenced Harvard’s expert’s simulation of model 1 which had an increase from 18% to 48% – exactly 30 percentage points. Your link shows almost the same increase for this model and greater increases than 30 percentage points for the other models. Specific numbers are below.

From your link showing the Plantiff’s simulation of the Plantiff’s proposed race neutral admissions policy:
Model 1: 18% → 46% (28 percentage point increase)
Model 4: 18% → 54% (36 percentage point increase)

From my link showing the Harvard’s simulation of the Plantiff’s proposed race neutral admissions policy:
Model 1: 18% → 48% (30 percentage point increase)
Model 4: 18% → 56% (38 percentage point increase)

Recall that the “disadvantaged” flag threshold was approximately $60k to $70k, under typical situtations. Harvard’s threshold for no cost to parents is also $60 to $70k, under typical situations. This is also the threshold Harvard OIR calls “low income” in the lawsuit. I have used “low income”, “disadvantaged” and “no cost to parents” to indicate this threshold in this thread. However, the specific criteria used in the models above was the “disadvantaged” flag.