Pages

Thursday, June 08, 2017

When the bad faith of schools is identical to their good faith

Cheryl has been sending me news about the latest in hare-brained schemes to reform college admissions, including this NR article detailing its hare-brainedness. In summary, this particular proposal, generated by 100 super fancy prep schools, follows the general trajectory of liberal reform ideas in repudiating all quantifiable measures of academic achievement in favor of some form of "holistic" evaluation that emphasizes non-academic qualities. In this case, the standard graded transcript will be abolished and students will instead be evaluated for their competence in a number of vague moral and intellectual qualities, not all of which are even clearly virtues: "leadership and teamwork," "adaptability and risk-taking," etc. It's all obviously a load of BS, and Rossman aptly describes why it's stupid and how it will only end up helping the wealthy students of these fancy prep schools at the expense of regular smart kids at public schools.

The question is, why do such proposals get made? Rossman implies that it's a matter of bad faith on the part of the would-be reformers, and I had a long Twitter exchange with Phoebe, who suggests the same thing (you can only read her side unless you are my special Twitter-friend). I think it's tempting to make this argument that places like Andover are feigning concern for the plight of the poor and disadvantaged to justify policies that they actually pursue to help their own students get (even further) ahead, and that Yale is feigning it just to get more Andover students. And I can't claim to be reading their minds. But I think there are two reasons this argument may not be correct:

First, there is the problem of secondary schools generalizing from limited experience. It is very possible when you work at an elite school (or are the parent of a child who attends one) to think that the biggest problem in education today is the hyper-competition among students for college admission that completely reduces education into a vehicle for resume-building and results in depression, exhaustion, etc. That probably is the biggest problem in your school. And trying to disarm the competitors in this war by banning their weapons (grades, rankings) is a reasonable response. You face a collective action problem if you're the only school that does this, so getting 100 similar schools on board is a big step towards solving that. But the error comes when you generalize from your experience to conclude that cutthroat academic competition must be the biggest problem in all of secondary education, and that your solution will therefore eventually benefit all, when it is eventually applied to all.

I think this error plagues a lot of the people involved in these sorts of reform measures: they really do have a poisonous hyper-competition problem in their schools, and they really can abolish traditional evaluative measures without risking any decline in their students' academic performance, at least in the short run. What they don't see, or don't care to see, is that this problem is pretty limited to, well, them.

Second, and more important, there is the problem faced by universities that sincerely do want to expand access to poor students who don't have the burnished resumes of Andover kids. How should they do it? Let's put aside the question of racial minorities for a minute, since that is subject to a whole set of legal as well as merely logistical limitations, and talk only about economic "minorities," the poor of whatever race. Here is the question: if elite universities were to acknowledge that "holistic" evaluation is basically subjective garbage and admit students based entirely on "hard" factors like grades, class ranks, and SATs, would this result in a bigger representation of the non-rich? I'm not sure if anyone has tried to empirically model this possibility, but if you know of any such efforts, send them my way.

But I suspect not, and the problem I see is this: class ranks across the different kinds of schools whose students apply to elite colleges are incommensurate. The valedictorian of a very bad high school may be academically less meritorious than a kid graduating in the bottom quarter of Stuyvesant's or Northside College Prep's class. An A in an AP course at the former school may be the equivalent of a B- or C at the latter. So the only consistent measure would be standardized tests. Now, even I don't think that determining college admissions entirely by test scores is a desirable policy, and I'm pretty close to a merit absolutist on these questions. But let's say we tried it. Well, wealth does correlate with test scores, in part because the wealthy are more likely to have preparation (whether in the form of pricey tutoring or just exposure to less formal prep through schools that expect most students to eventually sit for these tests) and in part because of the heritability of intelligence (the rich can even be smart). Does it correlate enough that the vast majority of the spots at top schools would be taken by the rich and upper middle class? Again, anyone with data should now speak up. My speculation, based on what happens at exam high schools in a somewhat different context, is that somewhat more working-class and poor kids than now are admitted would be admitted through a test-only system, but that elite universities would continue to be dominated by the rich, and more-or-less rich.

Even if a school committed to expanding access did want to go this route, imagine the public relations fiasco that would ensue: "Yale drops entire application in favor of exclusive reliance on SAT scores." All the people who believe the SAT is racist and that standardized testing doesn't measure anything except how much tutoring you got - which is pretty much everyone on the left - will immediately flip out, protest and boycott and threaten to withhold donations, and that would be the end of that.

So that's the bind that even a college sincerely committed to expanding access is in. It shouldn't be that surprising then that it will embrace reforms that expand its own discretion in the admissions process, because even if this expanded discretion also has the effect of allowing them to admit more rich kids, it's probably the only way they can admit more poor ones. There just aren't many (or any?) good discretion-limiting options. All I can think of is preference system that gives a "boost" to lower-scoring but low-income students, but that too introduces discretion. That's why, even though it's easy enough to suspect that everything these schools do is ultimately self-serving and never really public-spirited, we actually can't really say in the cases of these stupid reform schemes which it is, since both intentions would rationally lead them to the same policy.

A final, rather crass but nonetheless still probably true point is that elite colleges need rich students, and rather a lot of them, not just to keep themselves solvent but to serve the poor. The service - acculturation, social mobility - such colleges provide to poor students largely depends on the historical and continued presence of the rich. These people provide the scholarships the poor need to attend, and they provide the connections they need to then become rich themselves after college. There are some instances where this is not the case, like when the economy is expanding very rapidly and any college degree is a ticket to financial security, so that attending Brooklyn College in 1954 is roughly the same in terms of social mobility to attending Columbia at the same time. But I don't think that's the case now or will ever be the case again, except if some new certification that most people don't yet have replaces the college degree. In present circumstances, elite colleges need the rich to stay elite. The big question is, how many rich kids do they need? The apparent injustice is that the answer is always: more than the percent of the rich in the population at large. There are probably too few rich people as a percentage of the entire country to make Yale function as Yale if Yale could only admit as many as are in the population at large, proportionally. So even if only five percent of Americans have incomes above some high number ($166k, as it happens), it may still be necessary that 30 percent of Yale students be that rich to ensure that the 30 percent of Yalies who aren't become so. (Actual distributions here, and not too far off.) What the precisely optimal number of rich students is, I don't claim to know, but only bring this up to emphasize another limitation on colleges with sincere egalitarian intentions.

Now, you can conclude from all this that the solution is not to abolish grades or tests or holistic nonsense, but instead to abolish Yale. So, fine, go ahead and try, but under current conditions, you won't create a world of egalitarian 1950s CUNY colleges; the only effect you will have is to turn another school, now languishing just below it in the rankings, into New Yale.

18 comments:

Andrew Stevens said...

My favorite thing about your actual distribution is it shows that Yale is harmful. A higher percentage of its students come from the top 20% than end up there. Similarly, a larger percentage end up in the bottom 20% than came from there. This doesn't really surprise me, assuming the numbers are accurate (and I have no reason to doubt them).

This is probably why I've never cared that much about these sorts of debates which are really debates among the rich with each other. Smart poor kids don't actually need Yale and never have. (Poor kids who attend Yale seem to do well for themselves - they don't seem to be the ones who end up in the bottom 20% - but probably would have regardless of Yale.) Yale's purpose is to provide entry to various Ivy League welfare programs (such as public broadcasting) for children of the rich who fail to make it on their own so they don't become a very obvious drain on society. That's my (admittedly cynical) view anyway.

Miss Self-Important said...

I ignored the context of the data to stick to my point, but I also Have Opinions about the social mobility tracking that the NYT article does. My main objection is that it doesn't dis-aggregate life-failure outcomes from low-income but successful outcomes. That is, a substantial number of people who attend elite schools go on to relatively low-paid but still high-status careers, in social work, for example, or the arts, or various forms of political activism. So I'd want to know what those bottom 20% percentile alumni are actually doing before I could say that Yale has harmed them.

On the other side, the main point of this NYT project was to show that insufficient numbers of the poor students who attend elite schools end up straight-up rich as adults. (In the case of Yale, the NYT emphasizes that only 2.1% of the bottom-quintile students end up top-quintile adults.) That I find suspect as well. First, for the obvious reason that you don't have to have to go from rags to riches to be mobile, and I'd think that a more significant mark of mobility-failure is demonstrated by how many of the bottom-quintile students end up as bottom-quintile adults, which this article does not tell us. Second, there is a problem of job-selection that is not addressed at all here. If Yale wanted to maximize its performance on this metric, it would work as hard as it could to push all its low-income students into immediate high-earnings jobs like finance and management consulting. But that would require a lot of coercion and probably be counterproductive in the end. Obviously, it shouldn't discourage anyone from these jobs, and it may indeed be true that, due to campus culture, an artificially high number of low-income students are being informally "tracked" into career paths like social work or academia so they never maximize their earnings potential. (Like me! I have not optimized! This is partly the fault of my university's culture. But I don't my university owed me more.)

That said, no smart kid - rich or poor - needs Yale. That is probably the only saving grace of our higher education system from a political perspective, that it's decentralized. Yale's purpose is, like every elite college, to provide the best possible liberal education it can for the best students who are most capable of appreciating it that it can persuade to matriculate at it. That's all. It has all these other effects and has taken on all these other functions that we should not ignore, but they are not its purpose. But precisely because this is not an egalitarian purpose, it's never going to manage to become a truly egalitarian institution.

Alex Small said...

I think your point about trying to maximize discretion for admissions officers is important. It isn't solely about trying to get a "just" outcome that gets a lot of talented but disadvantaged students (for whatever definitions of "talented" and "disadvantaged"), but also about shaping a campus culture along many dimensions, whether getting enough "diverse" students to be acceptable or getting enough rich kids to provide valuable networking or getting enough "interesting" kids to keep the conversation a certain way or whatever else.

But nobody wants to defend discretion so they wind up insisting that they're REALLY measuring merit more fairly.

Alex Small said...

For a while now I've been meandering toward the conclusion that one outcome of an obsession with "fairness" (by whatever measure) is a belief that not only must institutions achieve certain outcomes, but they must achieve those outcomes with algorithms rather than discretion. (Or at least they must appear to use algorithms.) What I'm not sure about is whether it's because the outcome they're trying to get is controversial in some quarters (so they have to distance themselves from the appearance of trying to get it, even while boasting that they got it) or because discretion would undermine the legitimacy of the outcome ("They didn't REALLY earn it") or because discretion itself is shameful privilege of some sort.

I know it's a mix of the three, but I'm thinking that the first one is a pretty minor factor (otherwise people wouldn't boast about getting the diverse outcomes that they swear are due to an enlightened measure of merit) and the others are probably more powerful than people want to admit.

Miss Self-Important said...

I haven't really thought much about this, but I suspect #1 does matter b/c if you can say you got diversity through some impartial measure like an algorithm, you're saying you didn't have to lower your standards of merit for the sake of diversity, right? And the only reason you'd want to reassure people of that is because such a trade-off is controversial in many quarters, and the optimal situation would be one in which you could say you have the smartest students who also happen to be the most diverse.

But on the whole, I'd think the opposition to discretion in college admissions is based on the same revulsion that leads to opposition to it in all other institutions: liberals (of the Enlightenment, not as vs. conservatives) hate personal authority and personal rule of every kind, because it violates the principle of natural equality. Rule by algorithm closely approximates the perfect rule of law, according to which the sovereign would be completely impartial. But all mere persons are partial in some way, and so all personal rule has to be as much as possible depersonalized to correct for that. Even the most personal rule permitted by our Constitution is only an office, whose occupant is interchangeable and forcibly changed every eight yrs, max.

Andrew Stevens said...

I ignored the context of the data to stick to my point, but I also Have Opinions about the social mobility tracking that the NYT article does. My main objection is that it doesn't dis-aggregate life-failure outcomes from low-income but successful outcomes. That is, a substantial number of people who attend elite schools go on to relatively low-paid but still high-status careers, in social work, for example, or the arts, or various forms of political activism. So I'd want to know what those bottom 20% percentile alumni are actually doing before I could say that Yale has harmed them.

I am fully in agreement that economic success is not the only (or even a particularly important) component of success. I'd go further than you since I would be even more interested in finding out how happy their families are, how successful their marriages have been, etc. than I am in whether they are in a "high-status" career or not (especially since I question whether those careers really deserve the status they have and I very much question whether such status has any actual value - I'd certainly be happy to sell whatever social status I possess for cash).

I was taking the data on its own terms - i.e. that economic success is all that matters and all that Yale "should be" for - to help poor kids attain economic social mobility. What I thought was funny is that, based on this data, it appears we shouldn't want more poor kids to go to Yale. Apparently, they're more likely to be worse off than their parents if they do. But mostly I was just making a joke about Ivy League welfare.

Miss Self-Important said...

Well yeah, but all that is hard to measure. I don't mind looking at economic outcomes, but with the understanding that a good economic outcome is a phrase meaning something broader than Being In The Top 1%. By high status, I just mean respectable. Freelance writing is a reasonably respectable career but a pretty poorly-paid one. Drug-dealing is the opposite. Drug addiction and unemployment is neither. Besides, we can't look to see if their marriages are happy b/c, according to this data, they do not get married. So maybe that's the really important indicator in all of this. But I don't think this data shows that poor kids are made worse off by attending Yale. It says that the average adult income percentile of a poor students is 77th; that's not too bad, especially given that this data is based on the Class of 2002, 34 year-olds at time of data collection. (Which, incidentally, makes the NYT's use of the bottom-to-top quintile movement metric to determine how much a school contributes to mobility basically crazy because who even is at peak income or maximum mobility by age 34?)

Andrew Stevens said...

I do agree with you that poor kids seem to do pretty well if they go to Yale. On the other hand, I have seen data in the past that shows that kids who get accepted to the top Ivies, but choose to attend elsewhere, do even better, economically anyway, than those who do choose to attend. For all of their social cachet, I genuinely believe there are very real questions about whether the Ivies themselves add any value or if, in general, they are simply taking kids who, through drive, intelligence, family connections and/or wealth, or whatever, were going places anyway and then simply claim credit for it.

This is an interesting thought experiment though. How much money would you have to pay me for everyone in the world (other than my family) to believe I am a drug dealer (though with no actual criminal justice system repercussions since I wouldn't actually be one and so nobody would be able to prove it)? I think if I were just starting out, I'd settle for about $2 million up front. So I guess I'm not saying social status is worth nothing since that's a decently high price. And, to be scrupulously fair, there is no amount you could pay me to have my wife and daughter believe that of me.

Miss Self-Important said...

Yes, it's totally possible that the Ivies don't add anything to adult financial success when you control for intelligence. We can follow this girl's trajectory and find out. But that doesn't mean an Ivy damages your life chances either. I don't mean to argue that these are the greatest schools and you should attend them at all costs, just that the NYT's data-graphic is not very fair or really informative.

Social status can be a proxy for happiness, which is impossible to measure. Also, no one has any incentive to pay you anything to give up your social status since they would gain nothing from paying from the exchange. Only your grade school nemesis or someone like that would get some kind of utility from your public humiliation, and I doubt he'd get $2 million in utility.

Alex said...

but I suspect #1 does matter b/c if you can say you got diversity through some impartial measure like an algorithm, you're saying you didn't have to lower your standards of merit for the sake of diversity, right? And the only reason you'd want to reassure people of that is because such a trade-off is controversial in many quarters, and the optimal situation would be one in which you could say you have the smartest students who also happen to be the most diverse.

Certainly claiming a win-win is desirable. What could be better than win-win? What I'm less sure of is, in a world of trade-offs, to what extent the politically salient factor is opposition to affirmative action (prove to the opponents that you didn't unfairly favor anyone) or the dignity of those ostensibly being helped (don't saddle them with the stigma of getting in on something other than merit).

Legally, I know that it's best to insist that no unfair preferences were shown. But those claims are pretty laughable in light of how much people brag about their diverse outcomes. "We worked so hard to get diversity...but, of course, we never explicitly worked for it. But seriously, we worked so hard and we are so proud that we got it." So I think that appeasing opponents of affirmative action is a minor factor, a pro forma thing to keep some plausible deniability in court rather than a serious effort to head off controversy. Nobody brags that loudly about the resources devoted to diversity if they think that explicitly working for diversity could be politically dangerous.

What I think is more salient is the dignity of those who benefit. The amount of effort put into disclaiming any informative value in measures that might show the effects of disadvantage, the amount of effort put into highlighting privilege and its effects, suggest that people want to say "Just because the test scores and grades and whatnot were worse, that doesn't mean these admissions decisions were made on the basis of unfairly low standards."

And, as you say, nobody wants to admit that they have or exercise discretion. What's funny about that is that social science is claimed as the basis for legitimate policy these days ("Studies Have Shown...") but 20th century mathematical social science is full of negative results illustrating the difficulty of designing processes to get socially desirable outcomes. Arrow's Theorem, Gibbard-Satterthwaite Theorem, Sen's Theorem, Holmstrom's Theorem, etc. all show that processes are paradoxical and/or gameable despite our best efforts, and Coase's Theorem only gives us nice, ideal results in a world of perfect information, zero transaction costs, and well-defined property rights. (And even then the ideal results come from markets rather than central decision-makers.)

Andrew Stevens said...

Social status can be a proxy for happiness, which is impossible to measure. Also, no one has any incentive to pay you anything to give up your social status since they would gain nothing from paying from the exchange. Only your grade school nemesis or someone like that would get some kind of utility from your public humiliation, and I doubt he'd get $2 million in utility.

That's why it's a thought experiment. I did have some pretty quality nemeses in high school (not grade school though); it's at least possible one of them is a billionaire. On the other hand, one reason I'd be willing to sell my social status is because it would not in fact humiliate me (I wouldn't care) and any nemesis worth his salt would know that about me.

Andrew Stevens said...

Oh and my whole point about the Ivies is not that they harm you; I agree they probably don't. It's just that there is an awful lot of discussion and consternation about who is or who isn't admitted to the Ivies and I look at the entire debate and ask a single question, "Why should I care?" So far, nobody has answered that question to my satisfaction. That it is hugely important what kind of people are admitted to the Ivies (be it through merit or diversity or holistic or non-holistic or whatever) is taken as a given. I just don't think it's very important. Indeed, it seems that we mostly put people with Ivy League degrees in places like government, op-ed writing, public broadcasting, social work, and political activism. These are all places where they are likely to do very little harm. We also put them on the Supreme Court and in Wall Street banks where they often actually do do a lot of harm and we should stop doing that.

Miss Self-Important said...

Alex: Nobody brags that loudly about the resources devoted to diversity if they think that explicitly working for diversity could be politically dangerous.
Yes, that is true. So maybe you're right, and it's the other two reasons - proximately the second, distally the third. The deferral to social science is part of the deferral to "expertise" more broadly, which is one of the two ways that personal authority can be tolerated in a liberal regime. The other is charisma, or (less nefariously) affection. We accept the personal authority of various scientists b/c we believe it's not really them but a set of universal facts talking. We also accept the authority of our parents and schoolteachers and scary despots b/c we love and admire them. Admissions officers cannot plausibly fall into either category. So I guess the question is, is this situation a problem, not just with respect to college admissions, but in all fields where discretion could be applied or even is applied, but is masked by appeals to impersonal forces?

Miss Self-Important said...

AS: Ok, but you wouldn't want to be an actual drug dealer, or an actual drug addict with no job, so status (in terms of actual well-being) permitted by working in finance or whatever you do is worth something to you, and by extension, so is the college education that got you that status. Even if the outward trappings are inconsequential to you.

As to why you should care: well, for the most part, you shouldn't. I think one of the reasons that student activism has gotten so intense recently is because too many people care way too much about what goes on in colleges, and not even the elite colleges, but even places like Evergreen State. If students weren't guaranteed a national media platform for every lunatic demand they made, they'd be much less incentivized to keep making them, and colleges would be better able to maintain order if they could punish students whose behavior is out of line without countervailing media pressure. I care about higher ed and education generally b/c I work in it, but I don't think everyone needs to care because I do.

On the other hand, if you care about the health of the regime, you have to care about the education of its youth to some degree. And once you start becoming concerned with this, then elite universities become important b/c they are bellwethers. They're also important b/c, if Charles Murray is right and the intelligence-sorting they perform is getting ever more thorough, they educate a growing share of the smartest kids in America, so what they do with them has disproportionate consequences for everyone else, for the reasons you give. And if such sorting continues, we can't stop putting their graduates into Wall Street banks or the Supreme Court. But I'm not an alarmist about the efficiency of intelligence sorting by elite schools, so on the whole, if you're not worried, don't get worried.

Andrew Stevens said...

I care a very great deal about education; that's why we're home-schooling our daughter. But that's no reason to care about who gets admitted to the Ivies. If anything, it seems like I should be rooting for whatever hare-brained crackpot scheme they cook up to restrict it to children of the elites so the institutions don't corrupt poor smart children as well (i.e. in order to avoid Charles Murray's prophecy, in case he is correct).

Miss Self-Important said...

Well, look at it this way: when the Bakke decision prohibited the use of racial quotas in admissions, they explicitly praised Harvard's "holistic" approach as a replacement. And then everyone with a sufficiently competitive admissions pool started adopting this approach. Now it's de rigueur, and even selective high schools do it. So sometimes, what these schools do trickles down into the rest of the system, by means of pressure or emulation or whatever. But sometimes not. So whether you should care if you're not personally connected to it probably depends on how big you think the effect is.

You're right, it would be quite a good thing if the Ivies dropped all concern about "diversity" and "access" and just became unabashed finishing schools for the super-rich. That would strengthen educational decentralization in this country a great deal and revitalize a lot of flagging second- and third-tier institutions. Another thing that would be counterintuitively good is if costs rose so high at elite schools that reasonable people stopped bothering with them, and everyone started staying close to home for college. Also a boon for decentralization. But that's not in the offing.

Alex said...

You're right, it would be quite a good thing if the Ivies dropped all concern about "diversity" and "access" and just became unabashed finishing schools for the super-rich. That would strengthen educational decentralization in this country a great deal and revitalize a lot of flagging second- and third-tier institutions.

But if social acceptability requires that all other elite institutions pay (at least cosmetic) attention to "diversity" and "access" then the finishing schools for those who will run those institutions also need to pay (at least cosmetic) attention to those issues. Not just for consistency in the face of pressure but also to role model to their charges what it means to be in that leadership class.

I didn't go to an Ivy but I did go to a school that takes in a lot of rich kids and produces a lot of managers. In retrospect, the way that my institution talked about diversity feels like it was training for careers where we have to interact with the HR and PR professionals in our organizations.

Miss Self-Important said...

Yes, that's why this is not in the offing.