Fake Charity Nerd Girl

I’ve had a yfip: meritocracy tag since, well, yfip was a thing that people said. And these days I feel like hating meritocracy is getting increasingly trendy and I am increasingly like “nope actually it’s just good.”

Sure, the current system of college admissions creates an elite that is segregated from the rest of the country, as well as incentivizing lots of zero-sum arms-race behavior among high school students and their parents.

But I just think the basic function that elite colleges serve of just putting all the smart kids in the same place and making them interact is so valuable that nothing else matters.

(The kids that I want to meet each other aren’t exactly the kids with the highest SAT scores–it’s some combination of that and intellectual curiosity and creativity and ambition. But SAT scores get pretty close.)

It’s kind of terrifying to think about what my life would look like if I had gone to a different school. I probably wouldn’t have had the same jobs. I probably wouldn’t be an EA. I probably wouldn’t live here or have befriended or dated any of the same people. And I think in most cases, the outcome would have been worse. (And I didn’t even go to the optimal school.)

Someone asked on the CFAR mailing list about how to choose where to send their child to school and it turned into a discussion on the effect sizes of different pedagogical interventions and like … I feel like all of this is just rounding error and the only things that actually matter are 1) how smart are the other students and 2) how smart are the teachers. Thinking about interventions might be helpful at a policy level but you should really really just send your smart kid to where the other smart kids are.

accomplishments in the past month:

my high school AP stats final project (on college admissions rates by high school) convinced a private school not only to institute an official grade inflation policy, but to retroactively inflate all grades for upperclassmen

I gave an interview for my high school’s student magazine in which I mentioned that colleges do actually care about how many AP classes you take, causing the reporter to freak out over the realization that her guidance counselor has been lying to her for the past three years

one of the more annoying anti-SJ voices on my facebook newsfeed:

In fact, that trial showed that the University of Texas system is not racist in their affirmative action. They use residential segregation as a basis rather than race. It’s an example of a great affirmative action system to reduce residential segregation

like, just the fact that people who are opposed to affirmative action because it’s racial discrimination are fine with discriminating on other things whose sole purpose is to serve as a noisy proxy for race

I mean no one is pretending this is not the case. The Economist:

For nearly two decades, UT has filled some 75% of its seats with Texas public-high-school students who finished in the top 10% of their graduating classes. This rule, adopted by the Texas legislature in 1997, was intended to increase black and Hispanic enrolment in the UT system.

the idea that anything that has disparate impact on different groups is discriminatory is dumb but at least it’s less dumb than this

Random political opinion of the day, arrived at after five minutes of consideration: I think Common Core is stupid.

There seems to be this assumption behind it that it is bad if public schools in Alabama have less demanding standards than those in Massachusetts? But this seems obviously good. Schools in Massachusetts should teach more than schools in Alabama: since students in Massachusetts are on average smarter, the optimal curricula for the two states are different.

My friend is an elementary school teacher in an underprivileged area and she has this absurdly detailed list of things she has no choice but to teach, even if the students are still struggling with one-digit subtraction and have no business learning about the distributive property. But she can’t just go back and spend two weeks reviewing subtraction, because … we have to demand a lot of students? And this will somehow resolve the problems that different students (and, on average, different groups of students) learn at different paces?

I was at my graduation last weekend, and the commencement address was basically about twenty minutes of vitriolic insults directed at Trump. And in between burying my head in my friend’s shoulder in discomfort and laughing nervously, I was thinking about the family of this guy in my class.

He’s the first person in his family to go to college. He drove an hour every day to go to a somewhat better high school because there was an epidemic of gang violence at his local school. Against the odds, he did well, and got into college, where he has continued to get good grades and play sports and generally do things that make parents proud.

His family is not well off. They’re Mexican-American. And they’re Trump supporters.

Yeah, I’m kind of confused too. But they honestly are. (Not even reluctant Republicans supporting Trump–they voted for him in the primary. His aunt owns a Make America Great Again cap.) And all I could think about was how happy they must have been to be attending their son’s graduation from one of the best universities in the world [citation needed], only to have that happiness turn to bewilderment and anger as everyone around them cheered a series of caustic attacks against them and their values. The message couldn’t have been clearer: “You don’t belong here.”

My mom thought this speech was So Courageous. When I suggested that it might have been more courageous to say something that not everyone there agreed with, she replied, “the students maybe, but a lot of the parents looked unhappy.”

Seventy percent of the parents there had family incomes over six figures. (More, probably, since low-income parents are less likely to attend graduation.) A lot of them are members of the self-perpetuating intellectual/economic elite. This probably isn’t true of the few Trump supporters among them. 

So if we are going to single them out for judgment, force them to account for their support for an “infantile,” “bullying,” “proto-fascist” “charlatan” … can it not be on the day of their kids’ graduation?

Man I read this Vox article today and there’s just so much going on I feel like I have to say something about it.

• It claims “white women benefit most from affirmative action.” Its evidence for this is:
  
• Kimberle Crenshaw said “the primary beneficiaries of affirmative action have been Euro-American women”
• “A 1995 report by the California Senate Government Organization Committee found that white women held a majority of managerial jobs (57,250) compared with African Americans (10,500), Latinos (19,000), and Asian Americans (24,600) after the first two decades of affirmative action in the private sector.” 
• The wording of this sentence is somewhere between misleading and wrong: in fact what the report said is that there were 57,250 more women employed in managerial jobs in 1993 than in 1975, and similarly for the other groups. So this is actually slightly better evidence than I originally thought. 
• But still, there is no attempt to look at percentages? Like, 40% of California residents are white women, while 7% are black. Maybe this has something to do with the fact that white women gained five times as many jobs as African-Americans?
• Also of course there is a lot that happened between 1975 and 1993 and we can’t really say with any confidence that affirmative action caused anything here.
• The percentage of women physicians has gone from 8% to 25% since 1970. The article does not mention what has happened to the percentage of black physicians over time. It also does not give any evidence that this is due to affirmative action rather than medical schools and jobs just not actively discriminating against women. I mean, maybe it is, but the article gives no evidence for that.
• It then proposes this “paradox”: that white women benefit the most from affirmative action, but still oppose it. It then gives the examples of several white women who have been plaintiffs in affirmative action cases over admission to college and law school.
• Like. Maybe this is because white women don’t generally receive affirmative action in college and law school admissions?
• I understand that white women do benefit from affirmative action in contexts like tech hiring! But undergrad admissions is really not one of those contexts.
• It dismisses the mismatch hypothesis in a sentence with “Study after study shows there’s simply no evidence for the claim.”
• Linking to two sociology studies.
• The studies don’t look particularly bad (though admittedly my prior for sociology papers being good is lower than that for economics), and I’d like to read them when I don’t have finals. But I could just as easily link you two studies that do find evidence of mismatch. This is the problem with saying things are debunked based on two studies.
• It admits that the percentage of black students enrolled at UC Berkeley has gone down from 8% to 4% since admissions went race-blind. But it says this is not because of “underqualified students of color” but rather because 58% of black students reject Berkeley’s offer of admission.
• Okay, the 58% number is for black students from 2006-2010, and I only found Berkeley’s yield rate for 2012. But in 2012 it was 37%!
• That means if the yield rate for 2006-2010 was similar, black students were more likely to accept Berkeley’s offer of admission than students of other races.

there was a post going around about Social Matter Forum, so I was reading some of it and it occurred to me that both the bonobo left-libertarian whatevers and neoreactionaries I encounter seem to have converged on “public school is child abuse”

where are my fellow proponents of “public school is an effective way to socialize children into mainstream society and teach them compliance with authority”

So I was reading the Times yesterday and one of the letters went something like “It’s troubling that Scalia would cite debunked social science such as the mismatch hypothesis–haven’t you seen this blog post?”

So of course I had to check out the post, which led to a rabbit hole of links, so I thought I’d summarize what I found.

The first thing Chingos cites in his post is this paper which uses a scholarship program in Massachusetts in which the top 25% scoring students in each school district on a standardized test were given free tuition at any state university. They compared the students just above and just below the threshold, and found that getting the scholarship made students significantly more likely to attend state universities and that this made them less likely to graduate. Which is interesting, but seems not totally the same thing as mismatch from affirmative action? This is basically saying that (for kids around the 75th percentile of their district) going to a school they’re overqualified for makes them less likely to graduate, while mismatch theory says going to a school you’re underqualified for makes you less likely to graduate. These things could both be true.

And actually, as I read the paper more closely (this factoid is buried on page 23) it says that non-white students do not see a drop in their persistence rates from attending state universities. So if I’m understanding this correctly this effect only holds for white students? Uh, this doesn’t exactly seem like evidence against mismatch.

The second thing is two studies supporting mismatch that he reanalyzes. The evidence supporting mismatch is mostly based on studies that control for what colleges students were accepted to. The problem with this is that, of course, we don’t know if it’s causal or if the students who choose to go to a less selective school are different in some way. Most students just go to the most selective college they’re accepted to or somewhere of approximately the same selectiveness, so we’re trying to trying to draw conclusions from this kinda small and weird group of students.

So what he does instead is look at the graduation rates for students who go to different schools, controlling not for what schools they were accepted to but for SAT score, GPA, parental education, and family income. And he found the opposite result–that students going to better schools had higher graduation rates. This method doesn’t have the issues of the previous one, but it clearly has its own issues–when controlling for SAT score and GPA, what colleges you’re accepted to probably has a lot to do with motivation and other stuff that clearly has an effect on whether you graduate college.

The second study was one that found that for URM students who stated an intention to major in science, going to a more selective school made them less likely to major in science. He agrees with this in his reanalysis, but says that this relationship got stronger after California instituted its affirmative action ban, which is the opposite of what you’d expect if affirmative action were the cause. I don’t know what’s up with that tbh.

(An issue that occurs to me with all the studies focusing on California is that, well, California college admissions aren’t really race-blind. We’re comparing a system that does a lot of explicit race-based affirmative action to a system that does a smaller amount of covert and very subjective race-based affirmative action. I mean, I wouldn’t be surprised if “applicants get +5 points if they’re black” is a way better system in terms of matching students correctly than “well, it would be illegal to ask you to consider race when reading these applications … but uh, we wouldn’t usually give that sort of student that sort of score …”)

What I haven’t seen, and what I’d like to see, is studies that survey URM students about happiness/feelings of belonging. Just thinking about my school, almost everyone graduates–we have a good support system and plenty of easy classes for you to take if you fail calculus. I think the issues go beyond not graduating. Although, I guess given the earnings boost from attending selective colleges, maybe four years of unhappiness is worth it if you get your degree at the end? I did find an interesting-looking paper arguing that our current level of affirmative action is higher than the level that would maximize interracial interaction across all college students, since social segregation is more likely on campuses where there is a marked difference in academic preparation between students of different races.

So yeah. In conclusion, Chingos says that “To truly put the mismatch theory to rest, rigorous quasi-experimental evidence that focuses on the beneficiaries of preferential admissions policies is needed.” I agree with this. But I’m not quite as confident as him about which way the evidence will point.

So today I was talking to someone who works in admissions for an economics PhD program. The way they do admissions is that they go through every folder and give it a numerical score that’s as objective as possible, based on grades, GRE scores, quality of recommendations, undergrad research, etc. Then they go through all the folders, looking at the scores, and decide who to admit.

In the second step they always practice explicit affirmative action in favor of women; the score required to get in as a woman is significantly lower than the score required to get in as a man. And yet, consistently their female students do better on average than their male students.

I thought that was really interesting.

Hmm. What’s the evidence on selection vs signalling vs added value for education? My recollection is that people who are accepted to Harvard do about as well regardless of where they go, unless they’re poor, in which case Harvard does help.

I don’t know … It’s possible the benefit of the degree could outweigh the suffering they experience in college. I don’t know how valuable a nice-sounding degree with a terrible GPA is. I also think they’re not usually looking for jobs at like McKinsey where the name of your school really matters; the people I’m thinking of want to be a park ranger and an elementary school teacher.

The lack of even the most basic math class is sad. Do you not have mandatory general education classes? I thought you got a liberal arts education at your school? Or at least a pretense of it.

Also, in their defense, some schools are taking baby steps towards supporting students who didn’t get in on academic or sports merit. Assuming everyone has a working knowledge of (I’m guessing Algebra 2? That seems like a plausible thing some schools might not teach.) is terrible.

If you want to major in a technical subject you need calculus, and we don’t offer precalculus. To fulfill their math requirement humanities students often take the lowest level stats class, which only requires algebra 2, or something like that.

Also, there’s the even darker theory that affirmative action is good because it increases intellectual diversity, and as such might be good even if the poor saps who are accepted lose out from it. I admit, I’d be willing to trade “future presidents have actually known non-white non-asian people” for “some college students suffer and don’t learn as much but still have a fantastic (bad) signal of value.” Thinking of minority humans at top schools as ritual sacrifices on the altar of future leaders having some instinctive understanding of diversity sounds bad, I grant you, but politics really matters.

… yeah that’s just evil.

In all seriousness, I’m skeptical of supporting policies that have a lot of tangible identifiable victims in the name of some wishy-washy benefits we have no evidence they actually have. I mean, it seems just as plausible that affirmative action causes future presidents to become more racist because they spend four years in an environment where the average minority student is so noticeably different from the average white student.

I’ve always thought that the way we do affirmative action is stupid, that UC v Bakke was stupid, and that if colleges want to do affirmative action (which they always will for the forseeable future) they should have an explicit quota system and be transparent about it

I mean, the more transparent they are the easier it is to call them out for e.g. discriminating against Asians, and the less ~holistic~ their process for affirmative action is the less random noise gets introduced by admissions officers’ whims

but last night I talked to a girl who invoked her FERPA rights to read her admissions file and was so upset she almost didn’t come back to school this year, because the file made it clear that she was admitted on the basis of affirmative action

like, I always assumed she knew that already? she knows what her SAT scores are and what SAT scores are typically required to get in here, she knows that she’s a minority from a very disadvantaged background, she knows that our school practices affirmative action (and has always been vocally in favor of it)

but I guess she’s always genuinely believed, or at least told herself, that affirmative action played no role in her acceptance

polite fictions, man

A friend (who’s studying to be an elementary school teacher) and I were talking about ability grouping for elementary school students, which they’re now apparently being told not to do. The evidence they’ve been given for this recommendation by textbooks and professors has been mostly anecdotal, with maybe a few mentions of studies thrown in, which is depressing but not surprising.

I looked at a few meta-analyses and it seems like flexible ability groups for one or two subjects in the context of a heterogeneous classroom have a positive effect on average achievement, although I say this with low confidence.

I could tell her that, but I don’t think she’d change her mind, since the effects seem to vary by ability level and are negative for the lowest-achieving kids. She’d say what matters is that we help the kids who need it the most, the high-achieving kids will be fine on their own.

Which made me wonder, if we can choose to either improve the achievement of low-ability or high-ability kids by a certain amount, which is a greater utilitarian good? My intuition says the high-ability kids, since they’re more likely to create lots of value with their education by e.g. curing cancer or inventing the computer. I’m not sure though. Would I be less successful now if my elementary school education had been 5% worse? I always had the resources and the motivation to learn stuff outside of school, which a lot of kids don’t.

Also I think we should put nonzero weight on what makes kids happiest while they’re in school. I know my happiest memories from elementary school are things like when I was in the highest spelling group and I got to pick words like “brachiosaurus” instead of “about” and research their Latin and Greek roots, when I got to do things that were actually interesting for a change. Maybe the lowest-achieving kids were made less happy by the ability grouping, I don’t know.

(I started elementary school in a very diverse school district. By third grade, the top ten kids in my grade had all left because it was clear the teachers didn’t care about teaching us, leaving our classmates without the benefits of heterogeneity. If you sacrifice the high-achieving kids on the altar of closing the achievement gap they will opt out of your educational system.)

dataandphilosophy said:Overcoming self-segregation based on different interests seems like a hard task.

I agree and was hoping no one would ask for my solutions!

I don’t really know how you can address stuff like this legally without unacceptable encroachment on personal freedom. Ron Unz suggests we admit people to Harvard by lottery, which just seems kind of crazy?

And I think my first reaction to Unz’s proposal was something along the lines of NO YOU CAN’T TAKE AWAY MY SAFE SPACE, which makes sense and it’s totally reasonable for people to want to hang out with people like them and marry people like them on the individual level. But then when those people hold all the positions of power in a country …

Colleges, at least in their purest form, take all the intelligent high schoolers and sort them into little enclaves with the other intelligent high schoolers. And some of them go on to become lawyers, some investment bankers, and some get PhDs and end up back in the enclaves teaching the next generation of intelligent high schoolers.

And the thing about meritocracy is that the smart kids aren’t like the other kids, in so many ways, and this system puts them in bubbles where they never have to interact with the rest of America again. And colleges recognize this, and try to look like America in the most superficial ways, namely skin color and parental income. (Of course, they’re doing much better on the first than the second.) But that doesn’t really help, it just means instead of the smartest kids you’re getting the smartest white kids and black kids and Latino kids and Asian kids with a few first-generation college students thrown in and it’s not really that different.

Because I know the Red Tribe exists only intellectually. Because I don’t know any evangelical Christians, or people who smoke cigarettes or have been to Branson or even know where Branson is (thanks Charles Murray). Because I asked my friend to guess who there are more of in the US, Jews or Catholics, and she guessed Jews. And I can’t even blame her, because searching “catholicism” and “catholic” in my school’s course catalog gives 8 results total, and “judaism” and “jewish” give 187.

(my dad is Jewish and my mom is Catholic so I think I’m allowed to make this comparison)

And of course if you really want to look like America, you’re going to have to discriminate against oppressed minority groups and so they do. And I guess college admissions officers seem to sleep fine at night despite this but I’m not sure if I’d be able to.

So we get an intellectual elite that focuses on some questions, perspectives, and experiences to the exclusion of others. And they talk to the rest of the elite and breed with the rest of the elite and are just increasingly detached from anything resembling the real world.