Not Catching Up: Affirmative Action at Duke University

A working paper titled What Happens After Enrollment: An Analysis of the Time Paths of Racial Difference in GPA and Major Choice by Duke university economists Peter Arcidiacono and Esteban Aucejo and sociologist Ken Spenner is creating a stir. The authors track a sample of Duke students from admissions to graduation in order to determine the effects of affirmative action.

Under one theory of affirmative action the goal is to give minority students an opportunity to catch-up to their peers once everyone is given access to the same quality of schooling. On a first-pass through the data, the authors find some support for catch-up at Duke. In year one, for example, the median GPA of a white student is 3.38, significantly higher than the black median GPA of 2.88. By year four, however, the differences have shrunk to 3.64 and 3.31 respectively.

Further analysis of the data, however, reveal some troubling issues. Most importantly, the authors find that all of the shrinking of the black-white gap can be explained by a shrinking variance of GPA over time (so GPA scores compress but class rankings remain as wide as ever) and by a very large movement of blacks from the natural sciences, engineering and economics to the humanities and the social sciences. It’s well known that grade inflation is higher in the humanities and the social sciences so the shift in college major can easily explain the shrinking black-white gap in GPA. (The authors show that grades are higher in the humanities holding SAT scores constant and also that students themselves report that classes in the sci/eng/econ are harder than classes in the humanities and that they study more for these classes).

The shift of black students across majors is dramatic. Prior to entering Duke, for example, 76.7% of black males expect to major in the natural sciences, engineering or economics but only 35% of them actually do major in these fields (almost all Duke students do graduate so this result is due to a shift in major not dropping out). In comparison, 68.7% of white males expect to major in sci/eng/econ and 63.6% of them actually do graduate with a major in these fields (this is from Table 9 and is of those students who had an expected major). White and black females also exit sci/eng/econ majors at high rates, although the race gap for females is not as large as for males. The authors do not discuss the consequences of dashed expectations.

An important finding is that the shift in major appear to be driven almost entirely by incoming SAT scores and the strength of the student’s high school curriculum. In other words, blacks and whites with similar academic backgrounds shift away from science, engineering and economics and towards the easier courses at similar rates.

I have argued that the United States would benefit from more majors in STEM fields but that is not the point of this paper. The point is that there is no evidence for catch-up at Duke and thus to the extent that affirmative action can work in that way it may have to occur much earlier.

Hat tip: Newmark’s Door.


"The point is that there is no evidence for catch-up at Duke and thus to the extent that affirmative action can work it may have to occur much earlier."

So get poor black kids out of their horrible high schools and middle schools and elementary schools? School choice ftw!

Where'd you put those black kids? Their schools horrible because they're full of black kids.

Stay classy, JL. At least no one needs to wonder about the sort of person you are.

So are you arguing that I'm factually incorrect? Schools full of black students tend to be bad schools because blacks tend to be weak students. It's simply an empirical fact.

You are factually incorrect.

Empirically, the observation is accurate, but your reasoning is specious.

The problem isn't genetics, it's cultural attitudes and practices. That's why N Koreans are some of the poorest people on Earth while S Koreans are some of the richest.

I said nothing about genetics, TallDave. What I said is true regardless of why it is so.

"So are you arguing that I’m factually incorrect? Schools full of black students tend to be bad schools because blacks tend to be weak students. It’s simply an empirical fact. I said nothing about genetics, TallDave. What I said is true regardless of why it is so."

He's making an observation and it's largely true. Ignoring data because it's unpleasant does not make the data go away and it does tend to make for poor decision making.

Meg, the tendency of African Americans to poor academic achievement is one of the most robust findings in American social science, so if you want to dispute it, you will have do more than just assert that I'm wrong.

JL -- I inferred a genetic argument from your statement that the black schools are poor "because they're full of black kids" (a cultural argument would assume the school could be full of black kids and be successful given a cultural shift, while a genetic argument would not allow this). If you weren't making a genetic argument, then my I stand corrected on my incorrect inference.

And why are you so sure that the issue is 100% nurture and 0% nature?

Social science supports the finding that many black students are in horrible schools.
It does not support the finding that schools are horrible because they have black kids.

The causal determinants of what makes a school or a student bad is not race. Race can be an effective proxy for a lot of other things, which actually do affect school and student quality.

But racists love their sophistry.

maguro -- You couldn't have asked for a better natural experiment than E Germany vs W Germany, China vs Taiwan and Hong Kong, North Korea vs. South Korea.

Just by crossing our border, Haitians become something like 100x as productive. Absent some sort of gene-altering border technology (which I'm not ruling out) it's the society that dominates.

GiT: And yet race is still a better predictor of SAT scores than socioeconomic status. Interesting that there are still people who deny its crucial role in explaining why bad neighbourhoods are bad, and why bad schools are bad.

I don't care what kind of person he is. It's actually a good point, minus the color variable. Having the serious students sit in bad schools for years and then attempting to rescue them to Duke is not only inefficient but possibly ineffective, and not only ineffective it might make Duke worse, because now the serious minority students are the 'last man' holding back the median Duke class. On the other hand colleges are more likely to boot blatant troublemakers and remedial students. This is quite likely what is happening with the major changes as people who thought they were ready and willing to go into sTEm are disabused of that notion. Also, since everyone has to move off to college it's a natural choice logistically.

It's clear that there are some really awful schools, in which learning anything is all but impossible, and that those schools are usually majority black or hispanic. And it seems inevitable that getting kids out of those schools and into decent ones will help. But I don't think there's any evidence that interventions we know how to do will close the black/white gap of school performance. So while it's worth closing down the worst schools, that's not going to resolve the problem that black and white students get to college with different levels of ability and preparation. Solving that problem is something lots of people have worked on, and I think it is just genuinely hard to solve.

I don't think the North-South Korea or East and West Germany examples are fair.

First, There are to my knowledge no areas where on one side of a divide you have an all black school that does terrible at school, and the other side an all black school that does amazing.

Second, We are talking about different countries with radically different policies and known policies that affect eduction. America for themost part has a common educational policy, and for the most part education dollars are spent MORE on black schools and than white schools. This is in direct contradiction to the country examples from above. A good example would be Washington DC, where the average black student gets well more average spending than similar non-black students in the surrounding Virginia and Maryland suburbs, and yet fail at drastically higher rates.

I can accept that a cultural explanation is possible, but at this point in time, the evidence strongly favors genetics.

At least no one needs to wonder about whether or not someone would resort to ad hominem in this thread...

But good students seem to do well in bad schools just as bad students do bad in good schools (e.g. the study above).

Perhaps due to the history of slavery blacks see less value in school or more resent the authority of teachers or have different attitude in school for some other reasons.

Which means it's a cultural issue that won't be overcome by moving students to better schools. And at this point blaming it on slavery is a stretch. The original cultural values certainly derive from slavery, but this many generations away from slavery, even a few away from Jim Crow laws, the culture is handed down, not instilled by the cultural divide.

Make T&E easier #487

Travel and Entertainment?

What's T&E?

But of course, on the theory that a university education plays primarily a signalling role, the failure to narrow the GPA in a meaningful way is not as alarming. It might be that the benefit of affirmative action in such cases is to allow a significant number of black students to one day say "Yeah, I went to Duke."

People look at majors, though

For your first job out of college, but after that it's about experience. I studied music at Northwestern. When educational background comes up in conversation, people seem to automatically assume I went to Kellog.

Except that people are aware of affirmative action in admission...

Unfortunately, that also attenuates the signal that a university education (brand name or otherwise) sends, and so forces everyone to get more education to signal the same thing.

I disagree that it's necessarily unfortunate. More university education is still a good thing on the wash. It's only unfortunate if crowded majors are the only additional degrees sought.

Yes, clearly the disadvantaged black kids need to be reached sooner. But these results hurt the efficiency case for affirmative action at the college level, but not (in some sense) the equity case - just sharpens the tradeoff. However, if you take a broader view of the efficiency, there are spillover effects and other benefits to having a non-homogenous college educated work force in a country. At the same time, of course it is unfair to the whites missing out on Duke (and other elite colleges) because of affirmative action.

Agree for the most part. But typically it's not whites being rejected in favor of under represented minorities, but Asians.

OK... er... what is the control in this study anyway???

Did I miss something?

Affirmative action is not supposed to erase deficits between groups its supposed to improve outcomes within groups.

The control has to be a group of women or minority students without affirmative action and I am not seeing that here.


I wonder about the nationwide effect of the smartest blacks mostly going to schools where they're overmatched. (Perhaps I'm wrong about this--is there good data on what fraction of blacks are admitted to universities with lower test scores, grades, etc., than the average student?) One thing I'd guess would happen, and it's something Thomas Sowell pointed out a long time ago, is this shift in majors. Plenty of people bright enough to get through a EE major at, say, University of Missouri, will wash out of engineering if they're trying to do it at Stanford, because the competition is a whole lot tougher. So if a lot of people who are well-matched to the EE program at Mizzou end up at Stanford instead, my guess is that they'll often find themselves majoring in sociology or education or something. And summed up over the whole country, it seems like that should have an interesting impact, with more of the smartest blacks ending up going into law or business administration than engineering or finance, for example. How does that affect things like politics?

this is an exceptional point. And there is another.
If there is a group of similarly minded, similar background people in one school, that "posse" can do very well for themselves. You can feed off one another, you can bolster one another when the going gets rough etc. So if you had a pocket of smart black students coming out of a highschool in State "X", and they all went to SXU, they'd likely have lots of cool, interesting dynamics between them.
However, with AA, that same peer cohort will experience one in Cambridge, MA, one in LA, one in NYC and they will be disproportionately "alone".
You can see this in the fate of Howard University. Under segregation (which mind you i am not advocating for) Howard was a prominent school doing interesting things that *mattered*. AA has relegated Howard into a provincial backwater.
If I was the uni of Mississippi or Alabama, i'd be livid at the poaching of my best and brightest

They are not your "best and brightest" though; they are over-matched minorities that have been promoted elsewhere. The only problem that poses to Alabama or Mississippi is that they are in turn forced to import their own over-matched minorities from an even lower rung.

I have tried to find it but it was actually a professor from one of the elite schools (and I think it was Stanford) who found grade inflation to be more rampant at elite schools. He surmised this was because kids able to get into elite schools had higher expectations of performence than kids who hadn't had such expectations. They, and their parents, were more likely to scream when little Jack or Jill got a C (while paying a small fortune in tuition) and professors reacted to such screams. Me personally, I have taught at lessor schools and if what is said about tenured professors teaching at higher rates at lower ranked schools, it would be expected thay would grade more fairly than a grad student at Stanford. A tenured professor could really care less what a student or parent thinks because they are virtually impossible to fire to maintain student and parental happiness. Grad students on the other hand are likely more sensitive to such comments.

"Affirmative action is not supposed to erase deficits between groups its supposed to improve outcomes within groups."

The entire purpose of the study was to determine whether affirmative action erases deficits between groups.

And that seems a wrong yardstick to judge affirmative action by.

That's been said.

As far as I know candidates selected because of AA are not randomly selected from an eligible pool - so how would you deal with the very serious selection issues ?

What exactly is the equity case? Whites deserve to be discriminated against, in the country they built because __________.

"...they used to own an entire class of people."

"And this was codified in the Constitution until the 1860s and and many state laws until very recently"

"Very recently being well before anyone involved was even born"

Actually no, it was not codified in the Constitution. Just because an amendment freed them did not mean it was codified in the Constitution. You need to remember the Consitution prohibits and demands only a few things and leaves the others up to lower levels of governments. The 13th Amendment was added exactly because it did not allow or prohibit slavery in the past.

"Representatives and direct Taxes shall be apportioned among the several States which may be included within this Union, according to their respective Numbers, which shall be determined by adding to the whole Number of free Persons, including those bound to Service for a Term of Years, and excluding Indians not taxed, three fifths of all other Persons."

"The Migration or Importation of such Persons as any of the States now existing shall think proper to admit, shall not be prohibited by the Congress prior to the Year one thousand eight hundred and eight, but a Tax or duty may be imposed on such Importation, not exceeding ten dollars for each Person."

"other/such Persons" = African Slaves


To clarify: +1 to Original D, and I'd add "... thereby causing damage that lasted for several generations, which was further aggravated by Jim Crow laws and other forms of discrimination, so the impact of those issues continues to this day."

Define "damage." American blacks are the richest, freest blacks in the world.

Even if this is so, why are African-Born and Carribean born blacks being favored? Diaspora blacks, surely we can agree, is a bizarre overstepping of the original claim.
Triply so for hispanics. Especially given it is possible to be a "descendant of white slave holder" AND be a hispanic

TallDave: I suspect African-Canadians might have something to say about that.

Still waiting for the Egyptians to provide affirmative actions to Jews, given that Egyptians used to own (literally) all the Jews in the world.

Except not really. Recent archealogists found evidence that they were more like low wage manual labor. Google is your friend.

Bleah. Slavery was common in preindustrial societies. We should really just leave it at that.

All whites?

Well, the Germans and their friends in other European countries got about halfway murdering an entire class of people.
But there is no affirmative action in Germany for Jews, and there should not be. Sorry about this Godwin.

Slavery 150 years ago is simply not a good justification for racial discrimination against my immigrant sons on their college applications.

This is a red herring because Asians are the ones bearing the brunt of affirmative action, not whites, as per Espenshade and Chung.

So we need a different line of snarky excuse-making for discriminating against Asians in favor of latinos and blacks.


I'll believe in Affirmative Action when I hear why my Asian-American daughters should pay for the sins of European-Americans that occurred long before they were ever born. Anyone?

Good point!

@Jill for example having black police in majority black areas can help police be more effective. Having black Doctors (MD) can help because they are more likely to serve blacks and that is even if they are slightly less qualified same for teachers and others.

Actually, from what I've read this largely turns out not to be true for police -- the same culture problem still exists.

What would really help crime is doing away with the war on drugs, which disproportionately attacks minorities, and esp.blacks.

Ex cop here, black cops do nothing to remedy this situation. New Orleans had/has one of the worst police-community relations in the country and it was probably one of the most diverse forces in the country.

I'd be interested in seeing what their education backgrounds were prior to entering college. What kind of high schools are these kids coming from? If rigorous math and science training are lacking in middle school and high school, it's hard to imagine people being able to pick it up along the way at the university level. No doubt that a lot of the inequality we see stems from inequality in school quality. And it's even likely that the kids who went to the "better" schools are also not getting their dime's worth from 13 years of filling a seat.

"No doubt that a lot of the inequality we see stems from inequality in school quality."

There must be some doubts, otherwise we wouldn't be asking for data on educational backgrounds, right? As JL rudely demonstrated the causes of school inequality can be pretty contentious issue, too. I don't really have any data myself, but there is an anecdotal story here with a few secondhand facts:

It’s a self-perpetuating cycle: With more middle-class parents sending their kids elsewhere, the parents who still have kids in the local schools are more likely to be the ones least able to make time for such meetings. “I’ve been to some PTA meetings,” says Jones. “Sometimes there may be less than 10 parents in attendance.”

But the real culprit is the flight-not-fight mentality prevalent in the black middle class. Experts have long complained that such departures lead to starving neighborhood schools of the brightest students. Earlier this month, the Washington Post reported that test scores of children in some of the city’s poorest neighborhoods, like Wards 7 and 8, trail those of their counterparts in Ward 3. It didn’t mention, however, that many of those Ward 3 students are, in fact, upper- or middle-class African Americans from outside that Upper Northwest community.

I'm trying to visualize how I would make that kind of argument to a black parent. "Yes, the school your child is attending is crappy, test scores are awful, half the kids don't read anywhere near grade level, there are a lot of fights and disruptions in class, and hardly anybody from this school ever goes to college. But you should keep your child in this school, because otherwise, you'll be contributing to the worsening of the school. "

It's not an argument that is made to the parents. It is an argument made _by_ the parents. If parents don't feel enough connection to their location that they want their child to be one of the ones to lift the rest up, then so be it. My hometown closed 3 schools, but then re-opened 2 of them, partially because the parents in those areas worked so hard to keep them open. The third school was the poorest and least performing school in the district. So no, it's not a surprise that parents want their kids out. But it is still the reason those schools are so terrible.

>to allow a significant number of black students to one day say “Yeah, I went to Duke.”

Precisely this.

And if any of them become high-profile and successful, they get to say it into a microphone, and folks at the New York Times will dutifully report how super-awesome and inclusive Duke must be. Mission accomplished!

And in the meantime, it does not matter to Duke if you are a white Engineering major or a minority Transgender Awareness Studies major. They are both paying the same obscenely high tuition rates (to some of the richest people in the country, I might add).

On net, did the black students benefit? Sure. They got a Duke degree. Maybe in a mediocre major but it still opens up a lot more (admittedly mediocre) jobs than no college. Better get a 3.5 in Humanities than an F in Engineering.

Why are these results so noteworthy? Nobody expects AA to churn out high-achieving engineers; merely to help compensate for a historic wrong. Naturally, students adjust majors to match their aptitude.

Leaving aside the issue of AA, are the marginal students better off for having an Art History degree from Duke over, say, an Economics degree from the University of Maryland?

I remember watching some people struggle at MIT and wind up with a degree in Music, all the while thinking they might have done okay at UMass.

Yeah - it doesn't really pay to go to those top tech schools if you're not going to be able to do decently well at them. The Ivies are a different story, you can major in 15th Century French Poetry at Harvard and still wind up with a Wall St. job earning more then most MIT engineers.

I don't think I agree with this. Ruining your life by getting sucked into a Philosophy degree at Harvard is a pretty common story. I've not seen data, but I personally know examples. It's hard to recover from a useless degree, even from Harvard.

Yeah well tell me about about useless degrees, I got a math degree from some school in Canada and I'll never recover - but I've also seen a lot of articles by Harvard Humanities grads whining about how "I wanted to save the world but a Wall St. job was just the only option after school"

I think most people here think that it's you and not the degree. I don't say that to be mean.

WRT Harvard, a Philosophy degree is a fast-track to the wonderful world of IT consulting. I think you overestimate how easy it is to get a Wall Street job from Harvard, and also how generous people are being in their personal definitions of "Wall Street job."

It might be hard to get a Wall St. job from Harvard but some of those Harvard Humanities grads make the jump - no poetry major from the Ohio State University is working at Goldman Sachs.

Don't most of the bright students STEM majors or not end in MIS these days?

"I got a math degree from some school in Canada and I’ll never recover"

I've always thought that math degrees were the penultimate very hard, yet conversely low worth degree.

I obtained a Finance degree. Got out for a few years. Realized that unless I was working in sales, my future income was going to be pretty mediocre. I went back to school and got an EE degree. Yes, it kind of sucked getting out of school at 29 with a BS in EE, but 13 years later I have no regrets.

What the hell is MIS?

Isn't this interesting I'm going back for an EE degree right now.

"Isn’t this interesting I’m going back for an EE degree right now."

Make sure you get Co-op experience (cooperative education, ie. work in the field during school) for as much time as possible. Get a co-op position and then volunteer to work for the company during summers. Some schools hate this. Ignore the school. A good co-op experience with good references is resume gold. And Engineering co-ops are generally paid much better than the average university intern.

> What the hell is MIS?


> Don’t most of the bright students STEM majors or not end in MIS these days?

I don't think so. I've never worked anywhere where it wasn't seen as a cost-center. Also, it's not what people recruit for at top schools.

MIS = Management Information Systems. This is a horrendous degree that business schools give to STEM students that dropped out or decided that moving into management early is a good idea. It wouldn't be so bad, but the degree is supposed to imply competence in Information Systems. Instead they mostly teach the ins and outs of Excel and VBScript. This major is the reason many IT workers hate their boss.

"I’ve always thought that math degrees were the penultimate very hard, yet conversely low worth degree."

From experience (not just personal but my entire graduating class of math majors), yes this is correct, unless you go further and supplement that BA with an MA or Ph. D.

Just how does it compensate for a historical wrong to have students who would otherwise be successful engineers graduate with a high status, low productivity "something something" studies degree from an elite university instead of from a respectable mid level institution? And what wrongs are done instead to the grandchildren of Asian immigrants who suffered historical discrimination (sometimes far worse than any suffered by Latinos), who are rejected from those same universities because their ethnic groups are overperformers?

Truly, the logic of compensatory discrimination is strange.

I don't think the debate here is whether affirmative action is fair; almost by definition it will be unfair (to some of the current peers). The question is whether it benefits the group it is intended to help and by how much.

I wouldn't be so sure. I think studies have been done on this subject.

"Why are these results so noteworthy? Nobody expects AA to churn out high-achieving engineers; merely to help compensate for a historic wrong. "

I think the state of the world is opposite as you have described it. Plenty of people expect AA to churn out high-achieving engineers, or at profess such a belief publicly. There are also many people that hold on to the notion that elite schools are places to train our best and brightest to be leaders and not some place to acquire a mere proof of elite status regardless of the means.

For these people the genuine historic wrong was that some of our best and brightest were being denied opportunities because of the color of their skin, but measuring how much we have righted this wrong may not be as simple as glancing at how colorful the lecture halls look, or using cheap science and rhetoric to challenge what is commonly considered to be intelligence and achievement.

Caveat: Even if blacks benefit on net, it doesn't in itself justify affirmative action since AA hurts Asians.

> it doesn’t in itself justify affirmative action

Sure it does, you just have to have the right utility function. If you believe that society benefits by having every significant minority have significant representation at each level of society, then one can easily make an argument for AA to ensure adequate representation at higher levels. Certainly AA hurts Asians, but taxes hurt wealthy people. You simply have to believe that the benefit outweighs the costs.

You mean AN utility function that favors racial balancing over judging people by the content of their skin instead of that of their character, and/or a function that prioritizes the welfare of blacks over that of Asians.

Whoops, flip "skin" and "character."

Sorry, an utility function just sounds wrong to my ear.

Anyway, yes, an utility function (*urk*) that favors a society where all significant minorities can see examples of themselves at all class levels. Reality is reality, and visuals matter (i.e. sex and race).

For example, I think a permanent underclass with few examples of success is far more harmful to all of society than, for example, the damage done to selected individuals (such as my sons) by AA. For me the trade-off is not about individual benefit, but the benefit to society as a whole. (And yes, competing against it is the cost to society as a whole of "judging men by the color of their skin" - I happen to think the trade-off is worth it, *especially* since a substantial decrease of certain minorities from higher positions in society *will* inevitably become a judgement against those minorities. Self-reinforcing systems and all that.

Of course, as always with competing utility functions, it's be decided at the voting booth. We'll see you there.

(Okay, you won't - I'm Canadian - but the general point still holds.)


I was disagreeing with the word choice of "right" before "utility function"... I wasn't promulgating my preferred indefinite article. I'm actually wrong, it should be "a utility function" which is why "an utility function" sounds like "urk" to you, for it does to me too.

Companies practice afffirmative action too. Does anyone seriously think that if Duke didn't let less qualified minorities into the school (and subsequently graduate) that the companies would refuse to go to University of North Carolina to achieve their diversity? They wouldn't and this would at the same time likely expose their recruiters to bright non-minority kids that either couldn't get into an elite school or flat out refused to pay the cost to attend Duke. In the end this would hurt Duke so Duke has a vested interest in affirmative action. Remember the difference between those who get into Duke and a lot of those who don't is often minimal. No one has ever been able to demonstrate that the outputs from an elite school is any different than that at a good non-elite school once you remove the affect of the input factors. One author even comically said the best thing for someone admitted to Harvard to do after acceptance was to go to a good state school and on his resume include his Harvard acceptance letter along with his diploma. I will give you an anecdotal experience of my own. Our younger daughter visited a well ranked regional university back in 2003. In between the time we registered to visit and the time we visited, the school came out in the USN&WR as a top ranked regional university. On the day of the visit more than twice as many visitors showed as had been expected. Was the school any better than it had been a month earlier? No, but people had something to stroke an ego once the rankings were published.

But the thing is if you don't go to Harvard or Princeton there really is a sort of glass ceiling there for you. What if you're really good but decide to go to a state school. Then no matter what you do or how good you are you'll always be limited - the upper echelons of society will always be out of reach.

CBBB, could you give some idea what you mean about "the upper echelons of society will always be out of reach"?

For example, what percent of the 1% (what is it, ~$250K/year?) did not go to Harvard or Princeton? You sound like you're claiming maybe 5% and that number feels extremely low to me. What percentage of congressmen and senators?

Oh I mean you'll never be appointed to the Supreme Court, never become a partner at any major law firm, never become a partner at any major investment bank, never get to be a professor at an elite university, etc.
Sure this might be far-fetched but if you can get into Harvard why put that ceiling up?

You could become a billionaire and buy yourself a bank and and a law firm.

These days the easiest way to become a Billionaire is to move up the ranks of the IBanking or Hedge Fund world - and you need to get your foot in the door somewhere and that requires an Ivy league degree.

The reaction to this study on campus has been negative to say the least. The Black Student Alliance actually organized a protest. What exactly they were protesting I'm not entirely sure, but for what seems like the 100th time in the past year, Duke is embroiled in a controversy regarding the current social order and culture. I'm glad I'll be graduating in May if only to escape the raving political correctness faction at every corner.

They are protesting the fact that they are lumped in with other people of color, and simply misdirecting their displeasure.

Humanities majors don't understand that protesting won't change the results of empirical research?

Might change the kind of research that gets funded.....

maybe this article helps explain what they are protesting

Pretty embarrassing article for them, in my opinion. I guess they are protesting that the study did not examine whether affirmative action for legacy students closes the gap between them and the rest of the student population? They are protesting that humanities majors are easier than engineering majors? That doesn't make much sense.

On the plus side, another sort of affirmative action was able to effectively ruin the lives of the Duke lacrosse team for being white. So, we can't say it was a complete failure.

Seriously though, I don't think we do kids any favors by putting them in situations their test scores say they can't handle. Rather than giving them a leg up, it engenders a sense that they can't succeed.

Sure they're getting a favor. Instead of getting a humanities degree from a third-tier school, they get a humanities degree from.Duke. That's worth a lot of money.

Are you sure it would be a humanities degree at the easier school?

"Instead of getting a humanities degree from a third-tier school, they get a humanities degree from.Duke. That’s worth a lot of money."

Not if your employer has a brain in his head, it isn't. A useless degree is a useless degree.

However, the one from Duke does come with an additional $100K worth of student loan debt. So our lucky recipient has that going for him.

Doubt that many minority kids end up with $100k in student load debt. The kids who end up at Duke are heavily recruited in competitive process among similar schools.

An employer would prefer a student from a more selective college, because while not being able to filter on IQ directly, then can certainly let the school do it.

Note that I'm ignoring what happens when the employer knows about AA at that college.

Well you're wrong though - getting a Humanities degree from an elite school is a totally different story then getting one from a mediocre school. Duke might not be QUITE elite enough for this to matter but a Humanities grad from Princeton or Harvard is going to do better then most Engineering grads from lesser schools.

Show me the numbers

You really think all these Harvard grads are ending up unemployed or in crappy jobs? I don't think so.

Well, that's a pretty different question from the one you asked originally. But, yes, I imagine many of them have crappy jobs.

Okay I mean crappy job - maybe they only made Partner at a 2nd-tier investment bank

Even better for those minorities is that they're ruining their employment prospects.

The undertaking comes as American higher education continues to experience declining enrollments in the humanities. Combined with a public perception that science and technology are the key drivers to a competitive economy, Duke faculty and administrators are committed to supporting the critical role of the humanities in a global world, said the initiative's principal investigator Srinivas Aravamudan, dean of the humanities and a professor of English, literature and Romance studies at Duke.

A good STEM student from an unknown school should be making six figures a few years out. While I'd like to see their surveys, I'd bet a Duke humanities graduate is most likely either an academic, at a nonprofit, working in a relatively low-paid humanities job, or unemployed.

Maybe for Engineering grads but I don't like this acronym STEM - Science and Math degrees are utterly useless. You don't actually learn anything practical in any of those programs, not much different from taking some kind of humanities program

CBBB, You do get something from Math and Science degree programs but only if you continue into graduate school.

More to the point: black kids that get into Duke via AA are probably smart enough to do just fine at a reasonable state university in a hard major. It's hard to imagine that a black studies degree from Duke is going to give you as good a set of prospects as a computer science degree from the University of Maryland, and it at least seems plausible that this may be the choice some of those black kids face.

I'll admit to a prejudice here: I've always suspected that X-studies (X=black, womens', native american, latino, etc.) majors are more-or-less the equivalent of subprime mortgages with balloon payments at year ten and no money down--a great way to screw someone relatively poor and unsophisticated out of their best shot at landing in the middle class.

My experience is the opposite is true -- except maybe for the top 5 schools in a given area where there are really challenging courses, it is harder to get the same tech degree from a state school because they tend not to practice the Harvard curve and you get less resources directed to you to help you succeed.

Asian Studies has always seemed to be a reasonably respectable major to me.

If you use an ASUS maroetbohrd then use the “EPU” utility. It drastically reduces the quad core power consumption. In my case it reduced from 125 watts to some 35 watts.More advanced option is to go to BIOS settings and decrease the CPU voltage to a stable voltage. Slowly work your way down into a stable one. This one is advanced and may damage your CPU but it will reduce CPU’s energy consumption by a lot.

"So get poor black kids out of their horrible high schools and middle schools and elementary schools? School choice ftw!"

You'd better show some evidence these particular black kids are all attending horrible high schools. Seems more likely that most black students at Duke go to the same good public and private high schools the white/Asian kids attended.

This is so silly it's a shame it gets such notoriety.

"Under one theory of affirmative action the goal is to give minority students an opportunity to catch-up to their peers once everyone is given access to the same quality of schooling." Well, I suppose that's "one theory," even though it's the dumbest theory I've read this week. Catch up? Peers???

The goal of affirmative action is to give minority people an opportunity their white counterparts (not "peers") already have. The proper study of success of the affirmative action program should be this:
(a) Measure the improvement in their lifetime success for white students who attended Duke. Compare that to their white peers who did not attend Duke (or some similar school).
(b) Measure the improvement in their lifetime success for minority students who attended Duke because of affirmative action. Compare that to their minority peers who did not attend Duke (or some similar school).

The degree of improvement would be the "benefit" of attending Duke. Which do you suppose would be greater, the Duke benefit to middle class white youths or the Duke benefit to underprivileged minority youths?

Good guess!

One question is, would you go to Duke if you were guaranteed to be the lowest ranking student? I'm not sure.

You mean the opportunity to attend a school they are not qualified for?

It's not so clear what the answer is here, right? If the alternative was a CS degree from University of Maryland or a Black Studies degree from Duke, is it really likely the Duke degree helped you out?

Nice critique, Ken. Could the opportunity lost here be the _ability_ to graduate with a STEM degree? To put it another way, AA is about providing equal opportunity to reach a desired outcome. If the outcome is to graduate with a STEM degree, but by and large, those who've been provided an opportunity to achieve such a degree don't make it and it's because they can't make the grades. Then, shouldn't we question whether AA is helping to achieve truly-equal, opportunities?

Goal is to provide opportunity of a college education to disadvantaged groups. Not eliminate GPA gaps within a couple of years.

At Duke?!

The assumption or hypothesis is they are disadvantaged because they have been denied opportunities but possess the potential to improve their outcomes when they are provided the opportunity.

If you reach down the application stack and say "this kid doesn't meet our standards, but that is only because they haven't been taught as well" you think that giving them your better product is going to help close the gap.

If the gap can never be closed then the assumption above is invalid and all you are doing is a handout. Alex claims this result suggests that the gap does not close at the highest college level. If students are denied the highest resources because they can never take advantage of those resources, then they are not really disadvantaged, that is resource allocation working as it should

OK, it's been a while since I've graduated, but median GPAs of 3.38, and 3.64 seem quite high. We're ALL exceptional.

can you sa grade inflation!

Note that these are non-cumulative GPAs -- i.e. they're not the GPAs they graduate with, they're GPAs for courses taken senior year.

Correct, Joe, they are not CGPA grades.

And they suggest almost precisely that the black students are getting B+ grades consistently in senior year and that the white students are getting A - grades consistently in senior year. I think that this is hardly abnormal/unusual. If you can't get between a B+ and an A- in 4th year, in your Major, after 3 years of high level instruction, you either aren't bright or you aren't working hard enough. Presumably at this level they are selecting their courses too, taking courses with content they are interested in and/or possibly knowledgeable about.

My senior year was one big party.

Clearly this black underperformance is caused by the white students stealthily exerting their white privilege to pressure blacks into performing poorly on tests and schoolwork, and Asians working too hard and ruining the curve and the fun for everyone. Thus, we need MOAR affirmation action so blacks can feel more comfortable and dilute the effect of evil whites and Asians at these schools.

I'm sure there's no way average IQ has anything to do with this. But just to be safe, let's start lobotomizing whites and Asians as children to close the racial achievement gap in the name of social justice.

Average IQ would have something to do with this if it were some fixed thing passed down to us from God. Yet there is plenty of data showing that IQ scores actually rise with additional education. On top of that, they fluctuate over time. So the premise is, under the right conditions and using the right tools, you can right the historical wrongs, regardless of the current level of IQ - because the right policies would raise IQs, scores, grades, etc.

I agree with everything you said, yet the recent report from Duke shows that we arent even close to being in possession of any of these 'right' policies or tools. There is plenty of evidence that IQ can be tweaked under certain conditions but there is just about as much evidence that hints at a biological origin. The last time I checked even policies dreamed up by some bureaucrat (perhaps influenced by sociologists/humanities professors with political agendas) STILL obey the laws of the hard sciences.

One thing to look for with these studies is if they used children/young adults or not. It is known that the heritability of IQ can be fairly low at young ages but when it is measured in adulthood it reaches very high levels. The development of the human brain before adulthood is still a mystery to be solved, and the article may be wrong if they meant to suggest that I could inflate 'my' IQ much more, since I am an adult already.

Then we have studies like this one:

Now a question that has been raised is, what value does the narrow-sense heritability even have when talking about it for human populations? I think normally it points to genes causing differences in individuals, but more complex models that would indicate exceptions to this rule have been proposed that need testing.

Ugh. Children/young adults tend towards developing brains while adults tend toward stale brains. Where as children are much more prone to curiousity and stimulating experiences, adults are prone to routine and sameness. There's nothing mysterious other than why so many adults allow themselves to get trapped in brain rotting monotony. The research is junk because our culture presupposes that adults have normal brains while children are developing into adult brains. All the research makes perfect sense once it is accepted that children have normal brains and adults need to relearn curiosity.

Genetics account for a huge amount of variation in IQ. I don't think there is much evidence of any intervention (in a developed country) that can reliably improve IQ scores. That would be a really big deal, right?

No creationist needed, as it's passed through genes, as you can easily see from Steve Hsu's Heritability 2.0. I'd link but i'm on my phone.

A liberal creationist would be needed though, to ensure uniformity between population groups.

Here you go:

Wow, this topic sure brings the racists out of the woodwork.

If you looked at average money spent on each student's education up to that point I bet you'd find most discrepancies explained. The White kids getting into Duke aren't coming from trailer parks or poor public schools.

Neither are the minorities. Read the study.

Yes, the people against a racial spoils system where we funnel opportunities from Asians to blacks are the racists.

What evidence do we have that IQ is the same across population groups? None, other than hopeful assumption. The evidence we have for intelligence being both highly hereditary and different across groups, however...

Monet spent doesn't seem to have much of an effect. Districts like Washington DC and Newark NJ are funded at a level similar to or better than their rich suburban neighbors.

"If you looked at average money spent on each student’s education up to that point I bet you’d find most discrepancies explained."

In the last 50 years, Americans tried to throw money onto the problems of their educational system. The problems persisted, so they throw more money, and, as Meg illustrates, are convinced that if the method did not work yet, throwing some more money would help.

Somehow, the fact that they now have extremely expensive public education which is at the same time one of the worst in the developed world, does not seem to put the "throw more money" theory in doubt. It is more of an article of faith.

Why the hell are you throwing in Econ with Science and Engineering? Getting an undergrad in Econ is joke it's nowhere near as tough as Sci/Eng

His footnotes are cute:

"If economics was classified as a social science, there is a slight decrease in the the study time for engineering and
natural sciences relative to humanities, social sciences, and economics. However, the coefficient remains statistically significant."

"If instead, economics is classified as a social science, the coefficient on female and black falls slightly but is still
statistically significant."

I thought that was funny too. I remember people who were having difficulty switching to economics, not the from economics.

My kingdom for the ability to edit...

I'm not even sure Economics is more difficult then a History or English program

It was harder than the math program, though.

No offense Tom, but that's pretty hard to believe since the first year of a Grad econ program is the third year of a math undergrad program. I.e. most econs have to catch up by taking mathematical analysis and grad level linear algebra.

I came for the CBBB rants. Was not disappointed.

I agree, will, I am staying for them too :)

At Duke (I'm a Duke Student who thought about but did not major in economics), Econ tends to be squarely in between the Sci/Engineering Majors and Humanities. Duke Economics Undergraduate Program is pretty rigorous. It requires up to Multivariable Calculus and Calculus-based Statistics, and has 5 core classes that are all reasonably difficult. It's certainly nowhere as difficult as Engineering or any of the major sciences (Chem, Bio, and Physics especially), but it's also much more rigorous than anything in the humanities. It's the awkward middle child in terms of difficulty IMO.

That's not bad actually - but far and away most universities just tend to require a basic math course plus a maybe 1-2 stats courses for Econ majors.

If the Duke Economics program is requiring Calculus then surely it's closer to a Sci/Eng major than Humanities.

At MIT, the Music major requires calculus.

I thought that was cute too. At MIT it was pretty common for engineers to take economics classes to pad their GPAs.

"76.7% of black males expect to major in the natural sciences, engineering or economics"

"68.7% of white males expect to major in sci/eng/econ"

Is Duke an outlier in this regard for a non-technical University? Both percentages seem extremely high to me.

Duke is primarily a sci-tech school. Its like Stanford, Rice, MIT etc.

Although I will say contrary to what Tyler Cowen seems to believe - people are streaming in to STEM programs and have been for years and years

Huh? Didn't he have a post showing the number of degrees in STEM has been flat for like 30 years or even declining in some majors?

No he never did, he's been making all this up

You are referring to this post.

I don't get what CBBB's deal is.

How is this an unexpected or controversial finding? If a group of people is on average educationally underprepared entering college, isn't it to be expected that rather than somehow magically catching up - which would itself negatively implicate the value of secondary education - they would instead switch to easier subjects and therefore still be able to receive a degree from a prestigious university?

That's what Alex is saying. In other words, it is to some degree not like back when black athletes just needed the opportunity to get on the field of play at the pro level.

We've already seen that pre-college preparation almost entirely explains success in STEM fields, and it is nearly impossible to transfer into engineering after the first year of college (especially if the student doesn't want to pay for extra years of schooling). Humanities have done a much better job of welcoming students who may never have encountered the subject before; if STEM fields want to graduate more students they are going to have to expand the potential base of graduates.

Luckily, some schools are doing this successfully, so we know it is possible. The most successful efforts involve that dreaded word "tracking", and have mostly been driven by gender disparities rather than racial.

The reason these findings are controversial is because it gives a bunch of people data that conforms to their pre-existing biases and lets the racists rant on about average IQ and so forth. It is really hard to stay racist in the modern world, so any data that may be interpreted as supporting the continued policy of giving privileged students more resources than poor, Black or rural students gets lots of air time.

I still think we should make public schooling mandatory at least for politicians' children. Right now the people with resources have no incentive to support public schools, and it is significantly harming America's economic competitiveness.

Humanities have done a much better job of welcoming students who may never have encountered the subject before;

In the sense of being easier, you mean?

Yes. But so what? Meg's more salient point is that STEM success is almost entirely explained by pre-college preparation.

In the real world some jobs are easier than others too. But I'll wager having a degree from Duke makes it easier to land one of those jobs than one from a state school.

For the equivalent degree, I agree that tends to be true. But the argument that STEM is going to be just as easy as humanities if they just "welcome" students better is as nonsensical as the notion opposition to racist affirmative action is primarily driven by racism.

"We've already seen that..."

Really, we have?? Doesn't this study say that test scores/IQ largely determine results? Those things don't depend on pre-college STEM preparation, do they?

You seriously think IQ plays no role in school success? Because that pretty much contradicts all available evidence.

Not to say tracking programs/extra help is not warranted, I have no problem with that.

It's only surprising if you start with usual assumption underlying affirmative action, which is that the test scores need to be adjusted because they're unconsciously racist and don't accurately reflect the talents of (certain) minorities.

I don't even wish politicians on their children.

Affirmative action doesn't need to make blacks catch up to whites to be socially desirable. If we think whites are unfairly privileged, then actions that help blacks to do better will counterbalance some of this privilege. The proper comparison group is blacks who did not benefit from affirmative action - who surely do worse than blacks who go to Duke only because of affirmative action.

The paper demonstrates that affirmative action may not maximize productive efficiency, and allocating college degrees to blacks who would not go to college absent affirmative action is not efficient. But it may still be socially desirable.

> The proper comparison group is blacks who did not benefit from affirmative action – who surely do worse than blacks
> who go to Duke only because of affirmative action.

This doesn't warrant a "surely." Who is better off: the Music major from MIT or the Mechanical Engineering major from UMass? There are a whole bunch of questions here without obvious answers.

From purely anecdotal experience, the Music majors from MIT that I know are all better off than the Mechanical Engineers from UMass. I doubt cause and effect, however, as if one chooses to be a Music major after getting into MIT it isn't because you couldn't have been a ME major at MIT but because being a Music major gives you additional utility above majoring in ME at MIT.

Students are still economic creatures making the most rational decisions. Which college you graduate from may well be worth more than what major you pursue there. I know I have done better in CS with an Art degree from a well-known major university than my friends who have CS degrees from state schools. A degree from MIT in any field will get someone an interview at a software development shop, possibly in preference to CS degrees from less-prestigious schools: it implies both intelligence and a well-rounded individual who can learn new things.

This study also assumes that the desired outcome of college is GPA. I'd love to see the earning difference between Black students who were and were not admitted to Duke vs. the same statistics for White students. If Black students, even with their lower GPAs, get more economic value out of a Duke degree than their White counterparts, it is optimal utility theory to prioritize them even if the absolute value might be less. Comparative advantage and all that.

That's kind of the opposite. If you love music, that's fine. But if you love engineering and Duke is too hard at engineering so you go into education, that's different. Are they really going to get in the door at Exxon with a Duke degree as opposed to the first choice school that they could have finished in their preferred major?

It seems to me an extreme signaling and human capital argument to say you spend 4 years majoring in something completely different than your ultimate job. Actually, my position is that I did major in my ultimate job and STILL spent 4 years doing completely different things.

I doubt cause and effect, however, as if one chooses to be a Music major after getting into MIT it isn’t because you couldn’t have been a ME major at MIT

The correlation with grades is coincidence, then? Really?

I know I have done better in CS with an Art degree from a well-known major university than my friends who have CS degrees from state schools.

Er, what sort of CS work does an Art degree prepare you for? Are you comparing yourself to web designers as opposed to analysts? I know CS people from state schools who are multimillionaires, others who are unemployed from top schools.

I think there's something to be said for top schools in terms of getting a foot in the door, esp. at consulting companies, but it doesn't appear to matter all that much ten years out.

Getting a foot in the door is kind of the whole point of a degree.

Sure, I'm only saying it's probably not that predictive of future success, all else being equal. Getting a foot in the door only helps you if you manage to squeeze through.

Well the thing about a CS degree is the vast majority of the content is not relevant to the day-to-day work of a software engineer. Some of the content might be helpful but a lot of it is pretty academic or can easily be looked up (like all the standard algorithms).

Depends on your school/program, I think. Some degrees are more focused on practical issues like solving real-world problems, value-added metrics, user acceptance, building real apps, etc.

There are two kinds of Music majors at MIT, the double majors and the "this is a form of dropping out that doesn't involve my dad killing me" majors. It's my perception that the second is a lot more common than the first. If you double at MIT between a tough field and Music, I respect you.

> being a Music major gives you additional utility above majoring in ME at MIT.

I just don't think this is often the case. Nobody goes to MIT for the music.

Music majors from MIT that I know

In 2001 (the easiest numbers I could find), there were two students in that major.

And they use infinite series instead of a staff.

Racial discrimination can be socially desirable you say? Do you really want to open that can of worms? After so much brutal work was undertaken at convincing almost everyone otherwise and even putting the opposite principle into law? I don't want to open it, and I think a some of our problems right now are caused by not going far enough in making it clear that reverse-discrimination is morally wrong in the private sphere too. If one group is privileged over another and none of this privilege can be traced to discrimination against other protected groups based on their membership to that group (current discrimination, not former), then there is nothing to be done about it because we are all treating each other like individuals. If we discover remedies that can improve the environment of individuals and help them succeed then these remedies should be available to everyone.

"Under one theory of affirmative action the goal is to give minority students an opportunity to catch-up to their peers once everyone is given access to the same quality of schooling."

Under another theory the goal is to make liberals feel good about themselves. This study, I suppose, also undermines that goal so it will be viciously attacked...RACISTS!!!!

Haven't seen many mentions of the misallocation problem this seems to suggest:

What would these disadvantaged minority students have majored in, and where, in the absence of affirmation action?

Would it have been humanities at a lower ranked college? Or engineering at a lower ranked college?

If the latter, would that scenario have been better or worse that a humanities degree from Duke?

The black students in the study are certainly not from disadvantaged backgrounds. Their parents' educational backgrounds are as follows (from p. 7):

Mother's education:
BA or more 65.6%
Doctorate 11.2%
n/a 12%

Father's education:
BA or more 68.8%
Doctorate 21%
n/a 12%

Better formatting:

Mother's education:

BA or more 65.6%

Doctorate 11.2%

n/a 12%

Father's education:

BA or more 68.8%

Doctorate 21%

n/a 12%

First, the paper tested the hypothesis Alex proposed--that grade differences between science and non-science would shift students--and controlled for it, and bascially found that pre-college preparation of minority students for the sciences explained the shift.

Second, the answer should be obvious, but not one likely to be seen on this site: 1) enhance pre-college science prep for monorities; and 2) develop programs after admission and during enrollment to assist minority students in the sciences.

Third, Alex's post precedes from the premise that a science major for a minority has the same value as for a majority student, and, more importantly, that a majority students market value for a non-science major is the same as a minority student. I would suspect that since there are few minority student graduates that the value of a non-science major for a minority is higher than for a majority student. I would like to have seen the paper track job placement and income over time to see whether the minority student, controlling for everything else, made a bad choice in switching to non-science graduating major. (By the way, a non-science graduating major may have enough science training to be admitted to a medical school, to be a manager in a software company, to be admitted to a physical therapy program, or to be a banker.)

You are so funny, Bill. How is it that you make a comment on the post where Alex says exactly what you are saying won't be seen on this site.

Andrew, I am sorry, but you didn't read Alex's site, so let me quote you what he said as if it was some hypothesis the paper did not test: "It’s well known that grade inflation is higher in the humanities and the social sciences so the shift in college major can easily explain the shrinking black-white gap in GPA. (The authors show that grades are higher in the humanities holding SAT scores constant and also that students themselves report that classes in the sci/eng/econ are harder than classes in the humanities and that they study more for these classes)

Now, go show me where Alex said that schools should develop programs after admission and during enrollment to assist minority students in the sciences.

And, Andrew, I would also not that there is no hypothesis testing for whether an enrichment/assistance program at Duke for minority (and even for majority students who were equivalently prepared) would not work. All there is is a test of the hypothesis that placing a student of similar backgrounds will lead to similar results.

One of my law partners daughters went to Yale, majoring in a non-science degree. She was smart, but couldn't get into medical school because she didn't have enough science credits. Dad spent some money for her to take a year at Northwestern for a pre-med program. Today she is an ob/gyn.

Just goes to show you what money can do and what luck you can have if your dad had the sperm of a wealthy lawyer.

His last line says that remediation should take place earlier. That is what you say, much more specifically, without any backing evidence I might add.

If Duke science training doesn't fill a potential gap, how is it so obvious to everyone not on this site that 3 months earlier of remedial science training will turn the trick? It's not. Alex says what you say, except that he says something that has support, linking to Heckman, who I think is probably mostly wrong, but that's beside the point.

So you are proposing that the sperm of wealthy lawyers be distributed more equitably? I am on board.

You beat me to it, Andrew. Bill again not being able to read plain English.


Despite what you say about me, I am so glad that you, Alex and Andrew are with me in supporting affirmative action, improving pre-college programs for minorities, and for creating in-college programs to assist minorities in improving math and other science pre-req's.

Stand up and take a bow.

Sadly, however, my sperm is not for sale or even free distribution. Sorry.

Oh, I didn't know you were a lawyer. I was talking about myself.

By the way, do you think affirmative action etc. should be on the basis of race and not income/background? If so, why?

Cliff, In answer to your question: 1) income is significant irrespective of race, so I am in favor of equalizing opportunity and 2) actually landing a job after being educated, or even getting accepted, is function of a network both in and outside of school: ask yourself a question: what is the network of a majority v. a minority person in getting/finding a job.


I think once you remove income from the equation there is not much difference. I don't know.

Well, I for one am impressed that Bill has tossed aside the racist portion of affirmative action. (Of course, the Democratic Party will never do this -- they have always been in the business of using the gov't to hand out rents on the basis of identity politics, only the recipients have changed over time.)

As to networking, there are a plethora of explicitly ethnic identity groups and associations for networking, and virtually none for whites qua whites.

And if Yancey bows, I will clap.

TallDave, the biggest networking opportunity for any minority student is the network the student developed in school, and the network of alumni outside of school. This means that to equalize opportunity, you would join with me, Yancey, Andrew and others in supporting the recruitment of minority students and their exposure to the network.

Welcome aboard.


I am really confused about why you want to preferentially admit "minorities" (presumably non-Asians). Is it meaningful to speak of "equality of opportunity" when you are elevating some people above others on the basis of skin color instead of merit? I mean if they are disadvantaged in some way I can see an argument, but just based on skin color alone? Isn't that the opposite of quality of opportunity?

Dear Confused aka Cliff,

I share your concern about the admission of those minority of students admitted on the basis of their heritage and ancestry.

Heritage and ancestry should not be a factor in admission.

So, please join me in opposing legacy admissions.

Is that the minority you are talking about?

Oh, and Cliff, for extra points, and regarding legacy admissions, guess in which year Duke became desegregated:

a) 1865

b) 1921

c) 1948

d) 1963.

"Oh, and Cliff, for extra points, and regarding legacy admissions, guess in which year Duke became desegregated:"

I'm going to go with E) Before the parents of the average Freshmen were born.


E is correct.

But, those parents were not Duke alumni because their parents were not alumni.

When you think legacy, think N + 1, and not N.

Good try though.

And the benefits blacks get from affirmative action far outweigh the benefits white get from legacies. Legacy admissions get you the equivalent of a 160 point boost on the SAT; whereas being latino gets you 185 and black 230 vs. being white. And being black is worth 280 points over being Asian (

White vs. black is a misdirection, as I've said elsewhere in this thread, since Asians suffer most from black favoritism.

Mily, go back and look at the study. They controlled for SAT. The studies that I've seen give greater weight to legacy. As to Asians, what do you suppose their legacy weight is?

Miley, the study you cite gives a -50 effect on Sat for Asians with a +160 for legacy by the way.


I'm just about wrapping up my internetting for the day, but off the top of my head to your latter inquiry, the -50 for Asians and +230 for blacks is vis a vis whites, so +280 for blacks against Asians.

Bill -- no, college admissions are a result, not an opportunity. If that wasn't true we could just throw out the whole admissions process and admit people randomly to equalize opportunity.

I'm fine with opposing legacies, that never made much sense to me anyway. Let's also get rid of athletic scholarships while we're at it, no one benefits from a classroom full of Communications majors.

Miley -- and don't forget, legacy advantages are much rarer, something like 10x iirc, so even if they exist they're so much less pervasive that it's hard to hold them as any kind of equivalent.

Miley, So, a differential of 210 for Asian against legacy.

Miley, I want to put you on notice that the study you cited did not control for citizenship. So, with respect to Asian admittees, US colleges admitted very high Asian SAT qualified non-citizens, which then distorts the Asian data--in other words, to leap over domestic bodies, Asians(the predominant minority for transfer to US schools) have to be above average. I am sure that you did not spot this defect in the study, but it was pointed out to me by a registrar of a East coast college.


If you do a quick back of the envelope thought experiment as to what kind of figures such as sheer number of Asian internationals vs. Asian Americans and/or how much "above average" Asian internationals would need to be to account for the discrepancy, it becomes obvious that Asian internationals being "above average" cannot account for Asian over-performance.

What's also very well known is that black students choose to take easier courses to boost their GPAs. The typical Freshman will take mostly required courses and by their senior year they are taking mostly electives which are arguably easier or at least better favored.

You never quite grasp this reality until you actually see black student organizations offering coaching sessions on how to boost the GPA in this fashion. They don't even pretend to hide the fact. Black students also have greater access to free tutors, especially if they are athletes.

Grade inflation is a reality, but one should also admit there is survivor bias among seniors. Many students wouldn't be seniors if they had crappy GPAs, so there is bound to be compression without rankings changing.

Course selection bias, sure. but the original post says you can ignore survivor bias: "almost all Duke students do graduate so this result is due to a shift in major not dropping out"

That's why I took Economics classes: they massively increased my GPA. The difference was that it was my advisor and parents, rather than a student group, providing those suggestions and coaching.

Students are rational beings; as long as GPA is what matters they will maximize GPA. If field of study were what mattered they'd be maximizing fields instead. How can economists become scornful of people responding to incentives?

Students are also extremely inexperienced beings, being advised in what to major in by parties whose interests are not aligned with theirs.

I'd be curious to see the effect on their CHILDREN. Does affirmative action help give students opportunities that they wouldn't otherwise have, allowing their children to have the opportunities that their parents should have had (ie effects from acculturation)?

Agreed...a study of cross-generation effects would determine whether the present efficiency cost of AA is worth it over time

Considering the very high educational level of the minority students' parents, that seems unlikely.

Well your argument is becoming circular.

What do you mean? Adam is proposing an inter-generational effect where educated parents beget more highly educated children. But the parents are already about as educated as you can get. Doesn't that imply that AA should no longer be necessary?

Why can Tyler play Alex, but Alex can't play Tyler?

It's okay--Alex and Tyler still both benefit from the collaboration, thanks to comparative advantage.

Why group econ with science and engineering? Econ classes are a joke compared to science and engineering. I took quite a few of both and the amount of study time needed to get an "A" in an econ class is about a fifth of what is required in a typical undergrad chem or physics class.

Quite a few undergrads move from science and engineering into econ as it is so much easier. If you an aptitude for math the average econ class requires very little work (sorry Prof. Tyson).

According to one of the posts above, an Economics Degree at Duke requires Calculus. Thus it's more aligned with Sci/Eng than humanities.

Wow, calculus. Then it must be just like electrical engineering or physics degrees.

BTW most econ departments require require calculus. Though it is usually the watered down version: "Calculus for Social Scientists" or some such puffery.

"Wow, calculus. Then it must be just like electrical engineering or physics degrees."

For the record, I have an EE degree. And I didn't say it was "just like EE", I said "it’s more aligned with Sci/Eng than humanities." Reading comprehension is important in all aspects of life.

"BTW most econ departments require require calculus."
Can you cite this?

I don't think most econ departments require calculus - they usually require some "Mathematics for Economics Students" course which is a hodgepodge of calculus and linear algebra but not a straight calculus course.

I actually found econ to be more difficult than everything except for chem. Econ is a social science with a lot of math. Physics was definitely the easiest of the three.
I'm sure there is a difference if any of these is your actual major.

Well that would put you in the distinct minority. I can tell you my physics classes at Berkeley were ten times harder than econ and every single person I knew who took both felt the same way. The math in undergrad econ classes is first year calculus most. Yes, some undergrads will take a lot of econometrics and mathematical econ. but most do not.

Until we're willing to tackle "inequality of upbringing" - which will mean all people are raised by the state - this probably not fixable. Make it so that everyone gets roughly equal education, nutrition, physical security, cultural exposure, a rotating cast of highly skilled caretakers, etc. and we will stop having these discussions (and won't need AA anymore.) People in this country value their children too much as status icons, mini-me's, pets, free labor, personal accessories for that to ever happen, though.

Well, we would still need AA, if AA is a response to unequal outcomes.

Okay, but it seems like a great deal of this inequality has already been leveled by legal and social forces at work in first world countries, especially for middle-class parents. I mean, if we're talking about kids raised in a shack in El Salvador or Haiti, yeah, their upbringing is probably so much poorer and harder than some middle class white kid that differences in outcomes for the two are very likely to be affected by that difference in upbringing. But if we're talking about the child of a black policeman/nurse couple and a white policeman/nurse couple, there are differences in culture and expectations, but they're way, way smaller, and it seems really unlikely that those differences will drive large differences in outcomes.

I'm guessing controls like: (i) whether the student worked during school; (ii) whether student had already taken the first year science course by doing an AP major; and (iii) access to outside support, were all taken into account?

Just because a student is admitted with a lower test score, doesn't mean the playing field has been leveled. Also, do employers actually care whether you've done a BSc. in chemistry vs. a BA in history? For a large number of service jobs, which still form the bulk of our economy, I'd still take the person with superior verbal skills.

I have argued that the United States would benefit from more majors in sci/eng/econ but that is not the point of this paper.

Do you think that the United States would benefit from more majors in sci/eng/econ or from having more people with the ability to graduate with majors in sci/eng/econ given the current level of difficulty and boringness of classes in sci/eng/econ?

Else we should make classes in sci/eng/econ easier and more interesting/fun.


IMO the easier life gets the less motivated people are to work real hard to get an advantage (by say studying engineering).

I mean perhaps you can make the argument for Engineering degrees but Science and Econ are not really in demand at all. And why the fuck do we need more Econ grads? That's pretty self-serving of you to say that.

I agree with you on econ being easy but Alex included it and so I just played along.

No! You don't play along! You have to denounce him

OK CBBB Mia culpa.

Why is economics lumped in with science and engineering, but mathematics is not? It's a weird choice. Maybe there's some pretense that economics is "harder" (in its reliance on data) than other social sciences. It still seems to have a lot more in common with, say, political science than, say, computer science.

Because Alex and Tyler are serious scientist, doing important scientific work. Otherwise they would be hypocritical hacks who get to sit in their ivory tower of tenure and look like jerks for steering everyone else away from the easy life of a genuine guild-lord economist-professor to the hard scrabble life of a mechanical or computer engineer who might be outsourced/made obsolete.

"The authors do not discuss the consequences of dashed expectations." -- This is an evocative comment. It implies an expectation that there are some negative consequences of dashed expectations. I am curious as to what those might be -- loss of morale, self-confidence, etc.?

And on the basketball teams it is the opposite.

What do you mean? Duke Basketball has been notoriously (and falsely) accused of being too white. Much of the Duke hatred is out of the perceived whiteness of it's basketball team.

How do you perceive white?

Yes, AA may need to occur much earlier. It may also need to occur much later, through, for example, the development of retention programs within the fields black students are apt to drop out of. This may also need to be accompanied by increased institutional support for students who, because of initial unpreparedness for a field, are not able to proceed as quickly through the fixed, tracked schedules many STEM majors operate on, relative to soc sci and humanities fields which are much more lax in their course requirements.

Affirmative action or any other measure designed to correct for past discrimination would be more effective if applied earlier in the lives of the target group. However, having it later is still better than not having it at all.

The unidirectional transfer of opportunities from Asians to blacks, known as affirmative action, isn't enough. As I said earlier, we need to "start lobotomizing whites and Asians as children to close the racial achievement gap in the name of social justice."

If we lobotomize Whites and Asians to close the racial gap, we would have taken away their natural right to the pursuit of happiness and so there would have to be affirmative action efforts made to remedy the wrong done by lobotomizing people.

"having it later is still better than not having it at all"

... said the doctor from 911 and started giving artificial breathing to a skeleton.

Not quite like CPR to a skeleton. Affirmative action for blacks may or may not be effective but more importantly, it is a question of justice. An entire nation was deprived of its fundamental human rights for close to two centuries in law and longer in practice. We haven't even come close to compensating them for the barbaric practice of slavery. I don't mind the abolition of affirmative action as long as it is replaced by a different (more effective?) method of compensation. Black Americans are not under-performing primarily due to inherent inadequacies in ability or intelligence. Rather, their self-image and self-confidence as a nation was shattered by being treated as subhuman and property. It will take at least two more centuries for the wounds and aftereffects to heal and disappear.

"We haven’t even come close to compensating them for the barbaric practice of slavery."

THEM is long dead. Something like 6 generations. In this logic, Mongolians should compensate the Russians for the violent raids in High Middle Ages, Germans should compensate all the nations around endlessly (German occupation of Poland in WWII was, although much shorter, waaaaaay more brutal than any kind of slavery in the USA) etc. Pretty much everyone has claim on everyone else, if you get back at least 200 years.

Also, by selecting by skin color, you'll 'compensate' black people who recently came from abroad, and whose ancestors had nothing to do with slavery, as well as suppress non-black people who recently came from abroad, and whose ancestors had nothing to do with slavery. Basically, an Albanian-American and a Laotian-American will compensate a Kenyan-American for the fact that a Scottish-American oppressed a Senegalese-American 150 years ago. I see no trace of justice in that, especially given the fact that many of the recent immigrants come from seriously impoverished or totalitarian nations themselves, and already do start with a significant disadvantages (no inherited property, lack of knowledge of English, lack of formal education). Should they also be put to further disadvantage because they were born with a non-entitled skin color? In America, of all countries?

As for the performance of blacks, your claim about "Black Americans are not under-performing primarily due to inherent inadequacies in ability or intelligence. Rather, their self-image and self-confidence as a nation was shattered by being treated as subhuman and property." is not a scientific assertion, but an opinion, and I would rely more on some anthropologists' explanations.

There are plenty of oppressed minorities (Jews, Arab Christians, Iranian Zoroastrians, Sikhs, the Irish in the old British Empire), who almost immediately jumped to high performance when they moved out of the oppressive law system, and it took them less than a generation, not 200 or 300 years, and you do not seem to have explained that.

1. "Them" is not long dead. The trauma and damage that was inflicted by slavery is passed down from generation to generation in the way of negative outlooks on life the possibility of success, persecution complex, fear/distrust/hatred of authority and a poor self-image. Just because other nations didn't compensate doesn't make it right.

2. The Germans lost Posen, West Prussia, East Prussia, Farther Pomerania, East Brandenburg and Lower Silesia between the two world wars. However, there is a difference between international events and purely domestic ones when it comes to compensation so the Holocaust is a more appropriate comparison than the German conflicts with the Poles and other Slavs. The Holocaust - without a doubt another example of barbarism in the West - led to the creation and sanction of the State of Israel and to this day, Jewish organizations are seeking reparation and/or the return of property seized by the Nazi regime and its sympathizers. The descendants of Blacks were not financially compensated and no part of US territory was surrendered for the purpose of creating a Black American republic.

3. Perhaps a recently arrived Laotian-American should not compensate a recently arrived Kenyan-American for crimes that affected neither of their people unless you make the argument that a newly-arrived citizen accepts responsibility - on equal footing with a native - for all the debts of the nation and benefits from - on an equal footing with a native - all the benefits of a nation. The Kenyan-America is of African descent and although not directly affected by the effects of slavery as were the Black Americans, s/he was nevertheless a member of the targeted race and is now in a country still plagued by racism - particularly toward Blacks.

4. Even if we assume that Blacks are genetically predisposed to be inferior to others in regards to intelligence, there is ample hard evidence to show that Blacks who immigrated by choice and who were not subject to slavery have descendants who are less antisocial, less likely to end up incarcerated and more likely to succeed both in education and in the workforce.

5. There are plenty of oppressed minorities. Jews must be given opportunities equal to others everywhere in the world, so should Arab Christians and Iranian Zoroastrians, Sikhs etc... None of the groups you mentioned endured a fate as damaging as what slavery imposed on Black Americans. The Holocaust was shameful to say the least but it lasted less than 20 years. Iranian Zoroastrians are protected by the constitution of the Islamic Republic and the are guaranteed representation in the Majlis, Arab Christians were never enslaved etc... Perhaps that is why recovery took less time in the examples that you cite.

I blame Harvard for this mess. Let's assume that each racial category has a normal distribution (normal curve) of intelligence for their entire racial population. Now, due to social-economic status, culture, and historical path dependency, we will assume that the black normal curve is slightly behind whites and asians in the present day.

Now this is where Harvard comes in. Schools like Harvard and their ilk MUST reach racial quotas regardless of the normal curve, in this case let's say 10% of Harvard's incoming class must be black. However, based on the black normal curve, maybe only 6% are capable of doing Harvard level work. What this means is that Harvard will let in 4% of blacks who will not be able to do its level work, HOWEVER, these 4% of blacks are capable of doing the work at say, University of Virginia, or even Duke, but now they go to Harvard. In turn, UVA and Duke will now take say NYU black students and the process trickles through out the chain.

The result is that EVERY university will have black students not capable of doing the work at that school. However, this is the "empirical" evidence that everyone sees to conclude that blacks are worst students in college. If Harvard and their ilk accept less black students (FOR NOW) then all black students will end up at schools where they match up with their peers and can lay to rest this notion that they are worst students. In maybe one or two more generations, when the black normal curve catches up with white and asian normal curves then schools like Harvard and their ilk can accept more blacks without having that downstream affect.

So in the end, blame Harvard and the quotas that they must fill in the current day having this trickle down effect throughout the entire system. Maybe if the 4% (example) of Harvard blacks that should not be there (based on statistical normal curve) went to Duke, maybe these "researchers" would have found a different more positive outcome.

Comments for this post are closed