The Jordan B. Peterson Podcast - 384. This Man Ended Affirmative Action ｜ Dr. Peter Arcidiacono

00:00:00.940 Hey everyone, real quick before you skip, I want to talk to you about something serious and important.

00:00:06.480 Dr. Jordan Peterson has created a new series that could be a lifeline for those battling depression and anxiety.

00:00:12.740 We know how isolating and overwhelming these conditions can be, and we wanted to take a moment to reach out to those listening who may be struggling.

00:00:20.100 With decades of experience helping patients, Dr. Peterson offers a unique understanding of why you might be feeling this way in his new series.

00:00:27.420 He provides a roadmap towards healing, showing that while the journey isn't easy, it's absolutely possible to find your way forward.

00:00:35.360 If you're suffering, please know you are not alone. There's hope, and there's a path to feeling better.

00:00:41.780 Go to Daily Wire Plus now and start watching Dr. Jordan B. Peterson on depression and anxiety.

00:00:47.460 Let this be the first step towards the brighter future you deserve.

00:00:57.420 Hello everyone watching and listening.

00:01:11.120 Today I'm speaking with professor, researcher, and econometrician, Peter Parsidio-Cono.

00:01:16.760 We discussed the recent landmark decision by the Supreme Court to end race-based affirmative action.

00:01:25.940 How Peter's research was instrumental in that outcome.

00:01:30.240 Why merit has repeatedly proven to be the best indicator of success, and what merit is, by the way.

00:01:36.960 How compassion is used to cloak racial discrimination, and what might actually yield results in service to the under-resourced communities across the United States.

00:01:48.760 So Peter, let's start with this.

00:01:52.380 Affirmative action has been in the news a lot, well, for a long time, but particularly in recent weeks, given the new Supreme Court decision.

00:02:01.020 I think we should, first of all, alert everybody watching and listening to who you are and why people should consider you a valid source of information and what you do.

00:02:14.480 So fill us in about who you are and what you do and why this is a topic of interest to you.

00:02:19.080 So, you know, I'm an economics professor who studies affirmative action in higher education, and sort of as a result of that, got the opportunity to be an expert witness in the two students for fair admissions cases that were recently decided in the Supreme Court, one with Harvard and one with UNC.

00:02:41.240 And I took the cases in part because, you know, for someone who studies affirmative action, we've never had the data to really look at it well.

00:02:53.900 Universities typically hide their data, probably as a result of these lawsuits.

00:03:00.520 So, you know, there's a large gap in racial preferences between it being a tiebreaker and being what somebody like Abram Kendi might be in favor of, of equal outcomes.

00:03:11.580 So understanding exactly how big the preferences are, to me, would move us in a direction thinking about optimal policy.

00:03:19.100 Now, as it stands, you know, the way the rulings went, we're not supposed to have affirmative action.

00:03:25.180 I think universities are probably going to look for ways to get around that.

00:03:28.460 But the thrill was the ability to actually see Harvard's admissions files.

00:03:35.040 You know, I got to actually see their full database across six years.

00:03:38.780 I got to look at the actual applications themselves, look at the reader comments for a subset of these things, you know, all the alumni interviews.

00:03:48.140 It was an amazing experience, a frightening experience, but an amazing experience to be able to see all that.

00:03:55.200 Okay, so let's start with two things.

00:03:58.760 Why don't you outline the cases for us?

00:04:01.960 You said there were two cases.

00:04:03.200 And then maybe you could step people through the typical admissions process at, let's say, at Harvard and UNC.

00:04:10.200 That's going to be similar for many universities.

00:04:13.620 But everybody watching and listening needs to know how universities do, how they claim they admit, how they actually admit, and how they should admit.

00:04:23.020 Those are three, obviously, separate questions.

00:04:25.080 But let's start with what exactly, why exactly were these universities in court to begin with?

00:04:32.400 So there's actually two different reasons.

00:04:36.360 On the Harvard side, it had a lot to do with Asian discrimination.

00:04:41.420 We think about affirmative action, and it really is designed to help African-American students and somewhat Hispanic students.

00:04:50.360 But Asian-Americans is a population doing incredibly well academically and also in other areas.

00:04:56.780 And so Harvard was a good place to look at the Asian discrimination side of things.

00:05:04.420 So in the Harvard case, you had both Asian discrimination.

00:05:07.500 You also had, are racial preferences narrowly tailored?

00:05:12.680 You know, how big are these racial preferences at Harvard?

00:05:15.620 So those are sort of my two parts of the case.

00:05:18.200 There was another expert, Rick Kallenberg, who covers race-neutral alternatives.

00:05:25.480 Because that's the other thing the Supreme Court has said, is you're supposed to look for ways to get diversity without using race explicitly.

00:05:35.740 On the UNC side, we didn't have the Asian discrimination claim.

00:05:40.500 It was more on, you know, how large are the racial preferences coupled with these race-neutral alternatives.

00:05:46.300 But then there's another aspect, too, which if you look back at the two Michigan Supreme Court cases, the one on the undergraduate side said you cannot use a formula.

00:06:00.100 But on the law school side, you can't use race as part of a formula.

00:06:03.700 But on the law school side, you could take it into account holistically.

00:06:07.360 I mean, as an economist, and I'm sure for you, too, if I just hide part of the formula, now I'm holistic.

00:06:16.260 You know, so the question is, you know, how formulaic were the admissions?

00:06:21.600 And, you know, my models could predict UNC admissions incredibly well.

00:06:26.640 So that would also sort of speak to, you know, what does it mean if we don't write down the formula?

00:06:33.160 Is it okay?

00:06:33.600 You know, but, so those are sort of the heart of the two issues.

00:06:38.940 If you don't write down the formula, and yet you couldn't derive an equation from analyzing the outcome, all that means is that the admission process is random.

00:06:50.420 And that seems, I mean, you could go to a random admission process, right?

00:06:53.940 You could let every comer in, and then you could fail everybody who doesn't do well in the first two years and let the survivors flourish.

00:07:02.320 And, I mean, you could make a case for that.

00:07:04.980 It's somewhat inefficient because all those people who fail have to go there and fail.

00:07:10.140 And that's not necessarily so easy on them.

00:07:12.700 But there are certain advantages to that because there will be surprising successes as well.

00:07:18.760 That isn't generally the direction that we've chosen to go.

00:07:21.920 And then if you do have a more rigorous policy, so that would be one you could hypothetically model, well then, in principle, in fact, this is actually by law, as far as I understand the law, is that my understanding of the law with regards to hiring and selection, at least, and I don't know if that pertains to academic admissions,

00:07:42.100 is that you have to use, you are required by law to use the most valid and reliable current means of evaluation available that don't produce counterproductive and illegal racial differences.

00:07:58.460 And so, and no one knows how to do that because, well, because no one knows how to do that.

00:08:04.340 It isn't obvious how you can do both simultaneous.

00:08:07.240 In fact, it doesn't look to me like you can, and that's a major conundrum.

00:08:10.640 So, okay, so back to UNC, so you modeled their equations, and you did the same thing for Harvard, and you found, and you phrased it in these terms too,

00:08:23.040 is that for every advantage that's given to one race, let's say, or one category, there's going to be a commensurate disadvantage applied to another.

00:08:32.040 And that's particularly egregious in the cases of Harvard, and also, I believe, the UNC, that's particularly egregious in relationship to Asians. 0.83

00:08:40.600 And how big are the advantages, and what do you think they signify?

00:08:45.700 Oh, I think the advantages are enormous.

00:08:48.300 You know, I went in a lot more optimistic about holistic admissions than I came out.

00:08:54.060 You know, to me, it really looks like we give very large preferences across the board.

00:08:58.140 It's not just on race, but also on legacies, whether you're going to be on the sailing team, you know.

00:09:05.280 Right, right.

00:09:05.840 Legacy, athletics, and race.

00:09:07.960 Anything else?

00:09:09.420 Oh, so they also have children of donors, and children of faculty and staff.

00:09:14.740 The last group isn't that big, but the children of donors, it seems sort of odd, because they're supposed to have, like, need-blind admissions.

00:09:22.600 But then there's a special list where we have children of potential donors.

00:09:29.900 Okay, so let me dive into that momentarily.

00:09:32.700 So, I spent a lot of time delving into the selection literature.

00:09:38.580 And by a lot of time, I mean, like, 15 years.

00:09:41.140 I mean, a lot of time.

00:09:42.780 We developed selection tests for corporations and for academic institutions.

00:09:47.500 And so, like, I knew the literature, and one of my students did an excellent PhD on selection mechanisms in general and developed an entire battery for selecting students.

00:09:57.640 But, and we looked at this initially. 0.95

00:10:00.060 I was an ignorant Canadian. 0.97

00:10:01.520 I didn't know this was a politically charged domain. 0.93

00:10:04.280 I was curious about something very, very practical, which was, well, are there efficient ways of selecting the best candidates for different positions, academic, creative, managerial?

00:10:14.660 And the answer to that is, well, yes, and they're very well documented, and they're relatively objective.

00:10:21.160 And so, I started looking at the data on objective testing, personality, and also cognitive, let's say.

00:10:27.400 There are other, you can measure interest, you can measure creative ability.

00:10:30.880 There are other objective ways of getting at this.

00:10:32.940 And then I looked at the history of objective testing, and I learned that it was the socialists that brought IQ tests to the UK, and then they spread from there into the rest of the Western world.

00:10:46.180 And the reason for that, and the Army did this as well, the reason for that was that the hypothesis was that if you used objective tests,

00:10:54.460 that you could identify people of disadvantaged economic background who had the ability to succeed academically and professionally,

00:11:03.960 and that would be good for them and fair, but it would also be good for everyone else,

00:11:07.460 because why the hell not draw on the full talent pool if you're trying to move people along the educational ladder, up the educational ladder?

00:11:18.120 And so, my conclusion from all this, and this was before all this became politicized,

00:11:22.400 was that there was absolutely no better way of serving disadvantaged communities than to stick 100% to objective tests,

00:11:30.460 not because they didn't produce differential outcomes, because they still do,

00:11:34.700 but because any other system you could possibly produce would produce much worse outcomes.

00:11:40.720 And then I read Adrian Woldridge as well, you know, and Adrian, who did it, he was an economist journalist for years,

00:11:46.700 a very, very careful historical researcher, he showed very clearly, as far as I'm concerned,

00:11:53.440 that the alternatives, the historical alternatives to objective testing have been dynasty and nepotism,

00:12:02.480 not equality of outcomes.

00:12:04.340 So, the thing about objective tests to me is, well, first of all, they actually predict they're reliable and valid,

00:12:10.940 they're better than any other method by a large margin,

00:12:13.200 and, and this is the crucial and, there's no way that you can do better than that.

00:12:18.480 Everything else you do will be worse.

00:12:20.120 Okay, so now we have this mishmash at universities that you just pointed out.

00:12:24.520 If you're a child of a faculty member, you get preferential access.

00:12:28.020 That's usually a hiring perk for the faculty members, and a major one.

00:12:31.980 And it keeps them at private schools, because a state school can't do that, but the private elite schools can do that.

00:12:40.160 Right, right. So, you can see what the rationale is there.

00:12:42.720 And then children of donors, well, you can also understand the rationale there,

00:12:47.340 but basically what that means is that rich people can buy preferential access.

00:12:50.860 Now, you could argue that that's acceptable if one of the things that the rich people are doing

00:12:55.780 is donating like $100 million to establish an entirely new research complex.

00:12:59.980 You could say, well, perhaps that's a price worth paying.

00:13:02.620 It's not fair at the admissions level.

00:13:06.240 And then, well, the next one is athletics.

00:13:09.340 And, you know, I worked for the U.S. Naval Academy for a while.

00:13:12.540 That's a whole story in and of itself. 0.99

00:13:14.420 And I found there, and this was quite shocking to me as, again, as an ignorant Canadian, 0.95

00:13:19.000 that they had a walloping preference for athletic ability at the Naval Academy. 0.97

00:13:25.100 And that struck me as, you know, pretty damn counterproductive,

00:13:28.580 given what they were training these people to do.

00:13:31.180 And then you also mentioned racial preference.

00:13:34.160 And so, there's a variety of ways that these admission systems get gerrymandered.

00:13:39.500 Do you see a solution that's even possible, you know, holistic? 0.99

00:13:43.760 I think that's just complete bloody rubbish. 0.98

00:13:45.580 That's just hidden prejudice and discrimination. 0.99

00:13:47.440 Do you see an alternative to the catastrophe of objective evaluation?

00:13:53.200 Because it's also a catastrophe.

00:13:55.040 Well, there's a clear alternative, but it won't be pursued.

00:13:58.500 You know, basically, every other system has test-based admissions.

00:14:03.640 And I think the key to pushing on for more for test-based admissions is exactly what you said.

00:14:09.800 We think about the tests as favoring the rich.

00:14:12.580 But the other stuff favors the rich even more.

00:14:16.140 Well, right.

00:14:16.580 That's exactly.

00:14:17.440 That's it.

00:14:17.720 Yeah.

00:14:18.220 You see, athletics.

00:14:19.960 You know, Harvard actually offers more varsity sports than any school in the country.

00:14:25.100 You know, when we think about sports as being an equalizer, you're thinking about football and basketball.

00:14:30.100 When we're talking about sailing, that's an entirely different matter.

00:14:33.360 So if you get rid of athletic preferences, whites go down, African Americans stay the same, and Asian Americans and Hispanics go up. 0.58

00:14:45.140 I'm sure that if you took football and basketball out of that, it'd be only whites who are going down. 0.99

00:14:49.940 And that fits in, you know, because Harvard also has an athletic rating as part of their admissions criteria beyond just the athletic preference.

00:15:00.680 So everybody gets an athletic rating.

00:15:02.820 And the people who do the best on Harvard's athletic rating are white legacies, then legacies of other races, then white non-legacies. 0.86

00:15:13.240 You know, how does that work?

00:15:15.420 One, it's the sports that Harvard offers, like sailing and such.

00:15:19.660 But sports actually favors people who go to small private schools.

00:15:24.300 So I went to suburban public school.

00:15:28.140 No way could I make the soccer team.

00:15:30.680 My kids, they made the soccer team at their school because they don't cut anybody.

00:15:35.920 You know, those types of things are all work to favor people of means.

00:15:40.520 Yeah, well, this is the thing.

00:15:42.140 That's why I want to reiterate that for people who are watching and listening.

00:15:46.180 Like, if you use an objective selection system that's based on merit, so let's define merit first.

00:15:52.260 You tell me if you think I've got this.

00:15:54.640 Going online without ExpressVPN is like not paying attention to the safety demonstration on a flight.

00:16:00.260 Most of the time, you'll probably be fine.

00:16:02.240 But what if one day that weird yellow mask drops down from overhead and you have no idea what to do?

00:16:07.940 In our hyper-connected world, your digital privacy isn't just a luxury.

00:16:11.740 It's a fundamental right.

00:16:13.060 Every time you connect to an unsecured network in a cafe, hotel, or airport,

00:16:17.320 you're essentially broadcasting your personal information to anyone with a technical know-how to intercept it.

00:16:22.260 And let's be clear, it doesn't take a genius hacker to do this.

00:16:25.260 With some off-the-shelf hardware, even a tech-savvy teenager could potentially access your passwords, bank logins, and credit card details.

00:16:33.020 Now, you might think, what's the big deal?

00:16:35.080 Who'd want my data anyway?

00:16:36.620 Well, on the dark web, your personal information could fetch up to $1,000.

00:16:41.220 That's right, there's a whole underground economy built on stolen identities.

00:16:44.860 Enter ExpressVPN.

00:16:47.060 It's like a digital fortress, creating an encrypted tunnel between your device and the internet.

00:16:51.760 Their encryption is so robust that it would take a hacker with a supercomputer over a billion years to crack it.

00:16:57.380 But don't let its power fool you.

00:16:59.080 ExpressVPN is incredibly user-friendly.

00:17:01.540 With just one click, you're protected across all your devices.

00:17:04.580 Phones, laptops, tablets, you name it.

00:17:06.740 That's why I use ExpressVPN whenever I'm traveling or working from a coffee shop.

00:17:10.760 It gives me peace of mind knowing that my research, communications, and personal data are shielded from prying eyes.

00:17:16.880 Secure your online data today by visiting expressvpn.com slash jordan.

00:17:21.620 That's E-X-P-R-E-S-S-V-P-N dot com slash jordan, and you can get an extra three months free.

00:17:27.980 ExpressVPN dot com slash jordan.

00:17:32.320 Wrong.

00:17:33.220 So here's the definition of merit.

00:17:35.820 All right, so in principle, the enterprise you're pursuing has an outcome.

00:17:40.760 So at university, the outcome would be grades and graduation.

00:17:44.940 If you go to work, the outcome would be job performance, and that can be evaluated a variety of ways.

00:17:50.120 If you're an entrepreneur, it would be success at business.

00:17:52.840 And so that's the outcome.

00:17:54.600 Now, merit in relationship to that outcome would be the documented relationship between a trait that you might have and that outcome.

00:18:02.740 And so one of the things you see on the academic front is that one of the biggest predictors of academic success is general cognitive ability.

00:18:12.020 And the reason for that is, in part, because there's actually no difference between general cognitive ability and academic success.

00:18:19.220 Right?

00:18:19.520 They're the same thing.

00:18:20.280 Now, it turns out you can measure general cognitive ability with incredible rapidity, very accurately.

00:18:26.820 And they are the most accurate tests ever designed by social scientists.

00:18:32.300 They're much more powerful than almost every medical test that we use.

00:18:35.900 And they're very predictive, not only of academic performance, as evidenced by grades, but also then long-term life performance.

00:18:42.680 And they're even more predictive of speed of learning.

00:18:44.840 And again, that's because general cognitive ability and speed of learning are the same thing.

00:18:49.620 So, you know, there is a debate about our society whether or not merit is in itself a, let's say, a racist and prejudicial construction. 0.92

00:18:57.880 But that's an idiot presumption because that's the same statement as what any enterprise is for is not relevant. 0.95

00:19:08.260 And that's completely preposterous because an enterprise exists because it's for whatever it does. 0.99

00:19:13.620 That's its justification.

00:19:15.280 So, you can't get anywhere by claiming there's no such thing as merit.

00:19:18.300 And you actually can't get anywhere by claiming we can't measure merit because if you accept that, that means we'd have to throw out all of medicine and the social sciences because we can actually measure merit better than we can measure anything else.

00:19:30.120 And then you add to that, if you don't measure merit, whatever you measure is going to disadvantage the disadvantaged people even more.

00:19:38.840 And that's the crucial issue as far as I'm concerned because it will revert to these invisible forms of prejudice that are always associated with nepotism and dynasty.

00:19:47.900 So, do you think I have any of that wrong?

00:19:50.060 So, I'm in total agreement with you.

00:19:53.380 I think the point of contention that the universities would say is that their objective function is different.

00:19:59.440 They want to create the future leaders of society.

00:20:02.240 And maybe that connects all with the cognitive ability.

00:20:03.980 But they're so vague about what the objective function is that actually pinning them down on that and being able to actually measure is tough.

00:20:14.380 I also know the leadership literature.

00:20:17.380 Okay. 0.99

00:20:18.320 Fundamentally, it's rubbish. 0.95

00:20:20.400 And the reason for that is that, first of all, there isn't any such thing as leadership. 0.97

00:20:26.020 Like, it's not a unitary phenomenon.

00:20:27.720 And it's partly because there are different ways of leading people.

00:20:33.200 And different circumstances call for different, say, styles of leadership or different abilities.

00:20:39.360 Now, having said all that, you can say that, in general, people who are highly intelligent, who are conscientious, which means they'll do what they say they'll do and they stick to the task, and who are somewhat extroverted, tend to tilt more towards being approved of leaders because, while they can figure out how to do things, they'll actually do them and they can communicate about them enthusiastically.

00:21:04.680 But if you take that as a base definition of leadership, you're still going to find out that general cognitive ability and conscientiousness predict that.

00:21:14.900 And that is also the same two sets of traits that predict success at universities.

00:21:19.460 So all that gerrymandering by the universities, that hand-waving about the notion that they can't measure their outcomes, is that, like, that either means they don't know anything about how to measure,

00:21:30.000 or that they're just obscuring the situation to their own, for their own purposes, whatever they might be.

00:21:38.420 Well, and that's the thing that drives me most crazy about universities, is their unwillingness to use their data.

00:21:44.380 You know, if you take COVID as a perfect example of this, Notre Dame had a very different COVID policy than the Ivy League schools.

00:21:52.900 We should know how that worked out.

00:21:55.980 We should know how the mental health rates were different across those things.

00:22:00.540 And then, so we know how to do it better in the future.

00:22:03.660 We don't do that.

00:22:06.040 And, you know, the other thing that's just, I find hilarious, is that a lot of universities have randomized roommates.

00:22:13.360 That's great.

00:22:14.720 For the purposes of studying, what the effect of different roommate characteristics have on you,

00:22:20.740 and what the match component might look like.

00:22:24.020 But they don't do that.

00:22:24.820 They just keep having the randomized roommates.

00:22:26.600 They don't actually look to see what pairs might actually work,

00:22:29.800 and how can we construct policies that might actually help their students.

00:22:33.700 Well, you know, it's really, it's really, it's really sad that universities don't do that,

00:22:38.080 because universities spearhead the social sciences research enterprise.

00:22:43.840 And the fact that they don't capitalize on their own expertise either indicates that the expertise isn't there,

00:22:50.180 or that it's untrustworthy, or that there are other reasons they don't bother,

00:22:55.740 which is either, you know, incompetence or some hidden agenda.

00:22:59.760 And none of that is excusable.

00:23:02.060 Oh, it's crazy.

00:23:02.880 Now, so let's go back to the court case, if you don't mind.

00:23:07.500 So what magnitude of advantage are we talking about with regards to these different categories?

00:23:14.480 We have athletics, we have legacy students, we have racial preference.

00:23:20.180 That was the main three categories, apart from children and faculties.

00:23:23.600 What kind of differential advantage are you looking at, say, on the racial front?

00:23:30.300 Well, so by far the biggest is the athletic preference.

00:23:33.800 Oh, that's the biggest one.

00:23:35.440 Oh, by far.

00:23:36.460 Yeah.

00:23:36.960 And, you know, part of that you could think is they negotiated ahead of time that they're getting in,

00:23:42.500 so maybe you don't see the full applicant pool for them.

00:23:45.980 But the characteristics of admitted athletes, so, I mean, the athletic admit rate is, you know,

00:23:51.680 over 85%, and the average athletic admit has academic characteristics that are much worse

00:23:58.820 than the average applicant to Harvard, and the average applicant to Harvard has a very low chance

00:24:05.680 of getting in.

00:24:06.540 So, you know, we're really talking about massive preferences there.

00:24:10.720 So, could you outline the advantages and disadvantages to the athletic preference?

00:24:17.160 Because you could say, well, the kids who've been good athletes,

00:24:20.780 they are stellar at something, so they have a demonstrated track record of, so to speak,

00:24:25.040 track record of accomplishment, and I think there's something to that.

00:24:28.140 I think you could infer, although it would be nice to prove, that having developed the ability

00:24:33.320 to be a disciplined specialist in a sport might make you a better team player and potentially

00:24:39.020 a better leader.

00:24:39.740 However, I don't know of any data pertaining to that.

00:24:44.000 Um, you might say as well that for a school like Harvard that faces an embarrassment of riches

00:24:50.200 on the applicant front, that using additional criteria of achievement is a useful screening mechanism.

00:24:58.140 Um, why does this athletic preference exist?

00:25:02.020 Now, you said it favors the rich, too, and that's a very interesting thing to look at.

00:25:05.360 But why does the athletic selection advantage exist, and what do you think the pros and cons are?

00:25:11.640 Well, I think it's a backdoor way, in Harvard's case, of getting people from rich families.

00:25:16.300 And I say that primarily because, you know, uh, over 16% of white, uh, admits are recruited athletes.

00:25:28.560 That's actually way bigger than what you see for African-Americans, Hispanics, or Asian-Americans.

00:25:35.200 And that's because, you know, they're choosing sports like sailing, skiing, fencing,

00:25:42.580 all these things that are associated more on the upper end.

00:25:47.100 But I think you end up with the way, you know, more university governance, where you end up with

00:25:52.100 these handouts.

00:25:52.760 Why, why isn't it the case?

00:25:54.300 Why would you have an athletic rating and not a music rating?

00:25:57.680 You know, it's, it's a very odd thing to me that we have this tie.

00:26:03.660 So, so at places like Harvard, they do, they do look at additional forms of attainment.

00:26:09.520 And you'd know more about this than me, but I, I, I used to know the Dean of Admissions

00:26:13.620 at Harvard and we talked a lot and that he was a very interesting character.

00:26:16.780 And, you know, they do, they basically pick kids who were at that point, that was back

00:26:21.100 in the nineties.

00:26:21.640 They pick kids who generally were stellar academically.

00:26:24.680 A lot of them were valedictorians or, or the next best thing.

00:26:27.980 Then generally they had to have at least one other relatively stellar talent.

00:26:33.840 And that was sometimes athletics, but it was, it could be, it could be proficiency in any

00:26:38.820 number of other domains.

00:26:39.800 But you're saying that the athletic preference far outweighs any of the other criteria that

00:26:45.060 are applied.

00:26:46.220 That's right.

00:26:46.600 Because they don't even have like a field in the data for these other things.

00:26:51.560 You know, here we have athletic, you know, we know you get a score for that.

00:26:56.400 The other things are going to matter.

00:26:58.160 It might show up through your extracurricular rating or things like that.

00:27:01.960 But you can see in the court record, the conversations that happen between athletic coaches and the

00:27:10.140 admissions in a way that I don't think happens in other, other areas.

00:27:14.900 Right.

00:27:15.180 So, so athletics is considered separately from extracurricular and everything else is lumpy

00:27:20.420 into that.

00:27:21.540 That's right.

00:27:21.980 Right, right.

00:27:22.880 Yeah.

00:27:23.100 Well, you see, that also seems to be very careless, you know, because you'd think if you were going

00:27:26.820 to set up an equation to admit that you do some work in delineating what other stellar

00:27:34.060 performance features are actually associated with academic success and, and later, even

00:27:40.060 for your own financial interest, because, I mean, one of the things the, the Ivy Leagues

00:27:43.680 are trying to do is to pull people in who will become rich enough to become alumni donors.

00:27:48.180 And they're actually pretty good at that.

00:27:50.080 And, and they have the reasons for it.

00:27:51.900 But you'd think that that would be of sufficient economic interest to actually try to, to actually

00:27:56.780 try to model.

00:27:58.340 But I think even those athletic preferences, why they had that was not so pure a motive

00:28:04.160 as, as that, but more relating to Jewish discrimination.

00:28:09.020 You know, I think that's part of why we have the holistic missions in the first place.

00:28:13.080 Okay.

00:28:13.320 So that's another, so you think not only the athletics admissions, not only privileged the

00:28:17.340 rich, but they're also a way of tilting the scales against Jewish applicants.

00:28:21.940 I think that's why they might have, that's right.

00:28:24.400 And so then we have a, oh yeah, that sort of carries forward to today, you know.

00:28:28.960 Right.

00:28:29.200 And that, but it would be the Asians who are more in that position now, or is it still Asians 0.63

00:28:32.920 and Jews?

00:28:34.300 The thing is, we wouldn't know about Jews today.

00:28:36.900 And part of that is Harvard actually doesn't download that information.

00:28:40.660 So on the common application, it actually asks your religion.

00:28:44.680 Harvard doesn't download that information in part because of the history of Jewish discrimination.

00:28:49.760 And I think that's one of the remedies in these cases is don't download the racial information.

00:28:56.800 That's not going to totally get rid of discrimination.

00:28:58.960 They may still discriminate against the Jews based on their last name, you know, those kinds

00:29:03.120 of things.

00:29:03.620 But it's not as easy as being able to point to the Jewish box on the application.

00:29:10.380 Now, on the athletic front again, do you know if, so look, one of the problems at a place

00:29:18.920 like Harvard is that if you let students in who aren't academically prepared, one of two

00:29:24.700 things inevitably happens.

00:29:27.360 Three things.

00:29:28.420 One is, it's not that much fun for the student, right?

00:29:31.980 I mean, I saw when I was at Harvard, the consequences of being less intellectually gifted than your

00:29:39.480 peers, right?

00:29:40.820 And that's not fun.

00:29:43.460 And especially many of the kids came from places where they were pretty stellar in their

00:29:48.740 local environment.

00:29:49.860 And then they'd come to a place like Harvard, which is hyper-selected, and they wouldn't be

00:29:54.100 so stellar.

00:29:54.660 And that was salutary in some ways because it kept ego down, but it was devastating in

00:30:00.480 other ways.

00:30:01.000 And it's no fun to be selected to a place like Harvard and then to fail.

00:30:04.740 And if you're not selected on academic grounds, and that would be for general cognitive ability,

00:30:09.920 and you come to a place that's full of people who are hyper-sophisticated cognitively, that's

00:30:14.980 going to be a pretty damn rough go.

00:30:16.380 And the probability that you're going to fail is quite high and be demoralized, and maybe 0.99

00:30:21.120 even end up concluding that you're stupider than you actually are. 0.99

00:30:24.400 Because it's such an artificial competition at a hyper-selected school. 0.99

00:30:28.520 And then the other remedy, of course, is that if you admit people whose general cognitive

00:30:32.540 ability isn't up to scratch, then you're going to decrease the academic requirements.

00:30:37.900 Because otherwise, everyone who's admitted who, on these somewhat fallacious grounds,

00:30:43.820 let's say, everyone's going to fail, and then it's going to look like your institution

00:30:47.340 is prejudiced.

00:30:49.200 That's right.

00:30:49.640 Do the athletic, do the athletes, how do they do, how do the athletes do in terms of

00:30:54.760 dropout and school success?

00:30:57.300 So, unfortunately, the lawsuit didn't give us outcome data.

00:31:01.740 But I will say, I don't think dropout's an issue.

00:31:05.600 I think Harvard always figures out a way to graduate you in ways that might not be true

00:31:10.660 of other universities.

00:31:11.920 What they graduate you in, you know, you're not going to be getting a computer science

00:31:18.820 degree coming in that way.

00:31:20.680 Right, right, right.

00:31:22.340 And that's what's really interesting.

00:31:24.260 You'd be slotted into one of the disciplines that require a somewhat lesser degree of sheer

00:31:30.620 intellectual horsepower.

00:31:32.280 That's right.

00:31:33.500 That doesn't build off your previous academic background.

00:31:36.400 And that's the beauty of this, right?

00:31:37.560 Because I actually think what's happened over time is that the returns to your college major

00:31:42.780 matter more and more.

00:31:44.680 And so now you have fields where there's very little demand for the subject.

00:31:50.900 Having the athletic preferences, those other preferences for the people who would really

00:31:55.560 like to be an econ major, but it's a ton of work given their background, then they end

00:32:03.560 up switching into this other field.

00:32:05.900 Whereas that same person might have been an econ major at a less prestigious school.

00:32:12.760 They could have cut it there.

00:32:14.440 Right, right, right.

00:32:15.460 Well, that's the other thing I think I observed both at Harvard and at the University of Toronto.

00:32:20.760 Like, if you have a child who could be a star at a state school, but was a third-tier performer

00:32:28.080 at an Ivy League, you're probably better off sending them to the state school.

00:32:32.480 That was my sense.

00:32:34.480 Because compared to the general population, they might still be stellar performers.

00:32:40.420 But if you put them in with people who are hyper-selected, especially when they're young,

00:32:44.240 they're going to draw the erroneous conclusion that they're not particularly talented.

00:32:48.220 Now, compared to, you know, people who are one in 10,000, the person who's one in 100

00:32:53.420 isn't particularly stellar.

00:32:55.400 But compared to the other 100, they're more than perfectly capable.

00:32:59.420 So I would say to parents, and I don't know what you think about this, I would say to parents,

00:33:04.080 you know, you should put your kid in a place where they're going to be challenged, but where

00:33:09.060 they're not at the bottom of the pool.

00:33:11.520 That's right.

00:33:12.240 And I think that that's more relevant in some majors than others.

00:33:15.500 Starting a business can be tough, but thanks to Shopify, running your online storefront

00:33:21.100 is easier than ever.

00:33:22.800 Shopify is the global commerce platform that helps you sell at every stage of your business.

00:33:27.040 From the launch your online shop stage, all the way to the did we just hit a million

00:33:30.760 orders stage, Shopify is here to help you grow.

00:33:34.160 Our marketing team uses Shopify every day to sell our merchandise, and we love how easy

00:33:38.320 it is to add more items, ship products, and track conversions.

00:33:41.400 With Shopify, customize your online store to your style with flexible templates and powerful

00:33:46.700 tools, alongside an endless list of integrations and third-party apps like on-demand printing,

00:33:51.940 accounting, and chatbots.

00:33:53.640 Shopify helps you turn browsers into buyers with the internet's best converting checkout,

00:33:57.960 up to 36% better compared to other leading e-commerce platforms.

00:34:02.000 No matter how big you want to grow, Shopify gives you everything you need to take control

00:34:05.860 and take your business to the next level.

00:34:07.780 Sign up for a $1 per month trial period at shopify.com slash jbp, all lowercase.

00:34:14.380 Go to shopify.com slash jbp now to grow your business, no matter what stage you're in.

00:34:19.620 That's shopify.com slash jbp.

00:34:24.600 So the bottom of the pool, I think, is sort of all relative to the sorting that happens

00:34:29.900 within college, and there's massive sorting.

00:34:32.380 Tons of people come in wanting to major in the sciences and such, and then they switch

00:34:37.740 out, and it's very predictable who switches out.

00:34:42.640 Those who are relatively lower on the math side tend to switch.

00:34:46.260 Yeah, yeah, yeah.

00:34:47.360 Well, that, okay, so from what I've been able to derive, that's particularly true.

00:34:51.340 It's like, I think the discipline where general cognitive ability horsepower is most necessary

00:34:59.340 is physics and mathematics.

00:35:02.240 That makes sense.

00:35:03.600 And then there's a hierarchy down from that.

00:35:06.700 And so, now I guess you could also make the case is that, well, it's not so bad that in

00:35:11.560 universities there's a range of disciplines to match different levels of general cognitive

00:35:16.820 ability.

00:35:17.340 But I would still say that at elite schools that have all those resources, that their

00:35:26.420 resources are best funneled towards those who are most able to benefit from them.

00:35:34.360 And that's clearly the people who have higher general cognitive ability.

00:35:38.720 And sort of related to that, what also bothers me is they're not honest with their students.

00:35:43.580 You know, so if you told somebody up front, yes, we're admitting you, given your initial

00:35:51.200 major interest in your test scores and grades, here's the probability you're going to complete

00:35:57.340 this degree.

00:35:58.100 Here's the probability you're going to complete some other degree.

00:36:00.120 Here's the probability you drop out.

00:36:01.620 Now, at least you've given them full information.

00:36:04.560 Yeah, yeah, yeah.

00:36:05.400 To me, it takes away a lot of the mismatch argument because, you know, individuals can make

00:36:10.080 up their own mind at that point and say, well, I'd rather go to Harvard and graduate in some

00:36:16.940 non-science field than go to UNC and graduate in physics, if that's sort of the way that

00:36:24.560 the trade-off works.

00:36:26.320 Yeah, well, you could argue that, you know, there's going to be kids that you don't think

00:36:31.740 will do very well on the basis of your testing who will go there and actually do quite well.

00:36:35.800 And so it would be okay not to inform them because that way those exceptions can flourish.

00:36:41.900 But I've thought this through, and this is my conclusion, and you can tell me what you

00:36:45.160 think about this, which is that, yeah, but for every kid like that, you're going to doom

00:36:50.400 like eight kids to failure.

00:36:53.380 And it seems like an inefficient use of that kid's time and the school's resources to have

00:37:01.120 a failure rate that high to have the odd exception.

00:37:05.520 I mean, and I see this if you're recruiting people for a management position, too.

00:37:09.620 You might say, well, why not give this person a chance?

00:37:12.840 You know, maybe they'll succeed.

00:37:13.940 And the answer is, well, yeah, but probably they won't.

00:37:18.640 And that's going to be really hard on them.

00:37:20.900 And not only that, if they turn out not to be able to do the job, it's going to be really

00:37:25.240 hard on everyone that they're supervising and working with.

00:37:28.000 And setting someone up for failure from compassion is not advisable as a management strategy.

00:37:36.200 It's a bad idea.

00:37:37.200 It's not advisable as an admission strategy as well.

00:37:40.760 I totally agree.

00:37:42.040 But I also like being able to tell people, look, it doesn't look like you can cut it,

00:37:47.420 because some of the people then can respond by working hard.

00:37:52.360 And for myself, I was a guy who just floated through, thinking I had it all together.

00:37:58.660 Then I got to graduate school and learned humility and learned I had to work a lot harder.

00:38:05.360 But having somebody tell me, you know, you're not good enough, for myself, that was actually

00:38:11.680 a push.

00:38:14.060 You know, I wasn't internally motivated enough.

00:38:16.240 I needed somebody to prove wrong in order to do well.

00:38:21.260 Now, I know I'm sort of weird in that way.

00:38:24.040 Well, yeah, but not absurdly weird in that way.

00:38:29.160 I mean, you see lots of smart kids who have floated by, who hit a wall where they're now

00:38:36.760 competing with kids who are just as smart, but like 10 times as disciplined.

00:38:40.480 And some of them fail and they're bitter as a consequence, but some of them like pull

00:38:44.960 up their socks and think, oh my God, looks like I'm actually going to have to work.

00:38:48.920 And there's nothing about that that isn't good for them, right?

00:38:52.020 Because then not only, then they're more like Asians, right? 1.00

00:38:55.300 Because the Asian advantage, I looked into the Asian advantage a lot.

00:38:59.760 And the Asian advantage, by the way, the Asian performance advantage in the US disappears

00:39:05.460 by the third generation.

00:39:07.300 So it turns out the more that Asian kids are like American kids, the less they're like 0.98

00:39:12.720 Asians, so to speak.

00:39:14.040 And Asians have this hyper-focus, like Jews, I would say, they have this hyper-focus on academic 1.00

00:39:21.100 success and academic success as a status marker at home and in the community.

00:39:27.500 And, you know, that's actually, I think, something that in principle might be remediable because

00:39:33.360 it implies that if you can teach people at any given cognitive level to work harder, that

00:39:41.320 is one pathway to genuine success.

00:39:43.500 But it's also appalling that the Asians get discriminated against because, you know, the

00:39:47.800 ones who are being discriminated against, they're hard workers.

00:39:51.920 And that's not a good thing to discriminate against.

00:39:54.380 Being an American, I wanted my kid to do sports.

00:39:59.080 You know, that was just part of the thing.

00:40:00.640 And he was uninterested in sports.

00:40:02.740 And he kept telling me that.

00:40:04.000 And he kept putting him out there in soccer, basketball, whatever.

00:40:08.200 And he kept saying, I want to do karate.

00:40:10.860 Eventually, he did karate.

00:40:11.820 It was fine.

00:40:13.300 The kid was just different in that regard.

00:40:15.280 If he'd been in a society where the norm was to do something else, to do science, I think,

00:40:22.000 you know, that would have been very nice for him socially and so on.

00:40:26.200 My understanding is that in a place like Hungary, they're much more centered on math.

00:40:30.520 In America, we're much more centered on sports.

00:40:33.580 That sports craze, I think, is actually, you know, it does build some good skills.

00:40:38.620 But I think we're too obsessed with it to the detriment of these other things.

00:40:46.400 And I think Asian American families have sort of figured out not to be obsessed about that,

00:40:51.340 but to be obsessed about other things.

00:40:54.640 Right, which is another reason why it would be nice to see the equations adjusted so that

00:40:59.540 other forms of excellence, extracurricular excellence, are given their due weight, right?

00:41:05.900 And so, why were you called upon to testify specifically?

00:41:11.260 So, I've written a lot of papers on affirmative action and with mixed results.

00:41:16.960 So, I think that there's actually a lot of pressure to say good things about affirmative

00:41:20.340 action.

00:41:21.260 And I didn't always say good things.

00:41:24.020 And in fact, in 2011, there was a protest over one of my papers.

00:41:30.920 It was sort of, I thought, a very innocuous paper.

00:41:34.360 However, we actually used Duke data and we're looking at persistence in science and economics.

00:41:42.680 And what you could see is that, you know, African American students came in wanting to major

00:41:48.740 in those subjects at the same rates as white students, but they were leaving at a much higher

00:41:53.540 rate.

00:41:54.100 So, like over 50% of black males who started in the sciences and economics switched out versus

00:42:01.280 8% of white males.

00:42:04.320 Right.

00:42:04.940 Well, then the, well, the advocates of systemic racism as an explanatory hypothesis would say

00:42:13.000 that the racism is so deep that not only does it discriminate on the admission side, but

00:42:19.440 it also discriminates on the performance side.

00:42:22.000 And so-

00:42:23.000 And all those things disappeared as soon as you control for academic backgrounds.

00:42:27.840 You know, you control for the test scores, you know, other Duke ratings and such.

00:42:33.440 You start off with these big racial gaps where Asian Americans were most likely to persist in the

00:42:38.560 sciences and African Americans were least likely.

00:42:41.060 Once you conditioned on differences in academic backgrounds, driven both by affirmative action and

00:42:45.980 what, you know, the educational experiences prior to college, that just disappears.

00:42:52.300 And I think that's what they didn't like.

00:42:54.800 Did you control for grades or for SAT scores or for both?

00:42:59.160 Do you remember?

00:43:00.160 I controlled for a whole different specification said different controls, but it goes away fairly

00:43:05.700 quickly.

00:43:06.160 Um, and you could look at, uh, and suggest performance in first year classes and look

00:43:12.740 at the, and that, that wipes it, wipes it out.

00:43:16.040 Right.

00:43:16.060 So, so, so you, so, so for, for everybody who's watching and listening and tell me if I get

00:43:21.020 this wrong.

00:43:21.680 So you could imagine that if you're doing a statistical analysis, you could determine whether

00:43:27.220 race was the predictor or academic ability was the predictor by modeling both of them and

00:43:34.080 seeing which predicted dropout better.

00:43:37.540 And if it's, if, if academic ability destroys the ability of race to predict dropout, but

00:43:44.080 not the reverse, then you know, it's academic ability and not race.

00:43:47.380 And in principle, unless you assume that the academic ability markers are also contaminated

00:43:52.960 with what would you call it systemic racism, then you eradicate the racist argument.

00:43:59.280 So, so what the, you know, what the radicals do is they just say everything systemic racism

00:44:03.900 and it's virtually impossible to mount an argument against that.

00:44:07.200 But then we have the other problem, which for all of you who are compassionate, who are

00:44:11.480 listening, you know, it's not compassionate to put people into situations where they

00:44:16.240 disproportionately fail.

00:44:19.380 Right.

00:44:20.080 It allows you to look good on the admission front, but it's not good for the people who

00:44:24.940 are involved.

00:44:25.920 As far as I can tell, do you think there's anything about it that's good for the people

00:44:29.200 that are involved?

00:44:30.740 I think it's tough.

00:44:31.840 You know, we could, we could actually say more because they asked the reason why you

00:44:36.140 switched your major.

00:44:37.960 And one of them was the difficulty of the courses or feeling not as prepared for those

00:44:42.520 courses.

00:44:43.480 And again, you could see that black students were much more likely to say it was because

00:44:48.620 of course difficulty.

00:44:50.520 And once you controlled for the SAT scores and such, that all went away.

00:44:56.020 So it really pointed towards these academic measures really mattering for your experience

00:45:03.740 in those classes.

00:45:05.820 Now, on that systemic racism front saying, well, look, the test scores are biased.

00:45:11.060 To me, I think that does such a disservice because then it sort of says what's happening

00:45:16.040 to prior to college, like in the K through 12 education, it's really not that bad because

00:45:22.340 the test scores are just misrepresenting it, you know, when reality, the fact that, you

00:45:29.300 know, 1% of black students score above 1390 in the SAT, that's 8% for whites and 24% for

00:45:41.300 Asian Americans.

00:45:43.280 That's reflecting something that we could fix.

00:45:46.120 That's what, that's where we need to spend our time on is fixing that.

00:45:49.720 Okay, so, so let me, let me make a case for them, for the universities and what they're

00:45:55.860 doing.

00:45:57.080 So what would rapidly happen if we went to a purely objective evaluation system is that

00:46:05.020 at elite level universities, there would be a disproportionate compared to the population,

00:46:11.800 there would be a radically disproportionate number of Asians and Jews.

00:46:15.100 And a radically disproportionate dearth of black Americans.

00:46:22.900 That would happen very rapidly.

00:46:25.200 And the universities are concerned about that.

00:46:28.860 And now, and, and it's, it's, it's definitely a very difficult nut to crack, right?

00:46:35.920 And so now it doesn't seem like the appropriate solution to that is to disadvantage

00:46:42.100 extremely competent Asians.

00:46:45.560 And I would say partly because, well, they deserve their shot at the target.

00:46:52.060 But also, you know, there, there aren't that many hyper exceptional people on the cognitive

00:46:59.360 front, right?

00:47:00.180 And if we're greedy as a society and sensible in that greed, we would say we should set up

00:47:06.380 our institutions to capitalize on every available bit of brain power and persistence.

00:47:12.520 So it's better socially if we can put the smartest people where they have the greatest

00:47:17.600 opportunities to learn and contribute.

00:47:19.780 So we're not just hurting the Asians, let's say, and the Jews.

00:47:22.760 We're also depriving ourselves in principle of what they could offer.

00:47:27.140 But then we're going to have the problem of, of this disproportionate racial and ethnic mix 1.00

00:47:35.620 in the universities.

00:47:36.660 And so what, and maybe this isn't a fair question to ask you, because it isn't necessary that

00:47:42.960 you have the solution to this.

00:47:44.700 It isn't obvious to me that anyone has, but you were testifying on behalf of, you were making

00:47:50.600 a case that the affirmative action systems as presently constituted are very badly flawed

00:47:56.680 and probably illegally flawed.

00:47:58.660 But what do you see forward, if anything, as a way out of this conundrum?

00:48:06.480 So, you know, really, I view it that I wasn't actually saying it was flawed, so I'm just

00:48:11.280 saying how big the preferences are.

00:48:13.580 I hate actually being characterized as an opponent of affirmative action, because then it makes

00:48:18.220 it seem like the research is biased.

00:48:20.220 I'm just trying to show you what the data show.

00:48:22.980 So to me, I think it allows, it works, affirmative action works as a Band-Aid, where it covers 0.69

00:48:28.460 up all the inequities that are happening prior to college, and then we don't focus on those

00:48:34.420 things.

00:48:36.100 Roland Fryer, I think, has done a lot of work on the no-excuse charter schools, showing

00:48:40.960 that those are actually pretty effective at closing those achievement gaps.

00:48:45.140 I think the work sort of shows it doesn't translate as well into college, but it's

00:48:49.400 a start.

00:48:51.160 You know, unfortunately, they ran him out of town.

00:48:55.080 But, you know, he was, you know, the most prominent black economist there, came from nothing,

00:49:01.240 and, you know, what I think has happened to him is absolutely horrible.

00:49:06.100 And he's one of the few who actually shows how to close the achievement gap.

00:49:09.320 Um, right, and that was Roland, Roland, Roland Fryer.

00:49:15.060 Oh, right.

00:49:15.820 Yes, yes, yes, yes.

00:49:17.240 I've, I've been, uh, seriously considering him, inviting him onto the podcast.

00:49:22.140 Oh, he is absolutely amazing.

00:49:24.060 His story is amazing.

00:49:25.120 And what's happened to him is, yeah, it's a criminal.

00:49:28.620 Shocking.

00:49:29.340 It's terrible.

00:49:30.060 Yeah.

00:49:30.240 Well, so, so this, this issue here, I think you, you touched on something very crucial,

00:49:34.320 which is that, right, if we pretend that the achievement tests are the problem, that enables

00:49:40.620 us to ignore the underlying problems.

00:49:43.300 Of course, that gets muddy and ugly very quickly, too, because one of the things I would say,

00:49:47.440 perhaps, as I've studied the development of antisocial personality, for example, and other

00:49:51.860 forms of psychopathology that would interfere with educational attainment and lifetime attainment

00:49:58.600 over the long run.

00:49:59.360 And it's certainly the case, for example, that fatherlessness is a contributing factor

00:50:05.160 in a major way, right, unstable, destabilized families.

00:50:08.400 And so, if you don't allow the achievement tests to be the villain, you have to look elsewhere

00:50:13.160 for the villains, and that becomes extraordinarily complex and murky and troublesome.

00:50:18.080 You know, we tried to prevent antisocial behavior, for example, in, in Quebec.

00:50:23.520 And what we learned was that if children are antisocial by the age of four, most of those

00:50:31.720 kids would have been aggressive at the age of two.

00:50:34.140 Most aggressive kids are socialized out of their aggression by the age of four.

00:50:39.020 If they're not socialized out of their aggression by the age of four, it doesn't look like there's

00:50:44.040 a damn thing you can do about it afterwards, or it's extraordinarily difficult, at least.

00:50:48.260 And so, that would mean you have to remediate it at the age of two. 0.96

00:50:52.080 But then, if you start producing government programs, let's say, to remediate antisocial

00:50:58.320 behavior before the age of two, you're in people's households, right?

00:51:03.600 It gets very invasive, right?

00:51:05.720 So, some of these underlying systemic problems, let's say, are extraordinarily difficult to

00:51:12.000 address without falling into a kind of colonial, neocolonial overreach.

00:51:16.720 That's one way of thinking about it.

00:51:18.540 And they're very persistent problems.

00:51:19.500 It's such a shame because, you know, as soon as you would propose something like that,

00:51:23.320 you'd be accused of victim blaming, you know?

00:51:25.880 But the reality is, there's been some great work done in other countries where you go into

00:51:31.860 a place like India, and they're talking to the mother there, and the mother's like,

00:51:38.460 like, so, I should be talking to my child?

00:51:41.600 Like, that wasn't passed on to them, that they should be talking to their child regularly.

00:51:45.880 That's not, I'm not judging you that you didn't get that information passed on.

00:51:49.940 You're saying, how do we fix this problem?

00:51:52.360 Well, I've also been interested in precursors to literacy, you know?

00:51:58.000 And if you look at the data, you find that kids from literate homes are exposed to books

00:52:04.720 and a wider range of vocabulary at a differential rate that's staggering by the time the kid

00:52:11.020 is like two and a half.

00:52:13.440 You know, and it is the case that most poor families, if you interview poor mothers and

00:52:17.740 fathers, and you ask them what they want for their children, and you put educational achievement

00:52:22.720 into the mix, they will indicate in a fully committed manner that they would like their

00:52:28.600 children to be educated.

00:52:29.560 But the problem is, is they don't know what the, say, the nonverbal precursors are.

00:52:34.320 You know, I had friends where I grew up, grew up in this little place and weigh the hell out

00:52:38.260 in the middle of the sticks.

00:52:39.340 There were no books in the house, like zero, right?

00:52:43.380 And that's way different than growing up in an environment where, like, my kids,

00:52:47.220 children at 18 months old, they're already dragging books around behind them and sitting

00:52:53.200 down and pretending to read them, right?

00:52:55.960 They have all that literacy, pre-literate literacy skills are already built in.

00:53:01.340 They value books.

00:53:02.380 They know what they are.

00:53:03.160 They've made friends with them.

00:53:04.400 They'll ask their parents to read them books.

00:53:06.320 And, you know, in a family that's very deprived and very poor without a history of literacy,

00:53:11.540 nobody in the family even knows that that's a possibility.

00:53:14.900 That's right.

00:53:15.420 And it's such a shame because I think there's now a way of packaging that because what you've

00:53:20.180 revealed is, you know, the parents love the kids.

00:53:22.700 They love them.

00:53:23.560 They want to do right by them.

00:53:25.620 But they're either under-resourced or don't know what those steps are.

00:53:30.900 How do you, if it's presented more that way, maybe we get around the victim blaming into

00:53:36.700 something where you could actually help them.

00:53:39.520 I worry about that because actually I think one of the, you know, I used to be think that

00:53:44.260 affirmative action for doctors, that seems like the worst idea to place to have affirmative

00:53:48.160 action.

00:53:49.680 But there is actually an argument for it in that if you look at whether black patients 1.00

00:53:55.960 will follow the instructions given to them by a white doctor versus a black doctor, they're 0.98

00:54:02.280 more likely to actually follow the instructions of a black doctor.

00:54:05.460 So that actually makes the case, there's two ways to fix that problem.

00:54:09.600 One would be to somehow rebuild the trust so that when the white doctor gives the script, 0.53

00:54:15.380 they find it credible.

00:54:18.200 Or the short-run fix is to get more black doctors so that we have people following the 0.88

00:54:24.700 scripts. 0.87

00:54:25.760 Somehow that trust needs to be restored so we don't have this, well, we're the white saviors

00:54:32.000 coming in and telling you how to parent your child.

00:54:34.120 Right, well, right, right.

00:54:37.040 Well, you know, I also looked at the Head Start literature for a long time because that

00:54:42.400 was, Head Start's a very interesting program for those of you who are watching and listening.

00:54:46.800 It was part of the American War on Poverty that started in the early 1960s and it was actually

00:54:51.500 a program that was optimistically viewed by conservatives and liberals alike, right?

00:54:58.340 And because first of all, you know, who likes poverty and the answer is nobody.

00:55:02.100 So everybody's against poverty that has even an iota of sense.

00:55:05.520 And so the liberals were happy because there were, you know, steps being taken to remediate

00:55:12.860 poverty hypothetically at its source.

00:55:14.780 And the conservatives were happy because, well, wouldn't it be better if, you know, poor

00:55:18.800 young people were educated so they could pick themselves up by their bootstraps and, you 0.98

00:55:22.880 know, make their way in the world.

00:55:23.860 And so everyone was hoping that Head Start would be a success.

00:55:28.040 And fundamentally, it wasn't.

00:55:31.600 So, and here's how it wasn't.

00:55:35.220 The goal was, the theory was that if you got to kids early before school and you gave them

00:55:40.580 an academic boost, that that would not only catch them up to their peers, but it would

00:55:44.920 give them the kind of permanent advantage that would grow across time, right?

00:55:49.700 Because now you're prepared to go to school, you can do better in school, and the advantages

00:55:53.400 would just accrue.

00:55:54.660 And that was a pretty good theory.

00:55:56.660 But it turned out not to be true because what happened, and this has been studied to

00:56:01.600 death, and there's no doubt about this, and people of every political stripe analyze the

00:56:06.200 data, is that the Head Start kids actually did do better in grade one and two and three.

00:56:11.340 Their grades were higher.

00:56:12.420 They were more likely to attend school and so on.

00:56:14.380 But all the other kids caught up to them by grade six.

00:56:17.920 So the cognitive advantage didn't accrue, and it didn't multiply.

00:56:22.500 In fact, it disappeared.

00:56:24.260 And now, what Head Start did do was more kids graduated who were Head Start alumni, and fewer

00:56:32.360 got pregnant in the teenage years, and there were fewer criminals in the Head Start groups.

00:56:38.520 And the reason for that, apparently, was that some children's environments were so toxic that

00:56:44.760 just taking them out of those environments for some period of time allowed them to be

00:56:49.160 more socialized.

00:56:50.500 And because they behaved better, they had better outcomes, but there was no effect whatsoever

00:56:54.480 on cognitive performance.

00:56:56.200 And that's very disenchanting, eh?

00:56:58.940 Because that was a major league program, and people put a lot of time and effort to it.

00:57:02.820 And there were every reason to hope that there would be some gains on the cognitive ability

00:57:07.360 front that were permanent, but that didn't happen.

00:57:09.740 So I thought the kids who went to, the Head Start kids who went to better schools, you

00:57:15.820 did see it continue.

00:57:17.060 But maybe I'm misremembering, misremembering.

00:57:20.040 I don't know.

00:57:20.620 But the kids who went to the bad schools, it sort of continued to be, like, you couldn't

00:57:25.320 just stop the investment there if they kept going on that track.

00:57:31.960 But I'm not an expert in that literature.

00:57:33.820 I'm sure you know it much better than me.

00:57:35.100 Well, no, not necessarily.

00:57:36.440 I don't know if I knew the differentiation at grade six between the kids who, you know,

00:57:42.520 went to better schools and the kids who went to worse schools.

00:57:45.760 What I did know was that overall, the cognitive advantages that had been accrued disappeared.

00:57:51.160 It didn't seem to have a permanent effect on IQ, for example, which was really disheartening,

00:57:56.600 right?

00:57:56.840 Because now I looked into it even more detail.

00:57:59.800 Part of the issue was, you know, Head Start was also used as an employment program.

00:58:03.920 And so it wasn't necessarily obvious that the kids were actually learning anything at Head Start.

00:58:10.360 They might have been being taken care of reasonably well.

00:58:13.420 And it's also very difficult to take a group of three-year-olds and teach them anything in an hour and a half,

00:58:20.600 because there are three and, you know, you have to, you have to give them juice and you have to give them food and you have to stop them from tearing each other into bits.

00:58:30.520 And like, and then you have to be trained enough to actually educate them.

00:58:35.000 And so it isn't obvious that Head Start was set up optimally as a cognitive retraining program.

00:58:41.100 But then it's very expensive to set up an optimal cognitive retraining program.

00:58:45.720 So that's also a major league problem.

00:58:49.260 So let's go back to the judgment.

00:58:51.840 So what exactly did the Supreme Court determine and what are the, what have the reactions of the universities been and what are the consequences of the decision?

00:59:02.780 So, you know, the Supreme Court, I mean, there was some sense of vindication because I really felt like the lower court rulings abused the statistical evidence.

00:59:14.980 Now, I think the statistics played a role in all this, but, you know, there are other aspects besides just the statistical side.

00:59:24.700 I think it's very clear, you know, that Harvard kept harping on me over and over again about the idea that, well, we only use race to help you, not to hurt you.

00:59:36.520 But it's a, it's a zero-sum game.

00:59:38.640 So, you know, a penalty for one group is equivalent to a bump for the other group.

00:59:45.740 You know, we can always write it, write it that way.

00:59:50.240 And, you know, I think Robert sort of summed up the decision well by saying the solution to eliminating discrimination is to eliminate all of it.

01:00:00.560 You know, so you're not supposed to use race directly in admissions.

01:00:04.520 Now, they left this bit of a loophole in terms of being able to talk about your experiences of prejudice and such.

01:00:12.780 And, you know, I think that could be a good thing.

01:00:16.460 But if it just gets abused the way I expect that it might.

01:00:19.920 Yeah, it will.

01:00:21.480 Then we've got a problem.

01:00:24.720 Okay, so they eradicated.

01:00:27.820 So they said straightforwardly that you are not to use race as a determining factor.

01:00:33.320 That's right.

01:00:34.980 That's right.

01:00:35.620 Race can enter in only through your experiences.

01:00:40.020 What that means for what college admissions places actually do is going to be interesting.

01:00:46.960 I mean, I think what California did, you know, we had Prop 16.

01:00:52.080 Prop 16 tried to put racial preferences back in place.

01:00:56.060 It got voted down in California by wide margins despite Trump being on the ballot.

01:01:01.020 And the reaction of the UC system was, we're going to no longer require the SAT because we want to figure out a way.

01:01:09.820 You know, I think that the idea of a university throwing away data seems just...

01:01:16.160 Especially the SAT.

01:01:17.880 That's just...

01:01:18.480 It's anti-educational, you know.

01:01:20.700 It's just beyond comprehension.

01:01:22.220 You know, it's to take the SAT for all of its flaws is a pretty damn good test of general cognitive ability.

01:01:28.740 And that's a great predictor of the potential for academic success.

01:01:32.340 And it's also an equalizer across schools, which is crucial, right?

01:01:36.820 When you're trying to contrast rich kids from great schools with poor kids from dismal schools.

01:01:42.500 You know, at least the poor kids from the dismal school with a great SAT can get into university and likely succeed.

01:01:48.540 You throw that out, man.

01:01:50.180 What are you left with?

01:01:51.740 You're left with these other criteria.

01:01:54.880 Yeah, I don't know how the UC admissions could possibly make decisions.

01:01:59.380 You know, like, if you're going to use the high school grades, grades are all relative to whatever high school...

01:02:05.280 Yeah, right.

01:02:06.180 ...you know, you're attending.

01:02:08.540 This is, again, where if you use the data, you could actually show and build that trust with the data.

01:02:14.440 So, you know, if the SAT is biased in favor of the rich and it's over-predicting the performance of the rich, then we could correct that.

01:02:24.220 You know, but you show exactly how it affects it.

01:02:26.820 Well, we could point this out with the SAT, too.

01:02:29.560 So, this has to be done on the racial front because if the SAT was prejudiced against black test takers, it would under-predict their performance.

01:02:42.220 And it doesn't.

01:02:43.280 It doesn't.

01:02:44.840 Right.

01:02:45.300 And that's a killer statistic, right?

01:02:47.720 I mean, it's quite the shock because now you could say the only way to get around that on the systemic racism front is to say, as we alluded to earlier, that the performance criteria are just as prejudiced as the admission criteria.

01:03:02.220 But they would have to be exactly as prejudiced, which seems extraordinarily unlikely.

01:03:10.260 And you just roll it forward to the next stage every time to say, well, then the next place is exactly the same level of racism as the previous stage.

01:03:17.580 You know, it's crazy.

01:03:20.080 Well, you know, the conclusion I came to, and this was a painful conclusion in some ways, it wasn't necessarily in accord with my moral sentiments, let's say, is that we simply can't do better than actual race and ethnicity-blind objective tests of predictive merit.

01:03:36.560 And those should be established, those should be applied uniformly because that's the best solution, even though it still has its flaws.

01:03:45.260 Now, I don't, I still have no idea what the consequence of that would be, given that the elite schools would rapidly fill up with Asians. 1.00

01:03:52.300 Maybe the consequence would be that the other relatively underperforming ethnic groups, including whites, would pull up their socks. 0.95

01:04:02.320 You know, that's a possibility.

01:04:03.760 And say, well, those people are obviously doing something that we're doing right that we're not doing.

01:04:09.100 That was my reaction when I was looking at the numbers, thinking, I want to know what they, what they're doing, you know, because it's, it's incredible.

01:04:18.000 Yeah, right, right.

01:04:19.100 There's a lesson to be learned there in principle.

01:04:21.100 Knowing what those best practices are would be incredibly helpful.

01:04:24.300 And I don't, I don't have a good sense of it.

01:04:27.140 Yeah, so what, what has been the consequence for you of being involved in this line of research?

01:04:34.780 You said there were protests about your work in 2011.

01:04:38.140 I guess you were lucky it was 2011 and not 2016 because it didn't take you out and might have later.

01:04:45.740 So, so what has been the consequence?

01:04:46.580 There are many people wondering how I survived, you know.

01:04:49.520 Yeah, well, I'm curious about that.

01:04:50.840 Continue.

01:04:51.040 Exactly.

01:04:52.140 Yeah, yeah.

01:04:52.940 And, you know, that whole experience really prepared me to take the case.

01:04:58.640 It was actually one of the most spiritual moments, you know, of my life.

01:05:02.540 Because when people are protesting, you know, it just made me realize how much I care what other people think about me.

01:05:10.500 You know, somebody writes an article from some satellite state school in an ethnic studies department in the state of Washington calling me a racist.

01:05:19.160 That really hurt at the time.

01:05:21.160 Yeah, yeah.

01:05:21.720 It's really hard on people.

01:05:22.940 You bet.

01:05:24.020 And so it was, you know, a week or two of like no sleep.

01:05:27.660 And then I woke up one morning, I view it by grace, was free and felt like I was able to love the people who were coming after me and saying, okay, you know, I think they're misinterpreting what I'm saying.

01:05:43.060 I'm going to give them the benefit of the doubt and explain what I actually, what I actually mean.

01:05:47.520 And I think that that, I don't know whether that's been what's, I'm sure the horrible stuff could still happen to me.

01:05:55.100 But, you know, if somebody gives, I can take a punch and respond in love with, and I feel like that might be seen as wimping out.

01:06:08.340 But I don't feel like I'm compromising on the truth, you know, I'll try to go as far as I can to meet them where they're at without compromising on what I know to be true.

01:06:19.360 So, and I think it makes a difference, you know, I do, you know, I try to treat people well in my interpersonal relationships.

01:06:28.800 Right, right, right, right.

01:06:30.120 And that.

01:06:30.580 So you're not, you're not adding being generally dislikable to the raft of sins that might be utilized, let's say, to take you down.

01:06:39.380 So far, that's right.

01:06:42.400 So far, yeah, so far, right, no kidding, no kidding.

01:06:45.480 But I've been amazed, maybe this podcast will change all this, but I've gotten zero hate mail since the Supreme Court decision.

01:06:52.360 Okay, so you made reference earlier to the fact that this Supreme Court decision overturned a number of lower court decisions.

01:07:01.800 And it's also the case that, so obviously, the lower courts were persuaded by evidence that wasn't the same as the evidence you were bringing forward.

01:07:11.920 And it's also the case that at the Supreme Court itself, there were other experts with a pedigree as credible as yours, let's say, who don't agree about the interpretation of the data.

01:07:26.160 So if you had to make a case against what you were offering as a witness, how would you make the case?

01:07:33.620 And how did the people who were brought in as experts make the case?

01:07:37.300 And what were they claiming?

01:07:39.160 So actually, I would say they have a much higher pedigree than I do.

01:07:43.520 David Card was one of them.

01:07:45.180 And soon after the case, he actually won a Nobel Prize.

01:07:48.020 He was the one in the Harvard case.

01:07:49.900 And Carolyn Hoxby at Stanford was my counterpart in the UNC case.

01:07:56.160 What was interesting to say, I had the benefit of working on both cases, so I could be reasonably consistent across the two.

01:08:03.780 They actually attacked me from other sides.

01:08:06.760 So Card's opinion was that I needed to control for more things.

01:08:11.940 And in particular, the personal rating was an example.

01:08:15.480 And I actually estimated models of the personal rating showing the Asian bias.

01:08:23.320 But the way he would argue it would be, we need to take sort of Harvard's word for it, that this is not a biased measure.

01:08:32.940 So that was one aspect to it.

01:08:35.280 The other aspect to it is...

01:08:36.760 Right, but you showed it was a biased measure.

01:08:39.160 So I don't understand his claim exactly.

01:08:41.800 So your claim, if I get it right, is that the prejudice against Asians, so to speak, was making itself manifest as invisible variables within the so-called personal or personality rating.

01:08:56.760 Now, there are objective ways of measuring personality, which we could also point out, which are quite valid, right?

01:09:02.500 So Harvard could take that route, and they don't.

01:09:06.800 They do a subjective evaluation.

01:09:10.300 But your point was that, well, that was hiding substantive anti-Asian bias.

01:09:15.300 So how did your opponents muster an argument against that?

01:09:20.240 So it's very convoluted how you do that, because you don't actually look at a model of the personal rating.

01:09:26.780 Any model of personal rating shows that Asian Americans do just as well. 1.00

01:09:31.320 Sorry, on the observables associated with the personal rating, Asian Americans have observables that are just as strong as whites or stronger, and yet get worse personal ratings.

01:09:43.140 Often with discrimination things, you're worried that, oh, every time I add a variable, the discrimination goes down.

01:09:50.020 What if I keep adding variables, it may go away?

01:09:53.200 Okay, that's the amazing thing about this case, is with Asian Americans, you put in more variables, it often goes up, you know, the discrimination, because they're stronger on those measures. 0.59

01:10:06.200 So in order to get there, it's pretty convoluted.

01:10:09.400 The first thing you have to do is take those special groups, the athletes, the legacies, the children, and donors, and you have to say that the discrimination has to be happening against Asian applicants that are there as well.

01:10:25.280 And actually, that turns out not to be the case.

01:10:28.780 So 98% of Asian American applicants are not athletes, legacies, children, and donors.

01:10:35.840 The 2% that are, they're not being discriminated against.

01:10:42.420 There's not evidence of that.

01:10:44.060 And I think that makes perfect sense.

01:10:45.340 Okay, so it's not precisely Asian discrimination, or if it is, it's not pervasive enough to cut across all categories.

01:10:54.420 That's right.

01:10:55.020 And I would say the same thing on affirmative action.

01:10:57.460 So Zion Williamson was an amazing Duke basketball player.

01:11:00.480 He didn't benefit from racial affirmative action.

01:11:04.440 He benefited because he was an amazing Duke basketball player.

01:11:07.820 You know, when you're talking about those things, that's just not relevant.

01:11:12.260 You know, for a long time, they would talk about discrimination against black quarterbacks.

01:11:16.620 We shouldn't be able to say, we're not discriminating against linemen, so therefore we're not discriminating against black quarterbacks.

01:11:23.480 That doesn't make any sense.

01:11:25.400 All you can say is that the discrimination isn't universally pervasive.

01:11:34.220 You could say that some forms of discrimination will trump others.

01:11:37.740 That's right.

01:11:38.360 We discriminate against the Asian Americans who don't have those connections to Harvard through the legacy and the recruited athlete process.

01:11:48.060 So that was sort of how that case sort of worked.

01:11:53.480 And much of the focus in the Harvard trial was all about the Asian American discrimination.

01:11:59.380 I actually don't think David Carden and I had very different things to say about the racial preferences.

01:12:05.640 I would say that it quadruples their chance of getting admitted.

01:12:10.720 According to Carden's results, it would triple the chance of being admitted for black applicants.

01:12:15.720 What was interesting in the UNC case is things operated very differently.

01:12:23.220 In Harvard, I wrote a report, then Card saw my report and built off of that.

01:12:29.020 In the UNC case, we wrote simultaneous reports and then did that two more times.

01:12:35.400 So our starting places were completely different.

01:12:38.480 And the other expert basically took the position that we can't really model Harvard's UNC's admissions because it's holistic, you know, and that models sort of of admissions failed to predict the decisions well.

01:13:00.640 Now, I think that the criteria that she used to evaluate that was nonsensical and that if we started with me writing the first report, we wouldn't have got to that position.

01:13:15.000 But in this case, you can just keep doubling down, doubling down.

01:13:18.600 I mean, as an example, you know, her criteria was based on what's called the pseudo-R-squared. 0.85

01:13:26.700 These are nonlinear models.

01:13:28.500 So an R-squared in a linear model sort of tells you how much of the variation the data is explained.

01:13:35.160 When you have a pseudo-R-squared, it doesn't really work that way in quite the same way, you know. 0.79

01:13:40.780 So one of the comments that was made was a pseudo-R-squared of 0.5 means that we get half the admissions decisions correct.

01:13:51.060 Well, that, of course, is nonsense, right?

01:13:53.460 Because by flipping a coin, I could get half the decisions.

01:13:57.340 Right, right, right, right.

01:13:58.660 Correct.

01:13:59.060 You really need to be looking at accuracy.

01:14:01.560 And if you look at accuracy, you know, the models would predict the correct decision over 90% of the time.

01:14:07.120 So it was a very funny experience.

01:14:12.620 In the UNC case, they basically thought I didn't control for enough things.

01:14:18.360 Sorry, that I controlled for too many things.

01:14:20.400 In the Harvard case, they said I didn't control for enough things.

01:14:24.400 It wasn't coherent.

01:14:27.900 Okay, so why do you think—okay, so, okay, fair enough.

01:14:30.980 So why do you think that your arguments, so to speak, or the side of the cases that you were testifying on behalf of,

01:14:40.960 why do you think that that carried the day at the Supreme Court level and not in relationship to the lower courts?

01:14:48.700 And what do you think of—I mean, one of the response patterns, especially from the radical types,

01:14:54.160 is that, well, you know, the Supreme Court is stacked with reprehensible conservatives,

01:14:57.960 and, of course, that's how they voted.

01:15:00.980 And that was why the anti-affirmative actions say pro-merit—that's another way of looking at it—side carried the day.

01:15:11.020 What's your sense of that?

01:15:12.360 I think that, you know, I certainly came out of those cases a bit cynical about the role of the statistics here.

01:15:19.100 You know, both judges at the lower court could have called in a third expert to say, you know, which one of us is right.

01:15:28.200 But that would have limited their ability to rule how they wanted to rule.

01:15:32.080 So it's funny how that works, right, where we say, well, it's a conservative court.

01:15:36.280 They're going to rule however they want to rule.

01:15:39.720 Based on the statistical evidence, I feel quite strongly that that was what happened at the lower court case.

01:15:46.620 And then you're left with a really bad record, you know, in terms of what was admitted as evidence.

01:15:55.300 And do you think—well, do you think your—do you think that your political affiliation—

01:16:00.960 how do you control for the potential consequences of your political affiliation, which I don't know, by the way?

01:16:07.760 How do you control for the potential consequences of your political affiliation or viewpoint on the outcome of the studies that you've been running and the testimony that you provide?

01:16:19.320 Like, how do you protect yourself against your own bias?

01:16:22.360 Well, I'm very cautious on this front, partly because of that protest.

01:16:26.820 You know, so my burden of proof is, you know, I know I can get in big trouble, so I have to be very careful how I'm going to be talking about these things

01:16:35.820 and set a very high bar for coming to particular conclusions.

01:16:41.320 I mean, I feel this way sort of in general, like—

01:16:43.320 Right, the cost is high.

01:16:44.420 That's right.

01:16:44.900 So the evidence needs to be much better in that regard.

01:16:47.620 Yeah, yeah, yeah.

01:16:48.300 But my intent was always to get peer-reviewed papers out of this.

01:16:52.260 So I've published five peer-reviewed publications in good economics journals out of the case.

01:17:00.000 And on Asian-American discrimination, on racial preferences.

01:17:03.860 I want to elaborate on that for a second, too, because for everyone who's watching and listening,

01:17:08.480 it's like the social science domain is overwhelmingly tilted politically towards the left.

01:17:15.820 That may be less true for economics than it is for psychology, but it's true across the social sciences.

01:17:21.320 And so what that means is that if you publish a paper with so-called conservative implications in a decent peer-reviewed journal,

01:17:30.700 the probability that it's robust is extremely high, because if there are any reasons to thwart its progress forward,

01:17:38.360 those reasons will make themselves manifest.

01:17:41.240 I think that's right on.

01:17:42.680 And you could really see it, in my view, with, you know, I published one paper on legacy and athlete preferences

01:17:48.560 and another paper on Asian-American discrimination.

01:17:53.180 The same commentator will trash the Asian-American discrimination one and laud the legacy and athlete one.

01:18:00.800 And it's the same model.

01:18:02.000 Right, right.

01:18:02.440 It's the same model.

01:18:03.560 Right, right, right, right.

01:18:05.980 Yeah.

01:18:06.360 And that was actually stunning.

01:18:08.340 Harvard, the evidence against Harvard was so damning.

01:18:10.880 They had their own internal research team that had estimated models of Harvard's admissions.

01:18:18.080 And they found a big penalty against Asian-Americans.

01:18:22.380 And Harvard was like, oh, well, we don't know what to do with that.

01:18:25.620 But in that same model, they found that they were giving a bump to low-income students.

01:18:31.760 And they knew that that was true.

01:18:33.860 It's the same model.

01:18:34.980 You can't interpret it one way when you get the result you like and another way when you get the result you don't like.

01:18:41.280 You know, it doesn't make any sense.

01:18:43.180 And if anything, based upon the strength of the applicants, you should disbelieve the low-income result, not the Asian-American result. 0.53

01:18:51.020 Because all the observable factors are pointing to Asian-Americans being stronger.

01:18:55.700 They're probably stronger than the unobservable ones as well.

01:18:59.700 Right, right.

01:19:00.280 Okay, okay.

01:19:00.960 So you protect yourself against bias, first of all, just by fear.

01:19:04.500 Exactly.

01:19:04.900 Which is actually a good way of...

01:19:06.080 Well, it's a good...

01:19:06.740 I used to tell my students, look, guys, you don't publish unreproducible data, right, to further your career.

01:19:15.700 Well, why?

01:19:16.520 Well, do you want to study something that doesn't exist for the rest of your life? 1.00

01:19:19.780 Like, that's stupid. 1.00

01:19:20.920 Don't do that. 1.00

01:19:21.640 And you're going to be motivated to do it because it's hard to admit that a research project was a failure.

01:19:27.160 You're going to try to scrape something out of it.

01:19:28.900 So you have to be careful.

01:19:30.280 But then...

01:19:31.180 And so that's another, you know, use of fear.

01:19:34.040 It's like, be careful because you'll convince yourself of something that isn't true and waste your life. 1.00

01:19:38.320 That's stupid. 1.00

01:19:39.420 Your situation is, well, if you say anything even vaguely untoward or inaccurate, you're going to get slaughtered for it. 1.00

01:19:45.480 And then...

01:19:46.180 And that's a fair...

01:19:46.940 That's a fair protection.

01:19:47.640 And then also, well, if you have to publish these in peer-reviewed journals, you're going to hit the ideological vanguard and have to argue your way through it.

01:19:59.400 And if you can do that, well, obviously, the data has to be so credible that it can't be dismissed by people with an ideological bent.

01:20:06.340 Okay, so that's good.

01:20:07.460 So Harvard and UNC were hit hard by this ruling, practically.

01:20:16.580 And I would say morally, too.

01:20:17.880 They got walloped.

01:20:18.880 How have they reacted?

01:20:20.500 And how have other universities reacted?

01:20:22.980 Well, the public reaction is to send out emails how disappointed they are in the Supreme Court decision, how they're going to abide it.

01:20:31.020 But our commitment to diversity is unchanged.

01:20:34.000 So, I think that they're going to try to figure out legal ways or illegal ways where they're not going to get caught to somehow put the preferences back in.

01:20:49.320 Right, right, right.

01:20:50.320 Well, does that open them up on the liability front now?

01:20:52.920 I mean, they've been told in no uncertain terms that they can't do this anymore.

01:20:56.040 It's not...

01:20:56.540 This is pretty black and white.

01:20:57.640 So, it seems to me that if Harvard continues to gerrymander their admissions processes, that they're setting themselves up for a pretty walloping fall.

01:21:06.460 I don't know what that would mean on the liability front, for example.

01:21:09.740 I mean, there was a class action suit at one point, right?

01:21:12.280 Yeah, but it really wasn't about getting damages so much as changing the system.

01:21:16.820 But I think Harvard would have a very hard time because they just went through a whole trial of things saying that we need to explicitly consider race in order to get these levels of diversity.

01:21:29.120 If they somehow kept the same levels of diversity now, then when they're supposed to not be explicitly considering race, then it shows they must be cheating somehow in order to get there.

01:21:43.160 So, you almost have to see a drop or the whole record in their case was off.

01:21:51.140 Right, right.

01:21:51.840 Okay, okay.

01:21:52.560 Now, how do you think that universities should select in the aftermath of this decision?

01:22:01.060 Well, I think they should select just based on test scores.

01:22:05.000 But even within test scores...

01:22:06.480 Yeah, just with objective tests.

01:22:07.840 Yeah.

01:22:08.340 I mean, the problem is that the SAT, for a place like Harvard, doesn't test at a high enough level.

01:22:14.360 We actually need to go...

01:22:15.500 The tests like to get into college in Chile are at a much higher level than what we have in the U.S.

01:22:21.700 Right, so they should refine them and make them more demanding so they can discriminate. 0.63

01:22:27.720 So, listen, for everybody watching and listening, you know, imagine that you get people who score 95th percentile on the SAT and people who score 99th percentile.

01:22:38.740 And you might think, well, you know, what the hell's the difference?

01:22:41.740 And the difference is this.

01:22:43.080 If you score at the 95th percentile, you're the smartest person in a room full of 20 people.

01:22:48.680 And if you score at the 99th percentile, you're the smartest person in a room full of 100 people.

01:22:53.660 And then the difference is just as big from 99 to 99.9.

01:22:58.460 It's one in 100 versus one in 1,000 and so forth all the way up the scale.

01:23:02.360 And there's no indication that I can see in the psychometric literature that that ever stops being relevant.

01:23:08.700 And so it is important to differentiate at that upper end.

01:23:12.380 You know, and so, I mean, we've put together models to predict academic performance.

01:23:16.160 And you can predict academic performance quite accurately with general cognitive ability and trait conscientiousness.

01:23:24.560 And then low neuroticism helps.

01:23:26.940 And if you're looking to expand on the creative front, you could look at measures of creative ability.

01:23:33.060 And there are good measures.

01:23:34.400 And you can look at performance.

01:23:35.800 And you can make a formal model where you specify exactly what weight you put on all of those variables.

01:23:44.020 And you can easily test that against actual university performance, which is what you had recommended earlier.

01:23:50.040 Yeah.

01:23:50.500 I mean, clearly, for math, you want to put more weight.

01:23:54.560 On your math background, maybe less so in English.

01:23:58.940 That's not taken into account at all.

01:24:01.060 But I think those tests right now are too easy. 1.00

01:24:03.020 That distinction at that top level could be like, I just made a stupid mistake. 0.99

01:24:08.200 That's too bad. 0.97

01:24:09.180 You really want to test much deeper.

01:24:11.680 You bet.

01:24:12.180 You bet.

01:24:12.620 Yeah.

01:24:12.820 Okay.

01:24:13.140 Okay.

01:24:13.700 Now, we just outlined how universities should select.

01:24:18.840 This is how businesses should select, too, by the way.

01:24:20.920 And there's huge economic advantage in doing that, by the way.

01:24:23.540 Like, you don't have to increase your ability to select top people very much to benefit in a staggering manner from the consequence of that improvement.

01:24:33.100 I tried to convince corporations of this for, like, 15 years.

01:24:36.240 I went on the road to sell to corporations, which turned out to be absolutely 100% impossible.

01:24:42.200 That's when I first ran into HR departments, by the way.

01:24:44.880 And that was back in, you know, the early 2000s.

01:24:47.460 But we talked about how the universities should select in the aftermath of the decision.

01:24:54.180 The lurking question is, how will they?

01:24:56.920 I saw UC Davis, for example, at the medical school.

01:24:59.540 They're trying to produce an adversity quotient, which I think is just an absolutely catastrophically dreadful idea.

01:25:05.740 I mean, I can understand why they're doing it, but God, that's a dismal contest, right?

01:25:10.020 To match your misery against someone else's and to try to rank order, you know, who had the hardest lot.

01:25:17.760 I mean, that's a rough thing to adjudicate, man.

01:25:23.540 It's so prone to the political side.

01:25:24.700 The adversity I experience at a pro-life rally where people are harassing me, I don't think that's going to sell.

01:25:34.640 You know, fundamentally, it introduces ideological conformity, which I think a real outstanding question is,

01:25:44.180 does somebody like a Roland Fryer bring more or less diversity?

01:25:48.900 You know, he's not a progressive. 0.92

01:25:51.560 Black, not a progressive. 1.00

01:25:52.760 To me, that, in some sense, brings more diversity. 0.76

01:25:55.780 We don't want it to be the case that the conservatives are all white, you know? 0.82

01:26:01.920 But that's not really how I think it's seen.

01:26:05.780 I think it's just going to end up being...

01:26:07.560 Yeah, well, the thing, I think the whole diversity, shibboleth, is a front for posers to attain status they don't deserve, fundamentally.

01:26:16.540 Because I don't think you can do better than merit defined the way we already defined merit, which is you're a meritorious candidate if the features you bring to the position match the desired outcome of performance in the position.

01:26:31.700 And fundamentally, as we also pointed out earlier, you're actually bound by law to do that, especially if you're an employer.

01:26:39.920 You know, now, the law is tricky because it contains self-contradictory aspects.

01:26:44.580 So one counterexample for that, you know, there's a guy, Bassett Safar, University of Michigan.

01:26:50.740 He does amazing work, in my mind.

01:26:53.340 And he's from, he's a Muslim from Pakistan.

01:26:55.740 And he's able to go to these madrasas, these fundamentalist schools, and interview them and get data.

01:27:04.580 To me, that's actually, that's the kind of diversity that I think is good in the sense that actually he's opening up these new areas of research that there's no way the madrasa would ever give me data.

01:27:17.940 So there might be some, but it's a limited scope for, you know, getting certain questions answered that you might not fully answer.

01:27:28.400 You could argue that as a form of competence, like if you were going to hire someone and the person said, look, one of the things I bring to this position is that I have access to an ethnic group, say, or an anthropological group that another candidate won't have.

01:27:41.800 I mean, and I do think that that is actually a measure of merit and could be put under that rubric.

01:27:49.880 Yeah, and the key is not to have that measure of merit just be because I'm a particular race, right?

01:27:55.180 Because that's effectively how it's argued now, is that by your race, that gives you that, you do merit it.

01:28:04.480 Because you've got a unique insight into that community.

01:28:08.300 Right, right.

01:28:08.780 Well, and that's also a strange, that's also a strange claim too, right?

01:28:11.940 Is that merely because you have a very, what would you say, low resolution feature that you're somehow an emblem and standard bearer for that group.

01:28:25.320 You know, like I've never thought of myself as a representative of the white community.

01:28:29.540 You know, it's preposterous because, well, first of all, it's preposterous because we know as researchers into individual differences that the differences within any given ethnic or racial group are much larger than the differences between groups, right?

01:28:45.880 Which is actually the canonical non-racist statement, right?

01:28:51.220 Within group variance trumps between group variance.

01:28:54.140 It even does that between men and women in almost all domains.

01:28:58.040 In fact, I don't know of a single domain where that's not true.

01:29:00.740 Even in interest, where men and women differ most widely, the difference is one standard deviation.

01:29:05.160 There's still way more variance within the group than there is between the groups.

01:29:08.220 So, how do you think the universities will respond?

01:29:13.120 We said how they should respond.

01:29:14.640 Well, how will they respond?

01:29:16.700 Well, I'm hoping it will be a heterogeneous response.

01:29:19.540 I think some schools will go the UC route and get rid of the test scores, and they'll do other things like diversity statements and such.

01:29:27.180 I'm hoping that other schools will focus more on, you know, addressing some of the pipeline issues.

01:29:32.800 I think you will see a big movement against legacy admissions.

01:29:36.460 I was surprised at actually how quickly that happened.

01:29:41.380 So, I think that sort of stuff is going to go.

01:29:45.580 To me, you know, what I'm really pushing for is for them to use their data to be able to say,

01:29:51.180 look, I can tell you that you're going to really succeed here.

01:29:55.700 You know, and we've set up our program so that you can succeed here because the competition for those students is going to be fierce.

01:30:01.140 Well, right. So, what a university could do, you know, hypothetically, imagine that they set up actually rigorous objective testing models.

01:30:09.760 So, a university could say to a given candidate, your probability of succeeding in this discipline here is X percent.

01:30:17.540 And so, that would mean that qualified students could look at a range of universities and they could say, for example, if they were slightly lower performing on the academic front, what we're interested in sciences,

01:30:27.100 they could pick a university where they had, say, a 70% chance of graduating.

01:30:32.180 And so, then they'd know that if they were, they could pick the university that was the most challenging that would give them some reasonable opportunity of success, right?

01:30:43.060 That'd be a good deal, right?

01:30:44.220 Because there's a very wide range of universities.

01:30:46.820 And objective testing could establish that across time.

01:30:49.460 And you could see that happening in a state like Florida, you know, where the governor said, you know, some of the red states could move towards that model for their state system, at least.

01:30:59.860 So, that's where I think the best path forward is.

01:31:04.100 I think there are other ways, too.

01:31:06.200 You know, the whole tie to all this in terms of racial preferences was losing your government funding.

01:31:12.600 Obviously, on the admissions front, you're sort of stuck there.

01:31:15.400 But if you look at how these, the way we're paying college athletes now, you could easily have a separate organization that was sort of,

01:31:24.500 we're giving scholarships to black students who attend Duke.

01:31:29.320 You know, they're not tied with the government funding.

01:31:31.960 You could see, like, a huge competition emerging for that, where things happen more on the financial aid side, not associated with the university.

01:31:39.140 So, there's no scope to saying that it's illegal because the organization's not taking government funding.

01:31:44.700 Right, right, right.

01:31:45.720 Do you see any downside to that?

01:31:47.500 But, I'd have to think more about that.

01:31:52.520 I mean, to me, that's how the market works.

01:31:54.320 There's no, they're not obvious.

01:31:55.520 Yeah.

01:31:55.860 Yeah, yeah, yeah.

01:31:56.540 It's not, there's no, there's no downsides that leap obviously to mind.

01:32:00.620 So, now, is there anything else that we, we're running out of time here on the YouTube side?

01:32:04.980 I'm going to flip over to the Daily Wire side for everyone watching and listening and, and talk to Peter a bit more about how his interest in this domain and in economics in general made itself manifest.

01:32:16.400 So, you could, you're welcome to join us there if you're inclined to.

01:32:19.140 Is there anything else that we didn't cover today in relationship to the Supreme Court decision and in relationship to affirmative action and its complexities?

01:32:28.460 Oh, I, I know what, I know what we conclude with perhaps, you've spent a lot of time studying affirmative action.

01:32:37.360 Do you have, what would you say in relationship to its putative advantages?

01:32:44.700 Has it, has there been any manner in which affirmative action has actually been a policy success?

01:32:52.640 Well, it certainly has increased the share of minority students at top schools.

01:32:58.460 And I think those, when you're-

01:33:00.960 As applicants or graduates?

01:33:02.960 Even as graduates, because I think at places like Harvard now, they will graduate, great you.

01:33:09.460 They will graduate you.

01:33:11.120 So, that's always been sort of a gamble, right?

01:33:14.080 That by, yes, there might be these costs associated with it, but by getting them to Harvard, great things are going to happen later on in society.

01:33:23.420 You know, and that's, you know, they'll, they'll be the Supreme Court justices to get to those very top of the top positions.

01:33:33.720 That would be the argument.

01:33:36.000 To me, that's a little snobbish, but, but I understand the argument.

01:33:41.980 Do you, and do you, do you think there's any merit to that argument?

01:33:44.360 I mean, you see, my, the counter argument would be giving less qualified people a pedigree that indicates their qualifications and then putting them in ever more important situations that they're not qualified for is not a net good, right?

01:34:02.500 That's a rough argument to make.

01:34:03.840 And I, I wouldn't necessarily say I'm making that argument, but that is the counter argument.

01:34:09.820 That's right.

01:34:10.740 So, um, on the one hand, we, our Supreme Court is more diverse because of affirmative action.

01:34:18.180 On the other hand, I think that that is, you know, Clarence Thomas would say that it's done him a serious disservice because of being viewed that way.

01:34:28.000 And you don't know.

01:34:29.580 And that's, that's the problem, right?

01:34:31.980 Is, uh.

01:34:33.080 Yeah, well, that's one of the really ugly things about affirmative action, as far as I'm concerned, is that it, it, the question is then begged, and that's not good.

01:34:42.580 And that's particularly hard on people who are truly qualified, right?

01:34:45.840 Because what a bloody catastrophe that is, to have to face that additional level of doubt.

01:34:51.040 Like, who are you really?

01:34:52.820 Jesus, brutal, brutal, brutal.

01:34:54.560 I mean, you take something like Glenn Lowry, who I think you've had on.

01:34:57.840 He's one of the most brilliant people I've ever met.

01:35:00.240 You know, I think he's just fantastic.

01:35:03.080 And yet, he had to endure all of that, you know?

01:35:08.640 Yeah.

01:35:10.340 Yeah, that isn't something you'd wish on your worst enemy.

01:35:13.380 That's for sure.

01:35:14.280 To have your genuine competence in question.

01:35:17.320 Yeah.

01:35:18.020 All right.

01:35:18.680 Well, I guess we should wrap up on this side.

01:35:21.720 So, as I said to everyone watching and listening, I'm going to talk to Peter at some additional length on the Daily Wire Plus platform about the development of his career and his interests.

01:35:31.720 And so, if you want to join us there, then you'd be more than welcome to do so.

01:35:36.680 It's not a bad time to funnel some support the Daily Wire way, by the way, because YouTube, for example, has been pretty assiduously at war with the Daily Wire contributors over the last month.

01:35:48.420 Three of my podcasts have been taken down, and I suspect there's more in the pipeline that will suffer the same fate.

01:35:54.240 And that's not good.

01:35:55.340 And the other people who are using YouTube on the Daily Wire side have been hit harder than me by quite a substantial margin.

01:36:02.660 So, if you're ever thinking about subscribing to the Daily Wire platform, this isn't such a bad time to do it, let's say, on the moral front.

01:36:11.180 In any case, you can give that some consideration.

01:36:13.880 Thank you to the film crew here in Manhattan today for making this a pleasant experience and technically feasible.

01:36:20.840 And to the Daily Wire Plus for facilitating the conversation.

01:36:23.560 Peter, thank you very much for talking to me today and for your efforts on the research front.

01:36:29.960 You bet, man.

01:36:30.800 Good to see you.

01:36:32.820 Ciao, everybody.

01:36:33.920 Till next time.

The Jordan B. Peterson Podcast - September 21, 2023

384. This Man Ended Affirmative Action ｜ Dr. Peter Arcidiacono

Episode Stats

Summary

Transcript