����vlog����

- [Moderator] Welcome to the Wiener Conference Call series. These one hour on the record phone calls feature leading experts from Harvard Kennedy School who answer your questions on public policy and current events. Wiener Conference Calls recognize Malcolm Wiener's role in proposing and supporting this series, as well as the Wiener Center for Social Policy at Harvard Kennedy School.

- Hello everyone, welcome. I'm Ariadne Valsamis from the Office of Alumni Relations and Resource Development at the Harvard Kennedy School, and I'm very pleased to welcome you to this Wiener Conference Call, the first of our 2023-'24 academic year. These calls are sustained by Dr. Malcolm Wiener and his wife Carolyn, who support the Kennedy School in this and many other vital ways, and we are profoundly grateful. Today we welcome Dr. David Deming, who is the Isabel and Scott Black Professor of Political Economy and the academic dean at Harvard Kennedy School. David is also a professor of education and economics at the Harvard Graduate School of Education, and he's a faculty lead of the project on workforce across Harvard Initiative that focuses on building better pathways to economic mobility through the school to work transition. Today, he's gonna talk with us about his research on the relationship between admissions policies at highly selective universities and the composition of our national leadership. We're so fortunate that he's agreed to share his expertise today with the Kennedy School's alumni and friends. Professor Deming.

- Thank you Ariadne. Thanks everybody for being here. Before I get started, I want to give special thanks to Malcolm and Carolyn and the Wiener family for supporting this conference call and actually for supporting me over the many years. Some of you may know that prior to becoming the academic dean, I was the faculty director of the Malcolm Wiener Center and Malcolm and Carolyn's partnership was so important in helping me move forward, move that center forward, and it's such a vibrant intellectual community today, mostly because of their support and so I'm really grateful for that. All right, so I wanna talk about higher education. Let me put my slides up there, okay. And I'm gonna tell you all today about a specific project that I've worked on that you saw some materials about diversifying society's leaders. This is a project studying who gets in and why to elite private colleges in the US and then what are the consequences of admission? Are you actually better off for having gone to one of these schools? Before I do that though, I wanna motivate the topic by talking a little bit more broadly about mobility and economic opportunity and the role of higher education in that in the US. So if I say the phrase American dream, what does that mean to you? I think it means something a little different to everybody, but I'll tell you what I think it means. I think it means that the American dream is that through economic growth and shared prosperity, everybody in this country can do better than their parents did or has the opportunity to do so. And that therefore, everybody in society is kind of progressing or moving forward through shared prosperity. I think that's the American dream. What this figure shows you is that the American dream, if that's acceptable working definition, is under threat. So this figure shows by a child's year of birth, the probability or the percent of children by each of year birth who were earning more than their parents when they were adults in inflation adjusted terms. So a lot of mechanics behind this calculation, but what it shows is that for people who were born around World War II, nine in 10 were able to do better than their parents. And that percentage has declined over time until around the 1980s. It's basically become a coin flip. And the reason for that is not because the economy hasn't grown. It's not because we don't have any more prosperity. Actually real incomes, average real incomes have roughly doubled over this period. It's that the prosperity isn't widely shared. And the intuition for that is simple really. If you could imagine that all of the economic growth over the last 50 years accrued to a single person, let's call that person Jeff Bezos or Bill Gates or something, then what you would see is that you'd have large income growth, but only one family, only one person or however many kids that person has. That one person, sorry, Jeff Bezos or Bill Gates would've done better than their parents. Nobody else shared any of the gains. It was all concentrated in a single person. And so you'd have very low mobility even though you had high growth. And so that's obviously an extreme case, but what you can see in that example is that the more widely shared any amount of economic growth is the more people have the opportunity to progress generationally and do better than their parents. And so the reason why this number has declined is not because growth has declined, it's because income has become more concentrated and wealth has become more concentrated. Okay, so the question is what is the role of higher education in that? Before I became academic dean and I taught a large survey class here at the Kennedy School that that also drew on students in the undergraduates in the economic department and the business school and all around the university called the causes and consequences of inequality, where I basically cover topic by topic, what are the different contributors to the rise in economic inequality in the US and other advanced economies since around 1975 or 1980. And the interesting thing about that class is, or about this phenomenon more generally, is it teaches people to think in a very nuanced way, why? Because there's about 15 or 20 different explanations for why economic inequality has risen, but there's only one fact pattern. And so it's always overdetermined and there's always more candidate explanations than you can possibly fit to the data. And so what that means is it can't just be one thing, it's gotta be multiple things. You have to weigh the evidence across competing explanations. But whatever explanation you have for rising inequality has to be consistent with the evidence. And what this figure shows is that if you look at education, the rising returns to education, these figures show for men on the left hand panel and women on the right hand panel, changes in average earnings or real wages for full-time workers relative to 1963, right? So one is the 1963 line, and then an increase means that people's earnings are increasing relative to 1963. And what you see is that for both men and women, a fanning out of real wage growth by education. So for people with a bachelor's degree and more, their incomes have increased by 40% or more, 50, 60, 80% depending on the gender and the amount of education. But for people who have dropped out of high school or high school graduates earnings have barely increased at all. And the interesting thing about this pattern is that it coincides exactly with the overall rise in economic inequality. And that is not true for other characteristics. For example, the gender wage gap and the racial wage gap. Both of those things are large, but really not much larger than they were 50 years ago. Those things have been constant over time. So if you wanna explain rising inequality, you have to think about things that are changing. And a big candidate explanation is the return to education. It just turns out to be much more economically valuable to have a college degree than it used to be. And so why is that? Well, and why is higher education contributing to inequality? Well, one reason I think probably the most important reason is the distribution of resources. So you might say, well, it turns out that the reason why the returns to college are rising is because inequality in an early age is increasing and only some people can afford to go to college or benefit from college. Actually, what we see over time is that there hasn't been a change in inequality and academic achievement in the US. Actually, test score gaps have narrowed slightly since the 1970s for 17 year olds. And the reason for that is one reason that is this figure, which is that turns out since 1990, states have become more progressive in the way they distribute resources to schools. So actually, the poorest school districts, that's the blue line, districts in the bottom 20% of family income, spend the same amount per pupil as the richest districts. And that's kind of curious, right? But the reason for it is not kind of some accident of fate, it's that in states all across the country, you know, people sued school districts for violating the equal protection clause of the constitution and said, "Look, you need to guarantee equitable access to K through 12 schooling in your state." And so every state has its own system of redistributing resources, kinda like Robinhood, you know, taxing rich districts at higher rates and redistributing those resources for districts to equalize spending. So because of a policy choice or a series of court decisions and some policy choices that executed on them, we have pretty equalized funding at the K through 12 system. And therefore, I think we have inequalities and achievement that at least haven't risen and may have shrunk over time. What about higher ed? Not the case at all. So this is showing you the average expenditures per full-time equivalent student across colleges of different types. So if you look at research universities, that the line that go up really high, graph they're spending about $70,000 per pupil per year on an education. And quite a lot of that is devoted to instruction. Some of it's research and other things. But if you look at community colleges, which disproportionately enroll low-income students, they're spending about 10 to $15,000 per student per year on an education. And actually at Harvard, we're spending more than $100,000 per student per year on an education here, which I think it's hard for people to understand, but it's really true. You know, tuition bills are crazy at selective private colleges, but you're actually still getting a discount on the full resource cost of the education if you send your kid to one of these schools. And that reflects endowment accumulation, which allows schools to basically spend more than they charge year over year because of endowment returns. But only a few schools have endowments large enough to do that. So what you see is you've gone from a system K through 12 system where we equalize resources through policy to a system where the most academically prepared students go to the most selective schools which spend the most money. So it's really an inequality machine. And that's quite unique to the US higher education system. So that's why I think it contributes fundamentally to rising economic inequality. Okay, and so this is showing you, well, you know, what about the composition of students in them? This is from an older paper coming out of Opportunity Insights where we show that there's about as many students from the top 1% in Ivy Plus colleges, that's the Ivy League plus MIT, Stanford, Duke, and the University of Chicago. So those 12 schools enroll about 14.5% of their students from the top 1% of the family income distribution. And about 13.5% of students from the bottom half of the income distribution. Now because these are percentiles, they're all the same size. So that means you can actually compute the probability of attending an elite private college for people of different income ranks. And what you find is that if you are from the top 1% of the family income distribution, you are 77 times more likely to attend an Ivy Plus college than if you're from the bottom 20%, right? So we have highly unequal distribution of family income across colleges. And again, these are the colleges that are spending $100,000 per student per year. And the students at the bottom are going to community colleges or less selective four-year schools or no college at all. So they're receiving no resources or something less 10, $20,000 a year in resources. And what does that mean concretely? It means smaller classes, it means more attention, more counseling, more student support, all the things we do for students here at Harvard and places like it. We're doing that for a lot of wealthy students and not so much for lower income students, okay? All right, so why should we care about these institutions at all? Well, if you look at this top bar up here, that's an argument for not caring about them. Actually, less than 1% of all college students in the country attend an Ivy Plus institution, .8%, right? So this is a very small and stratified err of US higher education. So you might say, well look, and actually because of that, because it's only 0.8%, if we replaced every single student at every Ivy Plus college with a lower middle income student, it wouldn't make a dent in aggregate economic inequality, right? So in that sense, the number is small. But if we care about the composition of high earning and high status positions in society, because I think we do based on the amount of New York Times columns devoted to it, then we ought to care about what's happening in these schools. And this figure shows you that by outcome, so about, even though students at Ivy Plus schools are less than 1% of all students, they're about one in eight Fortune 500 CEOs, right? They're about one in eight of the top 0.1% earners in the country. They're more than a quarter of all MacArthur genius grant recipients. More than a quarter of all people who attend top 10 graduate schools in their respective fields went to an Ivy Plus college as an undergraduate. About a quarter of all US senators attended an Ivy Plus college as an undergraduate, about a quarter of journalists at major newspapers, half of all road scholars. And the Supreme Court justice figure is literally off the charts. We didn't have space for it in the chart. 71% of Supreme Court justices, since Thurgood Marshall, went to an Ivy Plus school as an undergraduate, not an Ivy Plus law school. Okay, so if we care about these positions and we think they matter for society, we should care about what's happening at Ivy Plus institutions. Okay, so that's my pitch for why you should care about what I'm about to say. Okay, so the question we're interested in this paper. I'm gonna give you a whirlwind tour through it. The paper is 140 pages long and has 53 main exhibits and a ton of appendix tables and figures. So there's a lot more where this came from. The question we're interested in is, do highly selective private colleges amplify the persistence of privilege across generations? And if they do, could they change that? Could they diversify society's leaders by changing who gets in? And in order to answer that question, you need to know two things. One is who gets in and why? And the one is, does it even matter? And so we answer both of those questions in this paper. And I'm gonna skip through a lot of the details. But one thing you might be interested in is, well, you know, maybe it's obvious that they do, but I think actually it's not. Because you could say, well, for example, that the reason why high income students attend Ivy Plus college is 'cause they're more likely to apply. Or is it because they're more likely to be admitted once they apply? Or is it because they're more likely to matriculate? That is they're more likely to accept and offer permissions perhaps because they don't need financial aid or something. In other words, where in the pipeline do we see these disparities persist? And that implies very different solutions, right? We might wanna encourage more applications or we might wanna increase financial aid, but if it's admissions, then it's really, in some sense the easiest lever to pull, 'cause it just means that schools ought to be admitting more or low middle income students among those who are similarly qualified, okay? So again, I'm gonna skip through a lot of the details here, but we're gonna do something that I wanna focus on because it's very important in this paper and you see it in this figure, we're going to explicitly condition on prior academic preparation. In other words, we're gonna say in all of the things that we do that I show you among people who have the exact same SAT and ACT scores. And the reason we're able to do that is 'cause we've assembled a dataset, which is the universe of all SAT and ACT test takers going back many years as well as data from US tax records. So we can link children to their parents, because parents claim kids on their tax returns and we can track them into adulthood. So we know your family income when you apply to and are admitted to college. And then we can track you into your early 30s in some cases and see what people are earning after they graduate. And then finally we've taken, we've assembled data from more than 400 colleges and universities across the country, including several Ivy Plus colleges who ask not to be identified. And we've matched that directly to the tax records and to the ACT and SAT data. So we have a really comprehensive picture of who attends these schools. And then for a subset, we know exactly how they were admitted, what their characteristics were, whether they were legacies, whether they were athletes, what their ratings were, their holistic ratings, their guidance council recommendation ratings, pretty much everything that a college is considering when they're thinking about admission, okay? So what we do in this figure is we say among people who have the same SAT score. So for example, I could show you this figure for people who have an SAT score of 1500 or an SAT score of 1450 or something. So for any individual score, what is the probability of admission for people at each of the parent income percentiles on the horizontal axis, right? And so if it's totally flat, that means basically for anybody who has the same SAT score, your chance of getting in are the same. And that actually is the pattern we see for public schools, believe it or not. And what this shows is that among people who have the same test scores, they're more likely to get in if they're poor and then especially if they're extremely rich. And the people who are least likely to get in, again, among those with the same test score are people from what we would call the college applying middle class. So the literal middle class is the 50th percentile, of course. But people who apply to college tend to be higher income, so 70 to 80 turns out to be about the middle class, the median among people who applied college. So it looks like there's a slight thumb on the scale for low income students who have high SAT scores, but a much bigger thumb on the scale for people from the very top of the income distribution. How big? Well, if you are from a family in the top 0.1% of the income distribution, that is a family with annual income of $2.9 million or above, you are about two and a half times, almost two and a half times more likely to be admitted, given your SAT score, than somebody from the middle class. We do that by 2.17 divided by .88, okay? So, and the reason why this says relative admissions rate is 'cause we're balancing across a bunch of different schools. So we norm it relative to one. So one means equal to the average person in the school, okay? Again, a lot more detail in the paper. So, and I'm not showing you the lines. We could show you the same line for application or for matriculation. Those lines are basically flat. So it looks like most of the reason for the disproportionate representation of high income students is due to admissions policies, not to application or matriculation. And so we can quantify that, we can say, look, in an average class, Ivy Plus plus class of 1,650 students, how many, quote, unquote, extra high income students do we have? Turns out that number is about 157. And so, sorry, yeah, 107 that's right. Could sometimes it's 158 depending on the merchant paper, so 157. And we say, look, you know, we kind of decomposed this into different elements of the pipeline and it turns out that about 2/3 is attributable to admissions policies. You'll notice that we have non-athletic and athletic broken out. Why is that? Well, athletic admissions is very different. We're talking about recruited athletes here. So that happens through a very different process. People apply only after they've essentially been informally accepted by the coach. And so the process looks really different and the way it works on the inside is not the same, and so we just kind of bracket the athletes and look at the non-academic wants. And so what we say is that, so we say, look, about 2/3 of the total extra representation is due to admissions. And then again, I'm skipping through all the details here. We think basically three factors account for that. The first one is athletics, I already mentioned that. And why is athletics important? Well, it's not because of athletes per se, it's that the athletes at Ivy Plus schools are higher income than the average student. So to make that precise, recruited athletes comprise a bit more than 10% of the average class in an Ivy Plus college, 10 to 15%. But the highest income students, among the highest income students, more than 15% are athletes. And among the lowest income students, only about 5% are athletes. And so because recruited athletes are disproportionately high income, they contribute to this extra representation of high income students. So athletes are about 16% of the total. The other part is legacy admissions. And this is the thing that's been covered in the news the most. And again, it's not because of something inherent about legacies, it's that legacies tend to be higher income. So when you have an overall admissions advantage for legacy students, you end up letting in more very high income students. So what we do is we simulate, well, what if you basically looked at every other aspect of a legacy applicant's portfolio, but you kind of turned off the legacy switch. So we basically predict the probability, somebody admitted as a function of everything we see about them and then we kind of flip legacy on or off and we see how much extra of a tip legacy gives. It turns out to be a lot among people with similar academic credentials. Legacies are about four times more likely to be admitted. And because they're higher high income that feeds through into this increasing representation. So the legacy preferences are about 30% of the total. And then the last share is what David Leonhardt in the episode of the daily called private school polish, which is a great term, I think, this is why he's a journalist and I'm an academic. It's a great term for this thing we call non-academic ratings. So what we find is that among students with very high academic ratings, you're not more likely to get in if you're high income. So academic rating that line I showed you looks pretty flat. But for non-academic ratings, and this is things like extracurriculars, the personal rating, other things, things that are anything but academics. And again, academic grad is a holistic rating. It's not just test scores or GPA, it's a sense of how difficult your classes were and all that kind of stuff. It seems like people, high income applicants are especially likely to get high non-academic ratings. And most of that is driven by high non-academic ratings coming from a small set of pretty elite private high schools, non-religious private high schools, often in major US cities. And it looks like students from those places are getting really high, teacher ratings and guidance council ratings and they seem to, you know, show up in these non-academic measures. So that's another 20%. So it's those three factors that make up almost the whole thing, okay? Now I wanna talk about the impact, okay? But before I show you the results, again for reasons of time, I'm not gonna get into all the details, but I'll give you intuition. So we're interested in the question of, well, if you go to a place like Harvard, does Harvard make you into a future leader? Or were you gonna be a future leader anyway? And Harvard just happens to claim the credit for you after the fact. So in other words, are these schools just identifying incredibly talented people who are gonna succeed wherever they go and skimming the cream off the top? And it's hard to answer that question just looking at data because the highest talent applicants can go wherever they want and they'll often choose to go to a place like Harvard. But what are we actually doing for them? What is the value added of Harvard, in other words? And the way we study that, we a couple different ways, but the main way we study it is by looking at wait lists, wait list applicants. So it turns out that almost all these colleges maintain a pretty serious sized wait list and most of the people on the wait list eventually get rejected, but a few get accepted. And it turns out, and we showed this in the paper, that the people who get accepted off the wait list and the people who get rejected off the wait list look very, very similar on academic ratings, non-academic ratings, test scores, GP, basically anything you can look at. And so one thing we do is we say, well look, if we think they're pretty similar, why are some people getting in and some not? And we think, and this comes in part from a lot of conversations with admissions folks sitting in on some meetings in some cases, that when you get to the very end of a class, you're often filling it based on idiosyncratic needs that are probably not related to future earnings or future graduate school tenants. So one example, this is a true story. We're sitting in on a meeting and a few candidates were being evaluated. It was the very end, maybe only one or two are gonna be accepted. So it was a question of who's gonna get in. And it turns out that one of the applicants was a pretty good oboe player, not a great like Carnegie Hall oboe player, but a pretty good oboe player. And the orchestra at this institution needed an oboe player 'cause the senior oboe player had just graduated. And so this student was admitted and we didn't actually see this, but you could imagine there's somebody else who's the same in all other respects except they played the saxophone. But we don't need a saxophone player so they don't get admitted. And so that's the kind of variation we're using when we look at the wait list. And the way we test for that is we say, "Look, if the reason you got in is 'cause you played the oboe instead of the saxophone, then your admission to one school should not predict your admission to another school." In other words, the oboe player at school A, let's say that school is Stanford, the oboe player who gets into Stanford because they need an oboe player, shouldn't be more likely to get into Yale than the saxophone player, right? It should be totally idiosyncratic. And so we use the fact that we have data from multiple schools and we ask this question directly. We say, does admission off the wait list to one school predict admission to other schools? And we find that it doesn't at all. And so that gives us some comfort and we do a bunch of other tests to try to convince people. That's kind of the intuition for it. And then we use some other approaches that other people have done in past work and find basically the same answers in all cases. And so we feel pretty confident that we've identified the true impact of being admitted just narrowly to one of these schools, okay? And then finally, we ask the question, well what is the impact for you if you get into a place like Harvard versus going to a place like a state flagship school, like for example, Ohio State where I went as an undergraduate? So what is the impact of, you know, let's say David Deming applied to Harvard when he was 17. I didn't, but let's say I did. And in one state of the world I was admitted off though I was wait list and once they were like, I got admitted off the wait list, another state I got rejected, and then I went to Ohio State instead 'cause I went to high school in Ohio. And so that's the comparison we're interested in. So here's what we find. And again, this is much broader than Ohio State higher And I'm gonna use that comparison just to make it more vivid in our minds. So if we look first of all the left hand charts, sorry, columns, showing you three things. One is the average income for people who attend to flagship public school. On average, they're in the 76th percentile of the income distribution. That's pretty good. That's upper middle class, right? The second bar is our estimate of what a person's income would be if they were the type of person who almost got into Harvard but didn't and went to Ohio State instead. And that bar is higher because you wanna think about it, it's like, well maybe David Deming was more qualified than the average applicant to Ohio State. And so if he went to Ohio State, he wouldn't earn the average, he'd earn something greater than the average. So we estimate he would earn 70, he would be in the 77.6 percentile. So 1.6 difference. But then the third bar says what would happen if David Deming went to Harvard? And then we say, well, he'd be in a 79 percentile. So that's a modest increase of about one and a half ranks, right? That's about five, six, $7,000 a year. So it's not nothing, but it's not overwhelming. So we find a small but not that large impact on mean income of being admitted to an Ivy Plus institution relative to going to a state school. But we find a very large impact on the probability that you'll be in the very top of the income distribution. So that's a second set of ours. This is earnings in the top 1%. We estimate that David Deming, again in my example, has a 10.4% chance of having income in the top 1% if he goes to Ohio State and a 15% chance of having top earnings if he goes to Harvard. All right, so that's an increase of about almost 50% in the probability that you'll be in the top 1% if you're barely admitted to one of these schools. And we see that directly in the income data. The people who are admitted off the wait list are much more likely to be at the very top. We also find there're about twice as likely to attend to top 10 graduate school. And they're about three times more likely to work in what we call in a leader of prestigious firm, which just for purposes of time I'll summarize by saying it's the type of company in various sectors that if you look up rankings, like the vault.com rankings of top consulting firms or US news rankings of top hospitals or top research universities. You're more likely to work at one of those very prestigious firms. You know, Goldman Sachs if you're in finance or McKinsey if you're consulting, and so on. So we find a very large increase in the probability of work for one of these firms, okay? Now the last thing we do, and then I'll take questions is we ask, well, given that there's such a large impact on top end outcomes, which by the way is very consistent with that first figure I showed you that these schools don't have that large of an impact on average income. It's there, but it's small, but a very big impact on being in high earning and high status positions in society. So we simulate what would happen if you changed the rules for who got in and why. And so we find basically to give you a shorthand, is that if you did the following three things, if you removed legacy preferences, and if you removed this extra preference for people with high non-academic ratings, and if you equalized athlete share, now that doesn't mean get rid of recruited athletes, it means recruited athletes have the same income distribution as everybody else. So you no longer have more high income athletes. If you did those three things, you could basically make admission to these schools much more representative. So with that number I showed you beginning was 157. By doing the three things, you could get 144 out of the 157 for low and middle income students. So you could do most of those things by changing those three policies, which would not be easy to do. There's another way you could get there though. You could actually keep legacy preferences. You could keep preferences for private school polish and you could keep the high income tip for athletes. If instead, you practice what we call need affirmative preferences for students with high academic ratings. So what we're simulating here is saying, look, and it turns out, I didn't show you a figure on this, but it turns out by the way, that having a high academic rating strongly predicts your benefit from going to one of these schools. So what I mean by that is if you look at legacies to attend. So in other words, is the impact of attending an Ivy Plus college greater for legacy students? No, it's about the same as for non-legacy students. Is it greater for students with private school polish? No, not really. It's about the same. Is it greater for athletes? No, it's about the same. It is however much greater for students with very high academic grading. So people who come in having gotten very high test scores and very high grades and taking very difficult classes, if they're given the opportunity to attend a school like Harvard, they benefit the most from it. And so we take that insight, we say, look, there's a lot of academic talent out there and a lot of it is in low and middle income families. So if instead we said, "Okay, look, we wanna take very talented low and middle income students, academically talented students. They're gonna do really well when they come here." And so because of that, we could actually keep all of the preferences that I mentioned, but put more of a thumb on the scale for low income academically talented students or middle income ones. And we could achieve more income diversity without sacrificing anything in terms of producing future leaders. So two ways to get there, one is getting rid of affirmative action for the rich. The other one is practicing a bit more affirmative action for low and middle income students with respect to academic talent. So my conclusion is it turns out that because all of it's about admission, at the end of the day, what's happening in 12 admissions offices across the country could have a meaningful impact change in that, could have a meaningful impact on socio-economic diversity. We could do that without sacrificing much in terms of future outcomes and future leadership, why? Because the country's a very big place. Harvard could fill its class five or six times over with students who have perfect grades and SAT scores. So no one's suggesting we do that. But because there's so much talent out there, you can make lots of changes without sacrificing the essential character and the commitment to academic excellence. There gives a way to increase economic diversity at Harvard without making a sacrifice. And that could happen in a handful of colleges. And because they're so important in producing society's leaders, it can make a real difference. Thanks a lot.

- Thank you, David. That was incredible. We'll open up this session now for questions. Please to ask a question, use the virtual hand raising feature of Zoom over there on your little Zoom bar of options. And in through Kennedy school fashion, please keep your question brief. Please end it with a question mark and we all would appreciate it if you would identify yourself and with your Kennedy School affiliation. You'll be notified via Zoom's chat feature when it's your turn to speak. And please be sure to unmute yourself when you hear from the staff. And in the meantime, while you're getting those questions ready, which we're excited to have, I'll start things off by asking one that was submitted earlier by Suzanne Anthony, Suzanne has an MPP 2009, and David, she wants to know what obligation do universities have to make their admission selection process more transparent? And if they did this, how might it level playing field of selection?

- Thank you for the question, Suzanne. I don't know what obligation they have, I mean, so legal obligation, none. These are private institutions that can sort of do things how they wish subject to the constitution. Moral obligation, I think some. And the reason is because I think that holistic admissions, like I understand the reasons why it exists, but I think it's become very socially damaging. This is just my personal view, having a kid who's in high school, maybe in a slightly informed personal view, but I think the idea that admissions is some kind of secret formula where you never know when enough is enough and you can't really point to a reason why one person who looks pretty similarly qualified gets admitted and another person who looks similar doesn't. I think that's poisonous for society, at least for people who care about these things because you have a sense of like, I need to just strive and strive and strive and the little details about how I write my personal statement could make a difference. And people are wondering, you know, does that person really deserve it? Do I deserve it? And I would love it if we could as a society come to a more transparent and fair way of doing admissions. Not because, not really for like, I guess I would call efficiency grounds, not because I think Harvard and schools like it are admitting the wrong students. I think a lot of our students are really talented and they deserve to be here, but so do lots of other people who don't get to come. And I would, because there's just no way we can be fair about it. I would prefer to make it more transparent and possibly even random. I would support, it'll never happen. But I would support saying, let's set a very high academic standard at a school like this. And then everybody who meets it gets a lottery ticket and we run a lottery. And some people get it and some people don't. Michael Sendell has proposed that too. Other people have proposed it over time. So it's not an idea that's unique to me. And again, I know it'll never happen, but I think it would be better for the health of society and the health of these conversations because if I knew that the reason you got in and I didn't was because you got a lottery ticket and I didn't, I wouldn't feel like, you know, you're more deserving than me, I wouldn't be jealous or envy. I'd say, well that was life and I got in somewhere else. So again, I have no illusions that this is a realistic proposal, but I wish that that were the world we lived in.

- Thank you for that answer. I have another pre-submitted question and I encourage people to pop up their hands. This one is from Adam Brummer. Adam has a mid-career MPA from 2022. And I'll summarize it, although Professor Deming got to see this question earlier, which is that Adam is interested in creating more diverse national leadership and advocates for admission practices like MIT's portfolio assessments and some alternative measures. For example, the Posse Foundation's Dynamic Assessment Process. And he wants to know, is there a world where the Ivy League could make their admissions policies more in line with these approaches and share applicant pools to create greater efficiencies for selecting their students. A variation perhaps on a lottery there.

- Yeah, thank you for the question, Adam. I mean, I think that'd be interesting. One thing that I think it's important to keep in mind is that the patterns, we identified this, you know, the influence of recruited athletes and private school polish and those things. I don't think that those are, it's not really about those things in particular. In other words, it's not like rich people are better athletes when they're born or something. What's really happening under the hood is that if you have disposable income and you have the means to do it, there's almost nothing you'd rather buy with it than an opportunity for your children. And we all feel that way, I think. You know, I feel that way for my own kids and I know everybody else who has kids does too. You want a better life for your kids. And so if you have means, you'll do almost anything to try to use it to get a leg up. And so what you see is knowing that there are only gonna be so many kids from your hometown who get to go to a place like Harvard or Stanford or something. You see recruited athletics as another way. And so from a very young age, you have your kid enrolling in club sports, travel teams, private coaching, playing the kinds of sports that not everybody plays that you need to access to special equipment for, and so on. And so again, it's not something about athletes in particular, it's that people will find whatever advantages they can get in a system and they'll use them 'cause they care about this outcome so much. And so I say that because my concern with things like portfolio assessments, others, whatever we pick, people are gonna find a way in. And that that's gonna be related to privilege and resources. And so I just think we have to be more mindful of it. My proposed solution to this would be that sunlight is the best disinfectant. I would like to see standard reporting for all US colleges of the income distribution of admitted classes every year. In the same way that we report race and gender and other characteristics, I'd like to see us report income and I'd like for each college to feel accountable to the public for that distribution and to find their own ways around it. Some of them might wanna get rid of legacies, some of them might wanna change the way they do athletes, some of 'em might wanna do something totally different. You know, I don't want it. Every college has a very hard problem they're trying to solve and I would rather just put more sunlight on this process and hold schools publicly accountable for economic diversity the way we hold them accountable for other elements of diversity.

- We now have a participant ready to ask their question and I'd like to invite them to identify themselves and state their affiliation with the Kennedy School and ask the question. Okay, just know the Zoom name was HKG.

- Hi HKG, what's your question?

- [HKG] Hi, you know, I literally, you I pretty much answered my question. It was about recruited athletes 'cause I didn't understand why they would have more disposable income than others. And actually, I would like to ask this question doesn't complete it. If it is the higher income people who are the recruited athletes, why are the recruited athletes such a huge thing at places like Harvard where a lot of people, I have a high school student as well, you see the kids, you know, racking their brains out to get a studying like crazy. And these recruited athletes just simply do not have the grades that other people do. So why is it so important, especially let's take football out of it, especially if you could get a more socioeconomic diversity, you know, by changing that. Why is that still there?

- Yeah, thank you for the question. So you know, I think it's really a question of mission and values, right? So what is the school? What are we trying to do? And I think, that's a question that like, just 'cause I'm a, you know, egghead PhD who studied this doesn't mean I have any claim on the truth for that relative to anybody else, 'cause it's a question of collective values. I'll give you my view, but I wanna be clear. It's just my view. I think it's probably gone too far. Again, in my view. I think athletics is important for well-roundedness. So like, you know, my kids are on sports teams and I think it's important. It does build skills too. There's a lot of evidence that being on a sports team builds teamwork skills and leadership skills and those are increasingly important in the economy. That's a big part of my other work that I'm not talking about today. So I do think that participation in group sports is valuable intrinsically. However, I also think it's not, you know, having a great squash team is not exactly aligned with a mission of a research university. And so I do think one could, again, not get rid of recruited athletes, 'cause I think athletes are an important part of our student body. And I love our athletes. I just think why not make sure, again, through sunlight, through reporting, make sure that the students, like, in other words, let me say it this way. I think if the distribution of athletes, the income distribution of athletes was the same as the rest of the school. I don't think anybody would really care about this. It's the fact that it's a backdoor into some process that gives an advantage to high income students. That's the problem. So if we just fix that, I don't think we would need to change much else. So that'd be my solution. Although, I do accept the premise of the question, which is what does this really have to do with the mission of a research university? I hear that.

- Know a number of people are chatting us with questions and we'll wait for those to come in. But I'll go to one pre-submitted by Armando De La Libertad. Armando has an MPP of 1996 and he writes the college educated Latino community in the US, and in particular, Mexican Americans have gained some financial stability. However, because their children are long, longer from a low income background necessarily, they may not benefit from diverse outreach that may target more hard to reach populations. As a result, the socioeconomic ascendancy of the next generation may be stunted despite real barriers that might still face this population. And he wants to know what can be done.

- Thank you for the question, Armando. That's a broad question. I think this is a place where what happens at elite colleges probably doesn't have much to do with, you know, it's probably second or third order relative to broader policy solutions. I think what you're talking about, Armando, if I understand correctly, is how to improve economic mobility more broadly for immigrants, for first and second generation, Americans from Hispanic or Latina origin. And I think that's really important, but I think it's a broad suite of things starting much earlier in the pipeline, you know. Ensuring, for example, more funding for English is a second language programs in schools. I have some students working on promising ways to do that. Thinking about representation. So how do we get more folks like that to the Kennedy School? That's been a big priority for the dean and I know a big priority for Ariadne and her colleagues and something I'd like to see continue to happen. So I think it takes a village kind of solution. I do think that we need more representation of those communities at Harvard, but I would view that as like an outcome, not an intervention, if that makes sense. I would view that as a measure of the success of interventions that happen much earlier in the pipeline.

- Understood, I'll go through another piece minute question. This is from Eric Elba. Eric has an mid-career MPA of 1982, and he wants to know what obligation universities have to make their admissions selection process more transparent. And if they did this, how might it level the playing field?

- Yeah, thanks for the question, Eric. I think I sort of answered that question already. I could like repeat, But I think, you know, transparency is an important value to me for moral reasons. I would like for people to have a clearer sense of not just why they got in or didn't get in and somebody else did, but like, at what point is it enough? Okay, at what point have I established that I'm the type of person who could succeed at a school like this and then let's leave it to chance. And I think there'd be some ethical value to having a system like that. So that's not quite transparency. Transparency would be like, you know, you can observe the process and figure out. I think if you did that, you'd find out that once we get down to wait lists, people are just throwing darts at a board. Like nobody wants to think they're picking randomly. But I think, you know, there's so much talent out there and how do we know when somebody's 17 whether they're gonna be able to take advantage of a place like this? You're never gonna get it all right. And so I wish we could be more forthright as a society about that.

- David, I wonder if you would share a little bit about how you got this information. I'm particular, I noted in the paper that you created the CLIMB initiative with some others. And which is a partnership among researchers. And I wondered if you could talk a little bit about that and why you had to create it.

- Sure, yeah, thanks for the questionnaire, Ariadne. So Raj and John and I in 2017 decided to start working on this project. And we spent the better part of, I would say, two years almost convincing college presidents, chief data analytics officers, human resources, basically c-suite people at the universities that became our partners to share data with us. And we pitched it as, you know, we're gonna link your data to tax records. We, in many cases, gave them some information that they wanted to know. And one of the things that is important to know about our study is that we have better income data than the colleges do because they know who applies for financial aid and who gets federal aid, but they don't know be anything more than that about people's incomes except it be indirectly, you know, if you list your parents' occupation or if they know who you are or something. So, but so colleges have a very limited ability to tell the difference between a family that earns $300,000 a year or $3 million a year or $30 million a year. They don't really make those distinctions. And many of them were interested in it. They wanted to know about outcomes too. They wanted us to link their student data to things like college enrollment. And I should say we have that for every college, not just our partners. Because if you've ever paid college tuition, you get a form sent to you by the IRS called the 1098-T, which is a form certifying that your dependent has paid, you have paid tuition on behalf of your dependent at this school, and therefore you're eligible to claim a tax credit. So because of that, we actually have a record of college attendance nationwide in the data. And so some of our partners wanted to know more about who was going to their schools and why. So we kind of used that as a way in and, you know, we asked a lot of people to share data. Nobody ever says no, but they just drag their feet and never say yes. So we had a bunch of people who didn't give us data, but we didn't need everybody. We got eventually more than 400 colleges and universities to share data with us. That's about, and they encompass like three and a half million students a year. It's about 15% of the undergraduate population. So we got a lot of data and it was, you know, like a boulder rolling downhill. We had a few early partners who were very committed and then once we told them, "Hey, that University of Texas is in and the University of California system is in." They said, "Oh, well we better be in too." So there's a lot of retail politicking and that stuff works. It just takes a while, so.

- Well, it's an incredible accomplishment. I wanna introduce David Rosenberg, who has a question.

- Hi, David.

- [David] Hi, David. It is great to see another Ohioan who's done so well here.

- All right,

- I think you probably discussed this in your presentation, but it was going pretty quick. So I wanted to just hone in on a couple of points. Would a proposed change to the academic rating factor? Just displace the type of advantage toward tests, elite prep schools and so forth, you know, so, so are we really changing the characteristic or we're just kind of redefining the way that income works to advantage some over others? And then just related to that is, you know, what effect did you see in the likelihood of application across income levels? So I think you hit those, but it'd be good if.

- Yeah, sure. Yeah, no, thank you for the question, David. So let me answer the second one first. We found almost no difference in the probability of application by family income among those who were equally qualified. Because I think that's something that used to be an issue. So there's a lot of work done by institutions like the college board trying to convince low income academically talented students to apply. And actually that stuff happened like a decade ago and I think it worked. Also, many more schools use the common application than before. So if you take it, you know, if you take the SAT today and you are low income and you get a high score, you just get tons of marketing materials sent your way by the most selective schools in the country. And I think that kind of outreach, particularly at Ivy Plus schools, which everybody knows about, they have a kind of national market. I think most people are aware of those things at this point. And so we didn't find that applications were very important to contributing to the overall cap. Knowing on your, sorry, what was your first question again, David? I'm sorry.

- [David] It was really about whether changing the academic rating factor.

- Oh, right, right.

- [David] Just displace it to other advantages.

- Yeah, that's right, that's right. So one of the reasons why my preferred solution is reporting of income is precisely because I don't see any one of, even though we identified these three factors, legacy, private school polish, and athletes as being the contributors to the high income, disproportionate representation of high income students, that's a statement about the past. But I'm mindful that whatever, however the system changes, people with means and with the wherewithal will adjust their behavior and response. And so I don't view any of those. For example, like I'm not gonna defend legacy preferences, but I don't think if all we did,. If we outlawed legacy preferences tomorrow and did nothing else, I don't actually think that'd make a very big difference in income representation at these colleges. And that's because, you know, most of the high income students at places like Harvard are not legacies. And not all legacies are high income. And I think what you'd see is you would, no longer admit a bunch of high income legacies and you instead would admit a bunch of high income non-legacies for other grounds. And so I don't really see any of these things as being solutions by themselves. That's why I would like to have income data publicly reported at the school level and let everybody figure it out. I actually think these schools care deeply about their reputation and about their legitimacy in the eyes of the public. And so shining more light on that is the most effective and most direct way to make a difference.

- Thank you very much for that question and that answer. I should say that David is a proud mid-career MPA of 1986. I missed that in the beginning, but thank you again. If anybody would still like to ask a question, I would encourage you. I have one last question, which may be where we end. And that is that, David, you've worked on Head Start and you've worked on college, what do you take from what you've found in both these areas?

- Yeah.

- For economic mobility?

- Well, so I'm gonna broaden the question even further Ariadne, and say, you know, I've worked in all element, all ages, education at all ages. And one of the reasons I'm so focused on college and then the career transition is because I've actually come to believe that those are the most important sources of inequality in America. And again, it's not because the other things don't matter. They matter quite a lot. I just think particularly at the K through 12 level, I think the system is actually a net equalizer because most people go to public schools and the financing for public schools is equitable by law. And you don't see that in college. Not everybody goes to college. The people who don't go are low income, people who do go, you know, go to colleges that are strongly stratified by both selectivity and family income. And I think there's a little bit of that in early family policy. So I think Head Start is an example the other way. But I think most of the inequity in achievement and attainment happens very early and very late. And I think it's really hard to intervene with the family. And so, as a practical guy, at the end of the day, I think, well actually higher education and the workforce transition are places where it's easier to intervene in policy 'cause people are adults, and it's something where resources really matter and it's feasible to make changes. And so I've devoted my energy to that because I think that can make the biggest difference.

- That's great. I'm happy to say we have one last question. It's from Elizabeth Gibbons. Elizabeth has a mid-career MPA from 2019. And Elizabeth, please ask your question.

- [Elizabeth] Good afternoon. Really a segue question that was just asked. I'm really curious about the practical policy solutions that you would recommend to those who do work in the K through 12 space and any promising systemic reforms that you have seen that you think are particularly helpful in helping to close those equity gaps.

- Yeah, great question, Elizabeth, thank you. So in K through 12, I would pinpoint high school. I think that when I say about this, you know, equity and funding and like the K through 12 system is a equal force for equalization, I think that's especially true at the elementary school level and becomes less true over time for a variety of reasons. Some of it related to teacher workforce and like more specialized needs for older kids. And you know, wraparound services are more important as kids get older and they don't have to be in school and they're, you know, more autonomous, and so on. And so I think one thing that I'm really excited about is the number of high schools around the country that are doing early college programs, essentially allowing people to get college credit in high school for things. And sometimes it attend community colleges and take classes that are quite practical. It doesn't mean they need to stop with those practical classes, but I think particularly for boys, but for everybody really who's 15, 16, 17 years old sitting in a classroom and listening to a teacher, it becomes less appealing every day. And so if we can give people some exposure to hands-on learning earlier in life, again, not that they have to be in the trades, they can keep going to college, but it's a chance to, you know, not pay for it. But you're already in school, you're required to be in school, but you're starting to get a sense of college life. I think that's really promising and I would like to see a lot more of that. So that's one thing that comes to mind, but there's a lot of other things. Find more time, we can discuss it.

- Well, thank you so much for that question and that answer. And I'd like to thank everybody who asked questions, and particularly I would like to thank Professor David Deming for sharing this wisdom with us and just presenting such an amazing piece of research. Our next Wiener Conference Call on November 16th features Elizabeth Linos. Elizabeth is the Emma Bloomberg Associate Professor for Public Policy and Management, and she's faculty director of the School's People lab. And we'll look forward to having you back with us then.

- Ariadne, can I say one thing about Elizabeth?

- [Ariadne] Absolutely, please.

- Elizabeth, actually, in a couple of weeks, we'll be awarded something called the David Kershaw Prize, which is an award given to faculty under the age of 40 who make distinguished contributions to public policy and management. So she's one of our faculty that I recruited as academic dean and I'm very proud also, I, myself, was a Kershaw Award winner back when I was under 40, no longer true. But Elizabeth is an absolute star in the field and you will miss out if you don't come. So please come to see her. She's wonderful. And I'm just so delighted that we can claim credit for the award that she's gonna win at a Pam in a couple of weeks.

- Thank you, thank you for that endorsement and for congratulations on your recruitment there. We're glad to have her. Thank you all and have a good day, everybody.

- Take care, everyone.