Note: This episode of University Innovation Alliance’s Innovating Together Podcast originally aired on March 10, 2025, appearing live on YouTube, Facebook, Twitter, and LinkedIn. The transcript of this podcast episode is intended to serve as a guide to the entire conversation, and we encourage you to watch video of the keynote and Q&A. You can also access our summary, along with helpful links and audio from this episode.
Part One
Bridget Burns:
Welcome to another episode of Innovating Together Podcast. I'm your host, Bridget Burns with the University Innovation Alliance. I'm so excited to welcome you today because we have a fantastic new episode that is going to be breaking ground again with a new unveiling from the UIA National Summit. And I am so excited to share with you this video because it is featuring Dr. Raj Chetty, and he provided a keynote that has really gotten people talking and thinking differently about how we use data to drive social mobility and student success.
So, his keynote really provided an in-depth overview of where we are currently on the content or on the subject of social mobility. There's a lot of discussion out there and obviously he and his team at Opportunity Insights and at Harvard have been leading the way, but I think that what he did here was really distill what has happened, where we are, and the questions that we need to wrestle with as we go forward and try and advance social mobility for students regardless of their background.
So, I am just so excited to introduce this to you and I hope that you enjoy it. This first episode, we're going to share with you the actual keynote, and then the next episode, part two, is going to have a really engaging discussion and Q&A back and forth with him. So please join me in celebrating this incredible piece from the UIA National Summit featuring Dr. Raj Chetty.
Raj Chetty:
... with you all today. So, what I'm going to do is present what we're seeing in the data using large-scale data on specific changes that we can make in institutions like yours to increase economic opportunity, improve student success.
But before I get to that granular level, I want to talk about these issues at a much broader level, talking about the American Dream, because at some level I see the work that you all are doing as being fundamentally about advancing the American Dream. That may not be the way you think about your jobs on a day-to-day basis, but I think ultimately that's what it boils down to.
And so, to set the stage for that, let me start with this chart here which shows you what the American Dream looks like in historical context and at present. So, in a paper my colleagues and I wrote a few years ago, we set about to measure the extent to which kids in America actually achieve the American Dream of upward mobility.
So, of course there are many different aspects to the American Dream, but one key aspect of it, certainly what drew my own parents to come to this country – and countless others – is the idea that through hard work, any child can move up in the income distribution relative to their parents.
And so, what we're doing here is just measuring the extent to which America actually lives up to that aspiration. We're asking what fraction of kids went on to earn more than their parents did, measuring both kids’ and parents' incomes in their mid-30s and adjusting for inflation. And we're looking at that data by the year in which the child was born. And you can see that for kids born in the middle of the last century, it was a virtual guarantee that you were going to achieve the American Dream of moving up. 92% of kids, by our estimation, earn more than their parents did among those born in 1940.
But if you look at what has happened over time, you can see that there's been a dramatic fading of the American Dream, such that for children born in the 1980s who are turning 30 around now when we're measuring their incomes as adults, it's become a 50/50 shot, essentially a coin flip as to whether you're going to achieve the American Dream.
So, this broad trend is, of course, of fundamental interest to economists like myself, because it reflects a fundamental change in the U.S. economy that we'd like to understand. But I would argue it's also fundamental political and social interest. I think it's this very trend that underlies a lot of the frustration that people around America are expressing that this is no longer a country where it's easy to get ahead even through hard work.
And so, motivated by that, in our research team at Harvard, Opportunity Insights, we're interested in understanding what is driving the fading American Dream and, more importantly in relation to what Bridget said in her remarks, what we can do to create more social mobility, to create opportunity going forward.
So, what I'm going to do is share a little bit about what we've learned about the drivers of economic mobility and why we think institutions of higher education can be a key lever to create more mobility going forward.
So, the first step in understanding what's going on for us was a study that we also – so happens – released 10 years ago, back in 2014 when UIA was started, where we started to look at these data, not just at a national level, but starting to break the data down in various ways. And I'm going to start here with the geographic breakdown of economic mobility in America in this map. So let me first describe to you how we construct this map and then tell you what I think we learned from it. So, what we do here is take data on about 20 million kids, essentially all kids born in the late 1970s, early 1980s in America. We use information from anonymized tax returns to link those kids back to their parents and back to the specific area in which they grew up. And then we divide the U.S. into 740 different metro and rural areas. And in each of those areas we calculate a simple measure of upward mobility.
We ask, if you were a kid growing up in a low-income family, which I'm going to define here as a family making about $27,000 a year, which puts you at the 25th percentile of the U.S. national income distribution, what is your average income when we see your 1040 tax return when you're 35 years old?
So for example, if you look at the darkest red-colored parts of this map, you can see that the kids who are growing up in low-income families there, one generation later, they're making only $20 or 25,000 on average, actually less than their parents were making on average in the previous generation, which is kind of shocking if you think about it, given the amount of progress that has occurred in the U.S. over the past 30 years. On the other end of the spectrum, if you look at the blue-green colors, like much of the rural Midwest, for example, you can see that in much of rural Iowa, kids growing up in families at that exact same income level, on average, they're making $50 or even 55,000 a year in the next generation. So, a substantial amount of upward mobility within a single generation.
So, this map shows you the geography of opportunity in America, and it shows you that kids' chances of rising up, very sharply. Even in the present day, there are many places in America where the American Dream is well and alive, but unfortunately there are many, many others like much of the Southeast, many cities in the industrial Midwest and so on where kids don't have great chances of rising up.
So, these data, I think, are useful to ground the conversation because they help contextualize the settings in which we're working. And so, we can start to ask, from a policy perspective, what's different about rural Iowa or Salt Lake City, a city with very high levels of economic mobility, versus a place like Atlanta, Georgia or Charlotte, North Carolina, where kids have much poorer prospects of rising up. It's also very useful from a scientific perspective, because in a sense, these sorts of data give us for the first time a microscope to break down the aggregate national data that I started out with, and start to make comparisons between different places, different institutions to really understand the science of economic opportunity.
So, what I'm going to do in the next few minutes is share what we've learned about that science and then talk about what it tells us about interventions that can potentially be effective going forward, focusing in particular on higher education.
So, one key aspect of these data, which is very important in the context of the issues you all work on, and you might recognize if you just think about the demographic structure of the U.S., is that there's a connection to race here. So, if you look at the parts of the map that are in the red and orange colors, you'll notice that there are exactly the areas with the largest African American populations – the Southeast, places like Detroit, Cleveland, and so on. We all know that there's a long history of racial disparities in America. So how much of what we're seeing in this map is about place versus race?
So, to look at that, what we did next in a subsequent study was took the data from tax records and linked them to census data, which gave us information on everyone's race and ethnicity. And that allowed us to draw this pair of maps here. Exact same statistics that I was showing you before, but shown separately now for Black men on the left and white men on the right. Now, when you first look at these two maps, you might react by saying, "Oh, they've put these maps on two different color scales," kind of a red-orange color scale on the left and a blue-yellow color scale on the right. But in fact, if you look at the bottom of the slide, you can see that we have not done that. The maps are on exactly the same color scale, and what that's telling you is that there's an incredible difference in children's chances of upward mobility by race in America. Even conditional on growing up in families at the exact same level of income, Black men have much poorer chances of rising up across the entire country than white men do. It's basically like you have two non-overlapping distributions, the very best places for upward mobility for Black men, they actually have worse prospects there than the very worst places for white men. In that sense, it's almost like you have two different countries in terms of economic opportunity, even in the present day.
Now, I emphasize that because, as you all know, there have been lots of conversations about the role of race and ethnicity in college admissions. What these data highlight is that even at the present time, even conditional on income, there are enormous differences in outcomes by race. Something that I think we need to continue to pay attention to going forward in order to narrow disparities.
Now, one other point I'll note: you'll notice that when I cut the data by race, I also began to cut the data by sex, showing you the data only here for men. If I were to show you the exact same pair of maps for women, which you can look up on our team's website, you would see that they actually have very similar colors to each other. So, racial disparities really seem to intersect with gender. Racial disparities are strongest between Black and white men. Black women seem to have very similar outcomes to white women on average, controlling for family income. So, what we see here is that race remains incredibly important, but even conditional on race, there are big differences across places among white men. For example, if you look in the map on the right, there are much poorer outcomes in Appalachia and as we'll see in particular cities relative to other places. So, race matters, but so does place.
So, given this context, what I want to talk about now is what we can actually do to improve economic opportunity in the red-colored parts of the maps that I've been showing you. There are many different things you can do. Our team works on issues ranging from affordable housing to K through 12 education to preschool programs and things like that. I think a key aspect of where we can make significant progress is in institutions of higher education.
So, here what we've done is overlaid, on that same map that I started out with, the locations of the UIA – the current UIA partners, where they're physically located. And you can see that many UIA campuses are exactly in the red colored parts of this map, serving the communities where upward mobility is most lacking, which I think is a good thing. We're trying to tackle the problem where it needs to be tackled. It also – I think is very important for contextualizing the challenges you all face and other peer institutions that may be located in similar places where there's a broader context of a lack of social mobility. And this is the problem that I think you all can hopefully help solve, but it's also helpful to recognize that it's not emerging strictly from the higher education system. There are broader set of contextual factors that matter here.
So, to pick out one example: Georgia State, located in the Atlanta area, serving that community, Atlanta in our most recent data ranks 50th out of the 50 largest American cities in terms of rates of upward mobility. That's a city where kids have a particularly hard time rising up. And I think the types of programs that are being developed at Georgia State to really make a difference where they've made a tremendous amount of progress with Tim Renick, who you'll hear from later on, are the types of things that can change these trajectories that I've been talking about.
So, how can we do that more specifically at institutions of higher education? So, in order to study that, what our team has done over the past several years is built a really powerful longitudinal database that allows us to essentially track and analyze the outcomes of all college students in the U.S. So, we do that by linking four sets of data: data from tax records, which I described earlier; to data on college attendance from the Department of Education and certain tax forms that essentially allow you to see where everyone in America is going to college; to data on SAT and ACT scores, which is useful for certain selective colleges; as well as internal college admissions data for about 300 colleges around the United States. By combining all of those different data sets, you'll see that we can get a very rich portrait of how colleges are contributing to economic mobility. And here I'm going to focus on data for the entering classes up to 2015 or so. There's some limitations that limit us from looking at more recent cohorts, but our best sense, and I'm happy to talk more about this in the Q&A, is that these trends are not that different in the most recent years, although they may of course have changed in any given institution.
So, using that data, I'm going to start by showing you how colleges contribute to economic mobility with this chart here, which I think is a very useful way to think about the types of problems that colleges are trying to solve ultimately in creating social mobility. So, what we're doing here is representing each college in America. Each of your colleges is shown by a dot here, and I'm going to label some of these in a second. But first, just to show you what information is being shown, there are two key dimensions that I think you should think about in terms of understanding how your college is contributing to economic mobility in the U.S. to helping kids achieving the American Dream.
The first shown on the vertical axis is the upward mobility rate. So, what this is, is the fraction of students who come from low-income families and end up reaching the top 20% of the income distribution. So basically, what we're saying is among kids on your campus whose parents were in the bottom 20% of the income distribution, what fraction made the leap to the top 20% of the income distribution? What fraction achieved a substantial amount of upward mobility? So that's what's shown on the vertical axis.
The horizontal axis is a measure of access. How many low-income students do you actually have on campus to begin with? What fraction of the student body comes from parents in the bottom 20% of the income distribution?
So, to give you some concrete examples of where different institutions lie, I may take my own institution, Harvard University. Harvard looks great in terms of student outcomes. It does not look great in terms of access. And so, mechanically, Harvard cannot be contributing a whole lot to social mobility in the U.S. because it doesn't have that many low-income kids. So, obviously, you can't be helping lots of kids rise up if you don't have that many low-income kids to begin with. On the other end of the spectrum, we have a number of community colleges in the U.S.; we've highlighted a couple that serve a tremendous number of low-income kids but, at present, don't have the student success outcomes that are necessary in order to achieve upward economic mobility.
So, where did the current UIA schools lie? These are shown for selected schools where we have the right data by the green dots here, and you can see that they're sort of clustered in the middle. They have somewhat higher levels of access generally, as you'd expect, than the Stanfords and Harvards of the world. They have better outcomes than some of the community colleges I mentioned before. But I think, as you can see, there's also room for improvement on both margins, which is of course exactly why we are all here today.
And I'll note: for those of you whose colleges not shown on the slide, there's a very nice tool that's been created by the New York Times using our data. If you just look up “New York Times, College Mobility,” you can go type in your college's name, and you can look up all of these statistics for your own college that I'm going to highlight various colleges throughout this presentation. But you can see the data for your own college freely publicly available. And our team would also be happy to help if you have further questions.
Okay. So, with this background on how colleges are contributing to economic mobility, this both shows you that individual data for each college, it also frames at a system level what I see as the fundamental problem in terms of higher education and social mobility in the U.S., which is basically that there are no dots in the upper right side of this chart. It's basically blank. There are basically no colleges that both serve many low-income kids and deliver great outcomes in terms of their earnings after college. You need both of those things in order to have a high level of social mobility. We just don't have that at present in the U.S. And so, at some level, the way I see the work you all are doing is trying to create those dots in the upper right. That is what I think is going to fundamentally change the trajectory of social mobility in this country and reverse that fading American Dream that I started out with.
So, how do we go about doing that? With this visualization, I think you'd naturally think of two different strategies. The first is to move the dots that are in the upper left side of this chart to the right to improve access in schools that already have high levels of student success. Now, I want to spend a minute on that – because I think it's useful to think about, and it is a viable strategy for many colleges – before turning to my central focus, which is going to be on how to improve student success, how to improve outcomes at the colleges that already have high levels of access.
So, in thinking about how to improve access, I think it's, again, very important to understand the context in which you are operating. As I emphasized earlier, many institutions are operating in environments where there are many other structural factors at play, ranging from disparities in childhood nutrition to the quality of K through 12 schools, segregation in neighborhoods, racial disparities, and so on, that limit opportunities before you even come into contact with these students. And so, that is going to create challenges, not to make any excuses, but it's important to be aware of the context that you're working with in improving access. And so, just to make that very concrete, let me give you one example of how you see that playing out in the data.
Here's a chart showing the fraction of kids who score above a 1200 out of 1600 on the SAT plotted versus your parents' income. Okay. So, you achieve a relatively high SAT score that puts you near the top of the distribution competitive at many selective colleges. How common is that based on parent income? You can see that if your parents are at the bottom of the income distribution, it's virtually unheard of. 1% of kids with parents in the bottom 20% of the income distribution achieve an SAT score above 1200. If you look at parents at the top of the income distribution, that number is nearly 50%. Nearly half of those kids are scoring above 1200. So there's an enormously steep gradient here, which I think largely reflects – it's partly driven by test taking and test prep and things like that, but we think more fundamentally reflects the very deep disparities that exist in communities before kids even apply to college, which is obviously going to limit our ability to increase access without changing the selectivity of our institutions. If we're trying to admit students with, say, SAT scores in a certain range or certain qualifications, there's going to be a limit to how much we can do that without fundamentally changing the pool of students we're seeking to admit.
Now, that said, it doesn't mean universities are off the hook completely on the access dimension. I actually think, to the contrary, there's a lot that universities can do. And just to illustrate that in the data, I want to give you a couple of concrete examples.
So, what we're doing here is showing data for Michigan State, one of the UIA partners, where we're student attendance rates by parent income at Michigan State. So, what fraction of kids in the state of Michigan attend Michigan State versus their parents' income, but importantly controlling for their SAT scores? So, we're saying take a set of students who have the same SAT scores as the current Michigan State student body. How likely are you to attend Michigan State if you're from a low-income family, a middle-income family, a high-income family? And you can see here that relative to the average attendance rate, which is shown by the – line, kids from families at the 95th percentile of the income distribution – so, quite well-off families – they're nearly twice as likely to attend as the average kid. Whereas kids from the bottom 20% of the income distribution, they're only 60% as likely to attend as the average kid. So, there's nearly a three-fold difference in attendance rates between kids from high-income families and the lowest-income families at Michigan State. Critically, even controlling for SAT scores. So, kind of benchmarking it so all these kids have the same level of preparation at the point that they're applying to college, you're still seeing a three-fold difference in access at Michigan State between higher-income and lower-income kids. Clearly something that is in the scope of what colleges can address without changing selectivity.
Let's contrast that with another example, UT Austin, where you also see a gradient not as steep and a little bit more of a dip in the middle class. Now, what I think is very valuable to do, and I would encourage some of you to do, going back and looking at the data for your own institutions, which again is publicly available, is to ask: why is this happening? Where in the pipeline to college, to showing up on campus is this happening? Is it about kids not applying to these colleges? Is it about them not getting in? Obviously, that's going to matter for figuring out what we should do going forward, and so we can look at that data as well. Here are estimates of application rates, again, controlling for SAT scores by parent income for these two schools. And you can see that at UT Austin, this pattern for application rates with the dip in the middle class and then a surge afterward. It looks very similar to the trends in attendance rates that I showed you before, showing you that a lot of those differences in who shows up on campus at UT Austin are driven by differences in who's applying even among those kids who have the same SAT scores. In Michigan, the differences in application rates also contribute.
If you then look at attendance rates among applicants, so, think of this as basically your chances of being admitted to these institutions. At Austin, that's basically not a factor. That looks like a horizontal line. There's no difference by parent income and your chances of getting in, conditional on applying. Whereas at Michigan State, you're quite a bit more likely to get in and show up on campus if you're from a high-income family. So, clearly thinking about that margin is very important there as well.
So, I just wanted to give you these two examples. Again, you can look at various other colleges to show you how we can use data to really hone in on which parts of the pipeline we need to address in order to – to the extent we can improve access and give those opportunities to a broader set of qualified kids. So that's the first strategy that we can use to create more dots on the right side of the chart that I started out with.
The second, and I think in many ways more important, strategy is focusing on improving student success, which is of course what many of you here are interested in. So, how can we go about doing that? So let me share a few different pieces of data. I don't think I have the definitive answer, and this is part of what I hope will come out of events like this, but let me share a few pieces of data that I think indicate what kinds of things matter.
So, the first is to try to look at the few colleges that are outliers. You might notice that there are a few dots that are actually in the far right on this chart. What are those places? So, let's pick one of them.
Here's a dot that was in the far right. It's an unusual college that both has very high rates of access and has quite good outcomes. It's called the Vaughn College of Aeronautics and Technology. It's a small college in Queens that focuses – in Queens, New York that focuses specifically on aeronautics. So, why do I highlight this? It's a very small institution compared to the types of institutions you all work in, but I think there's a message in this one example, which is that often it's colleges or programs that are oriented around a specific career pathway. In this case aeronautics, which leads you to a pretty high-paying occupation, which many kids are able to get into and ends up providing them with a great career pathway, apparently, when you look at their tax records in subsequent years. So, that illustrates that the sort of modern vocational approach of equipping kids with not just a broad liberal arts education, which of course can have a tremendous value, but specific technical skills, we often see that those are the cases where you have the greatest impacts on economic mobility going forward.
So, that's borne out, not just in that example of that one college, but more generally what we see is what seems really critical is the quality of the first job that people get right out of college. So, to show you that, I'm going to show you a different piece of data from a study that some of my colleagues, in particular an economist named Jesse Bruhn at Brown University and collaborators have been working on, which shows you the importance of first jobs for people's future trajectories.
So, what they're doing here is taking a very interesting source of data. They're taking data on people who are in the military, where effectively you get assigned an occupation at the point of entry that depends in some ways on the test scores you get on the Armed Forces qualifying test. And so, there's some random variation essentially in which type of occupation you get assigned to in the military. So, think of a comparable set of kids. They've all joined the military, some of them end up working in warehouse operations, some of them end up working in IT, some of them end up working in infantry, et cetera. So, what they do is track these kids over time, again, using tax records, measure their incomes 15 years later. All comparable people because by the chance of a coin flip, they end up in one of these occupations versus another. You can see that they're enormously different career trajectories with more than 30% differences in earnings if you happen to end up in IT, which offers a much better trajectory than in warehouse operations. Really underscoring the critical importance of that first job you get out of college in particular, whether it puts you on a good career path.
So how can we equip more kids to get on those good career paths, focusing not just on completion of college, but actually getting to these careers that are going to change their lives? So, there've been a number of efforts to try to do this more systematically. Let me give you some examples of interventions where we're seeing strong records of success in randomized trials.
So, here's one. It's targeting a different population than what you all focus on. This is the Year Up sectoral job training program that actually focuses on high school graduates who haven't gone down the traditional college pathway. But I think there's an important lesson for all of us from this. So here, we're looking at data from people who participated in Year Up, this job training program in the green, and comparing them to a control group in a randomized experiment of people who didn't get access to the program just through the chance of a coin flip. And you can see that when you follow these folks’ earnings over time, again, using tax data, the people who participated in the Year Up program have systematically higher earnings for many years to come after they participated.
So, why is that happening and why am I highlighting this example? You might know that there's a long history of job training programs in the U.S. that have a record of being unsuccessful. Year Up and a number of new what are called sectoral job training programs differ from those programs in one critical respect. They really focus on social connection and social capital. They focus on not just teaching people things like IT skills, but connecting them with employers who want to hire for specific jobs and providing further mentoring and wraparound support. And this seems to make all the difference in terms of making them highly effective.
We see that kind of lesson, not just in job training programs, but also here's an example. On campus that I interpret similarly, suspect number of you are familiar with CUNY's ASAP, Accelerated Study and Associates Program that has been highly successful on that campus, again, in a randomized control trial. Here, they're just looking at rates of degree completion, and you could see, again, the people who got a chance to participate in that program are far more likely to complete their degrees than those who don't. There are many things going on in the CUNY ASAP program, but one key aspect of it is it provides those student supports, provides that social capital that helps people navigate the program and complete more successfully.
Here's a third example. This program called Bottom Line Advising focuses directly on access and success advising. Highly intensive advising both when kids are in high school and once they get to college. Again, has been evaluated using a randomized trial with 2,500 students. They find that without changing resources, without changing financial aid, anything else, just this advising program, increases college enrollment rates by 13% and, importantly, increases BA completion rates by 23% four years after high school. So again, an example of a program that effectively is providing social connection, providing social support, and seems to really change students' trajectories.
Now, those are examples of specific interventions that are trying to move the needle on that front. What I want to do now to conclude here is to zoom back out and show you how this theme of social capital and social connection matters, not just at this specific programmatic level, but more broadly in thinking about the issues of the American Dream that I started out with.
So, let me go back to the map that I showed you at the very beginning of this talk of upward mobility on the left based on tax records. That's exactly that initial map we looked at showing how kids' chances of rising up vary across different parts of the U.S. The map on the right which looks very similar is actually from a completely different source. This is a map that we drew in collaboration with Meta, the company that operates the Facebook social network platform. We set up a collaboration where we analyze data on 72 million users of Facebook and looked at various measures of social capital by county in the U.S. And what we're focusing on is one specific measure here that we call economic connectedness, which captures a very simple idea. All we're doing is measuring: if you are a low-income person on Facebook with below median income, what fraction of your Facebook friends have above median income?
In the map on the right, the blue-green colors are places where low-income folks are more connected to high-income folks. The red colors are places with more social disconnection across class lines. And you can see the result immediately for yourself. You don't need any fancy statistics here. It's totally obvious that the places in America where poor kids have the best chances of rising up and achieving the American Dream are exactly the places where there's more interaction across class lines, where there's more social capital, where there's more economic connectedness. That turns out to be the single strongest predictor of economic mobility that we or anybody else studying these issues over the past decade has identified. And so, that underscores, I think, the importance of thinking about how to create these kinds of social connections going forward.
Now, again, that can happen in many contexts. It could happen in your neighborhood, in your schools, in your religious institutions. Importantly, I think it can also really happen on college campuses. And so, here we're showing the same kind of plot that I was showing you before with measures of student success on the vertical axis, that upward mobility rate, what fraction of low-income kids reach the top 20% from the tax data, but plotted against these economic connectedness measures, which we were also able to construct for every college campus, and by the way, have made publicly available. So, you can look at how much kids of different backgrounds are interacting on your own campus by going to our team's website. And you can see that there's a strong link between these two things. It's the campuses where you do have more of that cross-class interaction where you're seeing kids from low-income families thrive more. Underscoring that this is not just about specific programs, I think trying to build the sort of social capital and cross-class connection connecting kids to opportunity can really make a broad difference in increasing student success going forward.
So, to conclude, what does all this mean in terms of what we can do to create more mobility going forward? I think what emerges is maybe a plan or a way to think about an evidence-based path to increase social mobility, which I hope all of you will participate in going forward in the next ten years. So, the key open question in my mind to be addressed based on what we've learned so far is, what interventions are most effective in creating social capital in particular as a pathway to increasing student success on scale?
I've shown you a couple of examples of interventions that, at a very specific level, have I think had quite a measurable impact. How do you do that on the very large campuses you all are working in? I think brainstorming about this as you all think about these issues in the next couple of days would be incredibly valuable. Now, I think it's important to do this as Chancellor Wilcox emphasized, not just promoting the ideas that you think are best, but also systematically evaluating what is actually working. And so, let me leave you with a couple of suggestions on how to do that concretely.
So, there's an issue here of time being a critical factor. It's going to take a while until you see the impacts of what you're doing; people are on these better earnings trajectories ten, 15 years after college. And of course, you don't want to wait 15 years to evaluate the impacts of what you're doing. So, I think there's a tendency to gravitate towards measures like college completion rates or how much are you making in the year after you graduate. All of that can be helpful and make sense, but let me suggest to you two other things to measure that I think could be very valuable to add to your toolkit.
First, as I was emphasizing earlier, what really seems to matter is the quality of that first job that people get, but critically, not just how much money they're making in that first job itself, but what the trajectory is looking coming out of that first job. So, what I would do is look at the historical record of people who had those jobs in different occupations. Look at how well they did in the past and see if people are being placed in jobs that traditionally, if you look at historical data 15 years later, have put people in much better positions. I think that can be a much more effective proxy for later success that can be measured essentially in real time.
Second, given the mechanism that I was emphasizing of social capital being really critical, I think understanding who your students on campus, particularly low-income students, are interacting with, who are they connected to, what kinds of opportunities they're thinking about, trying to measure that systematically can help you understand, maybe diagnose the pockets where you need to intervene the most. So that's in the short run. In the longer run, what I would suggest, and I'll leave you with this thought, is that it's very useful to think about how to build knowledge in the field by linking data from the types of interventions you all are doing and may think about doing in the coming years to information like what I showed you from the tax records.
And our team would certainly be happy to talk about that. If there are interventions you've done historically that you're particularly excited about, can we demonstrate that, ten years later, 15 years later, it's fundamentally changed students' lives? I think that would be an incredible way to build knowledge, to give us a scientific understanding of how to promote student success, and ultimately revive the American Dream for everyone. Thanks so much.
Part Two
Bridget Burns:
Welcome to another episode of the Innovating Together Podcast. I'm your host, Bridget Burns, with the University Innovation Alliance. I'm so excited for today's conversation, which is going to feature Dr. Raj Chetty from the UIA National Summit. This is a follow-up candid conversation with him where I first interview him about what he discussed, but also we have audience live Q&A. The conversation gets very rich, and we get a level of insight, I think, about how we can apply the concept of social mobility in new ways from the presentation that he provides.
I just want to first say this entire conversation is sponsored by Mainstay, which is a student engagement and retention platform that we know has proven that it works through empirical third-party research, which is something that is frankly too rare in the field, but is something that we value greatly, and we found that several of our campuses have had incredible results with them. Second, it's also supported generously by the Carnegie Corporation of New York, who has been a longtime partner with the UIA. But more than that, I also want to add that the Kresge Foundation uniquely supported Raj's presentation and his participation in the UIA Summit. We are really grateful for them. Thanks for giving me a chance to give some shout-outs for the support that enables us to do this kind of work. Please join me in a really rich conversation with Dr. Raj Chetty from Opportunity Insights.
I have a first question that comes anonymously. If you have my cell phone number, you may or may not have texted me a very fun question. And it was specifically around how ACE is going to be doing a new classification with Carnegie, and they're coming up with new ranking, or new ways to classify institutions. What specific... Have you seen what they're doing –
Raj Chetty:
Yes.
Bridget Burns:
– and what would you change to make it better so that institutions will actually improve on this?
Raj Chetty:
Yes. So, our team has been closely involved in conversation with folks at Carnegie, and we very much appreciate their rethinking how universities should be rated or classified. Traditionally, my view is those kinds of ratings have essentially had the reverse effect. If you think about many college rankings, setting aside Carnegie, having a more selected student body and not taking low-income kids actually can help you move up in the rankings, which are backwards from the perspective of thinking about how to create economic mobility.
The way the Carnegie folks are thinking about it, to my understanding, is exactly along the lines of that key chart that I emphasized of the two dimensions, upward mobility rate and access, and basically thinking about which universities are falling in the upper right side of that chart, which universities are really creating economic mobility bridges, giving those pathways for lots of people to climb from the bottom to the top of the income distribution. That's going to be one element of how Carnegie is going to classify institutions going forward, which I hope, and I'm starting to see signs with other more public ratings, U.S. News and so on, starting to put more weight on these kinds of mobility measures, which I hope will further basically reward the types of efforts you all are doing by helping those programs essentially help universities move up in the rankings.
Bridget Burns:
Great. I believe we have a question, or unless I can go to my secret text chat. Yes, that seat?
Aisha El-Amin:
Hi, thank you so much. Aisha El-Amin from the University of Illinois Chicago. Thank you for your talk and your work. I really was taken aback by your mapping the upward mobility heat map, and your positioning that race and place matters, and I absolutely agree. I'm wondering, when you talk about sectorial job training in the Year Up program, how does that overlay to that map?
Raj Chetty:
Yeah.
Aisha El-Amin:
You show that when they have that training, there's an upward mobility. But when you overlay race on top of that, is there any distinguishable factor there?
Raj Chetty:
Yeah. Thanks Aisha, great question. Let me say a little bit more about what we've learned on job training. So, I do think any one of these programs can make a sizable difference. We see that the Year Up program, for example, for Black kids, puts them on a significantly better trajectory. Does it completely close those gaps? And if we were to imagine a world where every Black kid had that access to that program, would the map for Black kids look the same as for white kids? I think the answer is no.
I think this is a complicated situation where there are many different things at play, and we need to have many different approaches to tackle these problems. In fact, the talk that I'm going to give tomorrow is to a completely different audience, folks interested in place-based investment. How do you transform communities and reduce segregation and so on, which I think can also be part of the toolkit, just like what all of you are doing in the higher education space. So, there's no silver bullet, but I also think there's no excuse, that doesn't mean – Sometimes people interpret that as saying, "Oh, I can't solve that problem by myself. All these things happened before kids got to college, somebody else should solve the problem." That's not the right way to look at it either. I think there are a bunch of things that matter, and we see that progress can be made in each one of those areas, but none of them is going to fix the problem entirely by itself.
Bridget Burns:
Dr. Kenneth Staples.
Steve Wuhs:
Thank you and good afternoon. I was really interested in your social capital argument. My name is Steve Wuhs and I'm at Oregon State University. Woo-hoo. I was looking at the maps also, looking at the upper Midwest and the Southeast, and I just wonder about levels of economic inequality, I wonder about residential segregation, and I wonder also about economic base when you turn to the argument about enhancing economic connectedness as a path forward. Are there limits to the viability of that strategy based on those underlying social conditions that would then shape how we as universities engage in the space?
Raj Chetty:
Yeah, great question. I think all of your observations are absolutely right, and in fact, if you were to read the paper called “Land of Opportunity” that analyzes those data carefully, which you can find on our Opportunity insights website, we identify a number of exactly the factors you pointed out on segregation, what's going on in the Midwest and so on. Let me just give you more of an anecdotal example that kind of captures the spirit of your comment. So, my wife grew up in a small town in southern Illinois called Salem, Illinois, which is near University of Illinois Urbana-Champaign. About 7,000 people, her dad was the town doctor. If you look at her network of friends, it was completely integrated by class, because there was basically no other game in town in terms of school you go to, who you interact with. There wasn't an opportunity to completely segregate by class, let alone by race as often occurs in bigger cities.
And I think that experience, to me, captures why if you look at the maps I was showing you, the rural Midwest in particular excels in terms of social capital, that cross-class connection, and also in terms of economic mobility, and you hear lots of stories like that. So, you are asking, I think, the right question of, "Well, we can't all be the rural Midwest. That's not a policy solution. What do you do with that?" But I think you can take something from that in terms of thinking about how you create that experience and imbue that logic, importantly, in the programs that we already implement.
So, I see many contexts in the U.S., both in the case of higher education and elsewhere, where the debate is always about how much money we spend on, say, financial aid, on resources for a housing voucher, on many different things like that. Obviously that matters, but we find in many cases that if you don't pair that with the social capital needed to use that support effectively, it has vastly less effect than it otherwise could. And so, I would emphasize that as something that we can change even if we can't recreate the conditions in some of these communities.
Bridget Burns:
Another, yes, I see one over here.
Bridget Burns:
Well, I'll go here for – because there's a – go ahead, and then I'll come to you.
Dr. Kimberly Lowry:
Good afternoon, Dr. Kimberly Lowry, thank you so much. I'm representing the NISTS, National Institute for Study of Transfer Students, and also Texas Association of Community Colleges. I'm so grateful to see that you actually identified some community colleges on your charts, and also thinking about the large number of students, particularly historically underserved and low-income that start at community colleges. I'm just wondering if based off of your research, if you saw any outcomes in particular for transfer students at four-year institutions? Or just any high flyers, I know you mentioned SUNY in the ASAP program, but any other regions or institutions that maybe could be a model that we could look to for some outcomes?
Raj Chetty:
Yeah, absolutely. Thank you for that. We certainly see that a number of the two-year institutions where you see terrific, better outcomes, are cases where there's transfer capacity. You see them, the UC system for example, where there are lots of students who end up doing extremely well through the transfer pathway. I think thinking about how to do that more systematically in other settings could be incredibly valuable.
Again, I would encourage you – of course I'm not familiar with every college, and the Texas context or elsewhere. All of that data is out there. So, trying to understand which of those colleges do seem to have very good outcomes, which don't in your own context in the colleges you're interested in, I think could be very valuable. But more broadly, absolutely. I think transfers are a big part of the answer here.
Vincent Del Casino:
Hi, thanks very much. This is really awesome, I appreciate it. Vincent Del Casino, San Jose State University, and social geographer by training, so the maps always geek me out. So, one of the things that's interesting being at San Jose State is the valley itself, and you use absolute income. I wonder if you ever talked about living wage instead, because if you look at the percent of students in that valley, it's a very different thing, 43% of them with families over $110,000.
Raj Chetty:
Yep.
Vincent Del Casino:
And if you look at the Gini coefficient of that region, we look like a failed state because wealth, inequity, in fact probably maps right up onto that. So, do those things play into how you start to think about how universities might think about their local communities and how you respond to those challenges?
Raj Chetty:
Yeah, great question. No doubt differences in cost of living across places, $40,000 means something very different in San Jose versus Iowa. Let me say two quick things on that, and I know we're running up on time here. One is that, empirically, it turns out if I take those maps that you like, and I'm delighted about that, and I adjust for cost of living, they would look pretty much the same. They're correlated 0.9 with the maps that I showed you, meaning that 90% similarity in some sense. Why is that conceptually?
So, the way we do that is, you can get data on local cost of living, as you know, based on rents and cost of groceries and things like that. Imagine adjusting both the parents’ and kids' income by those measures, and recreating the map. And what I'm saying is the map you get is basically the same. Why?
The reason is that most kids in the U.S. live very close to where their parents grew up. The median kid in America, when they're about 30 years old, lives 13 miles away from their mom. And so, the reason that matters in this context is when you adjust for cost of living, you're pulling the parent down in the income distribution. So, the parent making $27,000 effectively is much lower in the income distribution in San Jose, and you're also pulling the kid down, as you emphasize.
When you think about economic mobility, it's really fundamentally about how you are doing relative to your parents, and as a result, it doesn't end up changing the calculation that much. If people moved around a ton then it would matter, but in practice, that's not the case. I think the conclusions are not completely skewed by that, but it's certainly something to keep an eye on.
Bridget Burns:
Can we have one question?
Raj Chetty:
Sure.
Bridget Burns:
One last question, because it's the college president in the back of the room, I believe.
Lisa Vollendorf:
Thank you. Hi, Lisa Vollendorf from Empire State University. Thank you for all of this, you've given us a lot of food for thought. And of the many, many disconcerting things that you've given us to think about, for me, one of them that really stands out is that just as America is segregated in all the ways you have laid out and alluded to, so too are many, many of our colleges and universities.
Bridget Burns:
Yeah.
Lisa Vollendorf:
So, my institution, for example, serves 16,000 online students, primarily adult students, but they range in age from 15 to 88. But I would hazard to guess that very few are in the upper quintile. So, it for me is a deep challenge that you've laid for us about how to build social capital across class lines, let's say, when many of us are serving only the lower quintiles. I'm wondering what you think of that in ways that you think we might intervene.
Raj Chetty:
It's a great question, I think you've characterized the situation beautifully. If we look at the data that I was showing you in that scatter plot of the colleges, one point we make in the paper that analyzes that data called mobility report cards, which again you can find on our team's website, is that there's actually more segregation across colleges in America than there is across neighborhoods in America. You are more likely to encounter somebody from a different class, different income level walking down the street in your childhood neighborhood, in the high school you went to, than the college you attend, which is weird when you think about it, because our conception is that colleges are going to be the place that are the level or the mixing ground, but that's absolutely not the case at present.
I think you lay out the challenge correctly. How do we create that integration? One is to change who comes in the door. That's kind of moving the dots to the right in the chart that I was talking about. At institutions that are extremely selective, diversifying the student body is extremely important. That's going to have its limits, for the reasons that I said.
I also think it's interesting to think about an institution like yours, you emphasized the online aspect of it. An interesting area to maybe innovate is to think about whether there's a way to harness technology to create those connections. My view is that at present, if you think about the Facebook data that I showed you, if you think about the online community, there's potential to connect with people in very different settings from very different backgrounds. In principle, you're no longer bound by geography, which can be the greatest stratifier in many cases. But that doesn't happen automatically. If you look at the segregation that people have online, it's about as large as what you see offline. And so, I think it takes deliberate efforts to try to create those communities. I don't have the answer for you in terms of exactly how to do that. Hopefully some of these programs that are showing some signs of success are promising indicators, but I think having everyone here think more about these issues would be incredibly valuable.
Bridget Burns:
Please join me in thanking Raj Chetty
Bios of Guest and Co-Hosts
Dr. Raj Chetty is the William A. Ackman Professor of Economics at Harvard University and the Director of Opportunity Insights, which uses big data to study the science of economic opportunity: how we can give children from all backgrounds better chances of succeeding? Dr. Chetty’s work has been widely cited in academia, media outlets, and policy discussions in the United States and beyond. He received his Ph.D. from Harvard University in 2003 and is one of the youngest tenured professors in Harvard’s history. He has received numerous awards for his research, including a MacArthur “Genius” Fellowship, the John Bates Clark medal, given to the economist under 40 whose work is judged to have made the most significant contribution to the field, and Harvard’s George Ledlie prize, awarded for research that made the most valuable contribution to science, or in any way for the benefit of mankind.
Co-Host: Bridget Burns, Executive Director, University Innovation Alliance
As a trusted advisor to university presidents and policymakers, Dr. Bridget Burns is on a mission to transform the way institutions think about and act on behalf of low-income, first-generation, and students of color. She is the founding CEO of the University Innovation Alliance, a multi-campus laboratory for student success innovation that helps university leaders dramatically accelerate the implementation of scalable solutions to increase the number of college graduates.
About Innovating Together
Innovating Together is an event series that happens live on Facebook, Twitter, and LinkedIn. It also becomes a podcast episode. Every week, we join forces with Inside Higher Ed and talk with a sitting college president or chancellor about how they're specifically navigating the challenges of this moment. These conversations are filled with practicable things you can do right now by unpacking how and why college leaders are making decisions within higher education. Hopefully, these episodes will also leave you with a sense of optimism and a bit of inspiration.
Rate, Review & Subscribe
Learn why hundreds of people have rated the Innovating Together podcast 5 stars. Please join others and rate and review this podcast. This helps us reach and inform more people -- like you -- who are committed to helping more students succeed.
Click here, scroll to the bottom, tap to rate with five stars, and select “Write a Review.” Then be sure to let us know what you loved most about the episode. Also, if you haven’t done so already, subscribe to the podcast. You'll never miss an episode.
Follow us to Join LIVE:
