Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Episode #59: Beth Akers and Matt Chingos
172 Plays8 years ago
Welcome back to the PolicyViz Podcast! I’m taking a bit of a turn on the show. Instead of talking with people directly creating visualizations or building visualization tools, I talk to two people working in the field of education research...
The post Episode #59: Beth Akers and Matt Chingos appeared first on PolicyViz.
Recommended
Transcript
Introduction and Sponsorship
00:00:00
Speaker
This episode of the PolicyViz podcast is brought to you by Tableau Software. Tableau helps people see and understand their data. Tableau 10 is the latest version of the company's rapid-fire, easy-to-use visual analytics software. It includes a completely refreshed design, mobile enhancements, new options for preparing, integrating, and connecting to data, and a host of new enterprise capabilities. To learn more, visit tableau.com.
Podcast Accolades and Listener Engagement
00:00:28
Speaker
Hi everyone, just a quick note before we get to this week's episode. The PolicyViz podcast has been shortlisted for the Best Data Vis website of 2016 from the Information is Beautiful Awards. So if you can take just a second and go on over to the Information is Beautiful website and vote for the show, I'd really appreciate it. There are more instructions and the link to the website on the show page. Okay, onto this week's episode.
Introducing 'Game of Loans' with Authors Beth Akers and Matt Chingos
00:01:00
Speaker
Welcome back to the Policy This podcast. I'm your host, John Schwabish, taking a little change of pace this week to talk with two researchers who are actually using data, using data in an interesting way and finding a lot of challenges with their data. I'm here today with Beth Akers and Matt Chingos, Beth from the Brookings Institution, and Matt, a colleague here with me at the Urban Institute to talk about their forthcoming book, Game of Loans, The Rhetoric and Reality of Student Debt. Yes, that's right. It's a book on student loan.
00:01:28
Speaker
And it's called Game of Loans. So enjoy that for a moment, savor that. Beth, Matt, thanks for coming. Thanks for having us. Thank you. So really interesting book, really nice read, and I think really summarizes this whole discussion and debate in some really interesting ways and great ways. But before we talk about that and talk about some of the data challenges, I was hoping you could just sort of introduce yourselves for folks. You're probably a little bit different than a lot of my guests recently, but a nice change of pace, I think. So Beth, let me start with you.
00:01:54
Speaker
Sure, yeah, I'm an economist, as you said, working at the Brookings Institution, and I've been studying higher education for several years now and have been focused primarily on student debt for the past few of those years.
Authors' Backgrounds and Interest in Education Policy
00:02:06
Speaker
But broadly speaking, interested in how to make federal education policy better and using data to answer the questions that we need to answer to make that happen.
00:02:14
Speaker
And as John mentioned, I'm here at Urban, where I've been for the past year or so. Before that, I was at Brookings, where Beth and I did a lot of the research underlying this book. In addition to work on student loans, I do work on some other issues in higher education and in K-12 as well. I'm sort of interested broadly in using data and research to inform education policy discussions.
00:02:35
Speaker
So let's talk about the core message of the book, and then I want to talk about the data issues.
Misconceptions of the Student Debt Crisis
00:02:39
Speaker
So we have what is commonly called a student debt crisis in this country, but you sort of refute that sort of popular argument. So can you talk a little bit about how you view this student loan crisis?
00:02:50
Speaker
Sure. I think it's helpful to start with where did this research agenda come from? And for us, Matt and I come from more academic backgrounds rather than policymaking or in politics. And so I think we were looking at this question about whether or not there is a student loan crisis on the horizon from the perspective of having all the evidence that academic research has put together that there are very large returns to investments in higher education. And that sort of prompted us to think about this question a bit differently
00:03:17
Speaker
than the popular rhetoric was addressing it. And so we started poking and prodding at what's happening with the student lending system in the United States and just started putting some evidence behind it. And what we were finding was that the narrative that was dominating the media was really inconsistent with some of the facts that we were finding.
00:03:36
Speaker
Yeah, and I think kind of the bottom line is that if there is a student loan crisis, it's not the one that people think there is.
Comparison with the Housing Crisis
00:03:42
Speaker
When they open up the New York Times and read about the kid from an upper middle class family in New Jersey with $80,000 in debt living in mom's basement, that's not really the big problem. Most people graduate college with not much debt at all.
00:03:57
Speaker
an average of $30,000 among those who borrow. A lot of the big debts are people with graduate degrees. We really see big problems are often in cases of people who didn't borrow very much. They didn't borrow very much because they didn't stay in college that long. They went and tried a semester or two often at a for-profit college or community college, borrowed maybe $5,000, $10,000. And those are the people who are really struggling, those are the people who are defaulting. So the crisis isn't the one you think.
00:04:23
Speaker
Right. And you also talk about the sort of difference in the market and the different types of markets versus let's say the housing crisis versus the student loan crisis. Right. So people often like to say that student debt is going to be the next bubble, right? That the bubble is going to pop and something terrible is going to happen. It's going to take down
00:04:39
Speaker
And we document in the book how that specific argument is kind
Impact of Income on Student Debt
00:04:45
Speaker
of silly. It's a much smaller market. It's all owned by the government. It's not connected to the global economy. But then we also talk about the question of whether there's a crisis on the horizon for the borrowers themselves. And we also don't find much evidence there that on average there's a big problem. The problem is in pockets of the market.
00:05:06
Speaker
What I like to say is that there are many crises in our federal lending and student lending system more broadly, but it's not the macro crisis that's being described. I think what was important to both of us in writing this book was getting that message out there so that policy can start focusing on solving the real problems in the system rather than this fictional macro crisis.
00:05:25
Speaker
Because you do make a specific point about people sort of not knowing there's uncertainty about whether they're going to finish school and what they're going to be earning when they get to the end of school and what the balance is between what they'll make and what they earn.
Proposed Solutions for Higher Education Investment Risks
00:05:38
Speaker
So when you think about that, when you think about the individual who may have $8,000 or $30,000 in debt, how do you think about a policy response to those sorts of individuals?
00:05:51
Speaker
You know, I think that that's really the right way to be characterizing what's happening in this space. So much of the conversation has been focused on how expensive college is, how much debt people are taking out. I think the bigger concern is that there's a lot of uncertainty in this investment. And so policy solutions that address that uncertainty through something akin to an insurance policy issued by the government would be a step in the right direction. So we have a system that does that to a degree. I'd love to see that more explicitly done with a more robust income driven repayment system, one that's
00:06:20
Speaker
students are defaultly enrolled into and one that just works a little bit better than what we have today. Right.
Importance of Data in Understanding Student Loans
00:06:26
Speaker
Okay. I'm sure you want to sit down and talk all about student loans, but let's turn to the data because there are a number of places in here where you actually talk about the problems with the existing data, what you'd like to see done with the data and how more data would help you and others better understand the questions here. So can we start first with where do the data come from that you've used in the book?
00:06:48
Speaker
Sure, so there's a variety of data sources that are available publicly and to researchers like us that we draw on in this book. I mean, the big one we've used in a lot of our work is something called the Survey of Consumer Finances. It's put out every three years by the Federal Reserve Board here in DC. And it's not just about student loans. It's about family finances more broadly. And they ask a bunch of questions about how much debt people have about their income. And it's one of the few data sets where you can get both debt and income.
00:07:17
Speaker
And that's how we're able to come up with findings such as the share of people's income that they devote to their student loan payments on average hasn't changed that much over the last 20 years or so. But really the best data on this would not be to go and ask a representative sample of people. It would be to look at the records held by the federal government on student loans because the federal government holds 90% roughly of
00:07:39
Speaker
outstanding student loan debt in the country and link it to records the government also holds in another agency in the Department of Treasury at the IRS and how much money they make and you can learn a lot more. And about a year ago, a paper came out that did that and it was better data than had existed before, but up until now has still been a one-off effort. So my hope is that the government will continue to both release those aggregate statistics
00:08:04
Speaker
And my understanding is that they are going to move towards making some amount of the individual level records, at least a sample of them that researchers would need. I'm available to researchers beginning I think with the Federal Reserve and then possibly expanding beyond that, but I don't know about the specifics of that. Yeah, and we're talking about innovations from a research perspective for the availability of data. It's also good to note that there's been some important innovations on the consumer side in this space too. So historically speaking, if you were shopping for a college,
00:08:32
Speaker
It was nearly impossible to look back and see how previous graduates had done in terms of their labor market outcomes. So under the Obama administration, we had the innovation of that data becoming publicly available actually on the White House website to help people make better decision making. So there's been a lot of big steps forward and I think that there will be some more in the near future.
00:08:53
Speaker
Now, for people who may not be familiar with administrative, well, economic data, financial data, can you talk a little bit about what you see as the advantages, maybe the challenges of having either multiple admin data linked together or linking survey data to administrative data?
00:09:08
Speaker
So one of the challenges for the administrative data is right now the federal government in higher education doesn't have the legislative authority to do a big researchers call a unit record database individual student level database. There's actually a law Congress passed I think in 2008 that says explicitly they can't
00:09:26
Speaker
do it. So as part of this important effort that Beth just mentioned, about publishing average earnings for each college, they kind of needed a back door. And the back door they use was, they said, Well, we're only going to report it for people who participated in a federal aid program. So the student loan programs, the program, our program, so an important step forward.
00:09:45
Speaker
but still limited by the fact that by law, Congress said that the government can't do that for everybody. So a step forward would be to repeal that ban, and we propose in the book that the ban be repealed, and we have a robust unit record system, which would actually make things easier for colleges. They wouldn't have to do nearly as much of the reporting as they do today. But obviously, that faces political challenges, both from some of the college lobbying groups that don't want that kind of transparency and accountability.
00:10:12
Speaker
And then also for folks who are concerned about privacy who oppose any kind of effort to make this sort of administrative data linkage easier. So from a researcher's perspective, if you could build the ideal data set to do more of this work, what would the basic elements include?
Linking Data for Better Policy Making
00:10:29
Speaker
Well, as Matt said, the motivation for us using the Survey of Consumer Finances for our research several years ago was that it contained both income and borrowing information. And so that's really the key link.
00:10:40
Speaker
The best source of income from the administrative records is going to come from the IRS. Everybody's reporting their income through IRS records. And then Department of Education holds information on all of the lending that takes place through the federal lending system, which as Matt said, is the vast majority of outstanding student loan debt. And so linking those two would go really far away in helping us to understand what repayment looks like and what the financial circumstances are of these borrowers.
00:11:04
Speaker
And how do you emphasize the repayment piece, right? So it's about borrowing income and then how do people repay over time. Because one of the other data sets that is out there and has been used is data from credit agencies, which is pretty good for looking at people's, how much debt they have, different kinds of debt they have.
00:11:22
Speaker
kind of creditor characteristics, things like credit scores and how they repay over time, but it's not linked to income. So there's lots of data that's good for getting at pieces of this, but very little to kind of get at it all at once. The whole thing and over time for the individuals, right? And the thing is too, in this conversation, there's been a lot of emphasis on what has happened to debt over time, but not necessarily how debt has moved in tandem with income. And that's really the key question, right?
00:11:46
Speaker
If someone's going into debt for the purpose of making a really great investment that's going to lead to much higher income in the future, that's something to be less concerned about than if that debt was moving but income was not. So that's really the key to understanding how people's well-being has changed over time with student debt, not just student debt alone.
00:12:02
Speaker
Right. We've talked a lot about federal debt. What about are there other state or local data sets that you've seen or that you've heard about that can lend some insight into this but not at the sort of federal level?
State-Level Data Integration Efforts
00:12:13
Speaker
So states have taken the lead in the absence of government. We've seen more federal action in the last couple of years. But prior to that, states have taken the lead by linking their own, usually just public institutions, higher education databases to their unemployment insurance wage records.
00:12:28
Speaker
And that is better than nothing, but it leaves out certain classes of employees, such as federal employees, and more importantly, it misses students who leave the state. So you're trying to look at outcomes for different institutions or even for different programs of study, which many states have done. You only capture people who stay in the state. So if you're trying to compare two institutions, one is kind of a typical in-state university where people stay in the state, and another one is maybe a private institution. A lot of people come, and then most of them go to another state to work.
00:12:58
Speaker
It's really not a very impulse to apples comparison. Right. Okay. So we talked a little bit about uncertainty. So this I think is part of the core issue from an individual's perspective, right? Also from the government as a lender, but from an individual's perspective, it's how much do I take out for some trade off at the end of the day? So I guess my question is, do you think people understand uncertainty generally? And then do they understand uncertainty when it comes to this specific decision point in their lives?
00:13:26
Speaker
I would say, one, we don't know. But if I had to guess, I'd say no.
Student Awareness of Loan Amounts
00:13:30
Speaker
So there are a few data points, I think, that push me in that direction. So Matt and I did some work a few years ago that highlighted the fact that students don't really know how much they're borrowing. So even if you ask them just a few months after they've taken out their first loan, have you borrowed from the federal lending program? And how much have you borrowed? One, they don't know if they've borrowed many of them. And those that do know they've borrowed can't tell you how much they've borrowed. So how do they not know that they've borrowed?
00:13:54
Speaker
That's a little unclear. I think it's really important with emphasizing two key numbers I like to use among people who we know from the administrative data have federal loans. 28% say, I do not have a federal loan. And 14% say, I do not have any loan.
00:14:11
Speaker
And then the people who do say they have a loan, they're way off on average. Most people can't get within a reasonable margin of error on how much they have. And this is a survey done during their first year of college. I want to say again what Beth just said because it's crazy to me. So forget about it. And your question is, do people understand uncertainty and cost-benefit analysis and trade-offs? It's like they don't even know how much they've borrowed. Yeah. They don't even get the first step.
00:14:37
Speaker
Yeah. Yeah. And I think, you know, one data point that pushes in the other direction, just to be a little bit more optimistic for a moment, we have seen this recent innovation of a new financial product to support investments in higher ed called income share agreement. So an income share agreement really kind of functions like
00:14:54
Speaker
a loan that has an insurance policy layered on top of it, which is that if I borrow for school, but I don't end up making a lot of money in the future because the degree didn't pay off, then I'm not going to pay back very much or I'm going to pay back nothing at all. So you can see how that kind of works as insurance. And it does seem like there's some interest among students in this type of product, which would suggest that they understand that there is this trade-off and it's a trade-off that they would like to not have to make.
00:15:18
Speaker
Matt, you mentioned a little bit one of your policy recommendations, which is to repeal this law from
Streamlining Student Loan Programs
00:15:23
Speaker
2008. Are there other recommendations you were the legislator in charge that you would try to push through either at the federal level or the state level?
00:15:30
Speaker
At the federal level, there's just kind of a lot of policy cleanup that needs to happen. I mean, we document in the book how the current system of student loans and of higher education financing more broadly as the result of an accretion of well-intentioned legislative action since the 1960s. So since the original Higher Education Act in 1965 created the first big guaranteed student loan program in an early
00:15:53
Speaker
version of what later became the Pell program. We kind of have expanded and we've layered and now we have all these different programs, all these different repayment plans and the result is that it's confusing, it's hard to navigate. So one big thing we recommend in the book is just cleaning all that up. Have one grant program, one loan program, one repayment program or at least a very small number of repayment programs that are really easy for people to understand and to navigate so they don't get lost in the bureaucracy as people are now.
00:16:21
Speaker
And coming back to how is it possible that these students don't know that they have a federal loan, this kind of sheds some light on them. So maybe they're thinking, well, I had a Stafford loan, or I had this other kind of loan. What was the name of that program again? So because there's this huge mess of programs that exist in this space, I think students are really confused about what it is that they're engaging in. And yet, unfortunately, I have no choice but to engage in it. Right. And so if it's called a Stafford loan, maybe that doesn't have any meaning to me as the individual that that's somehow related to my education. Right.
00:16:49
Speaker
Great. Well, this is fascinating. I think you've sort of, at least for me, shed light on the student loan crisis, which I'm now putting in quotes in a very different way. So Beth, Matt, thanks for coming on the show. It's been great. Thanks so much for having us.
Conclusion and Book Promotion
00:17:02
Speaker
Thank you. And be sure to check out the new book, Game of Loans. It's from Princeton University Press, and will be out in bookshelves and actual paper copy this fall. So thanks for tuning in this week to the Policy Viz podcast. Until next time, thanks for listening. Bye-bye.
00:17:28
Speaker
This episode of the PolicyViz podcast is brought to you by Tableau Software. Tableau helps people see and understand their data. Tableau 10 is the latest version of the company's rapid-fire, easy-to-use visual analytics software. It includes a completely refreshed design, mobile enhancements, new options for preparing, integrating, and connecting to data, and a host of new enterprise capabilities. To learn more, visit tableau.com.