Bicentennial Symposium day 2 session 5

November 10, 2017 1:26:21
Kaltura Video

The University of Michigan has long been a leader in social science research on the many dimensions of social inequality. This bicentennial symposium will highlight these contributions by focusing on the work of distinguished social scientists who were trained at the University of Michigan.

Transcript:

0:00:00: Hi. I'm Mel Stephens. It's not Stevens, it's Stephens. I pointed it out 'cause my name has been butchered so often. And then prior to this, I was asking Maria how to pronounce her last name, and she said I could just call her Maria. Maybe everyone will be on a first-name basis. It's not on? Is it on? Mic is not close enough? Alright. Wow. You missed out on $20 worth of humor there.

[chuckle]

0:00:25: Apologies. In any event, we have a great session for you. It's great to have so many people back to Ann Arbor. Just quickly, I was here at the University of Michigan myself as a graduate student. I was then at Carnegie Mellon for a while before I came back to Ann Arbor. I had a colleague there, George Lewis, and he worked on this project, which is they tracked all the student visits to Carnegie Mellon's campus, and they figured out... The end of the story is that, they figured out that students that came and visited campus on the really terrible weather days ended up not coming to CMU. And so, I could imagine how terrible it would have been to have all of you come today for your first visit because we would lose such a resource to the University of Michigan. But in any event, I'd like to think the cold weather kept me in studying as opposed to out and about. But anyway, we have a great panel. I will turn it over to Maria to begin.

[pause]

0:01:24: Let's see. I'd love to give a talk on... If I give Vonnie slides, it will not be a good thing. I won't sound nearly as... That's okay. I think I know what to do from here. Awesome.

0:01:46: Good morning. I'm very happy to be here. My name for the record is Maria Cancian. It doesn't look at all like what it says up there. It's particularly great to be here because I have lots of memories of this building and the two most salient, one is sort of traumatic, and one is very happy. The first one is I spent the better part of two years in the reading rooms downstairs studying for the prelims in economics. That is not my fondest memory, but there's a few economics graduate students here who can share that memory. And then about five years ago, I had the privilege to help organize a conference for Sheldon Danziger who is near and dear to many of the people who are here, and has had just an extraordinary impact in terms of not only his own scholarship, but also his mentoring, and I still count him as my best and greatest mentor and cheerleader. And the work that I'm gonna do today is very consistent with the work that I did with him way back when. And also, with the work that I got to... I learned a lot from David Lamb and other people who are here. So it's really terrific to be back.

0:02:53: So, I'm gonna talk a little bit about family change and... First, I'm gonna start by trying to commit since not everybody here is a demographer. Convince people that when we think about families, poverty, and inequality, that things like marriage and living situations are really important. So, here's just two factors to kind of set the pace. There are over a thousand federal laws that make benefits, rights, and privileges contingent on marital status. So, we had some references to marriage equality in the last session. And sometimes people say, "Ugh. Why do we even have to think about that? This is private. The government shouldn't be involved." That's actually not an option. Even if you wanted, we're very involved, lots and lots of things depend on marital status. So that's one piece. The other thing is, the world is different, at least correlated with marriage status. We can have a conversation about whether it's a causal thing. And one nice example of that is that children that live with us and married mom are about five times as likely to be poor as children who live in married couples. So we clearly need to pay attention to this if we care about poverty and inequality.

0:03:58: One of the things that makes this kind of an interesting and challenging place topic for research and policy is that families are much more dynamic than is generally recognized. And I'm just gonna spend a few minutes that I have today talking about the way in which that is true. And when I say dynamic, I want to think about two different things. One is, one family, or let's pick the kid in that family or the mom in that family over the life course, lots of changes. Okay. So I might be married today and might not be married at some later point. The other thing is the cohorts of families. So, families in the 1960s versus families in the 1980s versus families in 2000. So kind of families in general at different points in periods in history. Both of those, we have a lot of change going on.

0:04:49: So, I'm gonna start. One of the things they didn't teach me when I was here was how to draw. So this is, when I do graphics, this is my graphic of a simple family. And a simple family has a dad and a mom and two kids in common. Okay. And so, often when we think about families and family policy, we have this picture in our head. And we also had a discussion about frameworks and pictures for policy and even for research, a lot of what guides us are these implicit assumptions that we have. So often when we think about families... Now, people have different ideas when they think about low-income families, and we can have a discussion about that. But here's our simple family. So, if you look at the simple family, one thing that might change over time is that that simple family might get divorced. And if that happened, you have to think about a number of questions about what would happen and one question you might ask is where did children live after their parents divorced? And typically, what we thought of, and what was the empirical reality was that children lived with their moms when they divorced, and maybe they visited with their dads and their dad's pay child support. Now, there's a bunch of reasons why that might have changed.

0:06:00: Married moms are much more likely to be working, so it used to be part of the reason that we expected kids to live with their moms as their mom was their primary caretaker, and moms weren't in the labor market, and that's really changed. From 30% to 64% in a fairly short period of time of moms of young children in the labor market. Another thing is that dads are much more likely to be taking time with their kids, and I guess I won't tell the story about... But Mel and I, were talking about some of his family care responsibilities during the break. And I won't embarrass his kids by going into the details. But he gets woken up at night to take care of the kids instead of his wife. But so, yeah, exactly that's what I said to Mel, I told him it was an active resistance to do this and to tell his class about it. I will make the point that while both of these things are true, men are still more likely to be working than women, and moms still spend a lot more time with their kids than dads on average. So, we're moving in towards equality, but we're not there.

0:06:58: At the same time, we really have had changes in social norms around this. What used to be kind of a concept when there was divorce is this concept of tender years. Kids needed to be with their moms, especially if their moms were young... Kids needed to be with their moms, especially if the kids were young. Then we moved to this idea of the best interest of the child, and we now are in a policy space where there is in many places, many jurisdictions, an explicit preference for shared parenting. For example, in Wisconsin, where I do a lot of my work, on the statue it says, "Unless they can demonstrate that it will be harmful to the child physically or emotionally, they need to allow for shared parenting." For the child to spend significant time with both the mom and the dad following divorce. And policy has changed in that way too. Policies around child support in 1970, there were just nine states that had child support guidelines that allowed for shared... That kind of said what would happen if parent shared... If parenting were shared. But virtually all states have those guidelines now.

0:08:00: What does it look empirically? What is the change? The change is really dramatic. Here is some information, and I gotta say this is for the State of Wisconsin, but I will also say this is the only jurisdiction for which we can do this analysis. And I wish if Garfinkel were here, because the early years of these data are... Were collected in a project that Erv started, which Dan Meyer and I are now PIs for, many years later. And what this picture shows you, the pink is mom's sole custody. And since it's hard to read even on my screen, 1989 is the first year, about three quarters of children after divorce, the arrangement was for them to be solely in the physical custody of their mom. And in 2010, the last year for which we have data, it was 42%. That is an enormous change in a very short by our standard, period of time. From it being clearly normative for kids to just be with their dads, to that being kind of a 40%. And the green is shared custody, where the kids are spending... The dark green has 50/50 with mom and dad, and the light green is at least 25 more than 25%, with each parent.

0:09:18: So, a huge change. That little bit of blue is father sole custody. That remains a pretty rare thing, but shared custody is really the dominant arrangement for divorce. Now, since on of our key issues here is inequality, I wanna show you that this is really different for lower and higher income families. The blue line is for families with a joint income between mom and dad, under 40,000, the green line is for over 90,000. And so two things, the slope is pretty similar. Okay, an increase in shared custody over time for both of these groups, but really, really different levels. Okay. Two-thirds of higher income and 90,000 is not super high income, but higher income families have shared custody, and less than a third or about a third for families have more modest means. And I think one of the things we have to think about is, if the courts, if statute says, "Children have a right to time with both parents." So, you lose economies of scale, you lose a whole bunch of things when you have to maintain two households that have room for kids, and things like that.

0:10:28: If kids have a right to time with both parents, what does that mean for lower income families who may not have the resources? What does that mean for things like housing policy? And whether we allow part time parents to have access to public housing? There's a whole set of policy things, but if we start, if it is, if it has become normative for children to have time with both parents, when their parents aren't in a couple, and what does that mean that we owe to lower income children in terms of giving them access to that? Okay. So now, I talked about simple families, I wanna spend the second part of my time, talking about complicated families. Here's my picture of a complex family. In this case, we've got a mom who has two kids, and one of those kids has one father, and one of those kids has the other father. And I'm gonna focus on the older kid, that child, one way to think about them is that they have one co-resident half sibling. In other words, they live with a sibling, who has the same mom, but a different dad.

0:11:30: And I'm gonna show you a more complex family, same thing going on here, but there's also that child, number one, that father of that child number one, also has had a child, with another mom. In this case, the focal child, for what I'm gonna show you in a minute, has two half siblings. One co-residing, that's their half sibling on their mom's side, and another who doesn't live with them, because it's a half sibling on their dad's side. Now, you know... My stick figures are no doubt, beautiful to look at. But the question really whether they're interesting is, "How common are these complicated families?" So to look at that, Dan Meyer and I and our colleague, Steve Cook used administrative data or records from the state, again, of Wisconsin, to look at how common this was and what we started with was a sample of moms who'd had their first child when they weren't married. Okay, so first non-marital child.

0:12:27: About 40% of all children born in the United States are born to unmarried moms, so this is not a tiny population. Here's my picture. This is not following different families over time, so I'm not doing cohorts here. This is following one child. And just to explain that picture, if you look at the far left where it says, "At birth." At birth about 80% of these children, these first born children had no siblings. Because I picked them to be the first born of their moms. So they could not have a sibling on their mom's side. That's my sample definition. About 20% of them already had siblings because their father had had children with another mother. Now go all the way out to age 15, the story would be more or less the same at age 10. You see most of the changes in the first five years.

0:13:17: By age 15, about 30% of them in that light-blue had half siblings on both sides. That is to say, 30% of them were in that more complicated situation. Where they had... Their mom had had children with other fathers and their father had had children with other mothers. And really only 40% of them or 30% had only full siblings or no siblings. So very few of them were in simple families. And this is only children who end up registered in the administrative data that we have. And we know from some survey data that we've done on the same population that there are some kids here that we are missing. In particular, we're missing any marital children where that marriage is still intact and there's not a child support order. So this is an under-count of complexity. If I tell you that 40% of children are born to unmarried moms, and I tell you that about three quarters of those moms, those kids will have complicated families. This is not some special case that we don't have to pay attention to. This is something that we really need to grapple with in child support policy and housing policy and education. This is... In tax when we think about the EITC. This is really a key phenomenon.

0:14:32: Now, I told you this is following one child for a long time. So this went back to 1997 'cause I had to go far back in order to be able to follow kids for 15 years. You might ask yourself, I asked myself, Dan and I asked ourselves, "Has that changed over cohorts? What about kids that were born more recently? Do they have even more complicated families?" What we did is we repeated this exercise, but with kids born five and 10 years later. And I'm gonna draw your attention this... I'm gonna show you just the picture for the most complicated. So the light blue. And this is kind of the risk for having half siblings on both sides. And the purple line is just kind of graphing the flip side of what I showed you on that other. Almost 30% of the 1997 cohort had half siblings on both their mom's side and their dad's side. Now if you look at the later cohort, it's actually lower.

0:15:32: Fewer people have the most complex families. Or at least they're not having as complex a family as early. And in other work that we've done, what we... Trying to understand what might be behind this, one key piece of it seems to be the decline in teen pregnancy. As you might expect, if you have your first child when you're a teen mom, because my sample is based on moms... Teen mom, that might not be the relationship that's most likely to stick. If you wait to have your first child, even if it's outside of marriage, if you had that later, that relationship might stick more or you might be less likely to be involved in more complicated relationships later on. And that seems to account for about a quarter to a third of this, and we're working on other analysis. So another way in which kind of time things have changed over time.

0:16:27: Okay, so what does this mean for policy? Well, it means that children in complicated... There are more children in complicated and dynamic family arrangements. So I think it becomes less and less tenable for us to have that simple family in mind when we're trying to design policies. And family complexity is more common for lower income families, as you might expect given some of the stresses and strains in lower income communities. And we know that less educated couples are less likely to marry. And that we know that divorce rates are higher. So there's lots of reasons to expect this to be higher. That means that when we're thinking about income support policies and programs, we really have to get that simple family out of our head when we're trying to make our rules. We need to understand the complexity of families and really grapple with the issues. What does that mean?

0:17:22: For example, how much child support can you really expect a father to pay if he has obligations to children in a number of different households? It might be reasonable for him to pay... In Wisconsin, he's supposed to pay 17% of his income for one child. But what if he has three children across three different households, and he owes three times that? Now on the one hand, if you're an economist, if he has three children in three different households, those three moms don't have economies of scale, so you don't necessarily wanna reduce what he owes. But on the other hand, there was a limit to what one person can pay.

0:18:00: If we decide that it's too much, then what is our obligation as a society to make up for that gap? What policies might we need to do that? So those are issues that I think we really need to grapple with. We need also public institutions and policies that support families with children. So I won't show you the data because I'm gonna be short on time. But we've done some analysis using data from random assignment experiment that we have going on, where we've had an opportunity to survey about 10,000 non-custodial parents who are behind in their child support. And one of the things that we find is that if you compare how much fathers contribute to their most recent child and their oldest child, they give informal support more to their most recent child, which makes sense, that's the relationship they've been in most recently. That's where they're mostly likely to have a strong connection. But if you look at their formal child support payments, there's no difference between what goes formally to their youngest and their oldest child.

0:19:04: And you might expect that 'cause the formal child support system is pretty routinized. And there's not really a lot of opportunity for fathers to have discretion. So if we believe all kids deserve the same thing, one way to think about that is, "I shouldn't get less child support from my dad, if he goes on to have another relationship with somebody else." So if that's something that we care about, then the formal child support system is doing a really good job of assuring that kind of equity. That's an example of a way in which our institutions can support certain outcomes. But... I can say it would be good, we should do this, but it's difficult to design and to implement effective solutions given this complexity, and given these dynamics. So it's hard to figure out what the right child support order is for a father who has two children in two different households. But then you also have to think about the fact that next year, there maybe a third household. And are you gonna revisit what the kids are getting in the first two or not? So it's very complicated and dynamic over time situation, which makes policy design both critical and challenging. And so I'll leave you with that puzzle. Thank you.

[applause]

0:20:25: Next we have Sasha Killewald. Oh, is the last name, sorry.

[pause]

0:20:38: Good morning. I'm really pleased to be here. I'm gonna talk about wealth and equality in the US. I think that we tend to talk both in academia and in popular discourse a little more about income and equality, but wealth and equality is even greater. Just to give you a sense of how the wealth distribution in the US has evolved in recent decades, this is a graph reproduced from work by Fabian Pfeffer and Bob Schoeni, both professors here at the University of Michigan. To help you make sense of what you're seeing, just focus first on that green line. That's the line that represents how median wealth in the US has changed since 1984. And it's scaled so that 100 means the level of wealth, that household, that the median of the distribution held in 1984. So you can see that line ticking up prior to the Great Recession. By 2007, the median household had about 40% more wealth than the median household did back in 1984. And then you see the decline through the Great Recession and a little bit of recovery afterwards. So that's the macrolevel broad trend. But you can also see all these other lines. That top line, the red line, is how the 95th percentile of the wealth distribution changed over the same period.

0:21:54: And you can see, again, it's scaled to the level in 1984 being 100, you can see that the 95th percentile increases more rapidly than any other part of the distribution. By 2007, the folks who were wealthy had more than twice as much in net worth as the wealthy folks at the same point in the distribution did back in 1984. And things look very different at the other end of the wealth distribution. So that bottom line, the gray purple line, is the line for the 25th percentile of the distribution. Even through the '80s and '90s where other parts of the distribution were gaining, those folks weren't really any better off than their counterparts were in 1984. And then after the economic crash, you can see large declines in that wealth position. So of course if you have the top of the distribution rising and spreading out faster than any other part of the distribution at the same time that you have the bottom falling away, this tends to increase wealth and equality. And as professor Williams alluded to in his presentation earlier, wealth and equality is not something that's just randomly distributed in our population.

0:23:01: One of the major cleavages is very variation by race and ethnicity. This is a graph reproduced from Pew that shows trends in the ratio of networth for the median white household to networth for the median African-American or Hispanic household. So I wanna highlight first of all just how big these disparities are. When we think about disparities in income, median earnings for African-Americans compared to whites and for Hispanics compared to whites, we often expect a ratio in the neighborhood of two to one. Whites are earning roughly twice as much. But for net worth, you see we're talking about ratios, more like 10 to one for African-Americans and eight to one for Hispanics. Furthermore you can see that these differences aren't narrowing over time. Sometimes we think that if we just wait long enough, these racial disparities will go away. And that's obviously not the pattern that you're seeing here, it's mostly steady and then during the Great Recession spike in those disparities, because African-Americans, and Hispanics were hit particularly hard by the recession.

0:24:00: So it's with this background, that I became interested in the structure of access to wealth in the United States. I'm particularly interested in how the circumstances of your birth shape your opportunities to accumulate wealth. And by circumstances of your birth, I focus particularly on variation by race and by social origins. And within social origins, I'm especially interested in how the wealth your parents had shapes the wealth that you obtain in your own adult life. So, I'm just gonna show one picture from work that's joined with Fabian Pfeffer who I just see now in the audience here at the University of Michigan. And here we're looking at how your parents' wealth position relates to your wealth position. This is a graph for roughly middle-aged whites in the US currently. What you see on the X axis is where your parents were in the wealth distribution, and on the Y axis where you are in the wealth distribution. Each one of those dots represents the average outcome for offspring from a given decile of the parental wealth distribution.

0:25:08: So that dot that you see on the far left-hand side indicates that if your parents were in the bottom 10% of the wealth distribution, then we expect on average that you end up at about the 40th percentile of the wealth distribution. Each of those dots is scaled in size to reflect how many, or what fraction of the parents belong to each of those deciles. So whites are the largest group in this population. So here you don't see much difference in the size of the dots. It's roughly 10% of white parents that are in each one of those deciles. Looking at the slope you see, not surprisingly, that having wealthy parents is better for your own wealth attainment. For every one percentile that your parents move up in the distribution, your own outcome is expected to go up, or predicted to go up by about 0.4 of a percentile. So a slope of about 0.4 is better to have wealthy parents. Now I'm gonna overlay the same picture, but for African-Americans. The first thing I want you to notice here, is just how big the black dots are on the far left of distribution. So whereas whites are sort of evenly spread across the distribution in terms of parental wealth, that is not the case for African-American middle-aged folks in the contemporary US.

0:26:21: In fact, a full 40% of the African-Americans in our sample had parents who came from that very bottom 10% of the parental wealth distribution. So, African-Americans absolutely are disadvantaged in their own wealth accumulation, in part due the disadvantaged wealth positions of their parents. But that's not the only thing that's different by race here. Another thing you might notice, is that the slope is much shallower for African-Americans than it is for whites. So for whites, we predict that if your parents go up one percentile in the distribution, you go up about 0.4. For African-Americans the relationship is only about 0.2. Now by itself, that's neither good nor bad. Having your outcomes less linked to your parents' wealth is not necessarily a bad thing. But that lack of association follows a very distinctive pattern of downward wealth mobility. So for every decile of the parental wealth distribution that you see here, every point on that X axis, the black dot always lies below the white dot. So what does that mean? That means that even if you have the same access to parental wealth when you're growing up, your own wealth is still predicted to be lower as an African-American.

0:27:31: And on average, that tends to spread out across the distribution. So African-Americans who had wealthy parents have a particularly large gap in general between their outcomes and the outcomes of their white counterparts. So what can we do with this information? Well, one other part of Fabian's and my work, is to look at the channels by which wealth is reproduced across generations. And we find that education and home ownership are really key channels of transmission. This doesn't necessarily mean that your parents are writing checks to pay for your home, or pay for your education. Although it can mean that. It can also be other more indirect processes, like buying a home in a neighborhood with high quality secondary schools that allow you to go to college. But if we know that these processes are one of the key ways that wealth is reproduced and are generationally, one thing we might do to intervene is to try to weaken the link between your social origins and these outcomes. So we heard yesterday in the panel on educational disparity, some of the policies that might help to make your educational attainment less dependent on the resources of your parents.

0:28:39: Something else that I think is important to keep in mind is that racial and ethnic wealth and equality is reproduced fresh every generation. So, as I said on the previous slide, yes African-Americans wealth position is, in part, due to a legacy of disadvantage, less parental wealth. But just narrowing parental wealth, just reducing gaps in social origins, is not sufficient to close the race gap in wealth. And we can see that played out in the slide I showed earlier, that these gaps aren't getting any smaller. So if we wanna make real progress in closing the race gap and wealth and equality, it's not enough just to wait for the past to fade away. One thing we can obviously do is try to target and reduce racial discrimination, either in asset markets themselves, so for example, mortgage lending markets, housing markets or in other domains that have consequences for your ability to accumulate wealth. So, for example, if we reduce hiring discrimination, we might reduce the race gap in income, which could then have implications for the race gap in wealth. You might also notice that I haven't said anything yet about direct transfers from parents to their kids, either during their lifetime or at their death. And these things do matter. So bequests and direct transfers do explain some of the inter-generational reproduction of wealth. But they're not as important as education or home ownership or even other mechanisms like marriage, or ownership of a business.

0:30:04: And one of the reasons you might think of, for this, is that bequest in particular tend to come to you relatively late in life. For folks receiving bequest often on their 40s and 50s, where these other processes like education, and home ownership, and marriage, those are often beginning and completing even for education much, much earlier. Those bequest that you get later in life are just the cherry on top for folks who have already accumulated substantial wealth, in part because of their advantages of their parents' wealth. With that said, certainly we can get rid of the estate tax, and that could have important implications for redistribution, but we shouldn't think of it as a major tool for reproducing the reproduction of inequality in wealth across the distribution. It could change some things for the ultra wealthy, but it wouldn't change the general pattern across the distribution. Last, David asked us to talk a little bit about the difference that Michigan has made in our own lives and our research. And for me, for this project, this is not a hard exercise.

0:31:03: The question of race gaps and wealth and how much they are, aren't due to differences in parental wealth is a topic that I got interested in while studying for my prelim exams as a PhD student in the Public Policy and Sociology Joint Doctoral Program here, reading the work of Dalton Conley, "Being Black Living in the Red." And furthermore, the work I just talked about, as I mentioned, is joint with Fabian Pfeffer, who is a Professor here at Michigan in Sociology and at ISR. And the data we're using come from the panel study of Income Dynamics. It's a data set that's collected here at Michigan and has been since 1968. It is one of the premier sources of wealth data in the United States and for looking at intergenerational patterns of wealth, it is really the best data set. I first got to use that data set while I was an undergraduate at Michigan, I was an Economics major, and I took a class from Frank Stafford, that I don't remember what the actual name was, but it was basically how to use the PSID. And in terms of investments I've made in my future career that maybe the best class I've ever taken 'cause it is not a DS set for sissies, it is a serious investment of your time. And now, it's one of the days that I feel most comfortable using because of that great training and experience.

0:32:19: In addition to my own academic career, of course, Michigan is also engaged in various policies that speak to these issues of inequality. One thing I'm particularly excited about is the new Go Blue Guarantee, which I'm sure many of you have heard about, which I believe begins next year, that guarantees free tuition to Michigan residents, who are coming from families that earn less that $65,000 a year. This is one way to try to reduce that link between that social origins and educational outcomes. And as was mentioned in the earlier panel, I'm also, of course, proud to be an alum of an institution that has defended the use of affirmative action by race and ethnicity in college admissions, which is yet another way to try to intervene in decoupling circumstances of birth from educational outcomes. And then last, not only would I not recognize my professional life without the University of Michigan's influence, I also wouldn't recognize my personal life. My family has an intergenerational transmission of Michigan issue. My parents both worked for the University of Michigan. I grew up here in Ann Arbor. My mom just retired from the University of Michigan this year. And furthermore, I met my husband at age 19 in a South quad dorm room. My entire life has really been shaped by Michigan, and I'm just so proud to be an alum of this great University. Thank you.

[applause]

0:33:41: And our last speaker for this session is Vonnie McLoyd.

[chuckle]

0:33:45: Okay.

0:33:45: Training...

0:33:46: What?

0:33:47: That was part of your PhD training?

0:33:48: No, it was not part of my training. Well, I have to say, it's a pleasure to be able to talk about some of the work that I've done, but I'm gonna talk in more general terms about a body of literature, some of which I have contributed to. My comments will be probably more general than they would have been, otherwise, had I known that Erv was not going to be here. We overlap a lot, and we are on a National Academy of Sciences, Consensus Committee on building an agenda to reduce the number of children in poverty by half in 10 years, and I knew that Erv was gonna talk about that, right? I did not include any information on that initiative. I apologize for that, but anyway, it is what it is, as they say. What I wanna do is talk about the issue of whether income really matters for children's development. There is an enormous amount of evidence that on average, children growing up in families with income below the poverty line have worst outcomes in virtually every dimension, including cognitive development, socio-emotional functioning, academic achievement, and labor market success compared to children who grow up in higher income families. And of course, having said that, it's important to make note of the fact that there are certainly children who grow up in poverty, who do as well or even better than children, who do not grow up in poverty.

0:36:01: But the lower average attainment of poor children means that the resilience of this kind, is very much the exception. The overwhelming majority of studies of linking income to children's development is correlational; that is income is found to be related to indicators of child functioning in naturally occurring circumstances. And these studies are, one of the limitations of these studies of course, as you learn in Psychology 101, is that correlation does not equal causation. So, these studies are susceptible to biases from unmeasured parent and family characteristics. Poor families differ from non-poor families in lots of ways, other than the fact that they have less income, including things like maternal age at first birth, family structure, maternal education, etcetera. So, it links between family income, and children's development could well reflect these unmeasured and measured differences, rather than income differences.

0:37:15: Now, most researchers of course, try to take these differences into account by controlling for them in their model estimates. And that typically will reduce the association of the link between income and child outcomes. And some of the studies that do this, that is that take into account these factors, are much better than studies that don't take into account those factors. But nevertheless, these correlational studies are insufficient to establish a causal link between poverty and children's development.

0:37:56: So, one of the goals, I think, of what we're doing on this national committee, National Academy of Sciences, partly is to convince policymakers that income really does matter. There are people who absolutely don't believe that. Matters in the sense that it has a causal impact on children's development. So, the big question is whether or not income has a causal impact on children's development. And in order to make that case, I think we have to have, or need experimental research. It's critically important in establishing this extent to which this association is indeed causal.

0:38:46: I wanna talk about some experiments in which families are randomly assigned to treatment and control groups. And the treatment group receives an income supplement, and people in the control group do not. One can also infer some causal impact of income from quasi-experimental research even when families are not randomly assigned to treatment and control groups, as when there is, for example, a change in policy. So, some people have looked at what happens when we see an increase in the EITC credit? Does that correspond to increases in children's outcomes, positive outcomes, or increases in the Child Tax Credit? I won't talk about that today. I'm trying to be very conscientious, and keep to my 15 minutes.

0:39:42: I wanna talk about these welfare experiments. Several of them were done, begun back in the late 1980s, early 1990s. Several states were granted waivers of welfare rules, in order to allow them to experiment with changes in the welfare provisions in their state. They were required to use a random assignment design, in order to get that waiver. Because the idea was that they needed to have evaluation research to indicate what were the effects of these changes. Some of the experiments had work requirements, and wage supplements. Others of them had work requirements, but no wage supplements. There were other variations, like time limits, and so forth. And I'm not gonna talk about those, 'cause I don't have time. But these earning supplement programs were intended to compensate for some of the shortcomings of the labor market, like low wages, no health insurance, and to make work more financially rewarding.

0:41:01: The other programs that I wanna contrast this with were, they had mandatory employment services, like education, and training, and immediate job search. And, there were about 10 of these, that are included in what I'm gonna summarize. And, as you can see here, the programs that had work requirements, and wage supplements, increase a mother's employment,'cause most of these women were mothers, most of these parents were mothers, increase employment, increase parental income, and they had positive effects on the well-being of children. Some of them, all of the programs I should say, had a positive effect on children's achievement outcomes, school achievement, by approximately 10-15% as compared to children in the control group.

0:42:04: And some, also reduce behavior problems, increase positive social behavior, and improve children's overall health. Those that had work requirements, but no waived supplements, increase parental employment, but they did not increase income, and they had no consistent, positive effects on child outcomes. And the second category of programs did not increase income even though mothers were working more because their income... They lost welfare benefits as their earnings increased. That's a reason that they were working more, but they did not see an increase in their income. I think that we can conclude, at least from these experiments, that there is a causal effect of family income on preschool and elementary school children's well-being. And as I said, there are other kinds of quasi-experimental data that reach the same conclusion. And some of you are probably aware of the casino studies that were done in North Carolina, showing positive effects of income, as well as, as I said the EITC and Child Care Tax Credit.

0:43:39: I just wanna tell you a little bit about one of the studies that I was involved in evaluating. It was one of these earning supplement programs. And as you can see from here, it was a New Hope Program in Milwaukee and the people who... It was a team of us who did the evaluation. Greg Duncan, Aletha Huston, Tom Weisner from UCLA, we were all part of this team that traveled a lot to Milwaukee back during that time and... What you see here are effect sizes. Children in the experimental group performed better on the Woodcock-Johnson Reading Test compared to the control group. Their parents also rated them as having higher levels of literacy skills. Their teachers rated them as having stronger academic skills and better behavior in the classroom. The thing about New Hope that complicates matters is that the effects were much stronger for boys than they were for girls. And I can talk about why I think that is, but before I do that, I wanna finish talking about what the general pattern of effects were.

0:45:14: This was a five-year follow-up. We followed these kids for eight years, actually, and I should say, this five-year follow-up, the income supplement, the program had ended at this point, but we continued to see positive effects mostly for boys. At the eight-year follow-up, we also see some positive effects. The children in the New Hope Program had less cynicism about work, they were less pessimistic about their future employment prospects, and they were more involved in employment and career preparation. We were very impressed by this and somewhat surprised; even though we did the follow-up, we weren't quite expecting to see a fairly consistent evidence of positive effects. The other thing that I forgot to mention is that, in fact, these programs were not conceived to have a positive effect on children. The programs were focused on getting parents into the workplace, increasing their employment and increasing income. There was no plan to have a positive effect on children per se.

0:46:44: I think with the New Hope Project, we just said, it's really important to evaluate how this program is affecting children. And fortunately, we're able to get funds to do that. It's all the more impressive that these programs had positive effects on children when they were not conceived to have positive effects on children. Okay. If income does have a causal effect on children's development, what is the pathway? I have done a lot of work looking at this second pathway. There are two major pathways that people have looked at to explain income effects on children's development. One is the investment pathway and the other is a stress pathway. The investment pathway essentially is about goods and services. It says, parents, children don't... Poor children do not function cognitively, academically, etcetera at the level of children, who are not poor, on average, because their parents don't have the resources to invest in them.

0:48:01: They do not have moneys to provide stimulating activities in the homes, stimulating educational materials. They don't have money to provide high-quality childcare. So that investment pathway really looks at resources and goods and services. The stress pathway on the other hand focuses more on parenting, with the idea that when you don't have enough money to pay your bills, it creates a lot of mental stress, anxiety, depression, conflict between parents and marital partners. And that in turn influences your parenting behaviors, and in turn influences parenting behavior specifically by increasing the tendency to be inconsistent, to be harsh, punitive, etcetera. And those are the ways in which... That is a pathway by which ultimately the child's development is compromised.

0:49:20: But the experimental findings clearly favor the investment pathway. Most of the research that's consistent with the stress pathway tends to be correlational work. There's a lot of work on the family stress model. I've done a lot of it, some of my students have done some of it too. But when when we look at these experiments and try to figure out what is the pathway by which these wage supplements had effects on children's development, in general we don't see that they have strong effects on, or consistent effects on parenting, on parents' mental health. What we do see, and Greg Duncan has done a lot of this work, we do see that consistently increased income and center-based childcare seems to be the key pathway by which these wage supplements affect the children's development.

0:50:24: So that's kind of the story around that. I just wanna show you for the New Hope Project in fact, this shows the effect of New Hope on childcare. So the experimental group, children in the experimental group or whose parents were in the experimental group, spent minimal hours in center-based care, and before and after school programs. They were less likely to be in home-based care and less likely to be in unsupervised care. So that's just one example from one of these projects, but Greg has done a synthesis of these programs looking at pathways. And on the basis of that, his analysis shows clearly that it's childcare that's driving this positive impact on children. So I was... In response to your request that we talk about trends that show good news, trends that show bad news, I wasn't sure whether you were talking about the phenomenon that we are looking at or the field itself.

0:51:46: So I decided to focus my comments on the, what's going on in the broader world, United States in terms of income. So I think the good news is that more states are setting a minimum wage that is higher than the federal minimum wage, I think it's outrageous that the minimum wage, the federal minimum wage is $7.25. I mean, it's just hard to imagine that that is the case, but it is. And I would say that I think it's great that states are setting a minimum that is higher, and that states are indexing the minimum wage for inflation, so that it is automatically adjusted each year for increases in price. One of the problems that low income... That women in particular with low levels of education have, is that they are concentrating in jobs in the service and the retail sector, where pay is low. But here's a statistic that is just very striking to me. In 2014 about a half of low-income children and about 30% of poor children under 18 lived with at least one parent who was employed full time, year round.

0:53:20: And what these statistics reflect is the fact that the current minimum wage of $7.25 is not a living wage. So I think it's good to see that more states are setting a minimum wage. I also think it's a great thing, a great pattern, a good pattern, that in recent years several states have enacted, earned income tax credits, or the have expanded their EITCs to bolster the wages of struggling families. On the other side though, a trend that I think shows bad news, is that more states of enacting laws that prohibit cities from increasing their minimum wage on the grounds that, having these within state variances in minimum wage, will pit local communities against each other and foster an unstable business climate.

0:54:15: There are a lot of other bad news trends that I could describe. I had a hard time coming up with the good news. Very hard. Michigan experience. I was not trained in policy. I was a traditional developmental psychologist. My training was in traditional developmental psychologist. I actually did laboratory research on children's play, as a graduate student. And for the first five years post-PhD, that was what I did. After I got tenure I decided, "I want to do something more interesting and important." And it turned out that I got involved in this Bush Program in child development and social policy, it was here at Michigan. And I had a lot of students who were interested in policy and poverty. And they corralled me into being involved in this. And I'm very thankful that they did.

0:55:21: And the other thing that was going on at the time, when I came back to Michigan on the faculty is that we were in the midst of this terrible recession. And we were constantly reading articles in the newspaper about how the recession was impacting families. So I got very interested in that, as a result of being the local economy, and was fortunate to be part of this Bush Program in child development and social policy. Unfortunately, it no longer exists. I was also fortunate to be part of an interdisciplinary poverty focus work group, at the Center for Human Growth and Development. And I think, in general, the big influence, I think on me, as a researcher was just the richness, the intellectual richness that exists at Michigan and the fact that there is always somebody you can go and talk to about an issue that you care about. And so I had long conversations with Greg Duncan years and years and years ago. So he was a terrific person, he still is. And I have enjoyed working with him as part of this National Academy of Sciences group. So I'm gonna end there and... That's it. Thank you.

[applause]

0:56:51: Was that what was left or I was over time?

0:56:53: You were over time, but it's okay.

0:56:53: Oh okay.

[laughter]

0:56:55: While everybody is coming up, I failed to mention that Erv Garfinkle unfortunately couldn't make it. He had a health problem just this week and wasn't able to make it. I encourage you to read about him in the program, if you don't already know all his great work. He's made many important contributions, as Vonnie alluded to.

[pause]

0:57:23: Hear this? So we are taking questions, if anybody has a question I guess raise a hand, grab a microphone. Don't put us on the spot. Thank you. [chuckle]

0:57:36: Thank you, I have a question for Maria. It was interesting to see these increases in the family complexity and I was wondering whether you've also taken a look at how that intersects, or how that relates to within household and out of household complexity. So as you pointed out, these complex families may be co-residing or not. And the cohort trends and the life-course trends would be quite interesting to see, sort of, broken out into co-residence or not co-residence. I think that's conceptually interesting, right? Because it's very different whether you have five half-siblings somewhere in the country, or in your household. It's interesting from a policy perspective, and it may even be interesting from a measurement perspective. I wonder whether that poses some challenges, to really capturing everyone depending on whether that intersects with co-residence.

0:58:28: So I'll agree with you that it's interesting and hard. And I have a couple of projects right now with Dan, and Lonny Berger and some other colleagues. One, trying to use SIP data to try and capture some of those things. The main thing we're doing on this, and I don't have results to share, but just an idea. So we're trying to understand resource transfers, both within the household, but then from people in the household to those outside and vice versa. So we've used the administrative data that I did the analysis with on complex families, to draw a sample and then survey moms four times in the course of a year and ask them about resource transfers from them to other people, and from other people to them. And what we've done is focus particularly on food. Because we care about other things like rent and other things, but those transfers tend to be lumpy and fairly infrequent.

0:59:28: So we actually are asking moms for every person who they've provided a meal, or taken out for a meal, or bought groceries for, and every person who has done the same for them. And we ask about all the people who are in the household. We ask about all the fathers of any children who are in the household, and then we ask about everybody else, very explicitly. So this is long for those who are survey methodologists. We had a lot of fights with our colleagues at the survey center.

0:59:53: And on the methods question, one of the things that the University of Wisconsin survey center developed, and this is Nora Schaeffer's work, is a kind of interactive collaborative interviewing technique where people tell us about all the people in their life, and then the surveyors kind of enter that information and it appears on a little laptop that the person who is talking to us can see. So they can see a representation of the people and then they can tell us about the transfers between them and those people. Nora has actually and her team have actually developed new survey techniques that are required in order for us to be able to get the kind of information that we want about this complex relationships. And we're going in and interviewing every four months because they're so dynamic.

[pause]

1:01:00: I have question for Alexandra. When you did your wealth curves and I saw the bottom 25% went down, I really would like to know what the wealth was at that 25th percentile and are you doing, including negative wealth, so is it possible for... Okay.

1:01:24: Yes. So I can just say briefly that one of the advantages I think of looking at percentiles, which we did in that graph is that it's easy to incorporate zeros and negatives. So the 25th percentile if I've been reading... Correct me if I get this wrong, but it stays positive over this entire period is that... Yeah, right. And so it is a low level... It's the fraction of Americans who are in debt when you add up all their assets and subtract what they owe is less than 25% for this entire period. But it is a real issue of how to think about those with zero and negative net worth, and I think there's some interesting methodological issues. People have often just done the log of wealth, but then there's this question of what to do with it, so that's one of the reasons we use the percentiles.

1:02:12: Big thank you to all of you for your great talks. Vonnie, this is David Williams, I have a question for you, it's been a long time. Could you comment a little bit more about the gender differences that were observed in the New Hope Experiment and just in a lot of these early childhood interventions like in the Perry Preschool, Study B finds stronger effects for girls and boys. In the Abecedarian Project, we find stronger effects for boys and girls. Just how do we make sense of some of these gender differences in New Hope, but more generally?

1:02:47: Unfortunately there was a childhood [1:02:57] ____.

1:04:33: I just wanted to push you all to talk a little bit more about the links between research and policy in this area. It's an interesting area and Vonnie you talked about it a lot and you all alluded to it to some degree, but these areas are areas where there's been a lot of... Vonnie as you said even the welfare to work programs have resorted to this mandate of having randomization so that things could be learned from it. It's an area where we've had a lot of at least intention of having research and forum policy. But I wonder if you could all just reflect on how well does that work, is it working, is the research really linking to policy, could more be done to do more in that regard?

1:05:21: I don't wanna start 'cause I'm a huge cynic about the possibilities for policy interventions and wealth and equality. I don't think that people really fail to know the relevant facts here. I think they just don't care that enough to change it. And so I was really impressed in yesterday's session. There were some question about what policymaker would you talk to and what would you say? And I was so impressed that everyone had a really great answer to that. And I thought you can't say to a policymaker, "Hey wealth and equality is really big and it matters who your parents were and also it's really different by race," they already know those things, they just choose I think not to care. That's why I didn't wanna go first.

1:06:04: Okay.

[overlapping conversation]

1:06:05: 'Cause I thought that it's gonna become more... Not usually the half full person, but I'll be the half full person.

1:06:07: Yeah, great. [chuckle]

1:06:09: So I agree that when somebody's not open to an answer, this very little that evidence can do. At the same time, I think that a tremendous number of the policy issues... Issues the policymakers face are just on the margin fairly small changes that can actually have a big impact. So it might be... And in those cases, I think research can really make a difference. So someone, I think Celia Munoz noted yesterday, that early in the Obama administration we realized that we were expelling a lot of African-American boys from pre-school from Head Start Program. And it occurred to folks that that wasn't what Head Start was supposed to be doing, and we changed policy. Now that's not a big... That's not re-structurally, but it's something on the margin and a lot of the work that I've done that has, I think had an impact say in Wisconsin in the last eight years under Governor Walker who maybe has a different perspective on some issues than I might, have been about saying we can save money and make things better for families on this little margin.

1:07:21: And so in that context, not overall change, 'cause I think there's not so much reason to be hopeful in light of that. But on the margin, I think things like the evidence-based policy commission, which is a bipartisan effort. Efforts to try to use data to actually inform policy and practice, I don't think it builds all that well for illuminating these structural oppression, which is probably not caused by people failing to understand that it exists, but at least making the programs that we have. That are designed to offset some of the negative consequences of those oppressions to be more effective and efficient. So it's a smaller thing [chuckle] but I do think that there's a role for that.

1:08:11: Yeah, I too am very cynical about this. Even though I'm spending time on this National Academy of Scientists Committee, we are mandated essentially to come up with proposals for reducing childhood poverty by half in the next 10 years. And I'm doing it, I'm involved in a... We're doing policy simulations to see if you with certain assumptions for different kinds of policies, how much does it reduce poverty, childhood poverty? I think I find this all very interesting, but how much of a difference it's gonna make is very uncertain, but I think it's useful to be able to give this information to people who do wanna make a change. You give them the scientific information and hopefully they're in a better position, or their staff, to use that to back up their argument. This committee was a established. It was before the election and like everybody else, we thought Hillary Clinton was going to win and so everybody was really enthusiastic that our report would get, would be received very favorably. It was mandated by Congress that this committee be set up and we're supposed to submit a report by the end of the summer. And then after the election, everybody was really depressed like, "Why are we doing this?" But I think people do feel very, very committed to seeing this through and optimistic that maybe something will come up with it.

1:10:21: I don't know, I think David's comments this morning were very much on target in terms of ways that you present the message. That you'd better likely to get a good hearing and people are more likely to listen to what you have to say. It also helped us that on our committee we have people who have tremendous experience in working with Congress.

1:10:53: I'm in for a moment. If you can you hear me? I'm actually a... I have nothing to do with this personally, but I'm on the Academic Research Council for the Consumer of Financial and Protection Bureau and one of the things by the bureau coming online is that they've been able to get access to a lot of administrative record. They have power to get these records then do lots of things. And so one of the things that you may have seen about recently is that after looking, doing the research and looking at a lot of facts about payday lending and who's re-updating their allowance etcetera. They've just entirely rewritten the financial rules for payday lending. And so I was talking to... Well, anyway, somebody who is a former [1:11:35] ____ alum, who is working for the CFPB and one of the things that you see is a lot of the money that the industry makes from payday lending. It's coming from the people who not surprisingly are keep getting the loans. These regulations are basically gonna cut that tail, which is a lot of their... It'll be interesting to see how this goes for, but I think that is one place.

1:11:54: So everything they're doing is informed by the research and having the access, so I think getting the access to the data related to this being able to use the gold standard of the randomized clinical trials that you were mentioning before, is also another way that I think, with big data, we're gonna see other differences that people can make doing research, but having the facts to do it is tough, because also getting the proprietary, firm data is always gonna be a challenge. So, it's nice that some agencies have this access, for as long as this agency is going to be around, I suppose, but...

1:12:36: [1:12:36] ____ PhD. I'm also a fourth generation activist for policy. The last time I was in this room, was a conference on how professors have a responsibility to cope with their own cynicism and not share it with their students, so that we end up graduating early cynics, because that's really a shame. And that we combine our research with examples of skills for democracy. We had that one session yesterday where they, "Here's what you say. Here's how you do it. Here's how you listen. Here's how you connect. Here's how you build coalitions. Here's how you use leverage points for change." And, I am hopeful that as you take this research you're doing, you combine it also with assignments, so it doesn't all have to be in student life and extra programs like project outreach and service learning. So that students actually come out with the skills they need to be not only systemic thinkers, but effective change agents.

1:13:44: Getting a little bit more specific on the question, 'cause I wanna know what you have at each of your institutions that are actually institutionalizing that, because that will institutionalize systemic positive change, but also combining your research on wealth, with the research on, and if you don't fix this, you're gonna have higher incarceration costs, and higher these costs, and higher those costs and this is just not cost-effective for those who can't empathize with human suffering, but can respond to where they come from in their value system.

[pause]

1:14:38: Well, I have to say, I was never told that that was part of my responsibility. And I think there's just such a distance from research to policy. But, depending upon one's values, you do what you can. So, I don't do very much work, research, in the community as such, but I do other things like, last year, year before, I chaired a committee that was established by the Provost office, to envision what a poverty initiative would look like at the University of Michigan. The Provost said, "A lot of faculty are really interested in this topic. What, might we do to make a difference?" And so, this was a committee of about 15 faculty from around campus, from various schools and colleges. From the medical school, to the law school, to the business school LS&A. And we came up with a proposal. We met, every week, for about six months. And we came up with a proposal that did eventuate in a new institute, called Poverty Solutions, that is chaired by Luke Shaefer, who's a Professor in social work and public policy. And so that's my contribution to this process, as well as working on this national committee, National Academy of Sciences Committee. But I think there's a sub-diversity of things that people can do and you use what skills you have, in the service of that. So, there's some things I don't know, don't have the skills to do, and I acknowledge that and I'm okay with that, so...

1:16:51: I would just add to that, that in addition to being good teachers, I think one of the things that institutions, academic institutions, can do, is to create environments that are hospitable to engaged research. And so, if we were in Wisconsin, I would talk about the Wisconsin idea, which is the idea that the boundaries of the university extend beyond the campus and engagement. And that's really nice, I think if you look at tenure rates, you would probably find that, maybe it wouldn't be a negative correlation between how engaged your research is and you're probability of getting tenure, but it certainly wouldn't be super positive. I think people are really taking that on. So, one of the things that I would highlight, the WT Grant Foundation, which is really focused on how research gets into practice, just announced a new competition for projects that bring academics and practitioners in close collaboration.

1:17:51: But, one of the things that's unique about it, is it requires the academic instituiton to talk about how it's changing the rules of the game, what resources it's putting on the table, for making work that leads to real change on the ground to be valued. And whether it's in the tenure role or in different places. So I think there's an increasing both of focus on, increasing sophistication on kind of, how to talk, like it's not getting over the idea that academics have all this wisdom that we will just rain down on people who do real work. So I think most of us have gotten over that, but now I think there's actually some more self-critical focus on what's the nature of the academic enterprise and how in the classroom as you've pointed to, in terms of service learning and things like that, but then also in terms of the fundamental structure of the university and the kind of scholarship that we value. How do we have to change the rules of the game of a university in order to encourage that kind of work.

1:18:56: I should also add that in terms of the policy solutions institute I was just describing, it was very important to us to make sure that that was a component that was dedicated to having involvement of undergraduate students. And so this center gives priority to demonstration projects with random assignment to show the effectiveness of different policies related to reducing poverty and related to reducing the effects of poverty. So we looked very closely at that and that was really important to the committee, to have something that focus specifically on undergraduate students, as well as graduate students and post docs but, yeah.

1:19:47: I'm not cynical about the idea that we can reduce wealth inequality. I think, to sort of what Maria was saying, and to David's point on the previous session, I don't think there's a fact I can give people that would change their mind about the policies they should adopt at wealth inequality. And so I think for me, I don't always see the role of my research and the things I believe in and advocating for, as in this sort of dissemination framework. I think there's other things that might be the most relevant strategies to pursue. In terms of the things that universities do, one of my sort of formative experiences here at Michigan as an undergrad was I worked for the Sexual Assault Prevention and Awareness Center and for reasons that escape me, the director let me redesign how we collected surveys about the effectiveness of our peer education program. And so I got to design these pre-tests and post-tests and cross designs so there wouldn't be priming effects.

1:20:46: And of course now I think, "Wow. That was really an amazing research opportunity for me." And something just told her that will be okay for me to do. And similarly at Harvard I've been on a committee that awards theses prizes and one of them went to a student who thought that the way that the university student health services was collecting surveys on student health throughout their time in Harvard wasn't done very well, and so she re-did it, and they partnered with her and they adopted it. And I think those are wonderful experiences for students to have also, where they get to really apply in the social sciences, the skills that they're learning to a very real, very immediate, very short-term thing on their own campus. It's a great kind of practice and motivator I think for then taking that out in their post-college years.

1:21:37: I'm going to go back to the question before, and I really want to do a shoutout sort of for Vonnie because I think she among all the scholars I know has... She's worked so hard in designing experiments and in figuring out how you actually change things. For example the example you gave about center-based care and after school care. There's a reason Obama had a preschool initiative. There's a reason that schools now are for very different options for after school care particularly for low-income families across the country. So I think it takes a while, but her work has really changed the world for smaller children in the US. She may not wanna say so but I do.

1:22:40: Hi. I'm Otis, I'm at WashU. Two things that I think the earlier session suggested that the way that we frame our work really may be the link to policy. And so now we might not be talking about wealth inequality per se. But most of the work including your own, Alexandra, suggests that housing is a real contributor to wealth. And so we also know that during the Great Recession, that's how lots of families especially of color lost wealth. A loss of status relative to other groups. One of those programs was HARP, which is the Housing Affordable Refinance Program, which is set to actually end next year. And so what happened with that program is that a lot of African-Americans, a lot of people who were in these high interest rate loans ended up refinancing, avoiding foreclosure. And so I'm curious about the opportunity costs. In other words what would have happened to black wealth in the absence of that policy? So I see direct linkages between what you do and some of the things that are contributing to wealth inequality.

1:24:02: And so, I'm just hoping that, and I'm very cynical. I mean I've been in Appropriations and Authorization Committee's meetings, and I know how they are. But at the same time, if we do not marshal what knowledge we have and apply it to the pressing policy problems, the other side, hate to sound polemical, but the other side will. And so we need you in the game and actually applying your great work to a lot of these issues.

1:24:37: Again, I think that my cynicism is a little bit... I didn't say it very well to explain what I mean, which is just that I was attempting to echo what was said earlier that I think the depth of inequality in America is not an unknown fact. And so I absolutely think there are places to intervene in policy for wealth, and I absolutely support that. So, an amazing example, which we're just talking about last night was that an early draft at the Republican tax bill proposed getting rid of the mortgage tax benefit for second homes and capping the mortgage deduction on first homes, at $500,000 rather than a million as is present.

1:25:21: Well, that's a sort of interesting, redistributive, progressive idea to be in the draft for the Republican tax plan. I don't think that's gonna be what's in the final version, but absolutely I think we can advocate on policies like that. That'd be one really important way to intervene. So, again, all I'm saying is I don't think my research's saying, "Hey, who your parents are really matters" is the fact that's missing from our policy discussions. I think we need different kinds of conversations and again back to Maria's point, I think when scholars come in and say, "Hey, you probably haven't thought of the right... My brilliant scholarship yet," I think that can be counterproductive and we may need to shift a little bit to say, "I work with Big Survey Data about general patterns in the population." That may not be the relevant way to get into the weeds and talk with policymakers about whether we can shift that deduction from a million to $500,000.

1:26:17: I think that's all of our time for our session. So will you please thank our panel members.

[applause]