A New Experiment on Cash Transfers & Childhood Development, feat. Greg Duncan

An exciting new experiment, led by Greg Duncan of UC Irvine, is launching to explore how cash transfers affect early childhood on biological, cognitive, and social metrics. Duncan joined the podcast to discuss the motivation for the experiment, what it will study, and the research to date on poverty and children.

——

Episode Transcript

Owen: Hello, and welcome to the Basic Income Podcast. I’m Owen Poindexter.

Jim: And I’m Jim Pugh.

Owen: So there is an exciting research experiment that just launched in the US called “Baby’s First Years” that will provide new mothers with unconditional cash, and the researchers are interested in everything from how the money is spent to how this affects neural development.

And so recently, Jim got to speak with one of the lead researchers. His name is Greg Duncan. He’s a Distinguished Professor at the University of California Irvine. So here’s Jim and Greg discussing Baby’s First Years.

Jim: Thanks so much for joining us. Now, to start with, can you just tell us what generally the work you’re doing focuses on?

Greg: I’m an academic economist, and I’m interested in social policy of various kinds, and I’ve long been interested in the effect of inequality and income inequality in general and poverty in particular on the development of children.

And I think one of the things we’ve learned over the last 20 years from neuroscience is how important the very early years of life are, infancy and toddlerhood, because that’s the time when a lot is happening with kids brains. And it’s a period when kids are particularly susceptible to enrichment in their environment as well as insults in their environment.

So, the idea that economic deprivation caused by poverty might be especially detrimental for early in life has been a focus of my research. And it’s also a very important policy issue because over the last 20 years since welfare reform, we’ve oriented our safety net more toward working parents — Earned Income Tax Credit, child tax credit, those are much larger now than they used to be, and they’re going to working families, generally families with somewhat older kids, and families who are further along in the life cycle then the families just starting out with newborns.

So there’s a substantial amount of poverty among low-income kids and especially young kids, and the question of whether income is an active ingredient in producing all the correlations that we see between poverty and child development.

One of the more noteworthy studies coming out of neuroscience recently shows that white matter in brains between kids who are in low-income families and higher income families is very similar to about six months of age. But by 36 months of age, there are these growing differentials, and we don’t really know whether it’s income poverty or whether it’s other conditions associated with poverty — things like single-parent family structure, low levels of parent education, neighborhood disadvantage.

What we’ve done with the project that we’re just starting is to try to isolate the impact of income itself on young childrens’ — infant and toddler — cognitive development and socioemotional development. Hence our experiment that’s just starting out now.

Jim: Now, I think when a lot of people think about the academic world, there is often an assumption that people are doing analysis but aren’t actually getting into really changing the way that our systems and policies operate. But this experiment you just referenced, Baby’s First Years, this is actually an instance where you are going to be providing unconditional cash payments to a number of low-income mothers and seeing what happens. So, can you just tell us more — what are the details of this experiment? What are you hoping to learn from it? What will that actually look like?

Greg: Sure. I mean, we think about clinical trials all the time in medicine, and we don’t trust drugs unless they’ve passed a muster with clinical trials. And with clinical trials, you have random assignment between some sort of treatment condition and some sort of control condition.

So we’re just taking that technology and applying it to income supplements. Specifically, we’re recruiting a thousand mothers in four different sites around the country. All the mothers have incomes below the poverty line. All have just given birth — we’re recruiting them in birth hospital shortly after they’ve given birth. Their participation is completely voluntary, but we ask if they’re interested in enrolling in a chance for a cash gift. And we’ve got two levels of that cash gift: either $333 a month for the first 40 months of the child’s life, or $20 a month for those 40 months. So 333 versus 20, it adds up to about four thousand dollars a year over these 40 months.

But the payment that’s monthly is trying to be aimed at the kind of instability, economic instability, that a lot of low-income families have. So the question is: to what extent are the babies going to be developing along more positive trajectories if they’re in this group of families where the mother is getting $4,000 a year versus $240 a year? And we’re interested not only in the comparison of how the child does. We’ve got, at age three, we’re planning on bringing the kids into neuroscience labs and doing all sorts of testing on their cognitive capacities and associate emotional behaviors, executive function, and things like that.

So we’re interested in the child, of course, but we’re also interested in the families. So what if the child does better with this higher income payment, what is it about the family that’s changing in a way that the produces that improvement? Is it the kinds of enrichment goods and services that parents buy, center-based childcare, for example? Is it the fact that this steady stream of income widens the bandwidth, cognitive bandwidth, of mothers and enables them to stay on top of things better? Can they avoid addictions? Can they avoid food insufficiency? The various things that can happen and often happen to low-income families that some sort of income floor might smooth out.

So the idea is to run this experiment, recruit the mothers over the course of 12 months and then visit them when their children are age 1, 2, and 3 and track what happens in these two groups, both in terms of how the kids are turning out as well as what sort of differences are showing up in family process.

Jim: Now, we often talk about motivations for programs like universal basic income and unconditional cash generally on the podcast, and a number of our past discussions have looked at it from different perspectives as to why that could be a particularly valuable approach to take. I’m curious for you, what were the main motivations to look at unconditional cash here as opposed to more traditional in-kind benefits like food or baby supplies?

Greg: Yeah. So we’ve got a safety net that consists of SNAP, which is the food stamp program, housing subsidies, various programs that subsidize different kinds of goods and services for parents.

But if you think about the way social benefits broadly conceived are structured, we already have a child allowance in the United States. We have in our income tax system, we’ve got an exemption for every child who’s a dependent. We have a child tax credit. We’ve got a additional child tax credit. We spend 100 billion dollars a year on what amounts to an annual child allowance, but it’s only reaching middle class families and richer families. Families that aren’t earning enough to be part of the income tax system don’t benefit from that.

So the idea really is to think of taking those benefits that are now concentrated not at the low end of the income distribution and thinking about a more general universal child allowance that would be received, you could imagine different forms that, you could imagine simply universal child allowance for families at all income levels are getting this in recognition of the fact that they’ve got extra expenses because they’ve got kids in their house.

So, you know, the more general universal basic income movement is really conceived with adults in mind, and we’re focusing on some kind of universal income that’s focused on kids. Although our test really isn’t of a universal kind of child allowance, but it’s more what might happen if you had that for low income families, for families with incomes below the poverty line.

Jim: So, can you tell us a bit more about what the research landscape in this area looks like to date? What previous findings were you drawing on or looking to when thinking about how to design and execute this study?

Greg: Sure. There have been hundreds and hundreds of studies linking poverty to child development in various ways. A lot of it is just correlational, where you observe whether the kids are poor or not as they are growing up and then you see whether they complete more education or they earn more or they’re healthier. And of course you get those correlations. You get particularly strong correlations with income in early childhood, which is kind of interesting.

There are what social scientists call quasi-experimental studies. There’s one that Hilary Hoynes did of the rollout of the food stamp program back in the 60s and 70s, and she tried to see to what extent for low socioeconomic status families kids who were conceived and born into counties that were already receiving food stamps were doing better in adulthood than kids who were conceived or born into counties where food stamps weren’t available — it was a county-by-county basis — and finds this remarkable relationship where adult health, cardiovascular health is substantially better for kids who are born in the counties where there’s already food stamps compared with kids not very far away across the county line where food stamps weren’t available.

So that’s very intriguing. And there are some genuine experiments in the 70s and 80s. There were called negative income tax experiments, where they actually randomly assign people to get different amounts of income. There were welfare reform experiments in the 90s.

But none of these studies really concentrated on very young kids. So if you’re particularly concerned with the effect of economic disadvantage on infants and toddlers — the food stamp study I mentioned is very much focused on early life, but all of the other kinds of experimental studies really are focused on older kids, so they often found differences, but you really can’t generalize down to very early life on the basis of studies that are conducted on kids who are already in elementary school or high school.

Jim: You mentioned a number of different areas that you’re planning to look at as far as what the impact of this treatment might be. Can you say a bit more about what specific data that you’ll be collecting in order to be able to assess that?

Greg: Sure. You’re always trying to be copycats when it comes to growing measures for our study. It’s very nice to be able to use measures that were used in national studies so we can compare our study to results from larger and more representative samples.

But there are standard cognitive tests, IQ type tests that we’ll be using for the kids. There are standard assessments of executive function, which is kind of working memory and impulse control.

At age 36 months, you can’t put kids into fMRI scanners because they don’t nap reliably, and they get frightened. But you can put caps on their head to measure electrical activity, and that’s what we plan to do. EEG and ERP are the initials for the kinds of electrical signal data that we’ll be collecting.

So the idea is, around 36 months, kids are starting to develop an ability to regulate themselves. So when there’s an impulse — the three-year-old is hit by a playmate in preschool — is that child going to hit back? Now the instinct in the amygdala is to strike back right away, but then the prefrontal cortex might say, well, maybe let’s wait on that. But that’s going to require a connection between the amygdala and the prefrontal cortex and that wiring starts to take place around this time. So the question is whether we’ll observe stronger connections in the kids who are in the higher income families relative to the kids who weren’t.

So that’s on the child side. And on the family side, there are just many, many different kinds of measures we’re going to be drawing. We want to know about expenditures, to what extent are family spending their money differently between the two different groups on things like childcare? Are they moving to a better neighborhood? Are mothers working less? Are they using the opportunity for this higher income to take more months out of the labor force before re-entering in order to be with their children for the first three or four months of their kids lives?

We’re also interested in stress. We’re measuring that both with questionnaires as well as biologically with the stress hormone cortisol, which you can get in saliva as well as hair.

We want to know about the sensitivity of parenting. And for that, there’s kind of a standard play task that you can give to mothers with their kids. This is going to be at 24 months, when the kids are 24 months of age. You videotape it, and then you code it up for language. You code it up for the sensitivity of parent interactions.

So, we have a lot of, we have ambitious plans to gather a lot of information, both about the kids as well as the family.

Jim: So, I think most people think of providing unconditional cash as a super simple process: you literally are just giving people money. But when you actually get down to the details, there are some obstacles and complications that can arise. Can you tell us about what challenges you faced in getting this experiment going and how you dealt with them?

Greg: Yeah. That’s… we’ve taken about six years almost to get this project off the ground. And some of it was fundraising, but it turns out, as you say, to be rather difficult to give people mone. And one reason is because of the safety net that’s in place. So if you give families an extra $100, and they report that to the food stamp, the SNAP program officials, food stamps are reduced by about $33 as result of the additional hundred dollars of income. And for housing subsidies, that’s the same kind of thing. You get this chunk of the benefits reduced as you get additional income.

So in order to make sure that the net amount of income increase for the families was the full $333, we had to work with the states with all of our sites as well as a lot of localities to get either legislation or administrative rules changed so that income from our payments did not count as countable income for the determination of any of these benefits, So these are cash gifts, and we’ve worked out details, sometimes painfully, to make sure that caseworkers know that it’s not proper to count our income as countable income for the termination of benefits.

So we enrolled our first mom yesterday, which is very exciting, and in the next couple of months, we’re going to be doing a lot of work on the ground to make sure that all of these exemptions that are being secured are working for the moms, so they’re getting the full amount of money that we’re providing to them on their debit cards.

Jim: Well, that’s great to hear that things are actually really off the ground. Those are all the questions that I had — is there anything else that you’d like to add here?

Greg: I might say, you know, the connection between our findings and public policy are tenuous. In public policy debates these days, one side has one set of evidence, the other side has another set of evidence. At least in the welfare reform debate in the 90s, evidence from social experiments, where there was the kind of random assignment that we’ve got, was widely respected by a substantial swath of legislators on both sides of the aisle.

Whether that’s still the case, we don’t know. But I think one needs a random assignment experiment in order to have a possibility for influencing policy. And once we get our results in hand, we’re very much committed to try to inject these findings into debates that are taking place every year in federal government as well as state legislators all around the country.

Owen: That was Jim Pugh and Greg Duncan on the Basic Income Podcast.

Jim: So, it’s great to see another experiment on unconditional cash being moved forward in the US. I think that the more pilots, the more experiments we can have, the more data we collect, and the more stories we can tell, the easier it’s going to be to make the case for more policy. That as this becomes more of just a thing people do, it’s going to, it keeps moving the ball forward.

Owen: Yeah, and I think this delves into territory that we touch on a little bit just through our work in the basic income space, but doesn’t get a ton of mainstream media exposure, that how you grow up, the conditions in which you grow up, and the effects of poverty can affect your actual brain, and you can see those effects very early on within the first few years.

Jim: Yeah, and I think the fact that there is such rapid development then means that the effects of just a couple years of additional support could really be massive here. And so often we talk about, well, if you’re an adult, if you get support for two or four years, how much difference does that make in the longer term? That even if it’s helpful, that if you look at where you are 10, 20 years down the road, I think there’s still some questions as to whether that impact is going to be terribly large.

But if we’re talking about children and we are thinking about the development aspect, then two years of a different environment — yeah, that can be a game-changer.

Owen: Yeah, and it completely gets around the whole idea of, well, did this person make some mistakes? Were they responsible with their money? All the things that you hear that are attempting to kind of skirt around the impacts of poverty and the effects of that cash could have. No one’s going to say that about a one-year-old, and we’re talking about one-year-olds here.

Jim: Hopefully not. Hopefully we’re not in that place.

I also thought — and we’ve talked about this before — but another great example of how complicated it can be to actually administer paying people money. And that to actually be able to move ahead with this experiment, they had to get, they have to work with state legislatures and get approvals so that they didn’t kick people off their are other benefits plans.

And so I think that’s just another good example of how, as we’re thinking about moving forward here, we need to be ready for it to be more complicated than we first assume it might be.

Owen: Sure, and also, huge credit to them. They got legislation passed in multiple states to make this happen. And obviously, that’s no small feat. Even if it’s good legislation, it’s hard to get legislators’ attention to get something through. And so, yeah, big credit to them, and also, yeah, a sign that this is going to be a long complicated fight.

Jim: Yeah, certainly. I also, and we’ll see what happens, but going back to your point about the different perceptions there are around children, I think that we’ve seen in the past that it tends to be much more politically viable to push for more supportive policy towards children.

And so, if there are some compelling findings here after this experiment is over in a few years, that gives a strong reason for enacting this sort of policy widely. Maybe that sets us up to actually move pretty quickly on some more ambitious legislation on this front.

Owen: Yeah, some people propose a child allowance as the first step toward a basic income. And I go back and forth on that because I think there’s so many reasons to give an adult a basic income. But, you know, I think that it might be the most politically viable path.

Jim: Right. There is the question of, how well does that set up for everyone else getting in money? But it is, maybe, a more easy policy to start giving money out to some folks at least.

Owen: Yeah, and would be hugely impactful.

Jim: Yeah.

Owen: Well, that’ll do it for this episode of the Basic Income Podcast. Thank you so much for listening. Thank you to our producer, Erick Davidson. If you like what you hear, please rate us and review us on Apple Podcast or the service of your choice. And do tell your friends — we’re always looking to bring more people into the conversation.

See you next week.

Leave a Reply