UNIDENTIFIED PERSON: This is PLANET MONEY from NPR.
KAREN DUFFIN, HOST:
Once upon a time, in a cubicle not so far away, sat a government bureaucrat in his government-issued chair, preparing for a very big meeting. It was 1994, and Mark Shroder was an economist at HUD, the Department of Housing and Urban Development. And his team in D.C. had just flown in local public housing authorities from five major cities.
MARK SHRODER: Baltimore, Boston, Chicago, Los Angeles and New York.
DUFFIN: The staff gathers in a basement conference room with about 30 people. These are people who run what's now known as the Section 8 voucher program, which is a voucher that subsidizes rent for low-income families. It's like a monthly payment to a landlord. Families can wait years on waiting lists just to get one.
Everyone settles into their chairs, you know, sipping their government-brewed coffee, and the meeting starts. They learn that they are about to join a test program, small in the scheme of HUD overall but a huge change. They're going to start handing out a new kind of voucher to a small group of their tenants. And then HUD is going to run an experiment on them.
SHRODER: Hasn't been an experiment quite like this ever before.
DUFFIN: With the old vouchers, people could, at least in theory, move anywhere they wanted. But people who get this new voucher, they are required to use it to move to what HUD calls low-poverty neighborhoods - so just a nicer neighborhood. This is sort of a radical change for HUD.
SHRODER: A big part of the program had always been unlimited choice of neighborhood.
DUFFIN: Unlimited in theory, at least. But what the majority of tenants have chosen was to use the vouchers to stay in high-poverty neighborhoods for a lot of complicated reasons, including discrimination. Landlords often refuse to take Section 8 tenants.
So back in the basement at HUD, as this big, new idea gets introduced, I imagine people, you know, put down their government-brewed coffee and start raising their hands. Like, why are we doing this? Our job is to move people into housing, period. Now HUD is basically asking them to reverse that, to move people out.
SHRODER: I couldn't believe that public housing authorities would go along with inducing turnover among their tenants.
DUFFIN: Mark and the people running this meeting, they're like, yeah, I know, but now our job is to move people to opportunity. You know, the government spends money across all kinds of programs just trying to improve people's lives, like give them better education or health or increase their income. And the big idea that they want to test is, can we improve all of those things and more all at once just by moving a family somewhere new? But, of course, to get the government to give money, you have to prove that it works.
SHRODER: If you want to really know whether this works or not, you need an experiment.
DUFFIN: You know, control group, independent variable. We will run an actual, like, scientific experiment. And we'll track the data for decades.
SHRODER: The only randomized experiment to test the impact of neighborhood on people's lives.
DUFFIN: Can changing someone's address change their life? The answer to that seems very obvious, but this experiment did not at all go as planned.
SHRODER: Almost everything about Moving to Opportunity has been a surprise not only to me but to practically everybody else.
DUFFIN: Hello, and welcome to PLANET MONEY. I'm Karen Duffin. Today on the show, we are riding a social policy roller coaster alongside the researchers who ran this experiment and the people they tested it on, people who were not always excited to be experimented on. It is a quest to make the American dream a reality, a quest that economists thought failed but is in the midst of an unexpected revival.
(SOUNDBITE OF JONATHAN ANTHONY JAMES AND SIMON PETER WARD'S "WAY HUGE")
DUFFIN: Mark and this tiny band of public housing staff leave that conference room, go forth from D.C. and start signing up tenants. The way it worked was this - tenants from the five chosen cities would be assigned by lottery to one of three groups. Like any good experiment, there would be a control group. This group already lives in public housing and they would stay in public housing, so no change to group one. Group two would get a regular Section 8 voucher. This is the rent subsidy that they can theoretically use with any landlord who will take it. And Group three - this group is the experimental group. They would also get the Section 8 voucher, but they had to use it to move out to a completely different neighborhood, one where just 10% or less of their neighbors were what the government has classified as poor. For context, most of them had been living in neighborhoods with about 50% poverty. So HUD was asking them to move to what should hopefully be a better neighborhood. But not everyone was excited about that.
SHRODER: There were some people who were praying, literally praying they would be in the regular voucher group because they did not want to have to find another place that far away either socially or physically.
DUFFIN: When HUD surveyed families about this, the families said, yes, we do want a voucher, but we want it just because we want to make sure our house is decent, and we want to get into a safer neighborhood.
SHRODER: Only like No. 3 or 4 was better schools for my children, better jobs for me. Those were way down there.
DUFFIN: For them, housing was just housing, not opportunity.
SHRODER: I think they didn't buy into the hypothesis we were testing (laughter).
DUFFIN: Was that disappointing?
SHRODER: I have to say that I wasn't that surprised. We were asking them to do something that people were not doing.
DUFFIN: The vast majority of people who get vouchers do not use them to move to areas with lower poverty. Also, Mark and his HUD crew were asking people to participate in a science experiment on themselves, asking them to be living test subjects for a theory they have not proven yet, which is an especially hard thing to take from the federal government - a government that has a history of intentionally segregating people of color into high-poverty neighborhoods.
The experiment took years to set up. But in the end, about 4,600 families were part of it. And in 1994, the moving began. We were not able to talk to the families that moved, but NPR's Morning Edition did cover the study while it was happening. Steve Inskeep spoke to a mother named Shirley Hudnall (ph) who had moved with her 15-year-old son Brian (ph). She talked to him about some of the things that do make these kinds of moves hard.
(SOUNDBITE OF ARCHIVED NPR BROADCAST)
STEVE INSKEEP: When Brian moved out of Baltimore, he lost touch with his friends. He was expelled from junior high school for fighting with his new classmates.
SHIRLEY HUDNALL: I was like, have I failed? Or what is really happening? And the thing that I found out that it was him having to adjust feeling that I was taking something away from him until I really sit down and talk to him that this was for his betterment.
DUFFIN: For four years, across the five test cities, people like Shirley moved, and the researchers waited. And in 2008, the researchers finished gathering the data. And what it told them surprised them.
SHRODER: Impacts we expected, in many ways, didn't happen. Impacts that we didn't expect that did happen.
DUFFIN: For one thing, moving was better for girls than for boys.
SHRODER: Which we certainly did not expect.
DUFFIN: Parents reported better health, something HUD hadn't even originally planned to measure - improved mental health and physical health.
SHRODER: We saw impacts on diabetes, obesity.
DUFFIN: All of which, of course, is wonderful, but this is not what the experiment was designed to test. What they wanted to test was the, quote, "long-term housing employment and educational achievements" of families involved. And this massive, scientifically designed and rigorously tested social experiment that had moved thousands of people just to answer the question - can changing someone's address changed the course of their economic life? The answer to that question was no.
SHRODER: We did not find any impact on children's test scores. We did not find any impacts on grown-ups' earnings.
DUFFIN: Basically, no impact on educational outcomes, employment or income. If you want to improve those things, the final report said, housing is not your answer. Housing is just housing. And with that, all of the hopes and dollars and research and programs that had been going towards this idea about housing, a lot of that just got rerouted to other ideas.
But then this thing happened. About four years after that final, sad MTO data came out, another researcher badged (ph) into his government cubicle - this time on the 10th floor of the IRS headquarters in D.C. And, like Mark, he wanted to understand how to improve people's economic lives.
NATHAN HENDREN: We had been working with data at the IRS.
DUFFIN: This is Nathan Hendren. He's an economist from Harvard. And for a few years, he and some of his research pals had been granted access to tax records at the IRS - with all kinds of privacy restrictions, of course.
HENDREN: Before this, I was working mainly with survey data sets where if you had, you know, 10,000 people in your data set, you were quite excited. In this data, you know, you're dealing with millions.
DUFFIN: Having that much data allows them to be so much more precise.
HENDREN: And what that allows you to do is really kind of put a sharp knife into your analysis and really uncover patterns.
DUFFIN: Nathan is trying to understand the impact of tax policies on upward mobility, which is economist speak for, can you achieve the American dream? Can a child go from the bottom income bracket to, over time, the top income bracket? And this is an issue that has become increasingly urgent. Right now, the data says that you are twice as likely to be able to achieve the American dream in Europe or at least a lot of countries in Europe.
So Nathan is looking at tax records from across the country. And he does start seeing a pattern, a pattern that surprised him. That huge MTO study, it looked like it had missed something, something enormous.
HENDREN: Actually, I remember the day vividly.
DUFFIN: After the break, the experiment that changed everything has to change.
(SOUNDBITE OF MUSIC)
DUFFIN: OK. So Nathan Hendren was knee-deep in all of this income tax data at the IRS alongside his researcher pal Raj Chetty.
HENDREN: So Raj and I were both working away at the Internal Revenue Service.
DUFFIN: They'd seen this data that showed a pattern. People who are rising from the bottom income bracket to the top income bracket aren't just randomly scattered across the country. They're clustered. And your chances of escaping poverty vary widely depending on which cluster you're in. So Nathan digs in deeper. He's starting to look at families who moved into higher-mobility neighborhoods.
HENDREN: The longer a child spent in a neighborhood with higher rates of upward mobility, the higher their outcomes were in adulthood.
DUFFIN: Like, let's say you have two kids, and you move them to this better neighborhood. And one of your kids is 4 when you move, and the other one is 8.
HENDREN: You'd actually see higher outcomes on average for the 4-year-old relative to the 8-year-old.
DUFFIN: Where you live matters for whether you have a shot at achieving the American dream, which sounds terribly obvious. But that contradicts what Moving to Opportunity found, that huge HUD study.
HENDREN: At that point, we were of the mindset of, well, geez, we should probably get the MTO data.
DUFFIN: Nathan starts to map income tax data from the actual MTO participants. And this more precise MTO data confirms the pattern that they saw nationally. In fact, the data is so strong that they can say the new neighborhood actually caused the person's economic improvements. That's how much a change of address mattered.
But they also started to realize that it really only mattered for a subset of the MTO participants, only for kids who were younger than 13 years old when they moved. And this explains why HUD missed this outcome. Nathan was looking at the data in 2014. HUD had last looked at this data six years before that in 2008.
HENDREN: Back in 2008, you wouldn't have seen it. The children - there just weren't enough young children into the labor market where you'd really be able to say, ah, it looks like there is an effect here. But, you know, wait five years, and all of a sudden you can really start to see these patterns emerge.
DUFFIN: Those younger kids had now grown up, started earning money, started producing income tax data. And they were finally grown up enough now for the data to reveal itself. And what the data was saying was that housing had been hope. Even though they had spent 14 years on this study, they had just counted it out too early.
It's so crazy how, just by happenstance, kind of, you guys were studying this at a moment where you could actually see that.
HENDREN: Yeah. It's nice to get lucky sometimes.
DUFFIN: What they found specifically is that kids who move under the age of 13, over time, they earn about 30% more.
HENDREN: Were four percentage points more likely to go to college.
DUFFIN: Teen pregnancies, less likely. The data also seems to indicate that these benefits will probably be passed on to their children. This all did come with a caveat - for kids who are moved when they're older than 13, they earn less over time. The move is actually slightly harmful for that group. But when you look at the younger kids, and over time, as they earn more money, they pay more income taxes. And with that, the program actually looks like it pays for itself. So Nathan and his research pals are looking at all of this really exciting data.
HENDREN: You could kind of feel that we had something that was going to change the way people thought about neighborhoods and change the way people thought about inequality of opportunity in the United States.
DUFFIN: They write all this up in a paper. They release it. And then...
HENDREN: Overnight, we were inundated. We got a lot of emails from people who were eager to think about what we should be doing to use housing policy as a way to think about improving upward mobility for children who are most disadvantaged.
DUFFIN: Did you get a lot of those emails before this?
HENDREN: No (laughter).
HENDREN: No. We did not.
DUFFIN: Many of these emails are coming from people who work in public housing who are saying, look, now that we know that this thing works, we should reboot the original HUD program. And Nathan is excited about that prospect. But he does know that if they reboot this, there needs to be an important addition because in the original program, you know, way back in the '90s, people who were given vouchers on the condition that they had to move to a better neighborhood, more than half of those people just, like, gave the voucher back, basically said, I would rather forfeit this than be told where I have to move.
So if Nathan wants this to be more than just a nice research paper with nice data about opportunity, someone will need to figure out how to get people to actually move to that opportunity. So Nathan and his team, they start working with cities to basically reboot Moving to Opportunity. The original MTO program, it required people who got the special vouchers to move to low-poverty areas. But in this reboot, families can choose to move wherever they want.
What Nathan and his team are trying to test is whether people have not been moving to opportunity because they don't want to, or is it just because they need a little support? So in their first test city, in Seattle, one group gets a voucher. And the other group gets a voucher but also support, a sort of housing counselor to help navigate things like transportation in the new neighborhood, even helps cover security deposits. These counselors also work with landlords.
Could this simple intervention get people to move voluntarily?
HENDREN: It didn't strike me as the most - what would have been the most effective going in.
DUFFIN: But they tried it out in Seattle. And after a year...
HENDREN: More than half of the families end up using the voucher in a neighborhood that has high upward mobility.
DUFFIN: More than three times the number of people chose to move to an area with higher opportunity when they were given support from a counselor. And all of this data from the original MTO study, this new one was so compelling that the unthinkable happened. Congress recently passed a bipartisan bill giving federal funds to replicate programs like this across the country. Nathan and that team will also be expanding to new cities.
And, yes, we do now know that housing policy can be hope. Housing policy, we are sorry we underestimated you. Thank you for your service. But for now, we can probably only call it a sliver of hope because the number of families whose lives are improving through this program is really just a tiny drop in a much bigger pool. The researchers' next challenge is to find out not just what happens if you move a family towards opportunity, but how they can take all the things that they are learning from this and move opportunity back into the neighborhoods that these families have left behind.
(SOUNDBITE OF INGO MOELL AND THOMAS STRAUCH'S "LEAVE IT ALONE")
DUFFIN: We always love to hear from you. You can email us at email@example.com. You can also find us on Facebook, Twitter or Instagram. We are @planetmoney. Our show today was produced by the amazing Aviva DeKornfeld and edited by Bryant Urstadt and Jessica Weisberg. Our supervising producer is Alex Goldmark. Special thanks - a very big thanks - to Stefanie DeLuca, also Shannon Felton Spence. And thank you to Pam Fessler, who has done some great coverage on this issue for NPR. Check out her stories as well. I am Karen Duffin. This is NPR. Thanks for listening.
(SOUNDBITE OF INGO MOELL AND THOMAS STRAUCH'S "LEAVE IT ALONE")
NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.