For about 40 years now, since the beginning of the end of the era of the high-rise public housing project, government programs have been experimenting with a different idea to alleviate poverty: If we can’t easily change poor neighborhoods, let poor people move out of them instead. Give them vouchers to rent homes on the private market, theoretically outside of the ghetto.

The evidence of what happens next has been mixed. Some of it says that the children of poor families perform better in school when they relocate to the suburbs, and that their mothers have better mental health. Other research has found that voucher programs have little lasting positive impact on the families that use them. Yet another strand warns that crime disperses when poverty does – but that conclusion has been disputed, too.

Part of the problem is that the design of well-intentioned programs has often been flawed, as has the design of some of the research studying it. But there is one recurring theme here: When minority families are given a chance to move out of poor, segregated inner-city neighborhoods, many wind up living in places that look an awful lot like where they started. Or, often, they wind up moving back.

So why does this happen? And what does it mean for how we might design housing vouchers that could actually alter a poor family’s prospects, substantially and permanently changing the kinds of places where they live?

“We say ‘people need choices,’ and that’s absolutely true,” says Stefanie DeLuca, an associate professor of sociology at Johns Hopkins University. “But part of what goes into choices is having had exposure to things that other people have had, to say, ‘now I want this, this is something I can see that’s possible for me.’ Being poor doesn’t just mean you didn’t have enough resources and you had barriers to opportunity – but the benefits of those opportunities are relatively unknown.”

How do you identify good school districts if your family has never lived in one? Or if you’ve never had the luxury of prioritizing high test scores over safe classrooms? How do you weigh the benefits of a three-bedroom in a housing project against a smaller home in a quieter neighborhood when no one you know lives in a place like that? How would you make plans to leave the city for the suburbs if you had never even been there?

In effect, families given housing vouchers often have “blind spots on their cognitive maps,” as DeLuca puts it, in thinking about where else they might go (no doubt we all have such blind spots of places we never ponder). And the implementation of voucher programs frequently serves to reinforce them. Families are often given lists of landlords who accept Section 8 vouchers (it's legal in many states for landlords to discriminate against subsidized tenants). That means poor families wind up relocating to new areas that are known to welcome housing vouchers – and that, by definition, are already full of families using them. Or, without more guidance on where they might move, they wind up settling in nearby neighborhoods that feel familiar. Neither scenario dramatically changes the high-poverty, segregated context of where these families live.

For about a decade now, DeLuca has been following a housing voucher program in Baltimore that looks different in some fundamental ways from what I’ve just described. Some of the results now coming out of the Baltimore Housing Mobility Program hint at how we might reconsider housing vouchers on a broader scale – even as it raises thorny question about who gets left behind any time we relocate families out of failing neighborhoods.

First, a brief word about those neighborhoods. We know that where families live matters, and that neighborhood context can explain some of the difference in life outcomes between a black child and a white child who come from otherwise identical family backgrounds. A child without access to a good school grows up to poor job prospects and lower wages, then as she becomes a parent, she has fewer resources to move her child out of the same neighborhood with high poverty rates, high crime and little opportunity. Systemic disadvantage accumulates across time in what NYU sociologist Patrick Sharkey calls the “inherited ghetto.”

It’s a place that’s extremely difficult to exit, and where inequality perpetually reproduces itself.

The Baltimore program was created out of a federal court ruling after the ACLU sued the Department of Housing and Urban Development and the local housing authority in 1995 for violating the Fair Housing Act. The lawsuit argued that those agencies had never truly desegregated Baltimore’s public housing, and that the program was administered in a way that kept poor black families in the most high-poverty, segregated parts of the city.

The court’s ultimate remedy created a new housing voucher program with very specific criteria: Families could only use the vouchers to move to neighborhoods where the poverty rate was less than 10 percent, where the population was less than 30 percent black, and where fewer than 5 percent of residents were already living in subsidized housing. Today, this describes virtually none of Baltimore City – or, at least, none of the census tracts within the city with a suitable stock of rental housing. In effect, the Baltimore program has now relocated more than 2,000 families to the city's suburbs (with a final agreement for several thousand more). And it has required them to stay there for at least one year, before the voucher becomes portable.

To accommodate the strict criteria, the program, which is run by a contractor, also provides extensive counseling to help families envision living in those "blind spots" on their mental maps. Counselors take residents on neighborhood tours. They coach them through identifying good school districts. They help them hunt for housing, even when a family later wants to relocate a second time. All of this is the kind of hand-holding that strapped housing authorities can seldom afford. And the program has the luxury of working across bureaucracies, throughout the entire metropolitan region.

“Here we have this legal remedy that provides a window into what things could look like of we were willing to put forward the resources to really support the voucher program,” DeLuca says.

She and fellow researcher Jennifer Darrah, from the University of Hawaii at Manoa, have followed and conducted in-depth interviews with 110 families who have either moved through the program or attended its orientation sessions. Two-thirds of the families who have moved are still living in these integrated, low-poverty neighborhoods one to eight years later, often despite initially hesitation. In an article in the Journal of Public Policy Analysis and Management, DeLuca and Darrah describe interviews with mothers that suggest “profound differences in the way many of the parents in the BMP thought about where they live now, where they want to live in the future, and where they never want to move again.”

These women describe feeling “proud” of their addresses on job applications. They describe children who look forward to going to school, and who sleep more soundly at night. They talk about planning future moves with school districts in mind. In contrast, this mother living in Harford County, Maryland describes how she looks back on her previous home growing up in a public housing project in the city:

I had to basically barricade myself in my home . . . I couldn’t go sit out my backyard because I didn’t have one . . . I’m not going to sit on my front porch cause you had so much going on in front of you, that I choose not to want to see. So it’s just like it was like I was in a jail, but not in a jail. I didn’t have bars I could still come and go as I please, but I still felt like I was in jail. So that was the difference.

Some of the more revealing quotes come from the mothers who still live in such settings, who give some insight into why a family might stay there despite all of the costs.

“I’m so used to what I know,” says one woman.

“The only place I’m going to know is the walk out my door to catch the first bus,” fears another mother thinking of the suburbs.

“I never lived in the county so I couldn’t even tell you how a county is . . . I just know you need a car to do anything.”

Repeatedly, these women describe having learned to cope in high-poverty neighborhoods, to the point where they believe they can control how much influence those neighborhoods have on their families:

I always was raised you live inside, not outside, so what went on outside really didn’t bother me. It bothered me as far as being a parent, don’t misunderstand, but I run my household how I want to run my household.

All of this is admittedly qualitative evidence. A lot more hard data will be needed over time to assess whether children in the Baltimore program perform better in suburban schools, and whether they ultimately have better employment outcomes and higher wages. But the simple fact that so many families who moved with the program have opted to stay in their new settings – adjusting their lives to the amenities there – reaffirms that vouchers might in fact create dramatic and sustained change for poor families leaving the worst urban neighborhoods.

The bigger question then becomes how you’d scale up a program like Baltimore's, and whether we’d even want to. Every dollar spent giving in-depth counseling to one family is a dollar not spent on a bare-bones voucher for a second family. The resources would no doubt be a problem. But then there’s an even harder question: “What do we do about everyone else?” DeLuca asks.

This is a tension that urban policy has wrestled with for years: Is it better to invest in community development in poor neighborhoods, or to enable greater mobility for people to leave them? Despite the mixed results, there is still much more evidence on the side of the second strategy than the first.