«Moving to opportunity voluMe 14, nuMber 2 • 2012 U.S. Department of Housing and Urban Development | Office of Policy Development and Research ...»
Over and above the considerable structural barriers outlined previously, a surprising set of preferences governed these subsequent moves—preferences honed over years of living in distressed public housing located in some of America’s poorest and most dangerous neighborhoods. Again, these preferences were not what readers might think: both our interviews, mentioned previously, and the followup survey for the interim impacts evaluation showed that families seldom left lowpoverty neighborhoods because they found such communities uncomfortable or disliked the economic mix or racial diversity there (Briggs, Popkin, and Goering, 2010; Rosenblatt and DeLuca, forthcoming). To the contrary, in other mobility programs such as Gautreaux Two, movers often spontaneously cited race and class diversity as something they enjoyed and were reluctant to leave behind (Boyd et al., 2010). Nor was the desire to live near kin always, or even often, a major factor;
in fact, sometimes people wanted to get away from the “needy” ties of their families (Boyd et al., 2010;
186 Moving to Opportunity Constrained Compliance: Solving the Puzzle of MTO’s Lease-Up Rates and Why Mobility Matters Briggs, Popkin, and Goering, 2010; Rosenblatt and DeLuca, forthcoming). Wood’s (2011) sample of Baltimore low-income renters, including voucher holders, held strong notions of what constitutes a suitable place to raise children: a private entrance to increase safety, a multilevel unit to manage noise and ensure a quiet place for children to sleep and do schoolwork, a basement for extra room and to house older (especially male) children who may be a deleterious influence on younger children, and a fenced-in back yard so children could play safely outdoors without extensive monitoring.
These unit considerations usually trumped neighborhood and school considerations. Retrospective interviews from MTO movers reflected some of the same themes (Rosenblatt and DeLuca, forthcoming).
DeLuca, Wood, and Rosenblatt (2011) and Rosenblatt and DeLuca (forthcoming) detailed the psychological and practical coping mechanisms that shape such residential decisions among voucher holders. On the psychological side, respondents who moved on to high-poverty neighborhoods often engaged in telescoping—defining the neighborhood as the block face and seeking visual evidence that the unit was at least on a good block, regardless of the surrounding neighborhood’s quality—and typically adopted the belief that the quality of the unit was more important for family well-being and child development than neighborhood surrounds (“we live in here, we don’t live out there”). On the practical side, they “kept to themselves”—avoiding contact with neighbors who might bring trouble and restricting the children to playing indoors, strategies that proved less effective for teens than for younger children. In short, their expectations about the quality of their neighborhoods were often quite low, and they believed that by keeping to themselves, they could ensure their children’s well-being.
Similarly limited expectations for their children’s schools were also pervasive in the interviews with parents in the experimental group who moved on; many mothers told us they believed that good schools were those that required uniforms and had security guards (qualities that made perfect sense in the chaotic schools we visited in the course of our research). Other parents told us that the school environment was irrelevant and that their children’s efforts were a more important determinant of their academic success. After their own experiences coming up in low-performing urban schools, these parents believed that their children could make their own way, just like they could survive unsafe neighborhoods (DeLuca and Rosenblatt, 2010).
When combined with the structural difficulties of obtaining affordable housing in opportunity areas, these strong beliefs about ideal home environments for raising children, along with a powerful set of adaptive coping skills that reduced expectations about what constituted suitable neighborhoods and schools, were probably strong enough forces to draw families in the experimental group back into higher poverty neighborhoods over time. These forces also likely kept those families in the Section 8 group and those in the control group who left public housing via a voucher from using the vouchers to move to very low-poverty neighborhoods.
Previously, we used the word preferences to describe these beliefs and coping strategies. This terminology, however, oversimplifies the process and implies freedom of choice. Instead, the preferences this body of qualitative work has revealed have been honed by years of enduring discrimination and significant neighborhood adversity. To make our point, we turn to the way courts perceived “freedom-of-choice” plans after Brown v. Board of Education. In a subsequent case in 1968, Green v.
County School Board of New Kent County, the court struck down freedom-of-choice plans because
Cityscape 187Edin, DeLuca, and Owens
they placed too much of the burden of integration on the families, many of whom were too intimidated to choose the White schools in the county. Other legal scholarship refers to the idea of tainted choice—choices that are restricted because of previous discrimination (Gewirtz, 1986).
In the case of MTO, many families—at least in Baltimore and Chicago, where participants were almost entirely African American—had been living for generations in neighborhoods that were heavily segregated (often by design). In the face of more information and a more diverse range of previous experiences, their preferences might have been different. Indeed, the Gautreaux Two study showed that families who had recent experience living outside of public housing were more likely to persist in opportunity areas (Boyd et al., 2010), and studies of Chicago’s original Gautreaux program found that families reported more positive assessments of low-poverty White neighborhoods after having lived there (Rosenbaum, DeLuca, and Tuck, 2005). One experimental group mover told us— If I had not had that opportunity to go into the MTO program, I would not have known what it would have been like to live in a house in a positive environment—to see how middle class people live…. It just made me want that. (DeLuca and Rosenblatt, 2010: 1468) Sticking to the Status Quo?
What happens if families do not have an opportunity to experience lower poverty, less segregated, safer communities? Children’s educational outcomes provide a good example. Recent observational studies suggest that neighborhood context can have profound effects on the cognitive development of young children. Burdick-Will et al. (2011) noted that, despite a lack of educational effects for the five-city study, experimental group families’ neighborhood changes in Baltimore and Chicago did improve children’s reading scores. In a nonexperimental study, Sampson, Sharkey, and Raudenbush (2008) compared the verbal cognitive ability of African-American children living in Chicago neighborhoods of concentrated disadvantage to that of African-American children with nearly identical individual characteristics and family backgrounds who live in Chicago neighborhoods with higher socioeconomic status. They found that living in a neighborhood of concentrated disadvantage reduces the verbal cognitive score of African-American children by 4 points, the equivalent of missing 1 year of schooling. They also found that this effect develops over time, emphasizing the durable role of neighborhoods in children’s early development. Wodtke, Harding, and Elwert (2011) also showed that negative effects of living in neighborhoods of concentrated poverty accrue over time and that sustained exposure can have devastating effects on the odds of graduating from high school.
Burdick-Will et al. (2011) identified two explanations for why some recent observational studies found more consistent evidence that neighborhoods matter for educational outcomes, whereas MTO found only limited evidence from two sites. First, it seems that changing a child’s neighborhood context matters more for children in the most disadvantaged neighborhoods. Second, and most striking, exposure to violence helps explain achievement differences across neighborhoods (see also Sharkey, 2010). Families in the experimental group in the Chicago and Baltimore sites left neighborhoods that were both more disadvantaged and more violent at baseline than at other sites, and as a result, their children benefited in school after these moves.
188 Moving to Opportunity Constrained Compliance: Solving the Puzzle of MTO’s Lease-Up Rates and Why Mobility Matters Neighborhoods of concentrated poverty negatively shape children’s outcomes, such as education, and the fact that these dynamics stand the test of time, across generations, is even more remarkable. Sharkey (2008) found that children who grow up in very poor neighborhoods live in similar neighborhoods as adults, and the cumulative neighborhood contexts of multiple generations can affect children’s outcomes. Sharkey and Elwert (2011) found that the neighborhood contexts in which parents grew up have a nearly equal effect on children’s cognitive abilities as that of the children’s own neighborhood contexts. These findings suggest that the stakes of changing a family’s neighborhood are high; neighborhood contexts matter not only for parents and children, but for the children’s children as well. Research shows that these dynamics are especially acute for AfricanAmerican children, whose parents are much less likely to move to a low-poverty neighborhood via a voucher than their White counterparts, as we describe in the following section.
The Cost of Doing Nothing In sum, despite MTO’s substantial early success—MTO managed to move one-half of the families in the experimental group to very low-poverty neighborhoods—a wide array of problematic program features and structural forces of highly segregated housing markets, together with underlying beliefs and coping strategies predicated on years of living in highly distressed neighborhoods, may have spelled disaster for the longer term neighborhood attainment of participants in the two treatment groups. These dynamics limited choice so as to funnel families in the experimental group into higher poverty neighborhoods over time. The behavior of families in the Section 8 group— who at median spent not a single month in a very low-poverty neighborhood (less than 10 percent poor) and only 9 months in a neighborhood that was less than 20 percent poor over the duration of the study—is further testament to this constrained choice. Evidence from the larger HCVP suggests that the dynamics affecting low-poverty and traditional voucher holders, who are largely African American or another minority, may be felt among African-American voucher holders nationally—they seldom lease up in neighborhoods with lower poverty rates, despite the fact that, given the voucher, they can theoretically afford to do so (Devine et al., 2003). These trends are part of what motivated MTO in the first place. In the absence of mobility programs, minority voucher holders rarely manage to leverage the subsidy to access neighborhoods of opportunity, leaving them uniquely disadvantaged by the program (DeLuca, Garboden, and Rosenblatt, forthcoming).
Exhibit 1 shows this striking pattern for the Chicago voucher program. In Cook County in 2000, 36 percent of African-American voucher holders lived in neighborhoods with poverty rates of more than 30 percent, and more than 60 percent lived in tracts where at least 20 percent of the residents were poor. Within Chicago city limits, these figures are even higher: nearly 50 percent of AfricanAmerican voucher holders lived in neighborhoods with poverty rates of more than 30 percent and more than 75 percent lived in neighborhoods with poverty rates of more than 20 percent. Residential racial segregation among voucher holders is also extremely high. Nearly two-thirds of AfricanAmerican voucher holders lived in neighborhoods composed of 90 percent or more AfricanAmerican residents; conversely, nearly two-thirds of White voucher holders lived in areas with
less than 10 percent African-American residents. None of these figures have improved over time:
a slightly higher proportion of African-American voucher holders were concentrated in poor, minority-dominated neighborhoods by 2008 (authors’ own calculations based on HUD, 2008, 2000, and on 2000 census and 2005–2009 American Community Survey data).
Note: Tracts with no symbol have fewer than 11 total vouchers; thus, data on the racial composition of voucher holders are not available.