Place-Based Poverty, Social Experimentation, and Child Outcomes: A Report of Mixed Effects

John Goering
The Graduate Center of The City University of New York
School of Public Affairs, Baruch College, The City University of New York

Citation: Goering, John. “Place-Based Poverty, Social Experimentation, and Child Outcomes: A Report of Mixed Effects.” Children, Youth and Environments 13(2), 2003. Retrieved [date] from http://colorado.edu/journals/cye.

Abstract

In the past two decades, a host of non-experimental and quasi-experimental studies have been conducted on how neighborhoods affect child and family well-being and social status. The federal Moving to Opportunity (MTO) program is the most important ongoing demonstration in American housing policy on the effects of neighborhood upon the life chances and futures of children and adults who had previously lived in low income public housing projects. Research indicates positive impacts upon the neighborhood contexts, their levels of fear and sense of security, and upon a range of major behavioral outcomes. Gender differences appear significant but are currently unexplained. A new phase of qualitative research currently underway is described as the next stage in this longitudinal panel study of neighborhoods' effects on poor children.

Keywords: urban poverty, concentrated poverty, neighborhood effects, housing policy

Introduction

Rainwater and Smeeding , in this issue, describe the high levels of child poverty within the United States in comparison with other societies and note the significant need for comprehensive policies to ensure that no child grows up poor. The purpose of this paper is to focus on one policy demonstration aimed at altering the concentration and harmful effects of poverty within the urban United States. The program is named Moving To Opportunity, or MTO.

Following a brief background section focused on policy constraints in the area of space-based poverty alleviation, the paper will describe what MTO was intended to accomplish and its major statistical effects upon children, and conclude by acknowledging how little is understood about how and why program effects have occurred. A new stage of qualitative and ethnographic research is presented as the next means for aiding our understanding of what has caused some children's lives to improve and others not.

The Persistence of Concentrated Poverty

Long-standing efforts of the federal government to address the poverty of individuals and families have been more consistent, better funded, and more carefully evaluated than those aimed at eliminating slums, “underclass ghettos,” or the undue concentrations of poor families within our inner cities. The spatial concentration of poverty in the U.S. indeed has had a long and politically sordid past. Sparse political and programmatic attention has been paid in no small part because of the complexity, scale, and seeming implacability of the problem. Given reported notable declines in the concentration of poverty over the decade of the 1990s, however, researchers reasonably wonder whether and how public policies may have helped cause this shrinkage in underclass ghettos (Jargowsky 2003).

There are three interrelated reasons for the government's constrained ability to significantly reduce the spatial concentrations of the poor within U.S. cities. First, the government has been reluctant to act due to the fact that federal programs themselves have been partially responsible for concentrating poor families in certain neighborhoods. The federal public housing program, as a prime example, created many housing “projects” within cities that by a matter of policy mandate, only housed the poor, often only the very poorest, of families. For decades, such projects also were racially exclusionary, serving either whites or blacks and other minorities but never both (Hirsch 1983). They thus generated, almost overnight, neighborhoods of federally subsidized poor persons of color whose housing futures, we have since learned, typically were intergenerationally linked to the same apartments in the same projects (Goetz 2003). Further, the practice of constructing multiple buildings within the same communities resulted in concentrations of poverty that were often huge; in the worst situations, residents would have to travel for blocks to find non-subsidized housing and community services (Popkin et al. 2000; Vale 2000). The federal housing agency thus ensured that generations of families living within them would be known and stigmatized as “kids from the projects.”

Second, the relatively flat level of expenditures for federal housing programs has imperiled anti-poverty programs aimed at the spatial isolation of poor families (Colton 2003, 236). Reductions in the size of Congressional allocations for the Section 8 program at the U.S. Department of Housing and Urban Development (HUD) have meant that there have been in recent years few options to initiate major new renewal or housing relocation programs.

The third factor that helps explain the persistence of poverty ghettos as a major feature of the urban cityscape is the fact that the federal government could find few communities that would accept the poor even when there was funding. Residents of "middle-income neighborhoods [have] proved quite capable of detecting and resisting even the smaller, scattered site developments…" (Hayes 1995, 144).

Community opposition to families dependent upon housing subsidies has taken a variety of forms, including fire bombings, political petitions and multifaceted forms of community-based resistance driven recently by the argument that only through their own hard work should families be permitted access to better-off neighborhoods (Hogan 1996; Ihlanfeldt 1999; Farley, Danziger and Holzer 2000, 149-161; Vale 2000, 279-282). The U.S. public has thus tacitly, and at times expressly, endorsed the ongoing de facto segregation of public housing, including both the siting and tenanting of public housing developments (Hirsch 1983). The residents of such projects, especially minority families, are usually also aware of the likelihood of white resistance and factor it into any housing search they undertake (Farley, Danziger and Holzer 2000, 185-204; Sjoquist 2000, 72-73). Community opposition to the poor has thus become one of the loadstones of U.S. antipoverty policy.

Effects of Concentrated Poverty

Tenants of public housing, interviewed over 40 years ago, already recognized that pubic housing projects handicapped stable family life and were places where “all manner of anti-social behavior runs rampant, a bad environment for bringing up children, a receptacle for the very lowest elements of society” (Hartman 1963, 286). By the late 1960s, most public housing developments for families in the larger cities had become largely racially and economically segregated, and by that time, Congress had halted virtually all funding for any similar such projects (Bickford and Massey 1991, 203-204; Newman and Schnare 1997). Research over the last dozen years has shown that segregated public housing projects have continued to offer fewer and fewer opportunities for residents, and are indeed often the sites for some of the highest levels of gang violence and drug activity (Galster 1996) in urban areas. Low-income communities, recent research has shown, can have negative behavioral effects on residents (Harding 2003). One critic has identified public housing projects as “latter-day poor houses” (Husock 2000, 53). The social and political stigmatization of families in public housing became the signature of federal housing policy.

Central to the experience of living within neighborhoods of concentrated poverty is exposure to crime. Crime grew in high-poverty communities from the late 1980s into the 1990s with a rate and intensity that was seldom matched in the outside streets and communities. Fear of crime became a sociologically inevitable core of the way of life of those who had no choice but to remain living in projects such as Robert Taylor Homes in Chicago, Lafayette Homes in Baltimore, or Mission Hill in Boston. Impacts of concentrated poverty have been variously described, deplored, or feared but not seriously remedied (Popkin et al. 2000; Wilson 1991).

Moving to Opportunity

Creating a Federal Option to Move out of the Ghetto

While poverty for Rainwater and Smeeding (in this issue) “has been a central and self-conscious concern in U.S. society...” for forty years, spatially concentrated poverty has also been a source of some policy worry, abundant frustration, and inadequate programs. In 1992, however, the U.S. Congress created two allied, but independent programs to address concentrated poverty affecting public housing residents. The first program, HOPE VI, was designed to demolish the worst of the projects and rebuild in their place a reduced number of "mixed income" housing units. Tearing down such projects had been until this time only seldom attempted (Rainwater 1970).¹ In place of the old buildings would be mixed-income housing including homes for a modest proportion of those who formerly lived there. Those families unable to fit back into the now reduced number of units for the poor were scattered to the existing projects which remained or into private rental housing (GAO 2003). Those who did return would find a mix of working families with whom to relate, and a much improved physical infrastructure.

The second program authorized by Congress in 1992, and the focus of attention in this paper, was named the Moving to Opportunity for Fair Housing demonstration (MTO). Congress allocated an initial $75 million dollars for MTO in order to answer the question of whether moving families with children out of concentrated slum projects would in any meaningful ways alter and improve their lives.

Through a five-city demonstration, MTO was designed to test the use of housing mobility as a way to redress segregation and concentrated poverty (Schill 1992, 811-821). It was intended to demonstrate that it was possible for HUD to promote area-wide mobility out of poor, inner city projects into better off neighborhoods. Most importantly, it was designed as a test of the neighborhood effects thesis that environmental context matters for the life chances of poor children (Jencks and Mayer 1990; Harding 2003). In each of the five cities, roughly 150 families would be offered the opportunity, through a lottery, to move into more affluent communities throughout their metropolitan areas.

A Key Research Source for MTO: The Gautreaux Program

The creation of this new federally-funded residential mobility program grew directly out of research into the effects on families produced by a housing mobility program begun in the city of Chicago in the 1980s- the Gautreaux program. MTO was based upon many of the characteristics of Gautreaux and intentionally designed to correct some of its fundamental methodological shortcomings.

Systematic racial isolation of low-income public housing families has been judged as unconstitutional and illegal by a number of federal courts. Beginning in the 1960s, a number of lawsuits were filed against the federal government charging it with illegally segregating minority public housing residents (Goering 1986). Among the housing segregation lawsuits against HUD was a suit brought in 1966 by a Chicago public housing resident which resulted in the courts ordering HUD and the local Chicago Housing Authority (CHA) to remedy the segregation by providing a housing mobility option throughout the Chicago region for 7,100 black families (Davis 1993; Rubinowitz and Rosenbaum 2000).

The Gautreaux Program provided families chosen for the program with Section 8 rental certificates that required them to move to either predominantly white or racially-mixed neighborhoods. Participating families were provided support and counseling by a local nonprofit organization, the Chicago Leadership Council for Metropolitan Open Communities (Rubinowitz and Rosenbaum 2000). The program ended in 1998 but research on its effects on the lives of participating families has continued from the 1980s into the present (Mendenhall, Duncan, and DeLuca 2004).

Research on the behavioral impact on families' lives of this court-ordered mobility program indicated that housing mobility could benefit the lives of the children and adults who moved. This was an enormously provocative finding that suggested that there might be a low-cost means of improving lives through modest alterations in the administration of an existing subsidy program. It appeared that if families moved out of the ghetto, children would gain greater access to educational opportunities. Children were not only less likely to drop out of school, but were more likely to take college-track classes than their peers in a comparison group who moved to poorer, more segregated parts of the city of Chicago. After graduating from high school, the Gautreaux children were also more likely to attend a four-year college or to become employed full-time at a job that included fringe benefits (Rosenbaum 1992). The Gautreaux researchers described changes in the job situations of adults, too. Adults who moved to the suburbs had higher employment rates than adults in the city, although there was no statistically significant difference in the wages they earned or the number of hours they worked.

While Gautreaux researchers provided useful insights, additional research proved essential, as several issues threatened the validity of the findings. Selection bias, a small sample size, and the absence of a complete longitudinal panel of information regarding families' lives limited Gautreaux's ability to answer the question of whether federal housing policy could generate significant neighborhood effects, and whether they would occur in other cities. Gautreaux was critically limited because researchers did not conduct a baseline study, use random assignment, or keep track of all of the families who moved away. MTO was designed as a social experiment to address these shortcomings.²

The Problem of Selection Bias

The problem of selection bias has been recognized by social scientists as a crucial concern in understanding whether neighborhoods, in their own right, affect the lives of residents (Jencks and Mayer 1990, 119; Harding 2003). The core of the methodological problem is that as families gain increased choice over the neighborhoods in which they live, it is easy to assume that any differences in outcomes have resulted from their new neighborhoods when in fact they could also be due to some "unmeasured family characteristics" (Lehman and Smeeding 1997, 262). The solution to this problem is well understood: take people's ability to select their neighborhoods out of their hands by randomly assigning them to a community. The researcher thereby separates out the role of personal characteristics on individual choice from the potential impacts of the neighborhood (Brooks-Gunn, Duncan, Leventhal, and Aber 1997, 286). Until MTO, however, there was no research mechanism to manage the random assignment of a large enough number of families to learn whether the choice of neighborhood significantly matters to the lives of children or adults.

MTO would not only provide for the necessary random sampling to test the neighborhood effects thesis, but it would also make it possible to track and study the lives of families for a long enough period to know when effects occurred. It would then allow researchers to learn whether effects occurred, when they appeared, and then why and how these effects had happened.

The Uncertain Temporal Sequencing of Effects

Some social science evidence pointed to the high probability that residential mobility would lead to reduced, rather than improved, performance among children, most notably in schooling. Pribesh and Downey (1999), for example, provide evidence that a modest proportion of declines in school performance can be traced to the loss of social capital that occurs as part of one or more residential moves for teenagers. This performance loss is noted in particular among poor families, however they “were unable to identify any group that consistently benefited from moving” (Pribesh and Downey 1999, 531). Nonetheless, after ten years or more living in more middle-income suburbs, Gautreaux children showed notable improvement in school achievement.

As a result of the absence of testable theoretical or empirical propositions from Gautreaux, MTO began with a host of policy and even political expectations that it might indeed influence a wide, diverse range of behaviors for all family members, and thus required a methodology that sought out effects in employment, health, schooling, attitudes, and expectations. MTO researchers began with many false expectations in part because of the absence of any social science road map into the temporal trajectory of behavioral consequences of neighborhood effects experimentation. A key influence on the expectations was the suggestion from Jencks and Mayer (1990, 119) that effects would only occur after some undefined, but “protracted” length of time:

From a scientific perspective, the best way to estimate neighborhood effects would be to conduct controlled experiments in which we assigned families randomly to different neighborhoods, persuaded each family to remain in its assigned neighborhood for a protracted period, and then measured each neighborhood's effects on the children involved.

Some effects of MTO did not, however, require a long period of time to occur, while it now appears clear that others may require a longer period. For other behavioral change issues, perhaps adult employment for instance, MTO may be the wrong experiment or conducted at too small a scale for significant effects to appear at all (Orr 2003).

How much time should it take before ten-year-old children begin to score better on academic achievement tests once they have left their old school and neighborhood? Does it not take time for teenagers to forget their old friends and neighborhoods and accept, and be accepted by, a new peer group? How much time should we expect it to take before a mother finds a better paying job or decide to go to school to improve her chances in the local job market? When, in brief, should researchers, policy makers, and the public legitimately expect to see significant alterations in the lives of adults and children as a result of joining MTO and then moving to a new home? Should we expect to see changes occurring after only a few months, years or, as in the case of the Perry School study (Schweinhart, Barnes and Weikart 1993), decades after families have left some of the most frightening, dangerous parts of U.S. cities?

Families move for a better job, a bigger home, to provide better schools for their children, or to escape the crowding and crime of cities (Rossi 1980; Wolfe 1998). Just as it takes time for the average family who wants to move to find the right house, best school, nearest shopping, and best routes to work, time appears necessary because MTO requires families to reestablish themselves physically and, more importantly, socially (Pribesh and Downey 1999).

Implementation of MTO

MTO was authorized by the Housing and Community Development Act of 1992 to

assist very low-income families with children who reside in public housing or [families] receiving project-based assistance under Section 8 of the Housing and Community Development Act of 1937 to move out of areas with high concentrations of persons living in poverty to areas with low concentrations of such persons.

The demonstration plan links Section 8 rental assistance with housing search and counseling services for a randomly selected group of public housing families from impoverished neighborhoods where at least 40 percent of the residents lived below the poverty line in 1990. The randomly assigned "treatment group" could only make use of their rental assistance in areas where less than 10 percent of the population lives in poverty (MTO's definition of a low-poverty area). The thresholds– 40 percent and 10 percent– were intended to mirror social science measurement standards for deeply poor ("underclass") neighborhoods in contrast to non-poor communities.

Beginning in September 1994 through July 1998, over 4,600 public and assisted housing families who volunteered and were found to be eligible were randomly assigned to one of three groups:

• The MTO treatment group, which received Section 8 certificates or vouchers useable only in areas with a poverty rate of 10 percent or less. They were also provided counseling assistance from a local nonprofit organization in finding a private rental unit;
• A Section 8 comparison group, which received regular Section 8 certificates or vouchers with no special geographic restrictions or counseling; and
• An in-place control group, which continued to receive their current project-based assistance.

In spring 1994, HUD selected five local public housing authorities (PHAs) for the MTO experiment (Baltimore, Boston, Chicago, Los Angeles, and New York), and required nonprofit agency collaborators to participate in the demonstration. The PHAs were required to select for potential enrollment only families with children under 18 living in public housing or Section 8 project-based developments in census tracts where at least 40 percent of the residents were poor. Families who enrolled filled out a lengthy baseline survey and agreed in writing to cooperate with the information gathering and research needed for the demonstration. Random assignment computer applications and staff instructions were established and monitored at each site by a HUD contractor, Abt Associates.

Approximately 5,300 families volunteered and 4,608 families were found eligible and given random assignments. A total of 1,820 families were assigned to the treatment group, of whom 860 (48 percent) moved into a low-poverty community. Another 1,350 families were randomly assigned to the Section 8 comparison group, of whom 816, or 60 percent, leased an apartment. Roughly 1,400 (1,439) families remained in public housing. Thus, out of the 4,608 families participating in MTO, 860 families moved into a low poverty neighborhood.

Preliminary Social and Behavioral Outcomes from MTO

Below is an overview of research on the experimentally based social and behavioral outcomes of MTO. These results are drawn from two sources. The first source is analyses conducted by seven teams of social scientists, each working in one of the five MTO locations. The methods used by these research teams varied and included linking baseline surveys to subsequent household surveys, individual level administrative or agency data, as well as field observation and qualitative interviews.³ The results from this first wave of research are summarized in Goering and Feins (2003). The second source used is a mid-term report on MTO prepared for HUD by a team of three contractors called the Interim Impacts Evaluation that appeared in Fall 2003.⁴

There are four closely linked, central research questions that are at the heart of the MTO longitudinal panel evaluation. They are:

1. Are there measurably important effects of neighborhood upon the lives of adults or children?
2. How quickly do such effects appear and how persistent are they? Are effects all equally long lasting and do they all occur at the same time?
3. What are the principal causes of these effects? What social mechanisms and mediators help to explain both why and how the effects have occurred in the lives of children, teenagers, or adults?
4. The last issue, outside the scope of the present study, is to try to measure the actual costs and net social benefits that cumulate all that has been learned.

Results

Through MTO, HUD has been able to administer a multi-site demonstration over several years that succeed in finding homes in low-poverty areas for all of the participating treatment-group families. This seemingly trivial, but not readily predictable, policy finding is the foundation for the substantive social science results which follow. That is, families that volunteered (and we note that they differ from families that did not), were able to rent apartments in more economically, although not much more racially, mixed communities. The average poverty rate of residential locations for both the MTO treatment group families and the Section 8 comparison group families was significantly lower (by 15-16 percentage points) than the poverty rates of areas where in-place control group families lived. Moreover, median incomes in the treatment group families' neighborhoods were 73 percent higher than median incomes in the control group neighborhoods.

Better Neighborhoods and the Loss of Fear

Kling and Liebman (2003) summarize the overall effects of MTO on the neighborhood context for youth:

Treatment group youth live in lower-poverty neighborhoods, with lower youth idleness, and more high status adults. They are less likely to be dissatisfied with their neighborhood, to feel unsafe, or to have seen or heard gunshots.

Among the lessons from MTO research to date is, also, that positive transformations in the lives of children have occurred earlier than prior social science research led us to expect. The first phase of MTO research reveals that households in the treatment, as well as some in the Section 8 comparison group, have experienced improvements in multiple measures of well being relative to the in-place control group. Among the most notable early effects were reductions in crime and the fear of it. MTO reduced the proportion of families that felt unsafe in the experimental group by 61 percent.

The majority of MTO study families indicated in their initial, baseline interviews that victimization and fear of violent crime were the primary reasons for participating in MTO. Not only were they afraid for their own and their children's safety, mothers also worried that their children would join gangs if they stayed in their original neighborhoods. Although crime rates have declined in most urban communities in the years since MTO's launch, evidence consistently confirms that MTO families moved to dramatically safer communities and that they see the greater security as a major improvement in their lives (Popkin, Harris, and Cunningham 2002). Families across all five sites reported higher levels of security as a major outcome of their moves from high poverty public housing, the evaluation report tells us. MTO provided, as a social experiment, the precise social outcome desired by program participants. From their perspective, and with their expectations in mind, MTO has succeeded.

In general, the places where MTO treatment group families and Section 8 comparison group families were living had significantly fewer crimes. Statistically, the share of treatment group families feeling safer increased significantly. Also there was a significant decline in the number of family members who reported being victims of crime and in those who reported witnessing drug dealing in the prior six months. These statistical impacts were compellingly revealed in qualitative interviews. After her move out, one mother in New York explained what life was like before: “It was like being in a war zone. …a lot of drug dealings. Shootouts. Everybody had such low self-esteem. Nobody looked out for each other. It was horrible.” A Baltimore mother described what her new home was like: “It's a totally different neighborhood because there is no drug activity, no kids hanging on the corner, no kids fighting each other.” And in Chicago another mother told researchers: “you can wake up very day and we're not worried about seeing anybody getting shot and no gang members, nothing like that and it's quiet and it's cool and calm up here.”⁵

Fewer families in the treatment group reported neighborhood problems with drugs and crime compared to reports from the Section 8 comparison group. They reported that they felt they had moved to very safe neighborhoods or reported significantly reduced risks in comparison with their old neighborhoods. One mother reported, "I get along with the people in the house, and, more importantly, the street is quiet, especially for the kids. If I want to go downstairs, walk around with them or sit outside, I can do that and feel safe."

Fear has either stopped or declined, physical and mental health has improved, and young girls' lives now involve fewer problems. These effects have all occurred during the first four to seven years after families left deeply poor public housing developments. This greater safety and security should, researchers hypothesize, help reduce anxiety and stress, contribute to better health outcomes, and enable parents to be calmer and more patient with their children. The research will further explore the developmental consequences of this shift in fear thresholds to learn its impacts on children and parents.

Outcomes for Children

New schools attended by MTO children were better. Research from both individual sites and the interim evaluation report reveal that the schools attended were somewhat better than those attended in the old neighborhoods. There were lower percentages of poor, minority and limited English proficiency children in the schools the treatment group children attended.

One set of results that appeared in Baltimore but that did not appear when the evaluation examined all five sites was in regard to educational outcomes for children. Ludwig, Ladd, and Duncan (2001), for example, reported evidence of the effect of MTO in Baltimore upon individual children's school performances. Using standardized reading and math scores matched to identifying information for MTO revealed statistically significant improvements for the treatment group. However, in this early research, no direct educational testing of children was conducted. When such testing was completed in the interim evaluation, no statistically significant effects of MTO appeared in educational performance. This limited set of educational performance outcomes may well be due to the fact that nearly three-quarters of experimental group children whose families had leased under MTO were attending schools in the same district as before. Some children did not change schools at all when their families moved, and many families, even those who moved considerable distances, remained within the same city school district.

It is also useful to focus on results that while not meeting standard measures of statistical importance nonetheless reflect impressive changes in educational performance. An additional 6 percent of the full experimental group had graduated from or were still in school, and mong the households that chose to move, a 16 percent increase in graduation and retention rates occurred. There was also a decline of 41 percent in non-graduation attrition among students. Kling and Liebman (2003) comment on the gendered aspect of these effects:

odds of graduating or staying in school are nearly twice as high for experimental …females as for females in the control group…This illustrates that the estimated magnitude of some of our estimates can be large even when not statistically significant at the five percent level.

Behavioral changes such as these serve as clear and important guideposts for the next and final impact evaluation. They also suggest that Gautreaux researchers may well be correct that it takes more time than MTO youth have had to date for more powerful effects to happen.

Girls Have Done Better than Boys

One of the unanticipated results from MTO research so far has been a finding of major differences in outcomes among youth by gender; girls and boys have responded to the treatment in sharply different ways that we currently cannot fully understand or explain. Girls in the experimental and Section 8 groups have perceived their chances of going to college and getting a good job as much higher than their counterparts in the control group. The “idleness rate” (neither in school nor working) was also substantially reduced for girls in the experimental families. Further, there was a 62 percent reduction in general anxiety among girls. They also experienced a major decline in smoking: a reduction of 46 percent for those in the experimental group. Accompanying this was a drop by 55 percent in the use of marijuana. These are statistically impressive and important changes in behavior that are not, however, mirrored in the lives of boys.

For boys in the experimental group there was generally an inverse, seemingly gender-based set of increases in smoking (up 400 percent); in behavioral problems, which increased 68 percent; and in arrests for property crimes, which rose by 185 percent. At the same time that effects for boys appear quite negative in data covering all five MTO sites, research on youth living in New York City (Leventhal and Brooks-Gunn 2003) shows that there were positive, and not negative, effects on the educational achievement of boys in that site. They note that moving to low-poverty neighborhoods had positive effects on the achievement scores of boys aged 11 to 18 compared with peers in high-poverty neighborhoods. It is not clear how and why New York youth differ from those in the other locations but further research is underway to pursue answers.

There are for us only unclear causal explanations for these pronounced gender differences for older children in their reaction to MTO moves. For example, the increases in arrests might reflect more stringent policing in new locations rather than more criminal behavior, and there is some evidence that the relatively higher rate of property crime arrests may not persist after the first four years. Similarly, the interim report showed boys drinking more and girls less. As we have no evidence on the normative level of either drinking or smoking in the new neighborhoods into which MTO children have moved, it is not clear what is "normal" for higher-income, more suburban areas.

Why Have Changes Occurred? Next Steps

Social science researchers have for a long while emphasized that “it matters how neighborhoods matter” (Lehman and Smeeding 1997, 259). For many of the statistical and quantitative findings of MTO research to date, our understanding of why and how they happened is quite limited. Quantitative measures of school, health, and criminal outcomes often do not tell convincingly and clearly the reasons for positive or negative change. Through what social mechanisms and process have positive effects occurred? What is it about their new neighborhoods that have caused these benefits? Why are MTO teenagers arrested for more property crimes? Why did asthma cases in Boston initially decline but significant effects not appear for the full five-site sample? Why is gender so important? What has caused children's and families' lives to change, and through what social mechanisms have their neighborhoods been the cause?

Qualitative interviewing and ethnographic research are tools central to getting inside the “black box” of known quantitative experimental effects and to learn about the institutions, networks, and processes that have led to change in the lives of adults, children, or both. A first round of qualitative research was completed for HUD in 2001 that suggests issues and problems that will need additional research. A number of parents, for example, kept their children in the schools serving their original high-poverty neighborhoods rather than moving them to new schools (Popkin, Harris, and Cunningham 2002). The extent to which families have moved their homes into low-poverty communities but have not taken advantage of “local” resources and institutions represents a crucial question for the next stage of MTO research. There are many more comparable questions about why experimental effects have occurred that researchers will investigate in the next stage in more qualitatively-focused inquiry. Central to the proposed qualitative research is to understand what is driving effects and non-effects, as well any move-back decisions of those who decide to return to higher-poverty neighborhoods.

Included in the inquiry will be a search for relevant institutional resources, social networks including peer groups, and a sense of collective efficacy or social engagement among neighbors that enables effective cooperation in monitoring neighborhood youth (Sampson, Earls, and Radenbush 1997; Sampson et al. 2002). Researchers expect to learn more about the effects of relocation from high- to low-poverty neighborhoods in terms of the adaptation to new “neighborhood cultures.” Ethnographic studies in the context of urban poverty (Anderson 1999; Bourgois 1995; Sullivan 1989) provide very little insight into how poor minority youth navigate in new, mixed-income or mixed-race environments (Briggs 1997). MTO offers the opportunity to examine whether and for how long “codes of the street” dominate the outlook and behavior of young people who have left threatening inner-city “street” contexts and how children's age and gender influence these new cultural exposures.

Central to this next phase of qualitative research will be to learn if there are clear, irreversible developmental trajectories that are seen or are felt by participating youth. Once a child, for example, has achieved some positive improvement in employment, health, or education, will changes continue in a predictable path? Will children's lives be permanently and irreversibly altered by MTO, or will there be substantial reversal or “backsliding”? Will, at some point, treatment group children's futures dramatically improve, with students moving on to college and better paying jobs compared with their control group counterparts or will the appeal of low-poverty areas wear thin, leading families to retreat to their former, more familiar communities? Will reversals in the growth rate of the economy shake the hold MTO families have on their new situations? The persistence or fading of effects is therefore also a key concern as we seek to understand and model both temporal and behavioral factors affecting MTO families (Garces, Thomas, and Currie 2002).

Closely allied with the above is the question of whether early positive changes in certain domains will help leverage change in other aspects of families' lives. We hypothesize that our proposed research will help determine whether there is an evolving, positive synergy of effects- a ripple effect. Does the sense of relief from fear of crime and violence after moving lead mothers to permit their children more time to play outdoors with other children in their new, lower-poverty neighborhoods? What, if any, measurable consequences will such greater peer interactions have on subsequent socialization and educational achievement? Does improvement in physical or mental health alter other behavior such as interest in education or employment? When does a child's improved performance at school reduce parents' anxiety, improve their mental health, and lead to additional positive choices and options? When and how does a child's feeling of being safe translate into better school performance or different peer network choices?

The key methods that researchers expect to make use of in the next research stage for MTO will include:

• In-depth qualitative interviews with a sample of adults and teens that will probe their views on how neighborhood environment has affected their lives, as well as probe to understand their behavior, particularly in the domains of employment, peer associations, education, and training;
• Ethnographic fieldwork consisting of direct observation and informal interviewing of a small sample of households, focusing on the experiences of adolescents and young adults as they pursue education, build social lives, and begin to transition into the labor market and adulthood more broadly; and
• Neighborhood scans providing objective, broad-based analyses of key contextual/environmental traits that may affect the well-being of MTO families. It will offer a basis for comparing the “apparent” neighborhood seen by planners with the “functional” neighborhood as residents use and experience it.

Regarding the neighborhood "scans," researchers have noted the utility of a small number of recent studies that use a variety of methods to provide micro-level analyses of the ecological character of services and facilities that “neighborhoods” offer residents. Such methods allow more flexible definitions of “neighborhood” to emerge and make possible the measurement of more nuanced neighborhood characteristics. The Project on Human Development in Chicago Neighborhoods, for example, (Sampson and Raudenbush 1999) made use of trained observers to record descriptions or scans of key local institutions, services, and amenities. Researchers expect to make use of existing “objective” neighborhood research coupled with the use of trained observers walking through a sample of neighborhoods to record descriptions of physical and social conditions. These descriptions of physical and ecological conditions and opportunities will be linked to ethnographic interviews to learn how children, youth and adults actually see and use their effective neighborhoods. This research is consistent with prior ethnographic studies that have shown the important connections between physical settings and peer networks that play an important role in pushing adolescents toward or away from criminal activities (Anderson 1990; Sullivan 1989).

Conclusions

While most Americans take for granted their right and option to change residences as often as they need and can afford, federal housing rules previously prevented the same flexibility for residents of public housing. They were, if not trapped, restricted to the housing benefit for which they applied. MTO was established as a small-scale demonstration, the target population of which included a maximum of 1,600 families already living in some of the country's most distressed public housing neighborhoods. They were offered vouchers (then known as Section 8 rental assistance and now called housing choice vouchers) which gave them the option of moving out of the projects.

MTO is one part of the puzzle of learning whether a place-based approach to poverty redevelopment complements or works better than a person-based, MTO-type strategy. How well, and at what cost, will these programs work in transforming the lives of former residents of federally funded slums? Might the federal government be able to relocate some number of the willing poor out of slum communities into better lives by offering them housing vouchers usable in a range of neighborhoods?⁶

Whether the outcomes of these MTO projects will result in a net advantage or disadvantage for the minority poor is yet to be determined. There are skeptics who will argue that these tools are too indifferently funded and insufficiently coordinated with the private market to achieve sustained improvement in the lives of residents (Goetz 2000; Dimond 2000, 260). Others argue that these programs will work only to the extent that larger changes, such as welfare reform and a strong economy, reinforce them (Weisberg 2000). Only time and carefully conducted research will provide answers. A core reason for MTO was a commitment to conduct the research that would enable policy makers to learn whether moving out of ghetto projects would be a source of both hope and positive behavioral change for those who participated.

Neighborhood-based fears and organizations can, however, effectively thwart weakly based policies aimed at dispersal or accommodation; they have done so fairly successfully for the entire history of HUD and its predecessor agencies. More robustly funded but small-scale programs could emerge, rooted in the willingness to support PHAs that wish to deconcentrate poverty. Heclo (1994, 427) suggests another part of the solution:

full-scale attacks on ghetto poverty will inevitably mean targeting resources disproportionately on minorities. Whether such efforts are seen as pro-black preferences or an act of solidarity with the country's children and its future will depend heavily on how political leaders help educate the public.

MTO, we believe, offers political leaders helpful evidence that children's lives have been changed; MTO is about children's futures and not racial preferences.⁷

Heclo (1994, 407) has also worried that federal policies all too easily can become "half-hearted, poorly administered, and politically orphaned gestures" Such worry might also aptly apply to HUD's efforts to systematically undo the segregation and economic isolation experienced by public housing families. The political commitment to authorizing and funding policy tools aimed at poverty reduction and racial desegregation has indeed been fragile. MTO was, however, a policy creation whose success in being established by Congress was then matched by HUD's operating a demonstration.

Research has now found evidence of positive effects, balanced by negative effects for boys, but all set within the context that the program succeeded in freeing children and their parents from crime and fear. These first systematic insights into housing mobility help, we believe, to establish the relevance of the choice to move out.

Endnotes

1. This research was conducted on the Pruit-Igoe project in St. Louis, which was among the very first U.S. housing projects demolished in 1972.

2. Gautreaux's methodological shortcomings were nonetheless, at the time, a notable improvement over typical cross-sectional, non-experimental studies of neighborhood effects in which omitted variable bias and the absence of data on the temporal sequencing of effects have been of major concern.

3. Several comments about measures are necessary. Experimental statistical effects are typically reported using "intent to treat" (ITT) as the central measure. MTO provides an estimate of the effectiveness of the offer of the experimental treatment in improving the lives of public housing residents as a group . The intent-to-treat estimates reported recognize that some members of the target group did not use the Section 8 subsidy. In experimental research, ITT effects reflect all those randomly assigned into a treatment or control group, whether or not they were able to take advantage of the treatment services provided. In this case, ITT effects are those measured for the entire population after the point of randomization, whether or not they took advantage of the chance to move out. The measured ITT effects include the outcomes not only for those who moved but also for those who were randomly assigned to receive the treatment but did not relocate. However, even if the ITT effects are statistically significant, it is still true that the larger the proportion of those who fail to move, the less effective a program like MTO would be in improving the lives of poor families.

Intent-to-treat (ITT) effects are measured by looking at the difference between the average outcome for the entire MTO treatment group or the entire Section 8 comparison group and the outcome for the control group. For example, if the average poverty rate for census tracts occupied by members of the treatment group was 10 percent, the intent-to-treat effect is the difference between that rate and the control group's average poverty rate (let's say 48 percent); thus, the ITT effect is 38 percent. The treatment-on-treated (TOT) effect— that is, the estimated effect on those persons who successfully moved under MTO— is generally higher, as it is measured for only those participants who actually participated in the treatment (moved with Section 8 vouchers). The analysis in this paper mainly focuses on intent-to-treat effects, noting that whenever ITT effects are statistically significant, TOT effects are also significant and are stronger.

Once having established that we will only focus upon ITT effects, or those harder to detect, it is then necessary to indicate how high the hurdle it is in order for MTO to achieve statistically meaningful impacts at standard levels of statistical importance. To get a sense of the likely effect of most demonstrations it is of some interest to recall Peter Rossi's skeptical judgment that most experiments have no impacts that can be detected. He has argued that “the best a priori estimate of the net impact assessment of any program is zero; that is, the program will have no effect” (Rossi 1987).

One of the reasons that many demonstrations are able to show little statistical impact is that the thresholds for measurable change are quite high. Larry Orr (2003), for example, notes that on many outcomes only quite large impacts from MTO would be detected with confidence. For example, "TANF benefits would have to be reduced by 50 percent; youth asthma attacks would have to drop by two-thirds, or 67 percent; and adult earnings would have to increased by about 40 percent in the experimental group" to reach conventional levels of significance. Such huge behavioral changes would likely only occur with a larger or more "powerfully" constructed experimental program.

An additional, final methodological concern is that most non-experimental estimates of neighborhood effects appear to differ from experimental ones. That is, there appear to be good reasons to question whether non-experimental estimates of neighborhood effects are sensible. For MTO, researchers' comparison of experimental and non-experimental estimates got quite different results: the experiment, for example, makes the neighborhood effects on boys look more adverse. Since experimental effects appear to differ from non-experimental findings, there is no easy option to rely, as some have suggested, on non-experimental studies of TOT effects as an effective surrogate for experimental research outcomes (Smith and Todd 2000; Leventhal and Brooks-Gunn 2003; Harding 2003). That is, MTO is still very much needed as a necessary complement to non-experimental estimates.

4. Larry Orr, et al. (2003). Moving to Opportunity Interim Impacts Evaluation. (September). Washington, D.C.: U.S. Department of Housing and Urban Development. See: http://www.huduser.org/publications/fairhsg/mtoFinal.html

5. Quoted in Susan Popkin, Laura Harris, and Mary Cunningham (2002). "Families in Transition: A Qualitative Analysis of the MTO Experience." Washington, D.C.: U.S. Department of Housing and Urban Development, May. See: http://www.huduser.org/publications/pubasst/mtoqualf.html

6. This policy recommendation has been put forward, in a fairly pronounced form, by Owen Fiss (2003).

7. MTO was not, of course, designed to be the silver bullet that will help kill off ghetto poverty. It is rather part of the spectrum of policy choices which included the option to stay in place in a public housing development that is less isolated, more habitable, and which better serves the needs of its residents. MTO addresses the additional option for poor families to move into a variety of kinds of other, non-poor neighborhoods (Brown and Richman 1997; Downs 1994, 112-114). MTO was not intended, and should not be, the only choice available to the poor. Ongoing research on MTO's families will help to make this case even more persuasively.

John Goering received his Ph.D. in sociology and demography from Brown University. His research focuses upon housing and neighborhood development, as well as race and ethnic issues. He is the author of several dozen articles as well as editor or author of The Best Eight Blocks in Harlem (University Press, 1977); Housing Desegregation and Federal Policy (University of North Carolina Press, 1986); Mortgage Lending, Racial Discrimination and Federal Policy (Urban Institute Press, 1996); and Choosing a Better Life? Evaluating the Moving to Opportunity Experiment (Urban Institute Press, 2003). He serves as a consultant on research and litigation for the U.S. Department of Housing and Urban Development.

References

Anderson, Elijah (1999). Code of the Street: Decency, Violence, and the Moral Life of the Inner City. New York: W.W. Norton.

Bickford, Adam and D. Massey (1991). "Segregation in the Second Ghetto: Racial and Ethnic Segregation in American Public Housing, 1977." Social Forces 69 (June): 1011-1036.

Bourgois, Philippe (1995). In Search of Respect: Selling Crack in El Barrio. New York: Cambridge.

Briggs, Xavier de Souza (1997). "Moving up Versus Moving Out: Neighborhood Effects in Housing Mobility Programs." Housing Policy Debate 8(1): 195-234.

Brooks-Gunn, Jeannie, et al. (1997). "Lessons Learned and Future Directions for Research on the Neighborhoods in Which Children Live." In Brooks-Gunn, Jeannie, et al., eds. Neighborhood Poverty: Contexts and Consequences for Children (Vol. 1). New York: Russell Sage, 279-287.

Brown, Prudence and Harold Richman (1997). "Neighborhood Effects and State and Local Policy." In Brooks-Gunn, Jeannie, et al., eds. Neighborhood Poverty: Policy Implications in Studying Neighborhoods (Vol. 2). New York: Russell Sage, 164-181.

Colton, Kent W. (2003). Housing in the Twenty-First Century: Achieving Common Ground . Cambridge, Mass: Harvard University Press.

Coulibaly, Modibo, Rodney Green and David James (1998). Segregation in Federally Subsidized Low-Income Housing in the United States. Westport: Praeger.

Davis, Mary (1993). "The Gautreaux Assisted Housing Program." In Kingsley, G. Thomas and Margery Austin Turner, eds. Housing Markets and Residential Mobility. Washington, D.C.: The Urban Institute Press.

Dimond, Paul (2000). "Empowering Families to Vote with Their Feet." In Katz, Bruce, ed. Reflections on Regionalism . Washington, D.C.: Brookings Institution, 249-271.

Downs, Anthony (1994). New Visions for Metropolitan America . Washington, D.C.: Brookings Institution.

Edin, Kathryn and Laura Lein (1997). Making Ends Meet: How Single Mothers Survive Welfare and Low-Wage Work. New York: Russell Sage.

Farley, Reynolds, Sheldon Danziger and Harry Holzer (2000). Detroit Divided. New York: Russell Sage.

Fiss, Owen (2003). A Way Out: America's Ghettos and the Legacy of Racism. Princeton: Princeton University.

Fountain, John (2001). "Violence Is Down, but Some Areas Still Suffer." The New York Times. New York, January 11.

Furstenberg, Frank and Mary Elizabeth Hughes (1997). "The Influence of Neighborhood on Children's Development: A Theoretical Perspective and a Research Agenda." In Brooks-Gunn, Jeannie, et al., eds. Neighborhood Poverty: Policy Implications in Studying Neighborhoods (Vol. 2). New York: Russell Sage, 23-47.

Galster, George , ed. (1996). Reality and Research: Social Science and U.S. Urban Policy since 1960. Washington, D.C.: The Urban Institute Press.

Garces, Eliana, Duncan Thomas and Janet Currie (2002). "Longer-Term Effects of Head Start." Poverty Research News 6 (March/April): 3-5.

General Accounting Office (2003). Public Housing: Hope Vi Resident Issues and Changes in Neighborhoods Surrounding Grant Sites . GAO-04-109. Washington, D.C.: GAO. November.

Gephart, Martha (1997). "Neighborhoods and Communities as Contexts for Development." In Brooks-Gunn, Jeannie, Greg Duncan, and J. Lawrence Aber, ed. Neighborhood Poverty: Contexts and Consequences for Children (Vol. 1). New York: Russell Sage, 1-43.

Goering, John , ed. (1986). Housing Desegregation and Federal Policy. Chapel Hill: University of North Carolina.

Goering, John and Judith Feins (2003). Choosing a Better Life? Evaluating the Moving to Opportunity Experiment. Washington, D.C.: The Urban Institute Press.

Goetz, Edward (2003). Clearing the Way: Deconcentrating the Poor in Urban America. Washington, D.C.: The Urban Institute Press.

Harding, David (2002). "Counterfactual Models of Neighborhood Effects: The Effect of Neighborhood Poverty on High School Dropout and Teenage Pregnancy." American Journal of Sociology 109(3): 676-719.

Hartman, Chester (1963). "The Limitations of Public Housing: Relocation Choices in a Working Class Community." Journal of the American Institute of Planners 29 (November): 283-296.

Haveman, Robert (1994). "The Nature, Causes, and Cures of Poverty: Accomplishments from Three Decades of Poverty Research and Policy." In Danziger, Sheldon, et al., eds. Confronting Poverty: Prescriptions for Change. New York: Russell Sage Foundation, 438-448.

Hays, R. Allen (1995). The Federal Government and Urban Housing . Albany: State University of New York Press.

Heclo, Hugh (1994). "Poverty Politics." In Danziger, Sheldon, et al., eds. Confronting Poverty: Prescriptions for Change . New York: Russell Sage Foundation, 396-437.

Hirsch, Arnold (1983). Making the Second Ghetto: Race and Housing in Chicago, 1940-1960. New York: Cambridge University Press.

Hogan, James (1996). Scattered Site Housing: Characteristics and Consequences. Washington, D.C.: U.S. Department of Housing and Urban Development.

Holzer, Harry (1991). "The Spatial Mismatch Hypotheses: What Has the Evidence Shown?" Urban Studies 28(1): 105-122.

Husock, Howard (2000). "How Charlotte Is Revolutionizing Public Housing." City Journal Spring: 52-59.

Ihlanfeldt, Keith (1999). "The Geography of Economic and Social Opportunity in Metropolitan Areas." In Altshuler, Alan, et al., eds. Governance and Opportunity in Metropolitan America. Washington, D.C.: National Academy Press.

Jargowsky, Paul (2003). Stunning Progress, Hidden Problems: The Dramatic Decline of Concentrated Poverty in the 1990s. Washington, D.C.: The Brookings Institution.

Jencks, Christopher and Susan Mayer (1990). "The Social Consequences of Growing up in a Poor Neighborhood." In Lynn, Laurence and Michael McGeary, eds. Inner-City Poverty in the United States. Washington, D.C.: National Academy Press, 111-186.

Katz, Bruce and Margery Turner (2000). Who Should Run the Housing Voucher Program? A Reform Proposal. Washington, D.C.: The Brookings Institution.

Kling, Jeffrey and Jeffrey Liebman (2003). Causal Effects on Youth Outcomes of Moving out of High Poverty Neighborhoods in the Moving to Opportunity Experiment. Princeton: Princeton University. November.

Lehman and Timothy Smeeding (1997). "Neighborhood Effects and Federal Policy." In Brooks-Gunn, Jeannie, et al., eds. Neighborhood Poverty: Context and Consequences for Children (Vol. 1). New York: Russell Sage, 251-278.

Leventhal, Tama and Jeannie Brooks-Gunn (2003). "A Randomized Study of Neighborhood Effects on Low-Income Children's Educational Outcomes." Developmental Psychology December.

Ludwig, Jens, Greg Duncan and Joshua Pinkston (1999). Neighborhood Effects on Economic Self-Sufficiency: Evidence from a Randomized Housing-Mobility Experiment. Working Paper Number 159. Northwestern University / University of Chicago Joint Center for Poverty Research. January 31.

Ludwig, Jens, Helen Ladd and Greg Duncan (2001). "Urban Poverty and Educational Outcomes." Brookings Papers on Urban Affairs 2001: 147-197.

Lupton, Ruth (2003). 'Neighborhood Effects': Can We Measure Them and Does It Matter? CASE paper no. 73. London: Centre for Social Exclusion, London School of Economics. September.

Massey, Douglas and Nancy A. Denton (1993). American Apartheid: Segregation and the Making of the Underclass. Cambridge, Mass: Harvard University Press.

Mendenhall, Ruby, Greg Duncan and Stefanie DeLuca (2004). "Neighborhood Resources and Economic Mobility: Results from the Gautreaux Program." Unpublished paper.

Newman, Sandra and Ann Schnare (1997). "'…and a Suitable Living Environment': The Failure of Housing Programs to Deliver on Neighborhood Quality." Housing Policy Debate 8(4): 703-741.

Orr, Larry (2003). Moving to Opportunity: Assessing and Interpreting the Findings. Annual Meetings of Appam, November 6, Washington, D.C.

Popkin, Susan, et al. (2000). The Hidden War: Crime and the Tragedy of Public Housing in Chicago. New Brunswick, NJ: Rutgers University.

Pribesh, Shana and Douglas Downey (1999). "Why Are Residential and School Moves Associated with Poor School Performance?" Demography 36(4): 521-534.

Rainwater, Lee (1970). Behind Ghetto Walls: Black Family Life in a Federal Slum. Chicago: Aldine.

Rosenbaum, James E. (1992). "Black Pioneers--Do Their Moves to the Suburbs Increase Economic Opportunity for Mothers and Children?" Housing Policy Debate 2: 1179-1213.

Rossi, Peter H. (1980). Why Families Move: A Study in the Social Psychology of Urban Residential Mobility. Beverly Hills: Sage Publications.

Rossi, Peter H. (1987). "The Iron Law of Evaluation and Other Metallic Rules." In Miller, Joann and Michael Lewis, eds. Research in Social Problems and Public Policy. Greenwich, CT: JAI Press.

Rubinowitz, Leonard and James Rosenbaum (2000). Crossing the Class and Color Lines: From Public Housing to White Suburbia. Chicago: University of Chicago Press.

Sampson, R.J., S.W. Raudenbush and F. Earls (1997). "Neighborhoods and Violent Crime: A Multilevel Study of Collective Efficacy." Science 277: 918-924.

Sampson, Robert, Jeffrey Morenoff and Thomas Gannon-Rowley (2002). "Assessing 'Neighborhood Effects': Social Processes and New Directions for Research." Annual Review of Sociology 28: 443-478.

Schill, Michael (1992). "Deconcentrating the Inner City Poor." Chicago-Kent Law Review 67(3): 795-853.

Schweinhart, Lawrence, Helen Barnes and David Weikart (1993). Significant Benefits: The High/Scope Perry Preschool Study through Age 27. Ypsilanti, Michigan: The High/Scope Press.

Sjoquist, David, ed. (2000). The Atlanta Paradox. New York: Russell Sage.

Smith, Jeffrey and Petra Todd (2000). "Does Matching Overcome Lalonde's Critique of Nonexperimental Estimators." Unpublished paper.

Sullivan, Mercer (1989). Getting Paid: Youth Crime and Work in the Inner City. Ithaca, New York: Cornell University Press.

Turner, Margery, Susan Popkin and Mary Cunningham (2000). Section 8 Mobility and Neighborhood Health: Emerging Issues and Policy Challenges. Washington, D.C.: The Urban Institute.

Vale, Lawrence (2000). From the Puritans to the Projects: Public Housing and Public Neighbors. Cambridge, Mass: Harvard University Press.

Wilson, William Julius (1987). The Truly Disadvantaged. Chicago: University of Chicago Press.

Wilson, William Julius (1991). "Public Policy Research and the Truly Disadvantaged." In Jencks, Christopher and Paul Peterson, eds. The Urban Underclass. Washington, D.C.: Brookings Institution, 460-481.

Wolfe, Alan (1998). One Nation, after All. New York: Viking.