«Moving to opportunity voluMe 14, nuMber 2 • 2012 U.S. Department of Housing and Urban Development | Office of Policy Development and Research ...»
A Journal of Policy
Development and Research
Moving to opportunity
voluMe 14, nuMber 2 • 2012
U.S. Department of Housing and Urban Development | Office of Policy Development and Research
Managing Editor: Mark D. Shroder
Associate Editor: Michelle P. Matuga
Richard K. Green
University of Southern California
Keith R. Ihlanfeldt
The Florida State University
Annette M. Kim
Massachusetts Institute of Technology
Carlos E. Martín
Abt Associates Inc.
Douglas S. Massey Princeton University Sandra J. Newman Johns Hopkins University Marybeth Shinn Vanderbilt University Raymond J. Struyk Paul Waddell University of California, Berkeley John C. Weicher Hudson Institute, Inc.
Cityscape A Journal of Policy Development and Research Moving to opportunity voluMe 14, nuMber 2 • 2012 U.S. Department of Housing and Urban Development Office of Policy Development and Research The goal of Cityscape is to bring high-quality original research on housing and community development issues to scholars, government officials, and practitioners. Cityscape is open to all relevant disciplines, including architecture, consumer research, demography, economics, engineering, ethnography, finance, geography, law, planning, political science, public policy, regional science, sociology, statistics, and urban studies.
Cityscape is published three times a year by the Office of Policy Development and Research (PD&R) of the U.S. Department of Housing and Urban Development (HUD). Subscriptions are available at no charge and single copies at a nominal fee. The journal is also available on line at http://www.huduser.org/periodicals/cityscape.html.
PD&R welcomes submissions to the Refereed Papers section of the journal. Our referee process is double blind and timely, and our referees are highly qualified. The managing editor will also respond to authors who submit outlines of proposed papers regarding the suitability of those proposals for inclusion in Cityscape. Send manuscripts or outlines to Cityscape@hud.gov.
Opinions expressed in the articles are those of the authors and do not necessarily reflect the views and policies of HUD or the U.S. government.
Visit PD&R’s websites, www.hud.gov/policy or www.huduser.org, to find this report and others sponsored by PD&R. Other services of HUD USER, PD&R’s Research and Information Service, include listservs, special interest and bimonthly publications (best practices, significant studies from other sources), access to public use databases, and a hotline (1–800–245–2691) for help with accessing the information you need.
Contents Symposium Moving to Opportunity Guest Editor: Jens Ludwig Guest Editor’s Introduction
Acknowledgment of Extraordinary Obligations
Moving to Opportunity: Why, How, and What Next?
by Mark D. Shroder and Larry L. Orr Achieving MTO’s High Effective Response Rates: Strategies and Tradeoffs
by Nancy Gebler, Lisa A. Gennetian, Margaret L. Hudson, Barbara Ward, and Matthew Sciandra MTO: A Successful Housing Intervention
by Jennifer Comey, Susan J. Popkin, and Kaitlin Franks The Long-Term Effects of Moving to Opportunity on Adult Health and Economic Self-Sufficiency
by Lisa Sanbonmatsu, Jordan Marvakov, Nicholas A. Potter, Fanghua Yang, Emma Adam, William J. Congdon, Greg J. Duncan, Lisa A. Gennetian, Lawrence F. Katz, Jeffrey R. Kling, Ronald C. Kessler, Stacy Tessler Lindau, Jens Ludwig, and Thomas W. McDade The Long-Term Effects of Moving to Opportunity on Youth Outcomes
by Lisa A. Gennetian, Matthew Sciandra, Lisa Sanbonmatsu, Jens Ludwig, Lawrence F. Katz, Greg J. Duncan, Jeffrey R. Kling, and Ronald C. Kessler Making MTO Health Results More Relevant to Current Housing Policy: Next Steps......... 169 by Thomas D. Cook and Coady Wing Constrained Compliance: Solving the Puzzle of MTO’s Lease-Up Rates and Why Mobility Matters
by Kathryn Edin, Stefanie DeLuca, and Ann Owens Increasing the Value of MTO Research for Housing Policy Development
by Edgar O. Olsen Moving Neighborhoods Versus Reforming Schools: A Canadian’s Perspective
by Philip Oreopoulos Commentary: MTO’s Contribution to a Virtuous Cycle of Policy Experimentation and Learning
by Margery Austin Turner iii Cityscape Contents Point of Contention: Defining Neighborhoods Guest Editor: Ron Wilson The Tyranny of Census Geography: Small-Area Data and Neighborhood Statistics............ 219 by Jonathan Sperling Defining Neighborhoods in Space and Time
by Ralph B. Taylor Defining Neighborhoods for Research and Policy
by Claudia Coulton Dynamic Geography: The Changing Definition of Neighborhood
by Marc S. Buslik Refereed Papers Geographic Patterns of Serious Mortgage Delinquency: Cross-MSA Comparisons............. 243 by Lariece M. Brown, Hui-Chin Chen, Melissa T. Narragon, and Paul S. Calem The Housing Needs of Rental Assistance Applicants
by Josh Leopold Departments Graphic Detail Geographic Patterns of Regional Unemployment Versus Unemployment Compensation in the United States—2009
by Ron Wilson Data Shop Introducing the Ohio New Establishment Dynamics Data
by Joel A. Elvery and Ellen Cyran Impact Impact Analysis of the Proposed Rule on Streamlining the Portability Process in the Housing Choice Voucher Program
by Yves Sopngwi Djoko Correction Comparing Public Housing and Housing Voucher Tenants With Bayesian Propensity Scores
by Brent D. Mast Referees 2011-12
iv Moving to Opportunity Guest Editor’s Introduction Jens Ludwig University of Chicago National Bureau of Economic Research The contents of this introduction are the views of the author and do not necessarily reflect the views or policies of the U.S. Department of Housing and Urban Development, the Congressional Budget Office, the U.S. government, or any state or local agency that provided data.
Residential segregation of America’s neighborhoods by income has been increasing over the past 40 years, with nearly 9 million people now living in census tracts with poverty rates of 40 percent or more (Kneebone, Nadeau, and Berube, 2011; Watson, 2009). Because housing policy affects the geographic concentration of poverty in a variety of ways, policymakers have long been concerned about the possibility that living in a distressed neighborhood could have some harmful effects on the life outcomes of adults and children. The list of plausible reasons why neighborhood poverty might adversely affect people’s well-being and behavior is long and includes limited exposure to peers and role models who support prosocial behaviors such as schooling and work; neighbors who are willing and able to cooperate and work together to improve community life; high-quality local public institutions such as schools, police, health care, and housing; and elevated exposure to risk factors like pollution or crime.1 Empirically isolating the independent effects of neighborhood environments on the life outcomes of residents turns out to be quite challenging in practice, because most people have at least some degree of choice regarding where they live. A large body of research dating back to the 17th century shows that people who live in relatively more distressed neighborhoods tend to have worse life outcomes than do those people living in less disadvantaged areas, even after statistically adjusting for characteristics of the individuals and their families. What remains unclear is the degree to which these patterns reflect true neighborhood effects—that is, the causal influence of neighborhood environments on the life outcomes of residents—or instead reflect the influence of hard-to-measure characteristics of people that lead them to wind up living in different types of neighborhoods—or what social scientists call selection bias.
To overcome concerns with selection bias and help isolate neighborhood effects on low-income families, in the early 1990s the U.S. Department of Housing and Urban Development (HUD) For excellent reviews of the theoretical and empirical literatures on neighborhood effects, see Ellen and Turner (1997), Jencks and Mayer (1990), Kawachi and Berkman (2003), Leventhal and Brooks-Gunn (2000), and Sampson, Morenoff, and Gannon-Rowley (2002).
launched one of the most ambitious social experiments ever carried out by the agency—the Moving to Opportunity (MTO) for Fair Housing demonstration. Via random lottery, MTO offered some public housing families, but not others, the chance to use a housing voucher to move from highpoverty to lower poverty neighborhoods. Some of the families who were offered the opportunity to relocate received special vouchers that initially could be used to move into only very low-poverty areas, although, after 1 year, families could use the vouchers to move again (including to higher poverty places).
This issue of Cityscape focuses on the long-term followup that measured outcomes of MTO families 10 to 15 years after random assignment. This long-term followup study was carried out by a research team assembled by the National Bureau of Economic Research (NBER) and was supported by a contract with HUD and additional grants to NBER from other agencies and private foundations. In this guest editor’s introduction, I provide some basic background about MTO that frames all of the articles that follow, and I offer some thoughts of my own about what lessons we might take from MTO for both social science and public policy.
In the second section, I briefly review the motivation for the MTO demonstration and the specifics of its design. A more detailed discussion of MTO’s rationale and design is in the article in this symposium by Mark D. Shroder and Larry L. Orr. The symposium article by Jennifer Comey, Susan J. Popkin, and Kaitlin Franks shows that MTO was successful in helping families move into higher quality housing units. The article by Edgar O. Olsen in this symposium notes that the cost to taxpayers of providing higher quality housing units to MTO voucher holders might actually be zero or negative, in the sense that previous research suggests that the cost of providing a given level of housing quality might be lower with vouchers than public housing. Olsen notes, however, that there would be great value in exploiting the MTO platform to learn even more about these cost-effectiveness issues.
In the same section, I also show that MTO was successful in getting families to move initially into very low-poverty areas. One year after randomization, the difference in tract poverty rates between the control group and those who were offered housing vouchers to move into low-poverty areas was about 35 percentage points, or fully 2.8 standard deviations in the nationwide census tract poverty rate distribution. Previous housing mobility programs have found that families initially relocated into low-poverty areas tend to “stick” (Keels et al., 2005). An open empirical question is whether the same would be true for MTO families.
In the third section, I review the evidence showing that the very large initial differences in average neighborhood conditions between the two MTO treatment groups and the control group narrowed over time. This convergence is commonly attributed to the tendency of families who move with MTO vouchers to make additional moves back to higher poverty areas and has led to calls for the government to provide additional supports to voucher recipients to help them stay in low-poverty areas once they have moved there. I show that, somewhat surprisingly, most of the convergence over time between MTO treatment and control groups in neighborhood poverty rates is actually due to improvements over time in the neighborhoods of the control group.
In the fourth section, I consider the key question of whether MTO generates enough sustained variation in neighborhood conditions to provide a useful test of the “neighborhood effects”
hypothesis. When we look across the entire 10- to 15-year followup study period, moving with an MTO voucher reduces average census tract poverty rates by about 18 percentage points, equal to nearly one-half of the control group’s average tract poverty rate of 40 percent. This is about as much variation in neighborhood poverty as we see in studies of African-American families in leading observational data sets like the Project on Human Development in Chicago Neighborhoods (PHDCN). MTO generates less pronounced differences across randomly assigned groups in racial segregation, although as Shroder and Orr discuss in their article, much of the discussion leading up to MTO was about neighborhood-effect theories that emphasized adverse effects from economic segregation more so than from racial segregation.
MTO also had large, sustained impacts on more subtle neighborhood attributes that are not readily measured with existing administrative data sources, such as social networks and neighborhood social processes and safety, and that require original in-person data collection from the MTO participants to measure. Because families were followed up over such a long time (10 to 15 years), and because low-income families tend to be very residentially mobile and hence difficult to track, no one would have been surprised if the long-term survey effort had wound up with a low response rate. As Nancy Gebler and her co-authors note in their article in this symposium, however, the team from the University of Michigan tasked with carrying out the surveys achieved remarkably high response rates to preserve the key strength of MTO’s experimental design: 90 percent for adults and 89 percent for youth, which were very similar across randomly assigned MTO groups. Gebler et al.’s article includes some useful lessons for future researchers about how to track similar populations, and it presents some interesting results about what we would have found in the MTO data had we run out of time and money and been forced to stop the data collection at a lower response rate.