Episode 44: HOPE VI Public Housing Redevelopment with Rebekah Levine Coley

Coley, Spielvogel, Hwang, Lown, and Teixeira, "Did HOPE VI Move Communities to Opportunity? How Public Housing Redevelopment Affected Neighborhood Poverty, Racial Composition, and Resources 1990–2016"

More

Episode Summary: HOPE VI was a federal program running from 1993–2010 that sought to redevelop distressed, poor, racially segregated public housing into mixed-income communities. In that time it helped build nearly 100,000 new homes for people of varying incomes, and with the involvement of both the public and private sectors. Its goal was to reduce concentrated poverty and racial segregation; so how did it do? Rebekah Levine Coley joins us to share her research into the impacts of HOPE VI redevelopment on neighborhood poverty, racial composition, and community resources. We also discuss the lessons from earlier generations of public housing and urban renewal that informed HOPE VI, and what the program can tell us about gentrification, displacement, the role of the private sector, and much more.

  • “Public housing is a key federal investment, albeit one that falls far short of meeting affordable housing needs in the U.S. In fiscal year 2021, more than $7.8 billion in federal funding was invested in public housing, which accommodated ∼1.2 million households across the U.S. (U.S. Department of Housing & Urban Development, 2021). The provision of public housing has undergone dramatic reenvisioning since the 1980s following years of disinvestment, neglect, and social problems that entrenched problematic housing and social conditions in public housing neighborhoods. In response to these challenges, policymakers developed the Housing Opportunities for People Everywhere (HOPE) VI program with the goal of transforming public housing by improving housing quality, increasing economic integration, revitalizing neighborhoods, and providing paths to resident self-sufficiency (U.S. Department of Housing & Urban Development, n.d.). The program was bold in its design and set out to deconcentrate poverty and improve neighborhood conditions by creating mixed-income communities, fostering public–private partnerships to leverage federal funds (U.S. Department of Housing & Urban Development, n.d.). Approximately $6.7 billion in public funding was spent on HOPE VI between 1993 and 2010, with billions more in public and private funds leveraged (U.S. Department of Housing & Urban Development, n.d.). These public–private partnerships were responsible for the redevelopment of 262 public housing sites across 34 states, Washington DC, and Puerto Rico into mixed-income communities (Aratani et al., 2019; Gress et al., 2016; Popkin et al., 2009). Despite this notable federal investment, many questions remain concerning the program’s effectiveness at meeting its goals, particularly in relation to achieving neighborhood-level change.”

 

  • “Despite years of federal investment and much rhetoric about the ways HOPE VI’s focus on improved urban design and poverty deconcentration would benefit public housing communities and the broader neighborhoods in which they are embedded, surprisingly little research has documented how redeveloping communities of concentrated poverty into mixed-income communities affects core neighborhood features. There is a particular dearth of research that addresses selection bias, uses representative samples across the country, and assesses the potential for subgroup differences across key geographic characteristics (with the notable exception of Tach & Emory, 2017) … Accordingly, in this article, we use rigorous quasi-experimental methods and a unique data set incorporating a broad array of geocoded national administrative data to better understand the impact of HOPE VI on surrounding neighborhoods while also addressing methodological and measurement limitations in prior research.”

 

  • Questions addressed in the article:
    • “How does redevelopment of public housing through HOPE VI affect concentrated poverty, racial demographics, and the availability of resources in the targeted neighborhoods?
    • What neighborhood characteristics shift most quickly, and are the shifts durable over time?
    • Do effects of HOPE VI redevelopment depend on the features of neighborhoods and of the public housing sites undergoing redevelopment?”

 

  • “HOPE VI policy has since undergone several iterations, with the overall goal of improving the physical and social infrastructure of public housing developments. The first iteration of the HOPE VI policy aimed to redevelop or demolish crumbling public housing units and provide paths to resident self-sufficiency. Self-sufficiency was defined loosely in the 1996 Notice of Funding Availability for HOPE VI, which encouraged applying housing authorities to incentivize “moving up and out of public housing” through services like child care and training programs to improve employability and economic self-reliance … In line with this thinking, later iterations of HOPE VI embodied a comprehensive approach that provided for supportive services for residents and an influx of funding to redevelop or replace severely distressed buildings using mixed-income models to deconcentrate poverty. Housing authorities were authorized to use 80% of federal funds for redevelopment, rehabilitation, and physical improvements and up to 20% for supportive services for residents.”

 

  • “Mixed-income development aims to reshape public housing and alleviate poverty and associated stressors by drawing higher income residents to communities of concentrated poverty, which was hoped to spur neighborhood investment and bring valuable neighborhood amenities (Bulger et al., 2021). Such policies operate from the premise that social mixing of residents across income strata will improve both the physical environment and the life chances and economic mobility of low-income residents through increasing resources and social and political capital, decreasing physical decay and crime, and stabilizing neighborhoods (Chaskin & Joseph, 2010; Dulin-Keita et al., 2016; Galster, 2007). However, questions have been raised about effective implementation and whether policies truly support the economic and social mobility of existing residents, or whether they simply spur gentrification of poor urban neighborhoods with little benefit to public housing residents (Goetz & Chapple, 2010).”

 

  • “Many of the prior neighborhood-level studies have focused on effects on neighborhood economic composition, finding that HOPE VI was successful in decreasing neighborhood poverty and increasing the economic diversity of residents (Zielenbach, 2003; Zielenbach & Voith, 2010). However, much of the evidence is descriptive, assessing pre/post differences or comparing changes in HOPE VI neighborhoods to city or regional changes, while typically assessing a single or a limited number of HOPE VI redevelopments … In short, prior research has not adequately controlled for heterogeneity bias to identify causal impacts of HOPE VI, nor assessed generalizable samples of HOPE VI sites across the U.S.”

 

  • “A handful of studies have used more rigorous analytic methods to address selection bias. Two studies estimated interrupted time series models to evaluate HOPE VI sites across six urban cities, finding decreased poverty rates and increased household incomes in the neighborhood (defined as a 200-foot radius around public housing property) following redevelopment (Zielenbach et al., 2010; Zielenbach & Voith, 2010). In the most comprehensive and rigorous analysis to date, Tach and Emory (2017) examined a national sample of HOPE VI revitalization and demolition grants … Combining public housing data and census data of block groups containing public housing from 1990, 2000, and 2010, they used difference-in-differences models that incorporated propensity score modeling of selection factors to compare neighborhoods with HOPE VI sites to those with non-HOPE VI public housing. Results indicated declining poverty rates and increasing economic diversity within HOPE VI block groups as well as surrounding block groups … Although far more rigorous than prior research on the effects of HOPE VI, this work has several important limitations.”

 

  • “Other research has assessed whether the HOPE VI policy led to shifting racial demographics in public housing neighborhoods … [P]rior studies have found racially disparate effects. Existing data suggest that residents who faced the most negative effect of HOPE VI, namely displacement, were predominantly families of color, with the most dramatic estimates suggesting that, at least in the first round of HOPE VI funding, as much as 95% of those displaced by HOPE VI were families of color (Keene & Geronimus, 2011; National Housing Law Project et al., 2002; Popkin et al., 2004) … Other studies evaluating the effects of HOPE VI on the racial composition of neighborhoods have found variable patterns of change across sites and across studies. For example, analyzing changes to racial composition in neighborhoods receiving HOPE VI demolition grants, Goetz (2011) found that the majority of sites experienced an increase in White population and decrease in Black population, although about one fifth of the sites experienced stable or increasing Black population (Goetz, 2011) … The presence of cross-site variation in descriptive or case study analyses suggests that the effects of HOPE VI on neighborhood racial composition may depend on pre-redevelopment neighborhood characteristics. The present study aims to extend prior work by assessing short- versus longer term changes in racial composition across HOPE VI revitalization sites, and by assessing specific neighborhood and development characteristics that may moderate this effect.”

 

  • “A handful of qualitative studies has suggested that economic investment and availability of social and educational services and local amenities, such as grocery stores, increased in some HOPE VI sites (Cloud & Roll, 2011; Levy & Gallagher, 2006). Meares and Riggs (2016) found some short-term growth in businesses around two HOPE VI sites, but the growth was inconsistent in later years. Zielenbach (2003) noted increased residential loan rates across eight HOPE VI sites, a proxy for economic investment in the area. Together, limited qualitative and quantitative evidence from small-scale studies has associated HOPE VI with increasing or improving neighborhood resources. No study has undertaken a comprehensive quantitative analysis of whether HOPE VI redevelopment affected the availability of diverse types of resources.”

 

  • “To address our research questions, we created a new integrated data set on HOPE VI redevelopment grants, public housing developments, and neighborhood characteristics. Because we were interested in understanding the effect of HOPE VI redevelopments on neighborhood characteristics, we decided to focus on census tracts as the unit of analysis … Similar to Tach and Emory (2017), we restricted our sample to urban areas (categories 1–3 of the 1993 rural–urban continuum) within the continental U.S. and to census tracts containing moderate or large public housing developments (over 25 units), excluding scattered sites and elder-only developments, as these were not the target of the HOPE VI policy. These exclusions (along with issues defined below) led to a total potential sample of 3,884 census tracts.”

 

  • “Data on public housing were drawn from HUD’s “Picture of Subsidized Households” database. Information on the locations of public housing developments, as well as characteristics of households living in those developments, were available in 1993, 1996–1998, 2000, and annually from 2004 to 2016. Using Project ID numbers, we were able to follow specific public housing developments from 1993 through 2007, and from 2008 through 2016, tracking changes in development and resident characteristics over time … Exclusion of scattered sites, nonurban areas, and developments with 25 or fewer units resulted in a set of 7,202 public housing developments across the U.S.”

 

  • “Sample descriptives are presented in Table 1 for HOPE VI tracts and all potential control tracts with data drawn from 2 years prior to grant receipt. HOPE VI tracts were generally more economically disadvantaged and had more Black residents than potential control tracts, and had larger public housing developments. For example, HOPE VI tracts had average poverty rates of 40.10% versus 25.51% in non-HOPE VI tracts (2 years prior to grant receipt), and contained 60.08% Black residents compared to 31.15%. HOPE VI tracts also had lower housing values, greater vacancy rates, and more public housing units: 369 public housing residents, on average, in HOPE VI tracts versus 180 in non-HOPE VI tracts. However, HOPE VI tracts also appear to have had slightly more resources and services, with 7.91 institutional resources and 4.86 restaurants compared to 7.34 and 4.18, respectively, in non-HOPE VI tracts.”

 

  • “In this model, data were reorganized and then treatment cases were matched with control cases based on a designated date in relation to treatment start. We matched 2 years prior to the initiation of HOPE VI grant start dates to sidestep concerns about Ashenfelter’s dip (Ashenfelter, 1978) and to capture characteristics of neighborhoods at the approximate time they were applying for HOPE VI grants. As such, for each tract containing a HOPE VI redevelopment, data from 2 years prior to the grant start date for that tract and for all eligible non-HOPE VI tracts were used to identify the best-matched control tract at that specific time point. Beyond the exact match on time, we matched on geographic region to control for macroeconomic and political contexts as well as a series of neighborhood characteristics, chosen based on literature concerning how HOPE VI sites were selected to capture factors that may have affected treatment assignment (Heckman et al., 1999). We used an exact match within geographic region (nine geographic areas, termed districts—e.g., New England, West North Central) and nearest-neighbor matching using a combined statistical distance function on percent poor, percent Black, percent non-Black people of color, median income, median home values, vacancy rate, institutional resources, and, if not already included, the outcome variable, conducting a separate matching for each analytic model (hence for each of our 11 outcome variables). Across all of the 11 models, this matching process led to a sample with 309 HOPE VI tracts and 261–269 control tracts (see Table 2).”

 

  • “After assessing the effects of HOPE VI on the full sample of HOPE VI and matched control sites, we assessed whether effects varied across diverse neighborhoods. For each of our moderators (neighborhood poverty rate, percent Black residents, number of public housing units, total redevelopment cost, mixed-income), we split the sample of HOPE VI neighborhoods (using a median split for continuous moderators) and matched HOPE VI to control neighborhoods within each subgroup (e.g., within the HOPE VI neighborhoods with <40% poverty 2 years prior to HOPE VI start), and then assessed treatment effects of HOPE VI within this subsample, followed by matching and assessing effects within the other subsample (e.g., within neighborhoods with ≥40% poverty). Again, these analyses showed successful matches, with the occasional exception of vacancy rates.”

 

  • “Results of the flexible conditional difference-in-differences models are presented in Table 3. Results show significant effects of HOPE VI on neighborhood poverty rates that declined only modestly up to 10 years following HOPE VI completion. Specifically, compared to matched controls, tracts with HOPE VI developments had an average 2.92 percentage point greater decrease in poverty rates through the end of treatment. This effect remained relatively stable over time, decreasing to a 2.52 percentage point difference in tract poverty through 2 years post-HOPE VI completion, to a 2.27 percentage point difference through 5 years postcompletion, and to a slightly larger but no longer statistically significant (due to greater variability) 2.46 percentage point difference in tract-level poverty through 10 years postcompletion. Similarly, HOPE VI led to an increase in median household income, with a significant increase of .08 log units through the completion of HOPE VI, an effect that stayed relatively stable through 2 years postcompletion prior to declining to nonsignificance. In comparison, HOPE VI had no significant effect on rates of affluence in the tract. Together, these results seem to suggest that HOPE VI redevelopment led to decreased poverty and perhaps a rise in moderate-income residents, but did not increase the prevalence of affluent residents in tracts experiencing public housing redevelopment through HOPE VI (see alternate specifications section for further insights).”

 

  • “In contrast, we found no significant effects of HOPE VI on the racial makeup of census tracts. Results show primarily negative but nonsignificant changes in the percentage of Black residents and in the percentage of residents who were non-Black people of color. Similarly, we found no significant effects of HOPE VI redevelopment on the presence of institutional resources, social services, grocery/convenience stores, restaurants, bars, or liquor stores in the neighborhood.”

 

  • “A second set of models assessed whether effects of HOPE VI on neighborhood makeup and resources varied depending upon the original characteristics of the neighborhood. Models first addressed variation by neighborhood poverty rates at the time of matching (2 years prior to the initiation of HOPE VI). Results (Table 4) show that HOPE VI effects on neighborhood poverty were driven by HOPE VI sites in high-poverty neighborhoods. For example, HOPE VI led to a nonsignificant 1.33% greater decline in poverty through the completion of the redevelopment process within low-moderate poverty neighborhoods, but to a significant 4.19 percentage point difference in poverty within high-poverty neighborhoods. Extending to 10 years postcompletion, this effect was even stronger, with HOPE VI leading to a nonsignificant .25 percentage point difference in poverty in neighborhoods that started as low/moderate poverty, but to a 5.21 percentage point difference in poverty within high-poverty neighborhoods. A similar pattern emerged in relation to effects on median household income. HOPE VI led to a .04 greater increase in log units of median household income in low/moderate poverty neighborhoods, a significant effect that decreased by 5 years postcompletion, but to a much larger .11 log unit greater increase in median household income within high-poverty neighborhoods. This effect stayed consistent in size, although it became less precise through 10 years postcompletion. Like the main effect results, no significant results emerged in relation to the percentage of affluent residents, the percentage of Black and non-Black people of color, or the multiple types of neighborhood resources and amenities we assessed.”

 

  • “We next assessed whether the impacts of HOPE VI varied across tracts that were primarily Black. Results (Table 5) show that HOPE VI led to decreased poverty and increased median household income primarily in majority Black neighborhoods, decreasing neighborhood poverty by 3.8 percentage points more and increasing median household income by 0.1 log unit more in HOPE VI versus control tracts through the end of treatment. As one exception, the percentage of affluent residents increased through 5 years after HOPE VI completion only within tracts that started with fewer Black residents, although this effect was not consistent over time periods. No significant effects emerged in relation to the neighborhood racial makeup or resources and amenities.”

 

  • “Finally, we tested variation in effects by the level of income mixing that resulted from HOPE VI redevelopment (Table 8). HOPE VI led to significant relative declines in neighborhood poverty only among tracts with mixed-income HOPE VI redevelopments. This effect grew over time, increasing from a 4.12 percentage point reduction at the end of redevelopment to a 6.99 point reduction through 10 years after redevelopment in HOPE VI versus control tracts. The increase in median household income was similarly driven by mixed-income redevelopment and increased slowly over time, ranging from a .14 log unit increase through the end of redevelopment to a .18 log increase through 10 years following redevelopment. Still, no effects of HOPE VI were found for racial makeup or any of our assessed types of resources even among tracts including more costly or mixed-income developments.”

 

  • “While acknowledging the methodological innovations and strengths of our multisource data set and incorporation of new modeling techniques, it is also essential to acknowledge the limitations in this work when interpreting results and considering implications. Many of the limitations are common across neighborhood research. For example, although we focused on census tracts as meaningful proxies for neighborhoods, tracts do not always match local definitions of neighborhoods and may misrepresent the day-to-day context of public housing residents (Coulton et al., 2013). Also, although we excluded very small public housing sites with fewer than 25 units, HOPE VI sites varied in size (and location within census tracts), and smaller or more peripheral sites may not have had the capacity to exert substantial effects on their broader neighborhoods (although our moderation analyses comparing HOPE VI redevelopments within neighborhoods with more versus fewer public housing units showed minimal differences in effects) … Although additional models removed census tracts that included public housing developments receiving demolition-only HOPE VI and CNI grants, we were not able to account for other redevelopment efforts that may have derived from state or local programs and policies or federal policies, such as Low-Income Housing Tax Credits (Vale & Shamsuddin, 2017). Similarly, the HOPE VI program evolved over the 17 years in which it was actively funding grantees, and each site had unique qualities and some level of local control over how funds were deployed (Curley, 2010).”

Shane Phillips 0:04
Hello, this is the UCLA Housing Voice podcast, and I'm your host, Shane Phillips. This week's episode is with Professor Rebekah Levine Coley, and we're talking about the Hope Six program, an effort led by the US Department of Housing and Urban Development to redevelop distressed public housing into mixed income communities. That program ran for about 17 years from 1993 to 2010, and it built roughly 100,000 new units during that time. Now a little more than 10 years later, it's time for a retrospective. As many listeners will know, public housing communities in the US have struggled with high rates of poverty and racial segregation, and so a few explicit goals of Hope Six were to reduce poverty and increase racial diversity. One important strategy for achieving that was to build housing that people across a range of incomes were eligible for which is in contrast to public housing's historic focus exclusively on very poor households. In our conversation, we get into how well the program actually achieved those goals, including the ways its impacts varied in different neighborhood contexts. And we spent a lot of time digging into interesting questions about gentrification, displacement, the role of the private sector, and a whole bunch else. For me, it was a super useful review of a program that impacted hundreds of thousands of people, but that I knew very little about. It was also a great reminder of some of the ways that we've really learned from our past, and other ways that we still haven't. Just a quick note that we're publishing today's episode a day late, and that is entirely my fault. Hopefully, no one was missing us too much on their Wednesday morning commute. I also mentioned my book during the interview, possibly for the first time in 44 episodes, and if you're interested in finding it at your local bookstore or library, the name is 'The Affordable City, and it's published by Island Press'. This podcast happens to be an excellent complement to the book by sheer coincidence I'm sure. The Housing Voice podcast is a production of the UCLA Lewis Center for Regional Policy Studies, with production support from Claudia Bustamante and Jason Cetacea. As always, feedback and show ideas can go to me at Shanephillips@ucla.edu. Now let's get to our conversation with Professor Rebekah Levine Coley. Rebekah Levine Coley is Professor of Counseling Developmental and Educational Psychology and Director of the Institute of Early Childhood Policy at Boston College, and she's joining us today to talk about Hope Six, a federal program that redevelops public housing into mixed income communities, and its impacts on neighborhood poverty, racial composition, and community resources. Rebekah, welcome to the Housing Voice podcast.

Rebekah Levine Coley 2:49
Thank you, it's great to be here.

Shane Phillips 2:51
My co-host today is Mike Lens. Hey, Mike.

Michael Lens 2:53
Hey, Shane, and great to meet you today Rebekah, and good to be back on the pod.

Shane Phillips 2:59
So as always, we start by asking our guests for a quick tour of their hometown or a place that is special to them. Rebekah , you said you want to do your hometown, Waterville Maine. What makes that place special other than you were born there, of course, and grew up.

Rebekah Levine Coley 3:14
That's the most special

Shane Phillips 3:16
Where do you want to take us as as guests here?

Rebekah Levine Coley 3:19
Well, I think it's just a small town in Central Maine, and it's I see it now from my adult viewpoint as a microcosm of the loss of the manufacturing economy in America. So through my childhood, my hometown lost multiple paper mills and clothing manufacturers that employed substantial numbers of town residents. And that was followed by many hospitals merging or closing, which really led to just a sort of overall decline in the economic vitality of the town and a loss of many local businesses. And they're still really, I think, struggling to retain their vitality. On the other hand, it's a lovely place - Colby College is there, has a fantastic art museum, and it's in a great sort of lakes region in Central Maine that's beautiful.

Michael Lens 4:12
That reminds me, I'm reading Bill Fulton's recent book "Place and Prosperity", and he starts by talking about his upstate New York hometown- I think It's called Auburn, and he makes this this point that seems so obvious. He talks about how difficult it is for people in the 21st century to understand just how self contained small towns were in like the 1950s 1960s right where like, you just did not need to leave this town for basic errands for your job, etc. Like he talks about generations of his family that commuted a mile to work, you know, within that town and it was just it's just

...Crazy.

You know pre Walmart supercenter era

Yeah, yeah. It just struck me like, Oh, we've I bet most of us have forgotten this right?

Rebekah Levine Coley 5:08
Right

Michael Lens 5:08
That like you did not need, we had the butcher we had you know, all everything was just right there, and it just seemed so obvious in retrospect but to see it in print, I almost needed that.

Rebekah Levine Coley 5:20
I think my hometown was 18,000, maybe 19,000 people when I was a kid, we had three hospitals, we had multiple department stores. It was considered the big city in the area with 18,000 people so it really, it went through just a dramatic change.

Shane Phillips 5:39
So the article we're talking about was published last year in Housing Policy Debate, and it's titled 'Did Hope Six Move Communities to Opportunity: How Public Housing Redevelopment Affected Neighborhood Poverty, Racial Composition, and Resources 1990 to 2016'. In it, Professor Coley and her co-authors Bryn Spielvogel, Dabin Hwang, Joshua Lown and Samantha Teixeira look at how the Hope Six program, which ran from 1993 to 2010, and redeveloped public housing with mixed-income housing, changed the neighborhoods and communities around them both in the short term and up to 10 years after being completed. And just so our listeners know, Hope Six is an acronym that stands for Housing Opportunities For People Everywhere, but we're just going to use the acronym for the rest of the discussion, because that's all I ever hear anyone say.

Michael Lens 6:33
So just not to digress, but Hope Six definitely made it into my dissertation, 'Housing Opportunities For People Everywhere' did not so I don't know how old I was when I learned what that acronym stood for but nobody uses it

Shane Phillips 6:50
I was going to say like immediately after this that it's not in my notes but I had to note that it definitely sounds like the kind of acronym where someone came up with the acronym first and then backfield.

Michael Lens 7:01
Also, what is the six?

Shane Phillips 7:02
I didn't find out what the six stands for, is it just like the sixth draft...

Rebekah Levine Coley 7:06
I don't know because I don't think that there was a Hope One, Hope Two, Hope Three, I don't believe...

Michael Lens 7:11
I once knew this and I'm not even gonna try; there is a reason yeah.

Shane Phillips 7:16
Okay so yeah, a little background for folks that gets you nowhere. Rebekah to set the stage here, could you tell us a little bit about the history of public housing? I realize that's a big question but you know, how it led to the creation of the Hope Six program and why policymakers believed that mixed-income redevelopment was a better strategy than say, redeveloping all the housing solely for low-income households. Like what lessons have been learned from previous public housing construction or rehabilitation programs that were incorporated into Hope Six?

Rebekah Levine Coley 7:51
Sure. I think, as you say, it's got a long complicated history, (and) I'll hit some of the highlights. So public housing was started as part of the New Deal, the Housing Act of 1937, and so the federal government started funding the creation of public housing developments primarily in urban areas. I think right from the get-go, there was deep racism and segregation built into the system, and inequities based on race, with black communities having fewer amenities, and often being located in less desirable areas. Starting in the 70s, under Nixon - President Nixon, funding for public housing really declined, flatlined or declined, and public housing development started falling into greater and greater disrepair. And I think through the last couple of centuries, the last couple of decades of the 20th century and into the 21st, there were more and more problems with the quality of housing with social problems and public housing communities. And I think at this time, we also had evidence from urban sociologists like William Julius Wilson, starting to really show the detrimental effects of concentrated poverty neighborhoods for children, adults and communities. So I think all that together led to a desire to try to improve public housing. So the federal government commissioned a big report and then started the Hope Six program to try to rebuild and improve public housing communities, both the housing quality but also to lower concentrated poverty and to increase economic opportunities for public housing residents.

Shane Phillips 9:30
And what role did the mixed income part of that play? Why did they feel, I mean I guess I can see you know, it's deconcentration but could you talk a little bit more about, sort of, the philosophy or approach there?

Rebekah Levine Coley 9:44
I think again, from work like William Julius Wilson's 'The Truly Disadvantaged' who showed how the loss of more educated and more economically secure members of a community and movement toward concentrated poverty was very detrimental. And I think, you know, evidence from many other sources as well started to show the benefits of more integrated communities, both racially integrated, (and) economically integrated, for providing economic vitality, for providing opportunities, for providing role models for children and youth about their potential, for improving school quality (etc). So I think for all of those reasons, the government reports and the policy that was developed, really sought to increase economic integration and to move away from very concentrated poverty neighborhoods that were populated entirely by families in poverty or families (of)people with disabilities.

Shane Phillips 10:44
Got it. And I think we'll almost certainly come back to this, but we'll need to kind of juxtapose that deconcentration against the more modern conversations about gentrification and sort of the flip side to that.

Rebekah Levine Coley 10:57
Exactly!

Shane Phillips 10:58
So you're researching the impacts of the Hope Six program, but you're not the first to look into it. It started 30 years ago now. What did the previous research have to say about the impact of the Hope Six program on the residents of public housing and their surrounding communities? And as part of your summary, I think it might be helpful to mention some of the shortcomings of that earlier work that you had hoped to improve upon with your own team's research.

Rebekah Levine Coley 11:25
Sure. So there certainly have been many evaluations and studies of the Hope Six program using different methodologies, different methods, different statistical analytic techniques, focusing on different locations. I think all of those studies had limitations and internal and external validity in different ways. So...

Shane Phillips 11:47
Which we should probably say, literally, every study has limitations.

Rebekah Levine Coley 11:51
Absolutely, absolutely.

Michael Lens 11:57
And for our listeners who are not researchers or social scientists, internal validity...

Rebekah Levine Coley 12:07
I'll explain what that means

Michael Lens 12:08
okay then I'll shut up.

Rebekah Levine Coley 12:10
You can go right ahead if you'd like. So I think many of the studies looked at just a handful of sites, of Hope Six sites, in just one or two cities, for example. So you couldn't really tell if what they found in those sites would generalize to other sites across the country

Michael Lens 12:28
External validity!

Exactly! Many of them used quite descriptive methods so just looked at the community before the Hope Six grant and then after the Hope Six grant so you couldn't really tell if the Hope Six redevelopment caused any changes that were found...

A quality lesson in internal validity!

Rebekah Levine Coley 12:46
Exactly, and also most studies, previous studies, focused on pretty short term outcomes. So we were interested in looking at a longer-term picture. There's certainly, you know, there is substantial body of research, there are some studies that are definitely more rigorous, and more inclusive of a wide variety of sites across the country. I would say the predominant findings were that Hope Six led to a decrease in neighborhood poverty, and an increase in sort of economic variability in neighborhoods. There, I would say is really mixed research about whether the racial composition of neighborhoods changed after Hope Six; so some earlier work suggested that families of color, I think particularly black families, were far more likely to be displaced from Hope Six redevelopment. So often, they sort of displaced everyone in that community and tore down the housing, and rebuilt it. Many of the people who were displaced were not allowed to come back but (there's) mixed research about whether there was overall racial displacement or changes in the racial makeup of neighborhoods in which Hope Six developments were located.

Shane Phillips 13:58
If I can jump in there really quickly, when you say, you know, many of these families were not allowed to come back; what form did that, you know, if it was a prohibition, what did that look like?

Rebekah Levine Coley 14:08
Different sort of policies and systems in different locations I think but in some places, so many Hope Six grants were simply demolition grants so they just tore down the public housing and did not rebuild it. So of course, in that case, there's not really anything to come back to. Some of them, many of them were rebuilding grants, so they tore down the housing and then rebuilt it. But many of the sites sort of put new rules or prohibitions into place which limited original residents' ability to come back; so drug testing, different sorts of policies. Many families also once they got relocated, they might have found a different neighborhood, you know, they became embedded within that neighborhood and perhaps didn't want to move yet again.

Shane Phillips 14:52
Right right

Rebekah Levine Coley 14:53
I think that the third area that we were looking at in our study is the issue of resources. So a goal of the Hope Six program was to increase economic opportunities for residents and increase services, amenities and communities. And there's been very little research on that area in the past, a couple of qualitative studies suggesting that there was some improvement in services over the short term, but no really systematic work in that area.

Shane Phillips 15:21
And in both the research community and the advocacy community, what's the current attitude toward Hope Six, you know, more than 10 years after it's ended? I gather that there are, you know, let's call them mixed feelings.

Rebekah Levine Coley 15:35
Right.

Shane Phillips 15:36
And you mentioned how there's a sense, among at least some people, that this program was really used as a way to bring private capital in and gentrify neighborhoods rather than benefit the people living in these public housing communities. I think Ed Goetz is one such researcher who is kind of made that claim.

Rebekah Levine Coley 15:53
Exactly, I think you're right, I think there are mixed feelings about the success of of Hope Six in meeting its goals, and certainly some scholars and advocates, and policymakers that think that it caused more harm than good; that many people, original residents, were displaced - many permanently displaced, and some like Goetz argued that there was, you know, I think he calls it a sort of "government system of gentrification" or something like that. And there's also evidence that it led to an overall decrease in affordable housing units overall so definitely mixed mixed feelings about its success.

Michael Lens 16:36
Yeah, and I think, you know, just to piggyback on some of those criticisms, another thing that people point out is, this came from Congress in the early 90s. In 1992, there was a commission to study, you know, distress public housing as a as a central problem. And, you know, they came up with this number of 86,000 units that we need to redevelop, which wouldn't pretend

Rebekah Levine Coley 17:01
Which is only a very small proportion of all of the public housing units.

Michael Lens 17:05
Yes, yes.

Shane Phillips 17:07
But around 1.2 million at the time, or maybe that's the number today?

Michael Lens 17:10
Yeah, yeah

Rebekah Levine Coley 17:11
That's the number today.

Michael Lens 17:12
Yeah, I think it was more like 1.4 or five at that point, and they demolished or redeveloped 250,000 inevitably through the Hope Six program. And so, you know, I think that's an easy target for people who think that the program overreached, right? That people found more problems in public housing than actually existed. But, you know, I'm not saying that's definitely true but like, that's I think one of the things that people point to, right.

Shane Phillips 17:45
Well, yeah the fact that we ended up with less public housing at the end is one.

Rebekah Levine Coley 17:50
Right, and I think the way, I mean I think well we'll talk later on I think about some of the lessons learned and where we're going in the future with public housing redevelopment policies, and I think one of the major lessons learned was the negative effects of this sort of wholesale at one period of time, demolition and redevelopment where by definition, you are displacing every single person in that community which means not only that each individual has to move, but that you're completely breaking all of the social ties that that community had. And that people are being displaced to different locations, children might have to switch schools, parents might have a more difficult time getting to their job. And I think we've we've learned that lesson, and new models of redevelopment are really taking that into account and trying to develop redevelopment systems that that very much limit external displacement of that sort.

Michael Lens 18:47
Right, and also, I think that's a great point, and we also didn't have one-for-one replacement built into this program, right. So in a lot of cases, housing authorities chose to replace fewer units than they demolished, right? So people inevitably couldn't necessarily come back uhm

Rebekah Levine Coley 19:07
Right or replace them with other types of affordable housing, not public housing development.

Shane Phillips 19:13
I knew very little about this program going into it other than kind of the name and the mixed income nature and so forth, and I'm kind of amazed at how much of this reminds me of urban renewal despite it coming, you know, 30,40,50 years later. And, you know, having a lot of lessons from Urban Renewal by the 1990s and 2000s, like we knew a lot of this stuff.

Rebekah Levine Coley 19:36
And I think the policy did take a sort of new urbanists design perspective of trying to move away from big high rises into more attractive buildings that were better integrated into the community so I think it did take into account some of those previous lessons.

Michael Lens 19:53
Yes, I agree for sure. So, you know, before we get too far into the study, you know, as a researcher myself, I'd like to hear more about how you and your research team came to this inquiry. So your own background is in developmental psychology and you're collaborating with folks in that field, and in social work. This is not your first time studying housing and neighborhood outcomes for sure but more of your work is, probably if I understand it, you know focused on family composition, adolescent risk behaviors, you know, more bread and butter developmental psychology topics. And then your research team is different as well from kind of the usual suspects in housing policy but I think this is a really outstanding paper and and so I'm really curious how this group kind of came together and how you came to study this topic in particular. And you know, if you might get into like how your different kinds of training or skills like brought a different perspective than we usually see.

Rebekah Levine Coley 20:59
Sure. So I am indeed a developmental psychologist although I did postdoctoral training and policy - public policy and demography. So most of my research focuses on economic and racial inequities, and how those affect children's development and family well being. I've done lots of work evaluating other types of policies or programs like welfare reform and early childhood education. But I think my interest in housing and in communities is a key part of developmental psychology. So a sort of a key tenant is that children and humans in general, develop in context, and that the context in which we live have a substantial impact on our developmental trajectories. So I've been engaged in research on Housing and Communities. For decades, actually, my first sort of research projects out of graduate school were in the South side of Chicago working in public housing developments in the South side of Chicago. But I think this type of policy analysis is not so common overall in the field of developmental psychology, and my colleague, Sam Teixeira on this project is a social worker who has done a lot of community development work but often from a sort of participatory perspective - working in communities. So this sort of perspective was was new for her as well but we started this project actually, a few years ago, we were invited to collaborate with a public housing redevelopment proposal, (a) proposed project in Boston where a large public housing development is going to be redeveloped into a mixed income community. And the public housing authority and the developer and social service agencies working with them really wanted an academic partner to try to understand the needs, the goals and strengths of the current community to help inform the redevelopment effort. So we got involved in that project and through that started really delving into the previous literature and the Hope Six literature and realize that there were some, what we saw as big, holes in the research base that we thought we might be able to fill. So we pulled some of our graduate students together and got some funding - a small grant from the Russell Sage Foundation which we appreciate and launched into this project.

Michael Lens 23:18
Cool!

Shane Phillips 23:19
So your research is basically asking the question of how neighborhoods changed in places where Hope Six redevelopment took place, relative to those where public housing existed but it was not redeveloped through Hope Six, during the study period, at least. And as the title of your article says, you're specifically looking at neighborhood poverty, racial composition, and community resources; you've mentioned a little bit about what those meant. So you know, maybe we don't have to go too far into detail for each of these and what they mean, because we will talk about them in the results as well. But I wonder if you could talk, maybe at least a little bit about what a good outcome represents for some of these things? I think for something like community resources, it's pretty straightforward. Like, are there more of them? Right. But you know, with lower poverty, partly because of the gentrification conversation, if the share of the population that is black decreases, is that Is that a good thing? Or what are the different ways that some of these outcomes could be interpreted?

Rebekah Levine Coley 24:21
Right, that's a great question and a challenging set of issues. So some of the research on Hope Six and similar policies take an individual perspective, so they follow individual residents of a community or of many communities over time to see if their lives have improved or if they were displaced, and then can look at the whole population they're studying and you know see if those changes change the tenor of of the neighborhood. In this study, we are using all administrative data; so we decided to focus on the neighborhood as the unit of analysis and look at changes in neighborhood characteristics so poverty rates, affluence rates, racial demographics and resources. And it's, so Hope Six had some clear goals to increase economic integration. So if you look at a goal like that, and say, "well, what does that really mean, and in what way or for whom would that benefit residents". So you could get economic integration by keeping all of the original residents of the community who are primarily poor and bringing in middle income or higher income neighbors. You could change the economic integration by pushing out some of the poor residents and either just pushing them out or pushing them out and then bringing in higher income neighbors. Or you could change the tenor of the neighborhood by keeping the original residents and helping them to improve their economic resources so they became less poor, less likely to be poor. It's a little hard to tear those possibilities apart with administrative data so we used...

Shane Phillips 25:59
Right, you're not you're not tracking the...

Rebekah Levine Coley 26:01
We're not tracking individuals

Shane Phillips 26:04
People, households, right. I mean, that's extraordinarily hard to do with the data we have available especially in the US .

Rebekah Levine Coley 26:11
It's difficult and very expensive and very time consuming. So using administrative data, though, what we tried to do is look at a number of different measures. So we looked at the percent of people in poverty, but also the percent of people in affluence, at like the top 5% of the income distribution, as well as median incomes. And we looked at the percentages, but we also looked at the number of people in each of those categories, so that we could try to dig in a little bit into this issue of, for example, gentrification and displacement.

Shane Phillips 26:44
Could you talk a little bit more about that 5% threshold or looking at people in that group, the top 5% of the income distribution, presumably, this is the top 5% of the income distribution for either the whole country or the region in which the public housing is located. But you're also measuring median household income, which is a common thing to do - including this top 5% of the income distribution as a measure of the number of affluent residents is something I hadn't seen before. I'm sure someone's done it, but it's certainly much less common. Like what distinction are you trying to make there? You know, in what ways do you see a rising median household income as maybe having different impacts on the neighborhood than a rising share of the households being affluent?

Rebekah Levine Coley 27:29
Right, so I think one reason was this gentrification issue. So I think, typically, we think of gentrification as both the displacement of lower resource families but also the, sort of, really higher resource families, high income (and) high education families, coming in and sort of taking over a neighborhood. So we wanted to get at not only was the median income shifting, but were the people at the upper echelons of the income distribution coming into the neighborhood. So we saw that as one way of trying to get at this gentrification issue. I think the other reason is, there is research in sociology and other fields, and developmental psychology, looking at the effects of poverty versus affluence in neighborhoods, and that they're not just flips of each other, right? Because there's also the whole middle. So you can have a decline, you can have an increase in affluence, but actually not have a decline in poverty, because you're displacing people in the middle. There's also evidence that even a small portion of affluent residents, 4% or 5%, is an important predictor of things like lowered adolescent risk behavior. So there's a study, I think it was the early 90s, from Crane, showing that just having 5% of people in the neighborhood with professional level jobs was a very substantive predictor of lower adolescent pregnancy rates. So it can be an indication of economic opportunities, again, to go back to William Julius Wilson's argument that people with economic power and professional level jobs are really important sources of resources for other people in the neighborhood, to have as role models, to have as access for information into how to move up the economic ladder.

Shane Phillips 29:21
Did you think about maybe setting a different threshold for that income distribution? When I'm looking at it, I feel like 5% is quite high.

Rebekah Levine Coley 29:29
5% is pretty high, yes

Shane Phillips 29:31
It's a very affluent household, especially for public housing community. I'm thinking like, you know, someone's in the top 20%, making $120,000 or whatever, you know, that level is, that's a very, you know, that's a professional, presumably they're having some influence, the same kind of influence as a person making 500,000 or whatever. How did you come to that threshold specifically?

Rebekah Levine Coley 29:53
That's a great question. That was a data limitation of administrative data. So we created that measure using census data, decennial census data and ACS data, and they have sort of categories of income, and that was a category that was easily distinguished that we could track over time; that equal to approximately 5% - the top 5% of the income distribution. So it was a data driven decision that exact counterpoint.

Shane Phillips 30:21
Got it. Could you speak to the issue of selection bias and how you tried to address it in your study, and what selection bias means in the context of the States?

Rebekah Levine Coley 30:30
So selection bias is sort of a perennial problem in almost all research that's not random experimental designs. So for example, if you want to understand the effects of say something like parental divorce on children, you can't simply compare parents who get divorced to parents who don't get divorced and look at their child outcomes. Because there's very likely to be different characteristics of those two groups of people that helped to select them into becoming divorced, that also are likely to affect their children. So the same thing with Hope Six neighborhoods, they weren't randomly assigned across the country, it was a selective process, communities had to apply for Hope Six grant. So we can't simply compare communities public housing communities that got Hope Six grants to those that did not because they're likely to be different, to have very different characteristics. And in fact, we show that they do have very different characteristics; that the public housing communities that got Hope Six grants had much higher poverty rates, they had much higher proportions of black residents than other public housing communities in urban areas. They actually had slightly higher rates of services and amenities before getting a Hope Six grant than did non-Hope Six sites.

Shane Phillips 31:48
Which sort of points toward that concern that this is like private capital, wanting to go into the places where there's already sort of right, the resources there and kind of capitalize on them right?

Rebekah Levine Coley 31:58
Exactly. Yep, so that that was part of the Hope Six programs agenda was to find communities that had the potential to improve and to improve their resources. So perhaps the argument is you needed some level of a decent base to be able to improve. So as I noted, we showed that there was very substantial selection bias or differences across places that got Hope Six grants and those that didn't. So we used a new analytic technique to address that. It's built on prior work, but it's called the 'Flex panel DID' system that's available in STATA, and it uses a matching algorithm to match "treatment groups" - so here the treatment is receiving a Hope Six grant to "control groups" which is a community or public housing community not receiving a Hope Six grant.

Shane Phillips 32:53
Just to clarify here, because I myself don't remember this from your paper, and I don't think I took a note on it, but for those controls, these are places that had public housing. Were they also places that applied to Hope Six and didn't receive it, or could they have applied but not received it or maybe some didn't apply at all? It's just like, is the control just the existence of public housing or is it the existence of public housing plus they applied and didn't get a grant?

Rebekah Levine Coley 33:19
It's just the existence of public housing.

Shane Phillips 33:22
And some probably applied or definitely applied, and others didn't.

Rebekah Levine Coley 33:25
And others didn't, right right, we did not have access to those data about whether they applied got. So they were all public housing developments in urban areas, in the continental United States, that had public housing developments with at least 25 units. So we excluded scattered sites, we excluded elderly only developments, because those were not the target of the Hope Six program. I think what makes this the technique that we used a little bit of an improvement on past research, is that when you're thinking of the "treatment of Hope Six" - getting a Hope Six grant, there was variation across time, right. So the program ran for many years so grants were given out over different years. There's variation in the timing of when the treatment started, and variations in the length of treatment. So in the matching process, we took into account the variation in time. So we matched Hope Six sites with control sites two years before they got their Hope Six grant; if we argued when they would be applying for their Hope Six grant and having to present neighborhood characteristics to show their eligibility, and we matched on year, we matched on the region in the country, and then we matched on a whole host of neighborhood characteristics like the poverty rate and the affluence rate and the racial demographics, (and) also things like housing vacancy rates, homeownership rates. So really getting a strong match - so for each Hope Six site, so the census track in which a Hope Six development was located, we found its best match census track at that time period - at exactly the two years before the Hope Six Grant was started in that community.

Shane Phillips 35:10
And part of the thinking there is just that, like, you know, if you complete a project in 1999, right before there's a big recession, you know...

Rebekah Levine Coley 35:20
Right

Shane Phillips 35:20
And you see the impacts of that, they might have much more to do with the national recession than anything specific to the Hope Six redevelopment compared to something that is completed in, say 2004, and then you're measuring the next two years, which look great.

Rebekah Levine Coley 35:33
Exactly exactly

Shane Phillips 35:33
Until until you get around 2008. And maybe you're, you're back down again,

Michael Lens 35:37
Right.

Rebekah Levine Coley 35:37
And also, the Hope Six program changed over time, a little bit, so their priorities shifted a little bit.

Shane Phillips 35:44
Yes

Michael Lens 35:44
Uhm..

Rebekah Levine Coley 35:44
And we were able to show that all of those preexisting differences between Hope Six communities and non Hope Six communities were erased after our matching. So there were no more significant well, with one or two little tiny exceptions, no more significant differences between our "treatment groups" and our "control groups".

Michael Lens 36:04
Yes, that's great, that's hard to do. So related, our audience is probably tired of hearing this phrase, but I'm working on a book on on black neighborhoods...

Shane Phillips 36:15
We will never tire of hearing about it Mike

Michael Lens 36:19
Sometime soon, we might have an entire podcast...

Shane Phillips 36:22
I'm gonna have to like actually mention the fact that I have a book at some point on this podcast

Michael Lens 36:26
That would be a good idea, that would be a good idea. Your publisher might appreciate that. Um, so anyhoo so my attention, of course, was was directed at the right away to the racial composition of Hope Six census tracts which you just talked about, but specifically, those census tracts were approximately 60%, black and the potential controls, you know, kind of all public housing neighborhoods in your sample, were 33%. Black. So this tells us, there's a few potential explanations for this, right. That either we let public housing become much more deteriorated when they were housing black people, or they were in black neighborhoods, or housing authorities wanted to redevelop public housing that was housing by people or in black neighborhoods for good or bad reasons. And I guess, you know, also, one quirk about Hope Six that I don't think have brought up is that a lot of the spending and redevelopment was pretty heavily concentrated in Chicago, and Atlanta, in particular, and those cities are two of the most black cities in the country. So having seen the data in more detail, like, what do you think are some of the most likely explanations for this concentration of Hope Six redevelopment in black neighborhoods?

Rebekah Levine Coley 37:55
Well, my understanding, my reading of the policies is that the Hope Six selection process did not assess or follow racial demographics of neighborhoods as part of its application process, or part of its sort of follow up process. So I, that suggests that it wasn't a specifically targeted selection, you know, purposely targeting black neighborhoods. And there were Hope Six sites and states and locations all across the country so you know, dozens of of different locations. So I don't think it's just sort of an argument of focusing on particular cities with high black populations. Having said all that, I mean, I think the Hope Six program did target very distressed public housing developments, and I think there is very clear evidence of racism and segregation in the public housing program writ large, and in lots of other housing policies that have led to racially segregated neighborhoods. So I think it's possible that it's just by targeting the most distressed public housing communities, those who are far more likely to have high black populations.

Shane Phillips 39:07
Makes sense. So moving on into the results now that we've kind of set up the study itself. When you looked at all of these tracks together, you found that poverty rates were about two to three percentage points lower in Hope Six neighborhoods compared to the controls in the years after redevelopment, and that's considerable, but by year 10, the difference was no longer statistically significant. And similarly, median household incomes went up in the Hope Six census tracts relative to the controls, but after the second year post completion, so after the projects were completed, that difference also became statistically insignificant. You did some additional analyses that we will talk about later, but for this full sample, are there any other details or nuances that you want to share on the poverty and income metrics specifically. And on that same track, how should we think about the fact that the impact on incomes and poverty in the Hope Six neighborhoods was no longer statistically significant or different by 10 years after project completion? If not earlier in some cases; like is that an indictment of the program and its impacts or should we interpret those results in a different way?

Rebekah Levine Coley 40:22
Well, I think clearly, some of the results did fade. So back to our analytic techniques. One of the other strengths of this technique is that you can look at the impacts of Hope Six over different lengths of time. So we looked at the impacts by the end of the Hope Six grant, and then through two years, after five years after and 10 years after. So as you noted, we found a three percentage point decrease in poverty, and Hope Six sites compared to their match control sites by the end of the Hope Six redevelopment. By 10 years later, it had declined to about 2.5% so not a huge shift, and that was just barely below sort of standard statistical significance.

Shane Phillips 41:05
But do you recall the base from which that reduction occurred - like was the poverty rate in these neighborhoods you know, 20%, 30%?

Rebekah Levine Coley 41:13
Well, the poverty rate in, I'd have to look at my tables.

Shane Phillips 41:17
I feel like maybe the threshold was around 40% even

Rebekah Levine Coley 41:21
Yeah, I think so. It was exactly it was 40% - it's an actual percentage point declined from right so I believe it went down to 36 and a half, 37%.

Shane Phillips 41:32
So three percentage point decline, roughly an eight-ish percent decline..

Rebekah Levine Coley 41:36
...of the original yes. So in Hope Six sites, it went down to about 36.5% poverty and in non-Hope Six sites, it went down, like point 0.4% so it went down to about 39.6%.

Shane Phillips 41:52
Okay, yeah, just wanted to kind of clarify that because, you know, a three percentage point decline in a neighborhood where the poverty rate is 15% is much more dramatic than say where it's 60%...

Rebekah Levine Coley 42:03
Yeah, exactly, so the median income increases did dissipate more quickly as I think we're going to talk about in a minute - that varied by the type of neighborhood by characteristics of neighborhoods, and by characteristics of the public housing developments themselves. And in some areas, those shifts were much more sustained than in other's.

Shane Phillips 42:24
And for that last question, does that mean that the program didn't work? You know, like, how do we interpret the fact that these impacts seem to fade away after either two years or five years or 10 years,

Rebekah Levine Coley 42:36
I think many would argue that the program showed limited effects, and decreasing poverty and increasing economic integration was only one of the goals of the program. And as we'll discuss more, the other goals do not from our data appear to have been mapped, right?

Shane Phillips 42:53
Yeah. Can you talk about the impact on the racial makeup and the availability of institutional resources, social services, grocery stores, other resources like those?

Rebekah Levine Coley 43:03
Yes. So those were two of the other sort of categories of outcomes we looked at, we looked at the percent of black residents, and the percent of non black people of color in the neighborhoods, and showed that those did not shift in response to Hope Six. So that again, addresses some of the gentrification and the risks of, of sort of whitening of neighborhoods and pushing out people of color, we did not show a substantial shift. And we looked at a range of services and amenities in the community. So we looked at basic things like grocery stores and restaurants, we looked at perhaps slightly less socially desirable amenities like bars and liquor stores, and we looked at social services. And then we looked at a whole host of other services and amenities that we thought spoke to Hope Six's goals of increasing economic and social opportunities for residents - so educational programs, cultural institutions, recreational institutions, medical services; across all of those kinds of services, and amenities, we saw no changes. So Hope Six did not appear to be successful at increasing the availability of supportive services or opportunity services in the communities in which it was played out.

Shane Phillips 44:21
So those were the results for the full sample of census tracts. You broke down the analysis and a bunch of other ways, though, comparing treatments and controls that started out above a threshold poverty rate to those below it, those were the black share of the population started above a specific threshold to those that started below it just to name a few examples. So the question you're getting at here, let's just use the poverty threshold as an example, is whether Hope Six had a greater impact on say high poverty neighborhoods where it was implemented than on lower poverty neighborhoods where it was implemented - there's some distinction between these places. We don't have to get into the results of every one of the comparisons you make, because you do a lot of them, but could you tell us what you were aiming for by taking that approach, and maybe just share some of the results of a few comparisons you found especially interesting or that seems surprising or important in some way.

Rebekah Levine Coley 45:14
Sure. So I think previous research has suggested that there was variability in the in the success of Hope Six in changing neighborhood contexts, and some are more successful than others in reaching the goals of the Hope Six program. So we sought to really systematically address that. What we found was that Hope Six redevelopments did appear to be notably more successful in the most distressed neighborhoods. So in the highest poverty neighborhoods, when we compare neighborhoods that started at above 40% poverty versus below, we found that pretty much all of the benefits for poverty rates and median incomes occurred in the high poverty neighborhoods. So through 10 years out after redevelopment, there was a five percentage point decrease in poverty in those really high poverty neighborhoods. A similar pattern for neighborhoods that were predominantly black - so more improvement in predominantly black neighborhoods than in neighborhoods with fewer black residents.

Michael Lens 46:17
I was really encouraged you know. There's some dissipation over time that we talked about, but I was encouraged by the findings that poverty declined in the highest poverty and most racially segregated neighborhoods, reducing concentrated poverty was just obviously an explicit goal of Hope Six, and so some people see that as gentrification. But you can't look at a decline in poverty in the neighborhood and say that it's a negative outcome in my view, obviously, context matters. But for me, the fact that it appears that this happened in the highest poverty, again most racially segregated neighborhoods, without reducing the share of African Americans living in those neighborhoods, you know, seems like a very good outcome, particularly, you know, the effects are not huge. So we're not seeing like mass displacement of people in poverty. So anyway, I mean, that's more editorial that you can choose or choose not to comment on but I really kind of see that as a very positive...

Rebekah Levine Coley 47:19
I agree, and I think that the effects actually grew over time in the highest-poverty neighborhoods.

Michael Lens 47:23
Okay okay, so I had mistaken that.

Rebekah Levine Coley 47:28
So the decrease in poverty rates within the highest poverty neighborhoods actually grew over time

Michael Lens 47:30
Okay, okay

Rebekah Levine Coley 47:32
About a whole percentage point by 10 years out, so it was four point, I mean, 5.2% decline by 10 years out for the highest poverty neighborhoods in the rate of poverty in that neighborhood. Of course, the highest poverty neighborhood has more, higher poverty neighborhoods have more room to decline, right? They have less of a floor effect but nonetheless, the most distressed neighborhoods are those in which we saw the greatest improvement without seeing declines in the black population, for example, or increases in affluence.

Shane Phillips 48:05
I also feel like it sort of shows the value of slicing this data and analyzing it in different ways. Because if you had only done the kind of full sample, you know, you can get yourself into a situation where you've got some neighborhoods that are maybe making dramatic improvements, some might be backsliding, and that could average out into like, wow, nothing happened. But then you're missing the fact that there were some places where you had real progress. And maybe there's something to learn from those. And like, you know, maybe we need to apply the lessons of those to the ones that didn't improve

Rebekah Levine Coley 48:38
RIght

Shane Phillips 48:38
And again, if you had not looked at these different thresholds and compared different kinds of Hope Six neighborhoods to each other; those differences could have been overlooked.

Rebekah Levine Coley 48:47
Correct, correct. And we also saw found evidence that there were characteristics of the policies themselves or of how they were implemented that made a difference. So for example, the benefits of Hope Six on these economic outcomes for neighborhoods were stronger for Hope Six redevelopment grants that were larger, that were more expensive. That wasn't just because it was a larger housing development, so it didn't vary very much depending on the number of public housing units to start with, but more expensive grants - where they spent more money. But also, I think (and) perhaps more importantly, is the effects were the strongest among public housing developments that switched over to mixed-income developments. So even though a goal of Hope Six was to create mixed-income communities, only a portion, only about a third of the Hope Six redevelopments switched into or transitioned into being mixed income developments. In that third, we saw notably larger effects so by 10 years out, there was a seven percentage point decline in poverty in those neighborhoods.

Shane Phillips 49:56
Thinking about the private capital aspect of this because it was another explicit goal of the program to not just have public funding, but also leverage that public funding to also secure private funding to kind of, you know, do more. And maybe there's other kinds of just philosophical goals about public private partnerships and that kind of thing. I think up to this point, we've kind of framed that mostly in a negative way, and you know, this interest of capitalists wanting to come in and kind of make money off of these gentrifying communities or whatever. But what's the positive story here? If there is one of that approach? You know, it certainly seems like more money spent on this housing is a good thing. You know, you look at the Low Income Housing Tax Credit and other housing programs in the US subsidy programs, they all leverage private funding as well. But like, is there anything to say about that aspect of this program?

Rebekah Levine Coley 50:50
Honestly, we did not directly address those issues in this work. So we didn't look specifically at the private funding or the public private partnerships so I don't think I can speak to it very deeply. I mean, I would reiterate that the goods is claims of decreasing black populations and increasing white populations did not play out in our data at the national level. So we did not find that racial demographic shift. Having said that, I mean, there certainly is evidence of both a loss of public housing units as an absolute number through the Hope Six program. So one estimate that came from the National Evaluation of Hope Six that was conducted with folks at Case Western Reserve University, found a loss of 43,000 public housing units from Hope Six, and that only about 20% of the new units were occupied by original residents. So there definitely is evidence of displacement. But I think our data show that there's not, individual residents may have been displaced but low-income residents overall were not displaced in a massive manner.

Shane Phillips 52:01
This question is outside of your research, I think, but you know, while I've got you here, I'm going to try to get a little bit more of a history lesson. So you talked about demolition through Hope Six, I'd love to just hear more about that aspect of the program. I know you were focused more on I think just the new construction, but there's also a rehabilitation element. How does demolition fit into this goal, though, if Hope Six is about building these mixed income communities. Why were they in some places just tearing this down, and kind of what happened after that?

Rebekah Levine Coley 52:32
That's a great question, I'm not sure I can fully answer. I think many communities have got demolition grants, some got demolition grants, and then got revitalization grants. So the demolition grant was sort of the first stage. So first, they tore down all of the housing, which displaced everyone, and then they tried to rebuild,

Shane Phillips 52:52
And probably some failed, and never took the next step, which, again, is just giving me flashbacks of urban renewal lectures

Rebekah Levine Coley 53:00
Some just didn't get the revitalization grants, some may have never intended to they just wanted to remove the severely distressed public housing and move on to something else.

Michael Lens 53:13
Yeah, so like the Robert Taylor homes in Chicago, right, that's like right along the expressway was possibly the most notorious dangerous public housing development in that country, at different points in the 80s and 90s. I think that just got demolished and everybody ran away. And I'm not talking about residents, I'm talking about...

Rebekah Levine Coley 53:36
Politicians

Michael Lens 53:37
... the politicians, and the housing authority that was actually taken over because of years and years of mismanagement and corruption

Rebekah Levine Coley 53:50
This is where I started my career, working in public housing, indeed,

Michael Lens 53:53
Yes, I'm telling you

Rebekah Levine Coley 53:54
So I think you're right. I think Chicago was sort of, you know, ground zero for some of these efforts, and some of the least successful efforts of revitalizing public housing communities. I mean, again, I think newer policies are trying to learn from what many would consider to be these somewhat egregious mistakes of displacement rather than redevelopment and rejuvenation, and of you know, really inadequately planning for and supporting the original residents and helping them to find new accommodations. And, you know, as I mentioned, splitting up communities really breaking social ties. So even in in communities where housing that quality of housing has declined and communities that may have substantial crime issues, drug issues, it's residents' homes, and many of them have lived there for decades, sometimes for generations. They have very strong social ties, they have a very strong sense of connection, of ownership over their neighborhood even if they they themselves indicate some of the significant problems in their neighborhood. So I think those issues really weren't given adequate, perhaps adequate attention, and concern from the get go of this policy.

Shane Phillips 55:12
So Hope six was discontinued in 2010 as we said. Looking ahead especially with your research in mind, what lessons do you think we should take from Hope Six and apply to any future efforts to either build or rebuild public housing. You talk about how the Choice Neighborhoods Initiative (CNI) and the Rental Assistance Demonstration (RAD) programs sort of took the place of Hope Six after it was discontinued. I was hoping you could talk a bit about those as well, and the role that Hope Six played in their development, and again, kind of lessons learned, and maybe just as a kind of, to signal this, I don't know where Jordan Downs fits into this - the redevelopment happening here in Los Angeles, or Yesler Terrace up in Seattle where I'm from. I feel like these are kind of the next generation of these redevelopments, and maybe are doing things a little better, if not perfectly, and maybe we can talk a little bit about those to close this out.

Rebekah Levine Coley 56:10
Sure, so I mean, I think one big lesson that was learned was this issue of displacement, and massive displacement, and also the relatedly - the right of return. The right for people if they are displaced to come back to their neighborhood. So the new policies, CNI, and RAD, both have some level of right of return or lower displacement. I believe CNI mandates, a one-to-one replacement of all affordable housing units that are torn down, but they don't require them on site. So in CNI, there's some emerging evidence that they require that one-to-one replacement, but many people are being moved off site. So they're developing new housing, but not necessarily in that same neighborhood so they're still breaking up neighborhoods and social connections.

Shane Phillips 56:59
Is there any kind of like proximity requirement where it's like, or can you be you have to be in the same city but it could be on the entirely opposite side, that kind of thing?

Rebekah Levine Coley 57:08
I don't know. I mean, I think most of them are in the same city but I'm not sure, how that, you know, what the exact policy parameters are, honestly. You know, most of these programs and policies, it's a particular public housing authority that receives that grant, which are typically rather geographically constrained within a particular city. So I would guess that, and the examples that I've seen of just specific sites, the new locations, (and) the new units that they're building through the Choice Initiative tend to be in the same city and in the same general area, but not in the same specific neighborhood.

Shane Phillips 57:44
Something I've seen with Jordan Downs, which I think is promising, Yesler might also be doing this, is the sort of phased redevelopment as a strategy.

Rebekah Levine Coley 57:53
Exactly!

Michael Lens 57:53
Yeah yeah

Shane Phillips 57:54
Could you talk about that?

Rebekah Levine Coley 57:55
Yes, yeah. So well, you plugged your book, so I'll plug my new study. So actually, Sam Texera, and I, who was my co author on this paper, we have a big new grant. And we are studying the redevelopment of the largest and I think probably one of the oldest public housing developments in New England that is going to be redeveloped into a mixed income community. And I think they are trying to incorporate a lot of lessons learned from prior research, and one is to do as much as they can to limit external displacement. And they do that by using a phased redevelopment plan. So essentially, what that means is they start building in one small part of the public housing community, maybe have to tear down a couple buildings and displace those people and then build a new building and then move people over. So it's only people in the very first phase that get externally displaced, and everyone else has to live with construction next door, but do not get externally displaced from their neighborhood so they got to move directly into a brand new apartment within their community.

Michael Lens 59:03
Right, and Shane brought up Jordan Downs in South Los Angeles or the Watts area, and that's been a similar process. It has not been something I would call smooth in part because of some long standing environmental issues in the the surrounding site really yeah so that's complicated

Rebekah Levine Coley 59:25
I think that's one lesson that really has been incorporated into newer models, lessening displacement. I think another is, you know, continued efforts to incorporate an integrated services approach and really consider the needs of the residents and the desires, the wants of the residents. What they would like to see in their neighborhood, what sorts of services they would find to be most helpful for their own particular needs, and coordinating with local social service agencies, public agencies like Public Library Raise. So I think there's really a renewed or an increased attention to those sorts of efforts and to also getting resonant feedback throughout the process, trying to incorporate residents voices in the redevelopment plans. How successfully that will all play out, you can invite me back in, you know, five or eight years, and I will tell you

Michael Lens 1:00:22
Exactly

Shane Phillips 1:00:23
That, we'll have another episode on the Choice Neighborhoods Initiative in decade 12, or 15, of the Housing Voice podcast.

Rebekah Levine Coley 1:00:30
Right right. I mean, I think that's one of the challenges,

Shane Phillips 1:00:39
Did I say decade? I meant year

Rebekah Levine Coley 1:00:40
That's one of the major challenges of this type of work is that one, just the policies themselves. So in any large public housing development, it can take 10 years to go through this process of redevelopment, and, you know, from a research perspective, you want to understand the impacts on residents and on the community, not only immediately but down the line because some things can can take a while to emerge and grow from these efforts. So it really is, it's a long term endeavor.

Shane Phillips 1:01:08
This has come up on previous episodes, but it's something I just I'm very interested in this idea of the right to return and, you know, allowing people to come back and how great that sounds in practice. And I can just imagine people thinking, you know, we're going to build this housing, yeah, people are going to be displaced for a year or two, you know, maybe we're going to provide them some resources to make sure they can afford rent in the neighborhood in the meantime. That's not always done but even if it is, the studies I'm aware of that have followed this, they find that most people just don't come back even you know, they're offered housing at a much lower than market rate price. They have this kind of priority access, and yet for a million different reasons, that would be very hard to anticipate ahead of time, you can really only find out about after you've tried it and it has failed. They're just finding that that kind of approach doesn't work. And so I do think it's promising even though it's taken a few decades of practice and learning to figure this out that you really have to address displacement at the front end, whenever possible, and trying to solve it after it's happened or mitigate the harms. It just doesn't work very well.

Rebekah Levine Coley 1:02:19
Not surprisingly, right. So if people have to move and find a new place to live, move into a new home, meet their neighbors, you know, learn where the local stores are, and resources are, they might not want to have to do it again two years later and uproot their family and uproot all their daily routines yet again.

Shane Phillips 1:02:38
Yeah. All right. Rebekah Levine Coley, thank you so much for coming on the housing voice podcast.

Rebekah Levine Coley 1:02:43
Thank you, it's been a pleasure.

Shane Phillips 1:02:48
You can read more about Rebekah's research on our website. lewis.ucla.edu. Show Notes and a transcript of the interview are there too. The UCLA Lewis Center is on Facebook and Twitter. I'm on Twitter at Shane D. Phillips and Mike is there at MC_lens. Thanks for listening. We'll see you next time.

About the Guest Speaker(s)

Rebekah Levine Coley

Rebekah Levine Coley, Ph.D. is Professor of Counseling, Developmental, and Educational Psychology and Director of the Institute of Early Childhood Policy at Boston College.
Learn More

Suggested Episodes