ATTRACTING MIDDLE-INCOME FAMILIES IN THE HOPE VI REVITALIZATION PROGRAM

DAVID P. VARADY University of JEFFREY A. RAFFEL University of Delaware STEPHANIE SWEENEY University of Cincinnati LATINA DENSON University of Delaware

ABSTRACT: While the growing literature on HOPE VI emphasizes the presumed benefits of income mixing these benefits are most likely to occur if middle-income families with children are drawn to these sites.But is this feasible? Our comparative case study analysis of four HOPE VI sites in Cincinnati, Louisville, Baltimore, and Washington, DC, suggests that it will be difficult to achieve the mixing of lower- and middle-income families with children.None of the four developments explicitly sought middle-income families with children as part of their marketing. Louisville’s HOPE VI site was the only one involving close collaboration between the school district, the housing authority, and city government from the beginning of the HOPE VI process.Moreover, the Louisville site was the only one successful in attracting middle-income (and not simply subsidized moderate-income) families with children.Strategies for making inner-city HOPE VI sites more attractive for middle-income families are discussed.

In 1992, HOPE VI (originally known as the Urban Revitalization Demonstration Program) was enacted to transform the physical and social shape of distressed public housing. Under the current programthis transformationincludes: changing the physical shape of public housing by replacing the worst public housing developments with less dense developments, reducing concentrations of poverty by encouraging a

*Direct Correspondence to: David P.Varady, School of Planning, University of Cincinnati, Cincinnati, OH 45221. E-mail: [email protected] or Jeffrey A. Raffel, School of Urban Affairs and Public Policy, University of Delaware, Newark, DE 19716.E-mail: [email protected]

JOURNAL OF URBAN AFFAIRS, Volume 27, Number 2, pages 149–164. Copyright # 2005 Urban Affairs Association All rights of reproduction in any form reserved. ISSN: 0735-2166. 150 | JOURNAL OF URBAN AFFAIRS | Vol. 27/No. 2/2005 greater income mix among public housing residents and promoting mixed-income communities by providing market-rate housing, creating positive incentives for resi- dents to achieve self-sufficiency, providing comprehensive services that assist residents in achieving self-sufficiency, and creating broad-based partnerships to leverage additional resources. HOPE VI site initiatives seek to create mixed-income communities by encouraging middle-income households into neighborhoods that previously were exclusively occupied by the poor. Exactly why and how this should be accomplished, and what the effects would be, has not always been clearly stated (see Brophy & Smith, 1997; Turner, Kingsley, Popkins, & Abravanel, 2004). Attracting a range of middle-income families including ones with children could help HOPE VI projects promote increased self-sufficiency in the next generation because income mixing in the classroom could improve the educational performance of low-income chil- dren. The Coleman et al. (1966) seminal study found that the socio-economic composition of the school was related to student achievement, that is, that lower income students had higher achievement in schools with more middle-class students. Several articles in Burtless’ (1996) edited volume also assert that the presence of higher-income and higher-class students at a school site improves the performance of lower-income students. Empirical results from the Gautreaux housing mobility program (Rubinowitz & Rosenbaum, 2000) emphasize the benefits of mixed-income classrooms for low-income children. A key finding of research on Gautreaux is that the children of suburban movers dropped out of school less frequently and were more likely to go on to college than those in households that remained in . Both HOPE VI and suburban dispersal (of which Gautreaux is an example) are parts of the federal government’s strategy to deconcentrate poverty. Findings from one should logically be applicable to the other. This means that if it were possible to create mixed- income schools at or near HOPE VI sites, the low-income children presumably would experience some of the same benefits that low-income children experienced in the Gautreaux program. The preceding suggests the need for a HOPE VI income mixing strategy that emphasizes the attraction of middle-income families with children. Because middle- and lower-income parents would share common interests such as the quality of play facilities, they would be more likely to interact socially and such interaction might increase the likelihood that middle-income people would serve as positive role models. However, safe, high-quality schools would be a prerequisite for attracting middle-income families with children (see Varady & Raffel, 1995 for a discussion of the residential choice literature and an empirical analysis of the impact of public school quality on city-suburban choices of middle-income families with children). Table 1, a logic or outcome sequence model, highlights the presumed causal relation- ships between this form of income mixing (i.e., involving a substantial number of middle- income families with children) and HOPE VI public housing revitalization. The presumed efficacy of this form of income mixing is based on seven assumptions, the first three being that (1) housing and schooling officials will market HOPE VI sites to middle-income families with children, (2) housing, schooling, and planning officials (and others) will work together to develop schools that address the needs of both middle- and lower-income families, and (3) middle-income families with children will be willing to move into HOPE VI neighborhoods. Each step in this model is a potential breaking point. For example, school and housing officials may be unwilling to devote time and resources to attracting middle-income families with children. TABLE 1

Logic Model. Anticipated Links Between HOPE VI-related Programs and Outcomes

Activities Outputs Short-term Outcomes Intermediate Outcomes Long-term Outcomes

Housing programs: (1) Provide financing for (1) The HOPE VI commu- (1) Middle-income families Middle-income house- (1) Tax base of city is homes for low and moder- nity contains a mix of both with children are attracted holders (with and without strengthened (2) Low- ate/middle-income buyers market- rate and below to HOPE VI market-rate children) serve as role income adults achieve (2) Market homes to market-rate buildings housing models for their lower- social mobility (reduced middle-income buyers (2) Individual buildings (2) The HOPE VI develop- income neighbors. unemployment, higher contain a mix of market- ment is able to retain incomes) rate and below-market rate middle-income families tenants with children as the chil- (3) HOPE VI officials dren get older actively recruit middle- (3) Middle-income house- income families with holders interact with their children low-income neighbors Schooling programs: Work across housing, (1) A high quality school (1) Middle-income families Schools serving mixed- Low-income children

planning and schooling addressing the needs of with children are attracted income children help raise experience social mobility | agencies to develop low- and middle-income to schools in the area low-income students’ (go to college, get good Families Middle-Income Attracting schools that will attract children is introduced into (2) Schools are able to achievement paying jobs) middle-income families and the HOPE VI neighborhood maintain a stable mixture meet the needs of low- (2) The school system of low- and middle-income income children actively recruits middle- children in all classes income children (3) Low- and middle- (3) School officials adopt a income children interact in policy of mixing low- and the classroom middle-income children in all classes | 151 152 | JOURNAL OF URBAN AFFAIRS | Vol. 27/No. 2/2005 This article examines the validity of the first three assumptions. Based on a comparative case study analysis of four HOPE VI developments we seek to answer three questions. First, what were the goals of these sites (i.e., to what extent was social mixing stressed?) and what approaches were used to market the four sites in order to attract and hold middle-income families, including families with children? Second, what approaches were used to achieve collaboration among the housing authority, the school district, city government, banks, and other entities? Third, to what extent have these different sites been successful in attracting middle-income families? To answer these questions we conducted a comparative case study analysis of the following HOPE VI sites: City West (immediately adjacent to Cincinnati’s CBD), Park DuValle (a suburban-type site in Louisville’s West End), The Terraces (an inner-city site in West Baltimore), and The Townhomes of Capitol Hill (Washington, DC, a gentrifying neighborhood). We selected the four sites based on several considerations. We wanted HOPE VI projects that had moved beyond planning to bricks and mortar. Thus, we selected projects that received early as opposed to later federal funding. We sought at least one project where there was preliminary indication that housing-school cooperation had taken place. We also sought to include at least one project where the housing market was tight and the attractiveness of the project for middle-income homebuyers would be relatively high. We selected this convenience sample within the constraint that two sites needed to be within a short driving distance of each of the two research teams (University of Cincinnati and University of Delaware). Given the above criteria, we selected the cities and project sites simultaneously. In each city we only had the resources to study one HOPE VI site. Background on each city is presented in Table 2. The reader who is interested in more details on the case studies should review the earlier, larger report on which this article is based (Raffel, Denson, Varady, & Sweeney, 2003). Data gathering techniques included semi-structured in-person and telephone interviews. Other techniques included document analysis of HOPE VI reports and newspaper articles and direct observation. Interviews were conducted with individuals who were familiar with the HOPE VI programin their city, particularly with respect to linkages with the schools. We conducted 28 interviews with housing authority staff, residents, school officials, city planners, developers, and other community members. Transcriptions of the interviews were coded and analyzed using QSR NVivo Software for Qualitative Data Analysis and QSR NUD*IST 4 Software. A key limitation of our article should be emphasized. We did not choose our HOPE VI sites so as to include ones that emphasized the attraction of middle-income families with children or ones that fromthe beginning emphasized close collaboration between local housing authorities and local school districts. Consequently, our answers to the above three questions should be considered preliminary in nature. We hope that this article will stimulate future research that will include sites emphasizing the attraction of middle- income families with children and that emphasize school-housing collaboration.

PREVIOUS WORK ON INNER-CITY INCOME MIXING

What Does Income Mixing in the HOPE VI Program Mean? Despite the growing literature on the HOPE VI program, the term ‘‘income mixing’’ is rarely if ever defined. (A search of HUD’s web site for an operational definition turned up empty.) Khadduri and Martin’s (1997) article is unique in discussing the type of income mix that is needed at HOPE VI sites. TABLE 2

Background Information on Cities with HOPE VI Sites Studied

Location

Cincinnati Louisville Baltimore Washington, DC City West Park DuValle The Townes at the Terraces Townhomes on Capitol Hill

Population: 1990 ¼ 364,553 1990 ¼ 269,838 1990 ¼ 736,014 1990 ¼ 606,900 2000 ¼ 331,285 2000 ¼ 693,604* 2000 ¼ 651,154 2000 ¼ 572,059 *result of merger between city and county government. Type of School System: Central city school system Countywide school system. Central city school system Central city school system with separate suburban with separate suburban with separate suburban school districts within school districts within school districts in Maryland Hamilton County. Baltimore County. and Virgina. Approaches to Racial Magnet schools along with Countywide busing with Magnet schools along with Magnet schools along with Integration: neighborhood schools. racial integration maintained neighborhood schools. neighborhood schools. in all schools. Numbers of students and 1985 ¼ 51,528; 1985 ¼ 93,534; 1998 ¼ 106,538; 1980 ¼ 100,049; change over time: 2001 ¼ 44,888 2001 ¼ 95,459 2001 ¼ 95,465 1990 ¼ 80,694; | 2000 ¼ 79,037 Families Middle-Income Attracting Indications of school Rated by Ohio Dept. of Brookings Institution (2001) Closing 12 schools in 2001, Reforms attempted, low test problems: Education as Academic calls for improvements. low test scores. scores. Emergency status. Name of housing authority Cincinnati Metropolitan Housing Authority of Housing Authority of District of Columbia Housing and # of units: Housing Authority (CMHA) Louisville (HAL) 4,600 units Baltimore City (HABC) Authority (DCHA) 11,000 5,420 units 17,656 units units Public Housing Assessment 92 mid-90s 79 In 1995 DCHA was rated System (PHAS) score, 100 is second worst in the nation highest possible score. and went into receivership. In July, 2003 DCHA received a score of 88. | 153 154 | JOURNAL OF URBAN AFFAIRS | Vol. 27/No. 2/2005

We start with a basic premise about mixed-income housing. Such housing must, at a minimum, give poor children an opportunity to live close to working families with incomes above the poverty level. Housing that serves primarily the elderly or other households without children, or that excludes the poor fromits range of incomes,does not fall within our definition of mixed-income housing (p. 34).

Why is it that HUD and local housing authorities have not used this stricter standard to assess the program? Is it because there is a lack of consensus around the goal of attracting middle-class families with children? Or is it because most housing officials consider the goal to be unattainable? Our four city comparative case study of HOPE VI sites should clarify why local housing authorities are unwilling to target middle-income families with children in their outreach and marketing efforts.

Linking Schools, Housing and Planning Agencies Although local governments and school districts have worked independently of one another in the past, there are two encouraging signs of improvement. Adequate public facilities ordinances for schools (APFO’s) are state-enabled, locally adopted requirements that halt new development approvals once a school’s enrollment exceeds capacity by 120%. APFO’s could help local governments control sprawl while helping school districts prevent overcrowding (Donnelly, 2003). Co-located community schools could enable school officials to address the problemof overcrowded schools while local planners address suburban sprawl (Romeo, 2004). Typically, a school is located with a variety of other facilities on one site (e.g., a public library, a community college branch, a museum). While increased attention is being given to ways to more closely link public schools with community development, much of this work focuses on ways to use improved coordina- tion to improve the life prospects of low-income families (Children’s Aid Society, 2001). There is also some promising news to report of housing developers, planners, and school officials working to attract moderate-income families with children to cities. In 1997, a developer, McCormack, Baron and Associates, partnered with St. Louis Public Schools to revitalize Jefferson Elementary School concurrent with developing the surrounding mixed- income housing (Chung, 2002; Lowell, 1999; Matthews, 2004; National Housing and Rehabilitation Association, n.d.). The research literature on collaborative efforts aimed at attracting and retaining middle- income families with children at inner-city locations is quite limited. One widely cited example of effective collaboration between the housing authority and the public schools is Centennial Place Elementary School. Located on the site of the HOPE VI development in it is a child-centered, community-based magnet school near Georgia Tech (National Housing and Rehabilitation Association, n.d.; B. Williams, personal communication, January 7, 2003). Meanwhile, Chattanooga has taken an innovative approach by opening two new magnet schools in the downtown area in the fall of 2002, specifically designed to attract families with children through an admissions preference for children of those who live or work downtown (Kreyling, 2002). However, it is probably too soon to say whether the above two projects will be successful in attracting middle-income families with children to live in the area. Willie’s (1997) case study of Dearborn Park, an urban redevelopment area south of Chicago’s Loop, highlights the political and bureaucratic difficulties of linking housing and schooling so as to attract middle-income families with children to closer-in parts of the central city including HOPE VI sites. Middle-income parents wanted a high quality school on the site where children from middle-income homes would be the dominant | Attracting Middle-Income Families | 155 group. The school district responded by building a school just for Dearborn Park. However, residents of a nearby public housing development sought to allow their children to attend this school. Eventually as a result of a series of compromises the proportion of the low-income children rose to the point that middle-income parents took their children out of the new school and either placed themin private/parochial schools or movedto the suburbs. Willie’s study suggests that the prospects are poor for successful collaboration and in turn for achieving meaningful social mixing. One of the main purposes of this article is to determine whether that pessimistic assessment applies to HOPE VI sites.

Attracting Middle-Income Families to HOPE VI Sites There has been relatively little scholarly research on the ability (or inability) of HOPE VI sites to attract middle-income families with children. The literature that is available indicates that HOPE VI sites and subsidized private housing market developments have been able to attract (1) middle-income singles and childless couples (especially in tight housing markets), and (2) moderate-income families with children (i.e., those with incomes between 60 and 80% of area median income) but that they have not been able to attract middle-income families with children (‘‘Boston war zone becomes public housing dream,’’ 1991; Brophy & Smith, 1997; Rosenbaum, Stroh, & Flynn, 1998; Schwartz & Tajbakhsh, 2001).

MARKETING THE HOPE VI SITES: TO WHAT EXTENT WERE MIDDLE-INCOME FAMILIES SOUGHT? As mentioned above, mixed-income communities have been an integral part of HOPE VI, and HUD mandates that each applicant indicate how such communities will be achieved. However, as we will see, where location and economics were viewed as sufficient to attract market-rate homebuyers, officials ignored the issue of the schools and middle- class families with children. For each of the developments in our case studies, income mixing was an explicit goal. However, none of the proposals listed attracting and holding middle-income families with children, or improving links between housing and schools, as a goal, at least in the beginning. Cincinnati officials reported that a high priority was developing home-ownership oppor- tunities for low-income former residents of public housing. Jackie Davis, Cincinnati Metropolitan Housing Authority’s (CMHA) HOPE VI Project Coordinator stated:

I think when they designed the homes they certainly talked about them being for families with children at home. [However], even empty nesters are looking at them as a place where their grandchildren can come for the weekend, or come for spring break. We always thought of themas being there as an opportunity for familieswho wanted to live downtown, but the market is such that there are singles looking for affordable homes. We were not limiting it to families, but we were certainly thinking that [CMHA] families could begin to own homes (personal communication, April 12, 2002).

CMHA told The Community Builders (TCB), the developer chosen for City West, that it was much more important to create self-sufficient families that are off government subsidies than to just sell some homes to suburbanites who want to live in the city.

CMHA was confident that market-rate homes would sell because other market-rate developments in or near downtown Cincinnati were thriving. However, the other 156 | JOURNAL OF URBAN AFFAIRS | Vol. 27/No. 2/2005

developments were attracting empty nesters, singles and couples with no children. While the HOPE VI grant applications (see Abt Associates, 1998, 1999) clearly proposed economic mixing there was no discussion of social mixing including middle-income families with children (C. Henderson, personal communication, April 12, 2002).

Louisville informants disavowed any explicit interest in attracting market-rate families with children. Although children figure prominently in the marketing materials and the Montessori school located on the site is mentioned, there were no specific efforts made to attract middle-income families with children. This probably reflects the existence of Jefferson County’s combined city-suburban school system. Most of the people thinking about moving into Park DuValle likely were aware that their children would attend public schools outside the neighborhood. Thus, it would have made no sense to highlight the benefits of Park DuValle’s schools. On the other hand, the benefits of the city-county school systemto Park DuValle’s stability should be emphasized here. Because of the existence of the countywide school systemparents were not deterred fromthe public housing revitalization site because of a perception that their children would need to attend substandard schools.

Because of the magnet schools, I do not think they expect their kids to go to school in the neighborhood. They expect them to be more mobile, and they expect them to travel some distance to go to school, because of what they are interested in (C. Cash, personal communication, December 18, 2001).

In Baltimore, in order to encourage families with mixed incomes, the Lexington Terrace development team used marketing strategies that emphasized its proximity to city attrac- tions (e.g., the Lexington Market), school improvements, and the community center, and the medical center of the University of Maryland. According to Bill Struever, a member of the Lexington Terraces development team and a member of the Baltimore City Public School Board, all of these factors plus financial incentives for Baltimore employees have encouraged families to move into this former public housing community. Residents received grants under both the ‘‘Live Near Your Work’’ (LNYW) programand the program. LNYW is a homebuyer incentive program managed by the Maryland Department of Housing and Community Development. However, the slow progress of developing the school and community center posed a concern for original and new residents Informants in Washington, DC, reported that at the Townhomes the location and design of the development garnered enough attention to attract families. Because the developers were certain of success in attracting market-rate families (whether they had children or not) no special effort was made to attract market-rate families with children (D. Perry, personal communication, September 12, 2002). Furthermore the developers may have deemphasized attracting middle-class families with children because of a desire to get the HOPE VI development built as quickly as possible and to have it begin to promote the upgrading of the surrounding area.

CITY-HOUSING AUTHORITY-PUBLIC SCHOOLS COLLABORATION: WHAT APPROACHES WERE USED AND HOW EFFECTIVE WERE THEY? Cincinnati Public Schools (CPS) was not involved at all in the beginning of the HOPE VI planning in Cincinnati. CPS became more involved during the implementation phase | Attracting Middle-Income Families | 157 of City West, to a large degree because CPS was conducting its own facilities master planning process. In 1999 when the state of Ohio agreed to fund 23% of the cost to upgrade deteriorating public schools throughout the state, CPS initiated its facilities master planning process, working closely with the Ohio School Facilities Commission. The 2002 plan for CPS recommended that over the next 10 years, 34 buildings be built, 32 be renovated, and several be decommissioned (Mrozowski, 2001). CPS called for the meetings with CMHA to discuss the impact the HOPE VI redevelopment was likely to have on demographics in the West End neighborhood. Collaboration between CMHA and CPS staff increased remarkably during the HOPE VI process.

The need is there to communicate and work together; it just turns out it is going to be better for the plan to do it. I think over the last year and a half, maybe even two or three years, the communication [with CMHA] has been great (K. Cashell, personal commu- nication, May 10, 2002).

In Cincinnati, one of the schools to be constructed adjacent to City West will be a prestigious magnet school, the Cincinnati Academy of Math and Sciences (CAMAS). CMHA asked the school board to consider a preference for City West children to be able to attend the CAMAS citywide magnet located on the City West site. CPS has not yet agreed to give residents preference for a magnet school but such a preference might help attract and hold middle-income families with children. It is important to note that in Louisville, Jefferson County Public Schools (JCPS) was highly involved fromthe beginning of the planning of Park DuValle because it was participating in the Empowerment Zone process. There was a high level of collaboration between housing and school administrators in the planning and implementation of Park DuValle because Louisville’s community leaders expected that the Housing Authority of Louisville (HAL) and JCPS would work closely together and involve other community organizations such as the Park DuValle Neighborhood Advisory Council (PDNAC). Mayor Abramson saw an opportunity in the empowerment zone process and brought together these organizations and many more in a community planning process that gave birth to the notion of improving the neighborhood of Park DuValle. As a result of the involvement of JCPS in the planning, it decided to spend $14 million to renovate Carter Elementary in response to the HOPE VI development that was going on around it. In Louisville a culture already existed where the city’s mayor and county executives including the school superintendent worked together. The city of Louisville, the Housing Authority of Louisville, and Jefferson County Public School planners all recognized the value of the county’s educational facilities as a focus for redevelopment. Perhaps this was easier to do in Louisville than elsewhere because there is only one Jefferson County Public School System (J. Charmoli, personal communication, December 18, 2001). In the other three metropolitan areas, there are a large number of school systems. Interestingly, The Community Builders (TCB) wanted to see Carter Elementary con- verted to a neighborhood-based school. Fromits experience in other cities, TCB staff believed that a neighborhood-based school would be valuable to attracting potential renters and buyers. However, JCPS refused to consider making Carter a neighborhood-based elementary school because it believed it would disrupt the racial balance in its schools.

The hope with TCB was that there would be a community school within this develop- ment. And, it would be a value point; people who moved into the area would know that their children would go to a certain school, and that they would be involved in that 158 | JOURNAL OF URBAN AFFAIRS | Vol. 27/No. 2/2005

school, and there would be some ownership in that school. So, the reason for getting Jefferson County Schools involved very early in the discussions was that community schools do not exist in Jefferson County. Unless we talk about parochial schools, there are no community schools in the public school system. And, because of court-ordered busing, there will not be. But the hope was that we could convince the Board of Education that creating a community school in Park DuValle would be the thing to do, and would be key to the success of this development. We are not there yet (D. Peterson, personal communication, December 18, 2001).

In Baltimore, the relationship between housing and school officials has developed over time. The Terraces site was the home to the old Lexington Terrace Middle School. Before it was demolished, the school served students from Lexington Terrace, Poe Homes, and surrounding communities. At Lexington Terraces, the goal was to replace the old elemen- tary school with a new K-8 Math, Science, and Technology School with an adjacent community center. To achieve this goal required collaboration between the public housing authority, the Baltimore City Public Schools (BCPS), the Maryland State Department of Education (MSDE), the development team, and residents. Although collaboration with the public schools system existed, the effort was minimal during the early development of the plan. In the original HOPE VI application, Lexington Terrace School was simply going to close temporarily and reopen two years later when the project was rebuilt. The creation of a community planning committee set up by the Poppleton Empowerment Zone helped to increase the level of collaboration. Collaboration between housing and educational policy increased after a school board teamvisited the Centennial Place HOPE VI developmentin Atlanta, which contains a unique community-based magnet school. The community planning group proposed a K-8 math, science, and technology curriculum-focused school that would both be zoned for the returning families, and to the extent that there was additional capacity, would be a citywide magnet. The new Lexington Terrace School would essentially be a model K-8 for the Baltimore City Public School system (B. Struever, personal communication, March 28, 2002). Thus, the concept of the new school was developed after the HOPE VI grant was received. While joint planning at the Townhomes on Capitol Hill was virtually nonexistent, the prospects for Washington, DC’s housing and school officials working together have improved. At other sites, (e.g., the Frederick Douglass and Stanton Dwellings HOPE VI site, now known as Henson Ridge) the DCHA is actively engaging residents and schools in the planning process to ensure quality schools in their neighborhood (B. Tetreaux, personal communication, March 14, 2002).

MEASURING EFFECTIVENESS: TO WHAT EXTENT HAVE THE HOPE VI SITES BEEN ABLE TO ATTRACT MIDDLE-INCOME FAMILIES? While middle-income couples and singles have been attracted to market-rate HOPE VI units at all four sites, only Park DuValle has succeeded in attracting middle-income families with children. At two of the sites (Louisville and Baltimore) some conflict across tenure status and income groups has occurred. As this is being written, City West is in the early stages of sales and leasing. All 35 homeownership units have been sold with the market-rate units leading the affordable homeownership units in the pace of sales. Conversely, the public housing rental units (56) have all been leased, while 39 of 58 market-rate apartments have been rented. While | Attracting Middle-Income Families | 159 families with children are prevalent among the affordable homeownership units and the public housing rental units, no market-rate families with school-aged children had rented a City West unit. At present there are no plans at City West to develop programs promoting social interac- tion between market-rate singles and childless couples on the one hand, and below market- rate families with children on the other. TCB’s manager of resident services sees her role as strictly working with public housing families, and not working with people in the market-rate units. Whether a more active strategy would promote social mixing in a meaningful sense (between owners and renters, between subsidized and non-subsidized families) is uncertain. One reason why middle-class families with children are absent from City West is the perceived lack of safety. Crime continues to be a problem in Cincinnati’s West End (with 731 serious crimes reported in 2002, see Wilkinson, 2003). Furthermore, middle-income parents have been deterred by the generally poor reputation of the public school system, instead focusing their housing search on the school districts with the best test scores and with graduates going on to Ivy League and similar high-status colleges and universities. Donald Troendle, Director of CMHA, (personal communication, April 12, 2002) argues that some of these public school oriented families might consider living near the down- town (including at the HOPE VI site) if they knew more about the schooling options available in the downtown area. He has developed a list of 28 educational facilities located within a two-mile radius of City West. While 14 of them are public schools, others are parochial schools, charter schools, and special schools. Whether greater awareness of these options would increase interest in living in the West End is an open question. In Louisville, the HOPE VI development is attracting middle-income families including those with children. Phase two, The Villages, is now completed, leased, and sold. Fifty-two of 147 units have been sold to persons at or above 100% of area median income (AMGI). While it has proved difficult to obtain the exact percentage of these middle-income families where there are school-aged children, informants reported that many of the market-rate buyers have children in Jefferson County Public Schools. Originally, the Housing Authority of Louisville was concerned about whether there was a market for the development. However, early on it was clear that market-rate families were interested. Sales of homes were rapid for middle-class families, whether they had children or not.

In fact, more than interest, we have a ton of families living there. We had things other places just did not have. People who live there include a [HAL] board member, a city official, bankers, two bank vice-presidents, and a physician, next to families who may only make forty-five to fifty thousand dollars a year (T. Barry, personal communica- tion, January 14, 2002).

Park DuValle currently has a long waiting list for both apartments and single-family homes. At first, it seemed surprising that people would buy $250,000 homes in a formerly high- crime central city area. In other cities, it might be a chancy thing because of the schools. However, in Jefferson County, school choice is eliminated from the home buying equa- tion; where people live has little to do with where their children will go to school.

I think having one school system does make a difference. Because, in communities that have multiple school systems, you have families that make a choice about where they live based on the school systemthat children are going to go to (J. Charmoli,personal communication, December 18, 2001). 160 | JOURNAL OF URBAN AFFAIRS | Vol. 27/No. 2/2005 Thus far, because the project has appealed primarily to people who already live in Louisville, the schools have not become an issue for buyers (C. Romero, personal com- munication, January 14, 2002). People who moved in Park DuValle were already familiar with the West End of Louisville and what the school systemhad to offer (D. C. Howard, Jr., personal communication, January 14, 2002.) In Louisville, people just do not expect their children to go to school near where they live.

Because of the magnet schools, I do not think they expect their kids to go to school in the neighborhood. They expect them to be more mobile, and they expect them to travel some distance to go to school, because of what they are interested in (C. Cash, personal communication, December 18, 2001).

Whereas Park DuValle has succeeded in the promotion of income mixing, it has failed so far in promoting racial integration. Ninety-nine percent of buyers and renters are black. This public housing revitalization project is a mirror of the larger West Louisville com- munity (predominantly black) where Park DuValle is located. Racial integration is a long- term goal, if it is a goal at all (T. Barry, personal communication, January 14, 2002). Efforts to promote a social mix between owners and renters have created conflicts that may make it difficult to achieve the presumed benefits of income and tenure mixing. During the early phases of the development, one of the neighborhood groups instrumental in the planning, the Park DuValle Neighborhood Advisory Council (PDNAC), stopped meeting because homeowners were bringing to the meetings their complaints about the construction of their homes; broader neighborhood issues of interest to renters were pushed to the side. Currently, separate associations exist for the homeowners and the renters. However, there is an effort going on to revitalize PDNAC (G. Eugene, personal communication, March 19, 2002). Meanwhile, TCB is actively coordinating social oppor- tunities for residents of different income levels to mingle, including a Christmas tour of the homes (featuring rental as well as ownership units) and a summer jazz picnic (Petersen, 2001). Statistics reveal that crime has been almost totally eliminated at Park DuValle, which previously was a high-crime neighborhood. Safety is an important consideration for middle-income families with children. Clearly many middle-income families who chose to make their homes in Park DuValle had the economic means to purchase a home in other areas of Louisville/Jefferson County, yet they chose the HOPE VI redevelopment because they felt secure about the safety of their children. All of the market-rate townhomes at the Terraces in Baltimore were sold after the completion of the development. However, given the fact that these homes were heavily subsidized (costing $130,000 to build but selling for between $65,000 and $85,000), almost all of the buyers were moderate-income families (with incomes at 60 to 80% of the area median income). According to the developer, all but 16 of the market-rate townhomes were sold to households with incomes under $40,000. We estimate that 80 to 85% of the market-rate buyers were moderate-income and 10 to 15% were middle-income families. A little more than half of the families had children; 98% of the buyers were African American. Families are attracted to the market-rate units in the Terraces by the modest prices and because the homes have amenities not available for similarly priced homes in the suburbs (D. Sowell, personal communication, February 28, 2002). The presence of major employment centers nearby and financial incentive programs (discussed earlier) have also facilitated the entry of these market-rate families into Lexington Terraces (B. Struever, personal communication, March 28, 2002). | Attracting Middle-Income Families | 161 Although redevelopment efforts have contributed to a new sense of the physical environment, a major rift has emerged between the public housing tenants and home- owners at the Terraces. Some of the initial homeowners at the site indicated that they were not aware that public housing residents would reoccupy the site. The tension was mani- fested and grew when after only three months of living in the new community, some members of the tenant council complained that one of the homeowners (who had pre- viously been a public housing resident at Lexington Terrace) decided to move out when public housing residents returned to the site. Some public housing residents complained that that the new homeowners believed that they were better than them (E. Craven, personal communication, February 28, 2002; D. Foster, personal communication, February 28, 2002; L. Leadbetter, personal communication, February 28, 2002). The management company staff, therefore, faces a challenge. How can they promote social interaction between members of the two groups in order to achieve the presumed benefits of mixed-income housing (E. Craven, personal communication, February 28, 2002)? Today the Townhomes on Capitol Hill is a successful mixed-income community (‘‘HOPE VI creates mixed-income developments on Capitol Hill,’’ 2001). Although former residents of Ellen Wilson and current residents of two nearby public housing develop- ments had priority over other applicants for the Townhomes, the highest income band was the first to sell out. However, few if any of the market-rate families have children. The absence of families with children at the Townhomes in Capitol Hill is typical of gentrifying communities.

CONCLUSIONS Three major conclusions emerge from our comparative case study analysis of HOPE VI sites in Cincinnati, Louisville, Baltimore, and Washington, DC. First, although all four HOPE VI developments aimed to create mixed-income housing, attracting and holding middle-class families with children was not a prominent goal (see Table 3 which sum- marizes the results across the four sites). Cincinnati officials emphasized developing home- ownership opportunities for low-income former residents of public housing as well as attracting upwardly mobile singles and childless couples because these were the types of families interested in downtown and near-to-downtown living. Louisville officials made no specific efforts to attract middle-class families with children because children who moved into Park DuValle would be bused to schools throughout Jefferson County. The Lexington Terrace (Baltimore) development team’s marketing strategy did emphasize public school improvements but the strategy also stressed the proximity to employment opportunities and the low cost of the housing units. The developers of The Townhomes (Washington, DC) were so interested in getting the HOPE VI development built as quickly as possible that they took advantage of the Capitol Hill’s appeal to middle-income singles and childless couples. Second, whereas Jefferson County Public Schools maintained a high level of collabora- tion throughout the HOPE VI planning and implementation process (reflecting a culture of housing-schools cooperation in that city), there was limited collaboration in the other three cities. However, the level of collaboration rose during the HOPE VI process. Cincinnati Public Schools became more involved when it began conducting its own facilities master plan and needed to consider what was likely to happen to enrollments in the West End. At The Terraces site (Baltimore) collaboration between educational and housing policy accelerated after a school board teamvisited the Centennial Place HOPE VI development in Atlanta, which contains a unique community-based magnet school. 162

TABLE 3 | ORA FUBNAFFAIRS URBAN OF JOURNAL Attracting Middle-Income Families at Four HOPE VI Sites

Cincinnati Louisville Baltimore Washington, DC City West Park DuValle The Townes at the Terraces Townhomes on Capitol Hill

Marketing: High priority given to homeowner- Middle-income families consid- Marketing emphasized proximity Design features helped to attract ship opportunities for former pub- ered site because of city- to schools and community center middle-income singles and young lic housing residents suburban school district as well as to employment couples opportunities Marketing studies indicated the Marketing to international com- Financial incentive program for site could attract middle-income munity had little impact Baltimore employees helped in singles and childless couples attracting families |

Collaboration: Cincinnati Public Schools not From the beginning close coop- Collaboration between Baltimore Joint planning (schools-housing) 2/2005 27/No. Vol. involved at the beginning eration between Jefferson County Public Schools and Housing virtually non-existent Public Schools and Louisville Authority of Baltimore City mini- Housing Authority mal at the beginning CPS became more active when it Proposal by developer for a com- Collaboration increased when the Good prospects for schools- began to develop its own facilities munity school at Park DuValle housing authority director took a housing collaboration at future plan rejected by JCPS group to Centennial Place, HOPE VI sites Atlanta, HOPE VI site Proposal to give City West chil- dren preference at a magnet school not approved by CPS Degree of success: All of market-rate units sold but Many of market-rate buyers All market rate units have sold; Highest income band of none to families with school age have children in JCPS almost all to moderate-income units was first to children system buyers sell out Has appealed to families who Only a small number of middle- Few of market-rate buyers have know about city-suburban school income buyers have been children district attracted Has achieved economic A major rift developed between but not racial homebuyers and public housing integration residents Separate owner and renter orga- nizations evolved over time | Attracting Middle-Income Families | 163 A K-8 math, science, and technology curriculum focused school is being built near The Terraces. Although there was virtually no school-housing collaboration in the planning at Townhomes on Capitol Hill in Washington, DC, the school district has been playing a more active role in recent HOPE VI developments. Third, our results highlight the difficulty of attaining an income mix that will lead to meaningful social interaction across social class lines. Because of their highly urban locations City West (Cincinnati) and Townhomes on Capitol Hill were successful in attracting childless middle-income families but not middle-income families with children. The Terraces (Baltimore) was successful in attracting low/moderate-income homeowners (some of whom previously lived in public housing) because of the heavy subsidies pro- vided for ownership housing. Park DuValle, Louisville’s ability to attract middle- and upper-middle-income families, including many with children, was a function of the attrac- tive, moderately priced suburban-type homes, and the city-suburban school system that takes schools out of the locational choice equation. At both the Louisville and Baltimore HOPE VI sites, there has been considerable tension between owners and renters implying that physical proximity does not necessarily lead to positive social interaction. Because this article has been an exploratory one, it is most appropriate to consider a future research agenda. We have articulated a logic model for the HOPE VI program, focusing on developing mixed-income neighborhoods. Given the previous lack of an explicit logic model for this program, we expect and welcome further refinement and debate centered on the explication of the model and testing of links in the model. With the additional development of the program, further research on best practices is warranted, with a high priority given to identifying successful efforts to attract middle-income families with children. Further analysis, based on survey research, of the motivations of those middle-income households who choose HOPE VI developments would be most helpful. Additional inquiry should be oriented toward questions asking about the role of financial incentives, access to downtown schools (including magnets and non public ones), and crime and safety issues. Finally, more extensive case studies of school-housing cooperation over longer time periods would aid in understanding the development of such relations, moving beyond the shorter timeframe this study provides. HOPE VI is an ambitious and significant program and certainly warrants an increased understanding of its logic, its development, its implementation, and its effectiveness.

ACKNOWLEDGEMENT: The research was made possible by a grant from the Fannie Mae Foundation. The authors thank our informants in the four cities we studied and Margaret Wilder for her sage publication advice.

REFERENCES

Abt Associates. (1998, June). Lincoln Court. HOPE VI grant application. Prepared for the Cincinnati Metropolitan Housing Authority. Abt Associates. (1999, May). Laurel Homes: Building on our heritage. HOPE VI Grant Application. Prepared for the Cincinnati Metropolitan Housing Authority. Boston war zone becomes public housing dream. (1991, November 23). New York Times,p.7. Brookings Institution. (2001, July). Beyond merger: A competitive vision for the regional city of Louisville. Washington, DC: Author. Brophy, P. C., & Smith, R. N. (1997). Mixed-income housing: Factors for success. Cityscape, 3, 3–32. 164 | JOURNAL OF URBAN AFFAIRS | Vol. 27/No. 2/2005

Burtless, G. (Ed.). (1996). Does money matter? The effects of school resources on student achievement and adult success. Washington, DC: Brookings Institution. Children’s Aid Society. (2001). Building a community school (3rd ed.). New York: Children’s Aid Society. Chung, C. (2002). Using public schools as community-development tools: Strategies for community-based devel- opers. Report for Harvard Joint Center for Housing Research and for Neighborhood Works. Available: http://www.nw.org/network/pubs/studies/documents/schoolsCommDevelop2002.pdf [December 2004]. Coleman, J., Campbell, E. Q., Hobson, C. J., McPartland, J., Mood, A. M., Weinfield, F. D., & York, R. L. (1966). Equality of educational opportunity. Washington, DC: U.S. Government Printing Office. Donnelly, S. (2003). A toolkit for tomorrow’s schools. Planning, 68(9), 4–9. HOPE VI creates mixed-income developments on Capitol Hill. (2001, September–October). U.S. Department of Housing and Urban Development. Fieldworks, pp. 1–2. Khadduri, J., & Martin, M. (1997). Mixed-income housing in the HUD multi-family stock. Cityscape, 3(2), 33–70. Kreyling, C. (2002). Schools for downtown Chattanooga. Planning, 68(7), 32–33. Lowell, A. K. (1999, August). Rebuilding a neighborhood—Inside and out. Oberlin Alumni Magazine. Available: http://www.oberlin.edu/alumnag/oamcurrent/oam_august99/rebuilding2.html [October 2004]. Matthews, S. (2004). Building a community, not just housing. Planning, 70(3), 10–13. Mrozowski, J. (2001, November 9). CPS plan details coming soon. Cincinnati Enquirer. Available: http:// enquirer.com/editions/2001/11/09/loc_cps_plan_details.html [August 2004]. National Housing and Rehabilitation Association. (n.d.). Public housing agencies see positive changes from HOPE VI program. Available: http://www.housingonline.com/news/pha_observations.htm [July 2004]. Raffel, J. A., Denson, L. R., Varady, D. P., & Sweeney, S. (2003). Linking housing and public schools in the HOPE VI public housing revitalization program: A case study analysis of four developments in four cities. School of Urban Affairs, University of Delaware. Available: http://www.UDel.edu/ccrs/pdf/ LinkingHousing.pdf [July 2004]. Romeo, J. (2004). The ABCs of mixed use schools. Planning, 70(7), 4–9. Rosenbaum, J. E., Stroh, L. K., & Flynn, C. A. (1998). Lake Parc Place: A study of mixed-income housing. Housing Policy Debate, 9(4), 703–40. Rubinowitz, L. S., & Rosenbaum, J. E. (2000). Crossing the class and color line: From public housing to white suburbia. Chicago: University of Chicago Press. Schwartz, A., & Tajbakhsh, K. (2001, November 8). Mixed-income housing as social policy: The case for diminished expectations. Paper presented at the 43rd Annual Conference Association of Collegiate Schools of Planning, Cleveland. Turner, M. A., Kingsley, G. T., Popkin, S. J., & Abravanel, M. D. (2004). What next for distressed public housing? Urban Institute. Available: http://www.urban.org/url.cfm?ID=1000654 [December 2004]. Varady, D. P., & Raffel, J. A. (1995). Selling cities: Attracting homebuyers through schools and housing programs. Albany, NY: State University of New York Press. Wilkinson, H. (2003, January 8). West End residents strive to reverse century of decline. Cincinnati Enquirer. Available: http://www.enquirer.com/editions/2003/01/08/loc_westdayton.html [July 2004]. Willie, L. J. (1997). At home in the loop: How clout and community built Chicago’s Dearborn Park. Carbondale and Edwardsville, IL: Southern Illinois University Press.