MIXED-INCOME REDEVELOPMENT IS THE ONLY WAY TO FIX FAILED PUBLIC HOUSING
Lawrence J. Vale
Public housing in the United States began as a mechanism to house an upwardly mobile segment of the working poor, evolved into a program that housed chiefly the least economically advantaged and the poorest welfare-dependent households and, since the 1990s, has been slowly replacing many failed low-income projects with new mixed-income communities.1 Mixed-income redevelopment, frequently touted as an approach that benefits everyone and often assumed to be an almost self-evident good thing, has nonetheless been subjected to ongoing skepticism, expressed through the academic literature, the popular press, and the views of affected tenants. Supporters contend that it yields safe and attractive neighborhoods that serve all incomes, including some former public housing residents from the site, while permitting other former public housing residents to gain from dispersal to other mixed-income neighborhoods with the assistance of subsidized housing vouchers. Detractors counter that mixed-income redevelopment usually entails a loss of much-needed “hard units” of public housing, and unjustly displaces too many low-income households to other poor neighborhoods where they lack their former social networks. Especially since “mixed-income” housing has no commonly shared definition and since many alleged benefits of income mixing remain undemonstrated in practice, it is hardly surprising that mystification and mythification have proliferated.2
At least since the New Deal, housing policymakers in the United States have attempted to redesign and reengineer distressed urban neighborhoods to foster particular social and economic aims, often verging on the utopian. Then, as now, they sought to achieve the triple objective of revitalizing an area, providing higher-quality housing, and uplifting low-income tenants. All too often, these grand intentions overestimated the capacity of housing projects to effect such total transformations.3 The earliest incarnations of American public housing—the pre–World War II efforts of the Public Works Administration (PWA) and the United States Housing Authority (USHA)—frequently attempted to reduce areas of concentrated poverty through slum clearance. Under USHA guidelines, any new public housing had to be accompanied by an “equivalent elimination” of slum areas. Private-sector real estate interests insisted on this to prevent an enhanced supply of government-subsidized affordable housing from unfairly competing with their own efforts. Public housing advocates of the 1930s and 1940s decried slums as the source of disease, dispiritedness, and delinquency and viewed modern public housing as the antidote to both unsanitary conditions and behavioral problems. They argued that constructing public housing would be in the best interest of the poor, but—like many advocates of mixed-income developments today—did not always examine the actual range of outcomes that often countered their assumptions.4
From the 1930s through the 1950s, city officials optimistically viewed public housing as part of a broader effort to spur investment in depressed urban areas and to curb poor behavior in low-income communities by replacing slum dwellers with a more carefully screened and upwardly mobile brand of tenant. These residents of early public housing surely had a range of incomes, though the range was set in a deliberately narrow way. If they were too poor, they would be judged too risky to admit; if they were too rich, they would be declared ineligible for continued occupancy and asked to move onward and upward. High levels of formal and informal social control accompanied the extensive social networks of early public housing. Management enforced the rules, and tenants frequently responded by taking personal responsibility for the outdoor common areas adjacent to their apartments. If residents wanted to look for role models, early public housing was full of households that used hard work and savings to move on quickly—often even to homeownership. Again, no one called this mixed-income housing then; it was just a collection of families seeking to better their circumstances. In these early years of public housing, the public sector could still take the lead in facilitating a mix of incomes that could yield safe and desirable planned communities. The intent here is not to wax nostalgic about an era of public housing that was segregated, brutally exclusive, and overwhelmingly paternalistic. It is merely to point out that there was a time in the United States when public housing authorities themselves tried to achieve all of the objectives now claimed as goals by advocates of mixed-income developments.
As public housing shifted into its welfare phase during the 1960s, 1970s and 1980s, policymakers increasingly came to blame the “concentration of poverty” as the source of all problems.5 As tenant incomes declined, public housing communities suffered from some combination of inadequate financial structure, excessive deferred maintenance, poor design, indifferent management, and protracted racial tensions. These were all accompanied by a gradual shift in the tenant population away from the carefully vetted collectives of the first twenty years, when public housing was regarded as a reward, to a system where public housing became the last stop short of homelessness (or the first stop back) for the least advantaged of U.S. households. Housing authorities responded to the increased desperation of those who applied for units by admitting them regardless of the problems they might bring. They did so consciously, even conscientiously.
Starting in the 1970s, local housing officials began to experiment with deliberately created mixed-income communities as an alternative to public housing that contained only those with extremely low incomes, and this trend has accelerated. Since the early 1990s, U.S. Congress has appropriated more than $6 billion under HOPE VI (currently translated as Housing Opportunities for People Everywhere), a program of the U.S. Department of Housing and Urban Development (HUD) designed to facilitate more than 250 efforts to redevelop “severely distressed” public housing. Much of this redevelopment, spread across nearly every major American city, has entailed the demolition of housing projects and their replacement with mixed-income communities. Proponents have eagerly proffered evidence to suggest declining crime and disorder, higher neighborhood maintenance, reduction of social stigma, escalation of housing values, enhanced neighborhood attractiveness, improved local services, and increased local investment by the private sector.6 Critics counter that income mixing has been a thinly veiled attempt by a neoliberal state to commit public funds to support gentrification, while displacing the poor and disrupting social networks.7 As the disputes continue, most public housing redevelopment has become almost synonymous with income mixing.8 Lost in the shuffle of incomes, however, is the possibility that public housing salvation by income mixing may be more myth than reality.
There are three principal reasons why it seems misguided to view mixed-income housing as either a panacea, or as the only viable strategy for revitalizing distressed public housing. First, mixed-income development has been too loosely defined as a category, and unhelpfully allowed to encompass a huge range of scenarios—everything from market-rate developments with a small remaining number of public housing units, to developments that remain chiefly public housing, to those that include few or no market-rate units and instead mix public housing with other subsidized housing produced with Low Income Housing Tax Credits. In other words, even if there are positive effects that follow from the introduction of mixed-income housing, important unresolved questions remain: Which kind of mixed-income housing is needed to generate such gains, and which kinds of residents actually benefit from the changes?
Second, scholars and practitioners continue to debate the merits of mixed-income approaches, having found that many of the assumed benefits of income mixing for low-income residents, such as role modeling, provide little positive impact on their lives. By contrast, other aspects seem more promising: enhanced security, increased investment in neighborhoods, and higher expectations for management.9 Given that some rationales seem to be much more persuasive than others, and that benefits are not evenly shared, how can mixed-income housing best be justified?
Finally, several cities have implemented successful public housing redevelopment efforts without mixing incomes or by keeping their entire engineered mix wholly confined to various levels of low-income households. Such initiatives have been implemented both before the HOPE VI program began, and under its auspices. Recognizing this kind of alternative, is all the cost and effort associated with mixed-income public housing really necessary?
Mixed Incomes, Mixed Definitions
Mixed-income public housing is an ill-defined subset of the equally ill-defined but much larger phenomenon of mixed-income neighborhoods. Such neighborhoods can either be seen as produced by deliberate social policy (as in the case of HOPE VI public housing redevelopment) or as emerging in a more organic fashion.10 The policy shift toward deliberately created mixed-income neighborhoods emerged out of a sense that neighborhoods of “concentrated poverty” had been dangerous failures;11 it did not emerge as a policy response to some clear empirical evidence about the positive value of income mixing.12 The latter, for the most part, has simply been assumed rather than tested, constrained by the fact that the broader phenomenon of mixed-income neighborhoods has been defined so loosely and inconsistently.
At base, such neighborhoods are difficult to situate both in time and in space. First, neighborhoods may exhibit income heterogeneity for only a short period of time, a period of transition between a place that is predominantly low-income and a gentrified locale that is dominated by high incomes, or a transition moving in the other direction.13 Second, neighborhoods often have manipulable or contestable boundaries, and the degree of income mixing may vary markedly depending on how and where such boundaries are drawn. Should the mixed-income nature of a mixed-income public housing development be measured on its own, for instance, or should assessment of the degree of income mix also take account of the socioeconomics of surrounding blocks? If the latter count as part of “the neighborhood,” this may skew the nature of the income mix either upward or downward in quite substantial ways.
This jurisdictional matter also cuts to the heart of the question of how one classifies “concentrated poverty.” For example, a public housing development with most residents living below the poverty line will presumably have a far smaller percentage of impoverished residents if it is redeveloped into a mixed-income community composed of one-third public housing, one-third shallow-subsidized “affordable” housing, and one-third market-rate housing. It also matters considerably whether the surrounding neighborhood is low poverty or high poverty, since some high-poverty public housing located in affluent areas might already be considered less affected by “concentrated poverty” if this poverty is measured to include that larger area of neighborhood affluence. Conversely, if the neighborhood boundaries are drawn more tightly to encompass only the housing development footprint itself, the resultant mixed-income development (even if two-thirds of its residents are not in public housing) may still exhibit a poverty rate of 33 percent, a figure considered “high poverty” by most social policy analysts. Because of slippery definitions that are easily subjected to manipulation for political purposes, it is possible to define a public housing project—such as Chicago’s Cabrini-Green—as an isolated enclave of “concentrated poverty” only if the surrounding affluence is not counted as part of the neighborhood being assessed. Policymakers can claim that they have benevolently supported most Cabrini residents by relocating them away from such poverty, but, if the neighborhood boundaries are framed differently, it is equally true to say that they have been forced out of a larger low-poverty neighborhood and into one that is very likely to have a much higher poverty rate.
Next, at least in an American context, the definition of “mixed income” also must encompass considerations of race and ethnicity. Mixed-income neighborhoods that vary by class but are more homogeneous in terms of race or ethnicity may well have a different social dynamic than those where substantial income differentials are correlated with distinct racial or ethnic groupings. Also, since different races and ethnicities exhibit different income profiles, the very definition of income mix can vary in relation to these sorts of factors. Clearly, there can never be some single standard for a mix.
Similarly, a mixed-income housing development or neighborhood may also operate differently as a social and political unit if it contains mixes of housing tenure. Inclusion of homeownership opportunities, especially if accompanied by separate institutional associations that exclude renters, may yield communities that operate differently from those that are rental only. Homeownership often signals a longer-term investment in a place than does market-rate rental housing, though subsidized low-income renters may also take a long-term stake in their home neighborhood, and some owners are interested only in short-term investment opportunities. Finally, deliberately produced mixed-income neighborhoods can vary widely by building type, by type of management, by location (from urban to suburban), and by housing market (from “tight” markets where renters have few alternatives to higher-vacancy places where competition is less fierce); all these important factors also detract from the value of the term, rendering it as little more than an ambiguous catch-all.
At its core, though, the definitional morass of mixed-income housing stems from the failure to distinguish among narrow mixes—mixes that cut across a broad continuum of incomes, and mixes that are starkly polarized between high-income and low-income contingents. Moreover, there is no consistency in how income groups are defined; those who are “high income” in one mixed-income context could even be considered “low-income” in another.14 As Alex Schwartz observed in 2010, “Virtually no research exists on the actual financial performance of mixed-income housing,” since financial reporting systems used by the operators of mixed-income housing fail to separate out the distinct contributions of each income group.15 Narrow-mix options have famously included Vincent Lane’s Mixed-Income New Communities Strategy (MINCS) deployed at Chicago’s Lake Parc Place in the early 1990s, which filled half of the apartments with very low-income public housing households and the other half with employed households earning 50–80 percent of the area median income. By pairing public housing residents with those whose earnings put them the upper end of what counts as “low income,” Lake Parc Place managed an income mix where all incomes were nonetheless some variant of “low.”16 Similarly, some HOPE VI initiatives, such as San Francisco’s North Beach Place, have successfully crafted mixed-income strategies that replaced all existing low-income public housing units on site, while adding in another tier of shallow subsidy housing, funded by Low Income Housing Tax Credits. At North Beach, policymakers managed to combine public and private investment to create a “mixed-income” community that nonetheless served an exclusively low-income clientele.
By contrast, the much-touted development of Boston’s Harbor Point during the 1980s (i.e., pre-HOPE VI) replaced 100 percent public housing with a predominantly market-rate community,17 a practice frequently echoed in HOPE VI mixed-income developments. Most HOPE VI communities defined as “mixed income” have had at least three tiers of income groups: a market-rate tier, a tier still composed of tenants receiving public housing subsidies (or similar project-based vouchers), and a middle tier of more moderately subsidized households.18 In practice, however, there is no commonly deployed terminology that separates out these very different concepts of “mixed income,” and there has been little systematic effort to measure the outcomes from each type.
Mixed Incomes, Mixed Outcomes
Just as the term “mixed income” has stubbornly resisted consistent definition, so too the expected advantages from adopting this approach have remained woefully unclear, often subject to blind assumptions and wishful thinking rather than empirical assessment. It remains crucial to gain clarity about what mixed-income housing is, but it is equally pressing to understand what it does, since the results of this approach to policy need to be measured. To date, empirical investigation of mixed-income communities, at least in the United States, suggests that their outcomes are at least as mixed as their incomes.19
Social policy analyst Mark Joseph and his colleagues have provided a particularly astute dissection of the diverse motives behind mixed-income development.20 Their research encompasses ethnographic work on several new mixed-income communities created to replace public housing in Chicago but also includes a broader analytical framework—one that looks across the findings of many researchers to explain and assess the diverse rationales that undergird the phenomenon of mixed-income housing. Until policymakers and developers can answer Joseph’s apt question about whether mixed-income housing is intended “to revitalize the local area and provide additional housing options for urban dwellers, to provide low-income residents with higher-quality housing, to help lift low-income families out of poverty, or some combination of the three,”21 it will be impossible to assess its effectiveness—especially if it turns out that some of these goals are mutually exclusive. By separating out several ways that mixed-income developments potentially could benefit the urban poor, Joseph usefully connects these assumptions to what is known from the literature and thereby allows those involved in housing redevelopment efforts to have a better way to situate claims and measure success.22
As Joseph’s framework reveals, proponents justify mixed-income developments (however variously the mix is defined) as facilitating four distinct kinds of gain. The presence of higher-income residents is said to provide: (1) increased social capital for low-income residents, (2) direct or indirect role-modeling of social norms for work and behavior, (3) informal social control leading to safer and more orderly communities for everyone, and (4) gains for the broader community through enhanced engagement of political and market forces. Analysis of a wide variety of mixed-income communities, including those studied by Joseph and by others, casts doubt on the first two forms of presumed success. These rationales, emphasizing social capital and role-modeling, seem to be the most frequently claimed benefits, but empirical social science researchers have rarely found much supporting evidence when analyzing the social practices of actual communities. By contrast, researchers have found more support for rationales that emphasize enhanced security and greater neighborhood investment.23 These last two rationales are certainly important factors for judging the desirability of a community, but they are less centrally or exclusively the kind of benefit that inures primarily to those with the lowest incomes.
More pointedly, James DeFilippis and Jim Fraser observe that all four of the most frequently voiced rationales for mixed-income policies are “largely based on the (hegemonic) mantra that low-income people themselves are the problem, and that a benevolent gentry needs to colonize their home space in order to create the conditions necessary to help the poor ‘bootstrap’ themselves into a better socioeconomic position.” Framed this way, “poor people . . . come to be simply ‘a problem’ that we need to spread out—and the language of ‘fair share’ or ‘regional equity’ that is often heard sounds remarkably similar to how people involved in environmental justice movements talk about things like waste transfer stations or incinerators.” DeFilippis and Fraser ask a pertinent question: “Why do we want mixed-income housing and neighborhoods?” Their point is not to denigrate the potential value of mixed-income places; they just question who should benefit from them, and how benefits to low-income households can be maximized. They ask how mixed-income communities can be better governed, and how power inequities that continue to penalize on the basis of race, class, and gender distinctions can be minimized. Implicitly, they also ask which “we” should be empowered to choose the rationales for establishing the mix.24
Closely related to the assessments of mixed-income communities created as part of the HOPE VI program, others have questioned whether the dispersal or “vouchering out” of low-income residents from public housing and into the world of private landlords has helped them achieve the same sorts of goals ascribed to intentionally designed mixed-income communities.25 Here, too, the realities have not often met the hopes of proponents who expected substantial socioeconomic gains. HOPE VI dispersal programs have rarely followed relocated residents in any systematic, long-term way, but more fully studied programs, such as the Moving to Opportunity (MTO) demonstration, have proved to be no more than a mixed success, at best. Although dispersal programs consistently reveal important gains in perceived safety in comparison to former neighborhoods of concentrated poverty, there is little evidence of comparable gains in terms of social networks, employment, or economic security. As Edward Goetz and Karen Chapple summarize, “There seems to be little in the record of MTO or forced displacement to suggest that relocated low-income families will fulfill the expectations of the dispersal model and form relationships with their (presumably higher-income) neighbors, thereby building ‘bridging’ social capital critical to finding employment activities.” More generally, “dispersal efforts have not had any demonstrable positive effect on employment, earnings, or income of individuals.” Similarly, their survey of the literature on voucher-based dispersal resulting from HOPE VI finds “limited benefits” to families. Although they typically move to places where they feel safer, most move to other poor neighborhoods rather than to low-poverty destinations, and “there are conspicuously no benefits in employment, income, or [reduced] welfare dependency. Furthermore, many of the families suffer significant interruptions in their social networks.” In short, the overwhelming negativity of the research emphasis on the contagion effects of concentrated poverty may well have undervalued the vital and compensatory role of locally rooted networks of support.26
Given that research findings consistently provide no more than limited support for most of the cherished assumptions about the value of mixed-income housing, this raises questions about the advisability of large public investment in mixed-income developments. This is a particular concern for mixed-income communities implemented through the HOPE VI program, since these tend to reduce the already scarce supply of units serving those with the lowest incomes. If the presence of higher-income residents does not consistently provide the assumed benefits for low-income residents in terms of enhanced social capital, behavioral improvement, or economic advancement, then one may ask whether these important social goals could be better met by means other than mixed-income developments. Likewise, it is not clear whether the modest support in the literature for the third goal—enhanced informal social control—really depends on having a mixed-income community to achieve it. Finally, this raises questions about whether the remaining social goal—the fourth justification centered on neighborhood political and economic gains—might be met equally well by other policies.
Fixing without Mixing
If mixtures of incomes are good because they leverage resources and enforce higher standards, it is worth pressing harder to identify exactly which kinds of mixes are necessary to achieve the desired objectives. Narrow-mix experiments like Chicago’s Lake Parc Place may actually provide more of the presumed social benefits of mixed-income housing than a wide-mix HOPE VI development.27 In the latter, the gaps between very low-income, transit-dependent, long-term public housing tenants with children versus childless, auto-equipped, transient households with market-rate apartments or owned homes may simply be too vast for any form of social capital to bridge. The situation may be even more difficult if there are linguistic barriers or racial tensions as well.28
If providing greater informal social control and stimulating higher quality goods and services are the main benefits that mixed-income developments can proffer, it is worth examining more closely just what institutional mechanisms are needed to achieve these benefits. The first goal would mainly seem to be a management and tenant selection challenge. Whatever else the construction of a HOPE VI mixed-income community may be, it is an excuse to rethink the terms of housing management and an occasion to re-select the members of a community.29 The return to more careful tenant screening today reflects a partial revival of the hard-headed realism of strict entry selection practiced during the 1930s and 1940s. The efforts to reassert such social control today are made much more difficult because HOPE VI redevelopment sites are not a social tabula rasa; every move to import a higher socioeconomic group is met with mistrust from at least some among the less upwardly mobile previous residents who understandably fear displacement.
It is sometimes possible to achieve many of the social goals that are wishfully ascribed to mixed-income public housing without actually allocating any units to market-rate tenants or homeowners. Most discussion of the Chicago Housing Authority’s immensely ambitious Plan for Transformation boldly features CHA-led efforts to construct new mixed-income communities on and adjacent to the sites of some of the city’s most notorious developments. Yet, well over a decade into the plan’s implementation, only about three thousand public housing households had been accommodated in such places. By contrast, the city has quietly rehabilitated about five thousand public housing apartments for families and about nine thousand for seniors, all without any mixing of incomes—and all without any evaluation.30 Before assuming that only income mixing can fix distressed public housing, closer examination of many other national efforts to improve public housing management and design ought to be considered as alternatives.
Fixing without Mixing, Boston Style
At Boston’s Commonwealth Development—redeveloped a decade before the HOPE VI program existed—the Boston Housing Authority, the Commonwealth Tenants Association, a private developer, and a private management company worked together to replace a severely distressed 648-unit housing project with a well-designed, safe, and supportive community (Figure 6.1). They did so while screening out “troublemakers” but nonetheless rehousing the majority of former residents. The housing authority continued to use a citywide public housing waiting list, and preserved virtually all of the redeveloped apartments for very low-income and extremely low-income renters.31
The tens of millions of dollars spent in the early 1980s to transform Commonwealth is a story of a strong core of residents, activated with the assistance of a hired community organizer, and backed by a vibrant network of neighborhood, city, and state organizations. The housing authority, developer, and management company worked with residents to negotiate the physical and social goals for the redevelopment, building a level of mutual trust among parties that had previously been perpetual adversaries. Residents, working with skilled architects, landscape architects, and urban designers—as well as with the housing authority redevelopment team and sympathetic legal assistance advisers—agreed on the priorities for the physical redevelopment, and helped craft a 223-page management plan that covered a wide variety of community rules and procedures. Remarkably, this plan even gave the tenants the right to fire the private management company with 30 days’ notice.
Figure 6.1 Commonwealth Development, Boston, as of 2013. Photograph by author.
Nearly thirty years after the redevelopment was completed, it is arguably still the best public housing in the city. Commonwealth’s success has taken at least seven different forms: the team 1) implemented the redevelopment smoothly, 2) designed a high-quality place, 3) improved the capacity of the tenant organization, 4) enhanced maintenance and management performance, 5) dramatically boosted security, 6) made progress on socioeconomic indicators, and 7) created a community that yields high rates of resident satisfaction.32 Writing more generally about housing redevelopment success, Joseph observes: “It may be that the combination of strong management and more active informal control by residents is the most effective means of maintaining social order.”33 At Commonwealth, this is just what happened, except that everyone there is a very low-income public housing resident.
If, in the end, one measure of success in American public housing is the rate at which its residents become economically able to leave it, Commonwealth’s success is more limited. If, on the other hand, success in public housing redevelopment is measured by the enhancement of an attractive, safe, and stable community, where even many of those who can afford to leave will choose to stay, then the effort at Commonwealth warrants the highest praise. Moreover, because Commonwealth’s success in these terms was achieved through tenant activism, responsive design, and effective management—without resorting to the income mixing so often touted as necessary to revitalize public housing—it raises important questions about current policy trends.
Despite the seeming promise of interventions such as the Commonwealth redevelopment, Edward Goetz correctly laments that “the possibility of public housing with excellent property management, good schools nearby, high quality public services, engaged and informed public-sector supervision of housing authorities, and private-sector investment providing jobs and retail opportunities for residents” seems to be “off the table, altogether” as a policy direction. Still, as Goetz points out, “There is very little inherent in the public housing model that precludes these outcomes; it is how our public and private institutions respond to public housing that has produced the negative outcomes we have seen in American public housing.”34
The Commonwealth example suggests that distressed public housing can sometimes be reclaimed for its residents and for its neighborhood without income mixing. Yet this model seems to have dropped off policymakers’ radar screen, presumably because it does not leverage the same sort of additional private investment that mixed-income developments do and because it fails to reach moderate-income households that are also legitimately struggling to find affordable housing in high-priced cities. Instead, our collective embrace of the mythic mixed-income model reflects the prevailing view that enhancing private investment is what matters most in public housing redevelopment. In other words, income-mixing may enhance the statistical profile of a given neighborhood, and it may be the trigger to investment that improves the appearance of the built environment, but it may not be the catalyst that improves the lives of low-income individuals and families, since many of these more mobile poor may find themselves forced out of a place that is new and attractive, but also newly and unattractively unaffordable.
If we cared about maximizing the utility of the public housing system for very low-income households, we might look at the income mix in developments in a different way. Right now, the deeply political implicit question that is asked when calculating an income mix is, How can the number of very low-income households that need to be accommodated be kept to a minimum so that redevelopment will remain financially appealing to private developers and investors? Combining all of the multiple goals for mixed-income housing, a more equitable approach would be to ask instead, What is the maximum number of equitably screened very low-income households that can be accommodated in a mixed-income development while still ensuring a safe and stable community? Sometimes, as is the case at Commonwealth, that number may reach 100 percent.
Conclusion
Analysis of the literature on mixed-income developments reveals different motives for its backing and casts significant doubt on key assumptions about its presumed benefits. This literature provides more support for the ability of mixed-income developments to enhance social control and help leverage neighborhood political and economic gains. However, some of those advantages could presumably be achieved for low-income households through well-managed housing, careful tenant selection, and good design—without income mixing. The inconclusive endorsement for mixed-income housing proffered by Mark Joseph and other scholars suggests the need for further ethnographic research on these communities, including an analysis of the importance of homeownership, the pattern of engagement with public schools, and the advantages of different kinds of income mixing.
Most important, it is worth systematically comparing those mixed-income HOPE VI communities that have achieved their deconcentration of poverty by development-based dilution of poverty (keeping the original number of public housing units on site but mixing in some additional units of affordable or market-rate units) versus those that have pursued mixed-income strategies by means of neighborhood-based dispersion (reducing the percentage of low-income housing units on the original site while scattering replacement housing into other communities). In rapidly gentrifying areas, the mixed-income approach can either join the trend and shift the remaining deeply subsidized housing elsewhere or, alternatively, it can serve as the last chance to preserve some affordable housing in an area where low-income households are being priced out of the market. If the goal is to assess the benefits and costs of mixed-income approaches to low-income residents, higher-income residents, and neighborhood dynamics, then all of these need to be carefully considered together.35
The many mixed-income developments built outside the HOPE VI system—assuming that they do not displace existing low-income housing and households—do not face quite the same equity hurdle. Even so, given the immense challenges facing low-income households seeking affordable housing in U.S. cities, we should still hold all mixed-income developments to high standards. If advocates of such developments do indeed want to pursue a triple goal of area revitalization, better housing quality, and support for low-income households, it may well be that only those narrow-mix models that attempt to maximize—rather than minimize—participation by low-income households have a chance of reaching all three goals simultaneously. Comparative neighborhood research could empirically show which kinds of income-mixing matters most for which kinds of outcomes. At the very least, policymakers should scrutinize the assumptions that are used to justify mixed-income developments and researchers should develop more comprehensive methods to assess empirically whether the assumed outcomes are actually being realized. Only then will the mythic promises of mixed-income housing be given a true test.