10.2 Public Housing: Housing, Race, and Urban Policy Move the People, Move the Problem?
Move the People, Move the Problem?
Sumner Field, Hollman, Heritage Park, and the strange American experiment in treating poverty as something that might be rearranged on a map.
There is a particular temptation in geography that is almost impossible to resist. When a problem appears on a map, move something. Factories too close to houses? Move the factories. Traffic downtown? Move the road. Flooding beside the river? Move the buildings. Poverty concentrated in one neighborhood? Well, now things get interesting, because people are not factories and poverty is not a building.
Yet from the late 1980s through the early twenty-first century, American housing policy increasingly behaved as though the geography of poverty might be altered much as we alter the geography of roads and land uses. If poor households were highly concentrated in one place, perhaps the concentration itself was making poverty worse. Break up the concentration, move families elsewhere, replace large public-housing developments with smaller mixed-income neighborhoods, and perhaps opportunity would follow.
Minneapolis became one of the great experiments in that idea. The experiment began on a patch of North Minneapolis ground where the federal government first built public housing during the New Deal, where a civil-rights lawsuit later challenged segregation, where bulldozers eventually erased 770 units of public housing, and where the mixed-income replacement is once again, in 2026, the subject of emergency intervention. The buildings changed. The vocabulary changed. The question kept coming back.
First, build the housing
Sumner Field Homes opened in 1938, Minnesota's first federally funded public-housing development. It was part of a New Deal moment when government was willing to answer a housing shortage with a remarkably direct instrument: build housing. That sounds obvious now only because we forget how radical it was. Instead of giving a family money and sending it back into the private market, public agencies acquired land, cleared older buildings, designed a new neighborhood, and became landlords.
Even that beginning contained a contradiction. The new housing replaced an older North Side landscape of modest homes, shops, synagogues, and immigrant institutions. Public housing was therefore born in Minneapolis through an earlier form of urban clearance. Then, once constructed, it created another social world of its own: lawns, rowhouses, courtyards, children walking across the grounds, neighbors sharing the same paths and institutions. By the time later generations called Sumner Field an obsolete project, the place already contained memories that could not be measured by a building inspector.



By 1990 the larger Sumner-Glenwood public-housing complex contained 770 units over roughly 73 acres and housed more than 3,300 people. Jeff Crump's account emphasizes that the population had changed substantially over time. The projects were heavily African American during the 1960s and 1970s; by 1990, after refugee resettlement from Southeast Asia, a majority of residents were recent Asian immigrants, many of them Hmong. The same buildings had become a different social landscape.
The physical problems were real. Crump describes overcrowding, high poverty, deteriorating structures, weak maintenance, and serious water and foundation problems on a site partly built over a buried creek. But bad housing and a valueless community are not the same thing. Hmong residents described childcare networks, family proximity, mutual support, and the ability to watch over teenagers. A planner can see concentration. A resident may see the people who make daily life possible.


The buildings changed. The location kept its pressures.
This north-up locator puts the former Sumner Field / current Heritage Park site in relation to downtown and the transportation corridors that shaped the North Side. Move through the layers: this is the same tract, but not the same policy problem.
Government builds a new public-housing site.
Sumner Field opens as Minnesota's first federally funded public housing. It stands on the near-North-Side land where Bassett Creek had already been covered, beside corridors that connect the area to downtown and also divide it.
Meanwhile, Washington changes its mind about public housing
The Minneapolis story makes more sense if we widen the map. For decades, federal housing policy had moved through several different answers to the same question: what should government do when low-income households cannot obtain decent housing? The New Deal answer was direct public construction. The postwar system expanded that stock, often inside racially segregated urban housing markets. By the 1970s, federal policy increasingly subsidized rents in private buildings through what became the Section 8 voucher system. Then, in the 1990s, the problem was redefined again. The issue was no longer only whether a unit was decent. The neighborhood itself became part of the diagnosis.
Congress created HOPE VI in 1992 to address what a national commission had called severely distressed public housing. The program could demolish, rehabilitate, or replace projects, and it increasingly became associated with lower-density, mixed-income redevelopment. In 1995 Congress suspended the old one-for-one replacement rule that had required a demolished public-housing unit to be replaced; the 1998 Quality Housing and Work Responsibility Act made that change permanent. Once demolition no longer automatically required a new public-housing unit somewhere else, the arithmetic of redevelopment changed profoundly.
Same housing problem, different policy machine
Move through the eras. The point is not that one policy instantly replaced another, but that the center of gravity in federal policy shifted.
The 1937 public-housing system treats inadequate housing as something government can answer by constructing and operating subsidized dwellings.
The unit itself is the core policy object.
The national transformation was enormous. By June 2003, GAO reported that HOPE VI grantees had completed demolition of more than 60,000 public-housing units while having completed about 19,000 replacement public-housing units. That did not mean all displaced families simply lost assistance; vouchers and other subsidized units were part of the new system. It did mean that the American strategy was shifting from a large, permanent stock of deeply subsidized public units toward a patchwork of vouchers, mixed-income redevelopment, and smaller replacement developments.
Then comes Hollman
In 1992 the Minneapolis Legal Aid Society filed a class-action lawsuit on behalf of Lucy Mae Hollman and other African American public-housing residents. The suit named HUD, the Minneapolis Public Housing Authority, the City, and other agencies, arguing that public policy had produced a racially segregated pattern of assisted housing. The original complaint therefore began with discrimination and constrained choice. It asked why minority families had been steered into a narrow geography of opportunity.
Then the intellectual weather changed. During negotiations, the language of concentrated poverty became more powerful nationally and locally. Political leaders such as Myron Orfield argued that poverty and minority residents had become too concentrated in the central cities while jobs and wealth moved outward. If concentration was part of the problem, dispersal could become part of the remedy.
By 1995 the Hollman settlement committed to demolishing Sumner-Glenwood and to using redevelopment, replacement housing, and vouchers to expand residential choice. A civil-rights case about segregation had become, in part, a metropolitan experiment in poverty deconcentration. That is the moment when Crump and Goetz become especially useful, because neither is satisfied with the easy version of the story.
Crump and Goetz look at the same machine from different ends
Jeff Crump's 2002 article is the sharper theoretical attack. His question is almost embarrassingly geographical: when we say poverty is concentrated, are we describing where poverty appears or explaining why it exists? If the map becomes the explanation, the solution follows neatly. Move the dots. Demolish the project. Disperse the residents. Crump argues that this spatial shortcut hides the institutional racism, labor-market change, housing discrimination, public disinvestment, and political decisions that helped create both poverty and segregation in the first place.
Edward Goetz is less interested in stopping the argument before the bulldozer and more interested in following the policy after it leaves the drawing board. His 2000 work is also critical of deconcentration politics, especially when they make affordable housing harder to build in both central-city and suburban communities. His later Hollman research asks the brutally practical questions: Where did people actually move? Did they experience safer neighborhoods? Did incomes improve? Did the regional dispersal imagined by the settlement really happen?
Crump and Goetz are not simple opposites
Click a lens. The site stays the same; the research question changes.
Who turned a social history into a spatial diagnosis?
Crump asks what disappears when segregation, disinvestment, and inequality are compressed into the phrase concentrated poverty. His worry is that demolition looks politically neutral once the neighborhood itself has been defined as the disease.
What happened after families were moved?
Goetz follows the households and the policy apparatus. He looks at relocation distances, neighborhood conditions, household preferences, suburban barriers, and whether dispersal actually produced the metropolitan geography promised by the remedy.
Put them together and the criticism becomes more powerful. Crump says the theory may confuse a map with a cause. Goetz says that even when policy acts on the map, households, landlords, transportation, family networks, discrimination, and the supply of affordable housing refuse to behave like movable symbols.
The resident has a map too
This is the part of deconcentration policy that diagrams often miss. A planner can draw a low-poverty suburb and call it opportunity. A household can look at the same place and ask different questions. Where does my mother live? Who watches my children? Can I reach work without a car? Will a landlord accept my voucher? Where is my church? Where do my children know people? Where will I be treated as though I belong?
Crump cites an early analysis by Goetz showing how limited the actual geographic dispersal initially was: most relocatees remained in Minneapolis and more than half remained within three miles of the old site. Goetz's later work likewise found that families sometimes gained safer or more satisfactory neighborhoods without experiencing the sweeping economic transformation implied by the strongest versions of neighborhood-effects theory. Mobility could matter. It was simply not magic.
Your diagnosis determines what you move
Select a cause. Watch the same neighborhood produce a different policy prescription.
HOUSING
If the central problem is deteriorated buildings, the obvious intervention is physical: repair what can be saved and replace what cannot.
Question to keep: Can the housing be fixed without erasing the community around it?
This was not just Minneapolis: America was tearing down projects
By the time Sumner-Glenwood came down, public-housing demolition was becoming a national landscape. Chicago launched its Plan for Transformation and set out to demolish thousands of obsolete units, including much of Cabrini-Green, Stateway Gardens, and the Robert Taylor Homes, while promising mixed-income redevelopment and renovated public housing elsewhere. Atlanta moved even further: the city that had opened Techwood Homes, the nation's first federally funded public-housing project, ultimately demolished all of its traditional family public-housing projects. New Orleans later demolished major developments after Hurricane Katrina.
But there was never one American path. New York did not simply erase the towers and superblocks. NYCHA remains one of the largest public-housing systems in the world, with more than 150,000 conventional public-housing apartments across hundreds of developments, plus tens of thousands of apartments converted to preservation programs such as PACT/RAD. Even Minneapolis retained thousands of public-housing apartments, including high-rises and townhouses elsewhere in the city. The national shift was powerful, but local political economies mattered.


National policy opened a door; cities walked through it differently
Choose a city. The visual is schematic, but each path reflects a real policy direction.
Sumner-Glenwood was demolished and replaced with a mixed-income, lower-density neighborhood, but Minneapolis did not eliminate public housing citywide. The city's public system continued to include high-rises, townhouses, and other developments.
The national demolition era did not produce one universal local strategy.


And then the bulldozers arrive
Crump's narrative becomes darkest once demolition begins. He argues that the language of deconcentration made clearance seem progressive even when replacement housing lagged far behind. By 1997, he reports, only a small fraction of the suburban replacement units envisioned under the settlement had actually been built. Affordable suburban rental housing was scarce, some landlords resisted voucher holders, and suburban communities often resisted low-income housing proposals. The metropolitan solution had run into metropolitan politics.
In September 1998 demolition of the historic Sumner Field rowhouses began despite a preservation recommendation. By June 1999, as the remaining buildings were being torn down, African American pastors and community activists stood in front of bulldozers; police arrested fourteen demonstrators. The civil-rights coalition itself had fractured. People who had originally supported demolition as part of desegregation now questioned whether the promised replacement system existed strongly enough to justify destroying the remaining units.
A bulldozer is very good at removing a building. It is less talented at building a housing system. The first task can happen in a morning. The second requires land, money, landlords, transit, political consent, construction schedules, and years of maintenance. The mismatch between those two clocks is one of the recurring facts of urban renewal.
Heritage Park: the promise of the opposite
The replacement was designed to embody nearly every lesson planners believed they had learned from the old projects. Heritage Park restored a conventional street grid, reduced the visual separation between subsidized and market housing, mixed incomes, added parks and pedestrian routes, and used lower-rise forms intended to look more like the surrounding city. The development included 440 rental homes, of which 200 were public-housing units, alongside a large for-sale component. The superblock was supposed to become a neighborhood.
There were real accomplishments. Badly deteriorated buildings were replaced. Some families obtained safer housing. Homeownership became part of the plan. Streets connected through the site again. The new landscape looked less like a separate project and more like the surrounding city. This is the strongest argument for the redevelopment, and it deserves to be taken seriously.
But Goetz's research also complicates the triumphal version. The settlement did not produce the scale of suburban dispersal its strongest advocates imagined. Families often remained close to Minneapolis because close is where their lives were. Housing mobility expanded choice for some households, yet the structures that shaped jobs, incomes, discrimination, and affordable-housing supply were far harder to relocate.
And then, twenty-five years later, the story comes back
Heritage Park is again in crisis. The City says Certus Financial Services is the court-appointed receiver and Property Solutions & Services handles day-to-day operations. MPHA owns the land but does not own, operate, or control the housing units; the 200 HUD-designated public-housing units are among the 440 rental units at the site.
The City's latest update says eligible households can now receive cash assistance for voluntary moves under the City-adopted relocation policy. The City continues to allocate $2.5 million for resident-support services. Earlier, on August 26, MPHA's board advanced a proposed $6 million stabilization package. These are active interventions, not a final resolution.
The historical symmetry is hard to miss. The original Hollman litigation began because government-assisted housing had become racially concentrated and physically troubled. The settlement demolished the old development in order to create a safer, more integrated, better-managed alternative. A generation later, the replacement itself requires emergency intervention, and some residents may consider voluntary transfer.
That does not prove mixed-income housing was a mistake. It proves something less satisfying and more important: no architectural form can substitute for durable institutions. A beautiful site plan still needs competent ownership, maintenance, financing, inspection, resident power, and public accountability year after year. If those systems fail, even the landscape designed as the cure can become another troubled place.
So who was right, Crump or Goetz?
The most useful answer is that the question is badly designed. Crump gives us the stronger warning about explanation. Poverty is produced through wages, employment, discrimination, education, housing markets, transportation, wealth, health, and history. Changing where poor people live cannot by itself change all those systems. When the phrase concentrated poverty becomes the whole explanation, a map can hide the processes that made the pattern.
Goetz gives us the stronger warning about implementation. People have preferences. Landlords discriminate or decline vouchers. Suburbs resist affordable housing. Relatives and childcare matter. Transit networks matter. A voucher does not create an apartment, and a low-poverty census tract does not automatically produce a job. Mobility can improve safety and neighborhood conditions for some households while still falling short of the promised metropolitan transformation.
Neither argument means place is irrelevant. Place matters enormously. A dangerous building matters. A high-crime environment matters. A segregated school matters. A long commute matters. A stable network of relatives matters too. The problem begins when one of these truths is made to swallow all the others.
The building is never just the building
In 1938 government built Sumner Field because the private market had failed low-income households. Decades later government was criticized for concentrating poor households there. In the 1990s government demolished that housing in order to disperse poverty and desegregate opportunity. A public-private partnership then constructed a mixed-income replacement. Twenty-five years later public agencies are again arranging repairs and relocation because the replacement has deteriorated.
The architectural form changed. The ownership structure changed. The planning philosophy changed. The language changed from slum clearance to public housing, from public housing to deconcentration, from deconcentration to mixed income, from mixed income to Heritage Park. But the basic question never quite moved.
Who gets a decent home, and who is responsible when they do not?
That question is considerably harder to bulldoze.
Quick check: read the housing landscape
Sources and further reading
Jeff Crump. "Deconcentration by demolition: public housing, poverty, and urban policy," Environment and Planning D: Society and Space 20 (2002): 581-596. This reading uses Crump for the structural critique of concentrated-poverty discourse and for the detailed Sumner-Glenwood/Hollman narrative.
Edward G. Goetz. "The Politics of Poverty Deconcentration and Housing Demolition," Journal of Urban Affairs 22(2) (2000); Clearing the Way: Deconcentrating the Poor in Urban America (2003); and "Desegregation Lawsuits and Public Housing Dispersal: The Case of Hollman v. Cisneros in Minneapolis," Journal of the American Planning Association 70(3) (2004). Goetz provides the political and empirical analysis of relocation and dispersal.
National policy: U.S. Government Accountability Office reports on HOPE VI; HUD/HUD USER materials on HOPE VI, Choice Neighborhoods, demolition/disposition, and public-housing preservation; federal regulations implementing the 1998 Quality Housing and Work Responsibility Act.
National comparisons: Atlanta Housing history; Chicago Housing Authority Plan for Transformation archives; New York City Housing Authority development statistics and preservation programs.
Site geography and relocation: Edward G. Goetz, Planning for North Side Redevelopment, CURA Report No. 2 (2002), and Relocation of Residents from North Side Public Housing, CURA Report No. 5 (2002); Hollman v. Cisneros, Consent Decree (D. Minn., 1995). The locator map is a schematic interpretation of the site's position northwest of downtown and its surrounding transportation corridors, not a parcel or household-movement map.
Current Heritage Park status: City of Minneapolis, "Heritage Park resident information and updates," last updated September 24, 2026; Minneapolis Public Housing Authority, "MPHA Board Advances $6M Investment for Heritage Park," August 26, 2026. Treat the current update as a dated status report, not a claim that the policy outcome is settled.
Images: Library of Congress Historic American Buildings Survey; Minnesota Digital Library/DPLA; Jewish Historical Society of the Upper Midwest via Wikimedia Commons; Wikimedia Commons public-housing collections.
Comments
Post a Comment