Distressed, dilapidated, deserted, and depressed, San Francisco’s Hunters View public housing project was seen in 2008 by then-mayor Gavin Newsom as one of the worst examples of concentrated poverty in public housing. By the time the mayor proposed the city’s own version of the federal HOPE VI public housing redevelopment program, called HOPE SF, only about 150 of its 267 units were still occupied, and many of the others were uninhabitable due to mold, deferred maintenance, and sewer problems. In the broader area, over 40 percent of the people live below the federal poverty level of $22,300 for a family of four; in Hunters View, annual incomes average $16,000, and only 20 percent of residents are fully employed. Illegal activity among rival gangs plagued the area and the project.
Anne Torney and Daniel Solomon, architects with the San Francisco office of Mithun | Solomon, designed a complete transformation of the area. They think the source of the area’s distress comes not from its location, but from its isolation and planning. In fact, the project sits on an elevated 22-acre (8.9 ha) bluff in southeastern San Francisco overlooking the Hunters Point Shipyard, India Basin, and San Francisco Bay. Downtown is visible toward the northwest.
But the history of the site and its development led Hunters View in a very different direction from what might have been expected given its favorable topography and views.
The superimposition of the grid over San Francisco’s hilly topography is the single most compelling feature giving the city its character.
The site is named after the Hunter brothers, agents of early real estate developers who in 1849 built a dairy farm on the land. The area below became home to an array of slaughterhouses and related facilities for tanning leather, making fertilizer and tallow, and processing wool. By 1867, shipyards dominated the waterfront, and Bethlehem Steel Shipyards built barracks on the site to house 30,000 workers.
In 1943, when the shipyards were taken over and expanded as Hunters Point Naval Shipyards by the federal government during World War II, thousands more workers, a large proportion of whom were African American, occupied the barracks housing, and the larger area had become known as Bayview–Hunter’s Point and was populated largely by minorities. A race riot inflamed the area in 1966, and closure of the naval shipyard and associated shipbuilding facilities in 1994 led to widespread unemployment.
While economic and social issues might have doomed the area no matter how it had been designed, the physical layout and facilities crystallized its marginalization and segregation. To build with minimal excavation and expense, the barracks had been designed to follow the curvilinear lines of the topography. There were few streets, and the long barracks buildings blocked almost all views. One street provided the only two ways into or out of Hunters View, and even that roadway severed the area into two culs-de-sac, further isolating the area.
After the war, employment declined, and in 1954 the San Francisco Public Housing Authority converted the barracks into public housing, with infill buildings following the same layout. Now the area was not only socially segregated; it was even more physically segregated into an institutional ghetto of racial minorities.
To address this morass of problems, Solomon and Torney were hired in 2008 while at WRT/Solomon ETC to conceive a development plan for Hunters View, which needed a design solution that was larger than the site itself.
The design team chose first to distill the essence of what made San Francisco unique and determine how and why Hunters View had violated each of the principles upon which the city had been based, thereby ensuring its isolation.
Thanks to William Penn in Philadelphia in 1682, James Oglethorpe in Savannah, Georgia, in 1733, and Thomas Jefferson with the Northwest Ordinances of 1787 and 1789, the United States, unlike other countries, has the distinction of having the layout of its cities during the 19th century based largely on a rectilinear street grid.
The San Francisco street grid dates to the 1839 plan of Swiss ship captain and surveyor Jean-Jacques Vioget laying out the city on a north–south, east–west grid without regard to its topography. Subsequent plans extended the grid, except for its inflection south of Market Street. Torney and Solomon believe that the superimposition of the grid over San Francisco’s hilly topography is the single most compelling feature giving the city its character, so brutally violated by Hunters View.
What does the grid do to create character? Except for places like Hunters View, the streets interconnect all neighborhoods. Even where the streets are too steep for vehicles other than cable cars to pass, public stairways continue to connect the streets. Moreover, the orientation along the cardinal directions over the peninsula provides view corridors east and north to San Francisco Bay and west to the Pacific Ocean. Buildings largely block such views at Hunters View.
San Francisco’s narrow 25-foot-wide (7.6 m) lots essentially dictate party wall buildings, which create continuous street walls that not only distinguish the public realm from the private, but also put eyes on the streets, rendering them safer. The relatively narrow streets, reinforced by minimal or no building setbacks, also provide some public, shared parking that also is in full view from the buildings and safer to navigate.
The narrow lots, build-to lines near the street, and street slopes together create an urban fabric of low-rise buildings stepping up and down the hilly streets. Larger lots would have led to lower density and a disconnected urban fabric. The highly efficient and rational system of street grids and narrow lots leads to more efficient land use and higher densities, thereby reducing the cost of infrastructure connection on a per-unit basis.
San Francisco also has a tradition of placing its parks on hilltops, maximizing the shared public use of the view-rich topography. The tight building mass around them at lower levels permits those parks to anchor neighborhoods.
Torney and Solomon concluded that Hunters View violated each of six San Francisco urban design principles:
Street grids superimposed irrespective of hills. There is no grid and no connection to the Bayview grid to the west. And the few internal streets follow the land’s contours, not any grid.
Street walls defining the public realm and providing eyes on the street. Buildings in Hunters View are scattered and twist in all directions around the contours.
Stairs connecting steep streets to each other. No such stairs exist at Hunters View, and the whole project connects to adjacent areas at only two places on one street.
Narrow parcels creating low-rise density supervised by eyes on the street. There are no separate parcels in the institutional environment, and the buildings are large and long and lack a fine-grained definition of units and building masses. Yet the overall density is only 12 units per acre (30 per ha), a fraction of what such a site could hold.
Streets and stairways serving as view corridors. The indiscriminate placement of long barracks buildings without regard to views results in most of the views being blocked.
Hilltop parks serving as public gathering places. The considerable open space in Hunters View is amorphous and treated as leftover areas between buildings. Certainly, it is not aggregated to take advantage of the views. Nor does it serve to anchor the neighborhood as shared public space surrounded by supervising eyes from neighboring housing.
Two major parks were placed to take advantage of the views from the hilltops: Promontory Park in the northwest corner of the site is placed on a hilltop facing the clear view of downtown San Francisco to the northwest; Panhandle Park is a linear park running east-west through three blocks of Fairfax Avenue. In addition, there are four miniparks. (Photo credit: Mithun l Solomon)
The Mithun | Solomon plan to redevelop Hunters View, currently completing its first phase, derives from rebuilding the infrastructure and buildings according to these six San Francisco urban design principles. The site is being divided into 18 new city blocks by four north–south streets and three east–west streets. The blocks are of somewhat varied size primarily because it was necessary to bend the extension of Fairfax Avenue through the site in a northwest–southeast direction.
Two major parks are placed to take maximal advantage of the views from the hilltops. One, Promontory Park, is placed on a hilltop facing the clear view of downtown San Francisco to the northwest. It is also the visual terminus of Fairfax Avenue where it bends. It has both an upper level, which is accessible to people and cars, and a lower level, where extensive usable grassy green space is connected by crosswalks to residences across the street.
The other major green space park is Panhandle Park, running east–west through three blocks, bordered on either side by Fairfax Avenue, which splits into two tree-lined streets. Those streets also provide extensive shared public parking for those not residing close to the park. There also will be as many as 816 off-street parking spaces.
There are also six miniparks, including one each at the north and south ends of New Street and one each inside blocks 5 and 6. Hudson Overlook, connecting to Hudson Avenue on the southeast, and the West Willis minipark create connections on the east and west sides of the site.
Despite building a gridded network of seven new streets and nine new parks, the developers are more than tripling the density of the site, to 740 to 800 units from 267, including the replacement of all public housing units on a one-for-one basis. This achieves a density of about 35 units per acre (87.5 per ha) and overcomes a frequent criticism of the federal HOPE VI program, which often created fewer units of public housing than were on such sites before redevelopment.
The project is intentionally split into three phases. The first, encompassing the eight acres (3.2 ha) in the northwest portion of the site, was chosen to give a fresh start at the edge untainted by the stigma of the earlier troubled project. The fact the public housing project was half empty was used as an advantage for the HOPE SF program because it gives those currently living in the project the right to stay there during construction and then relocate within the new redeveloped site.
This first-phase area surrounds the extension of Fairfax Avenue into the site, linking it to the neighborhood to the west and to the Third Street light rail. The housing is planned to be split between 120 rental units and 160 for-sale units. To date, $84 million has been spent building 107 units for both public housing replacement units and new affordable rental units—as well as a park, infrastructure, utilities, sewers, and pedestrian walkways. In an effort to diversify the architecture, the project uses more than one architect. Daniel Solomon designed Block 4 of Phase I as stacked flats that appear to step up the hillside with staggered window heights; Paulett Taggart Architects designed townhouses stepped up the hillside on blocks 5 and 6.
The second phase lies immediately south of Phase I on four acres (1.6 ha) equally split between rental and ownership housing. Plans call for daycare and elder care facilities, with links to nearby schools and bus lines. The third phase, which will be on the remaining eastern portion of the site, will contain 23 percent more owned than rental units. It is planned to surround the new Panhandle Park, a major recreation and place-making feature of the redeveloped project. Buildout of the project is expected to occur as market conditions and financing allow.
A project of this scale, which is expected to cost more than $450 million, requires a carefully crafted set of public/private partnerships. Leading the team on the development side is San Francisco–based developer John Stewart Company, founded in 1978 to develop and manage affordable housing throughout California and now the state’s largest affordable housing manager, with 31,000 multifamily units under management. Development partner Devine & Gong Inc. (DGI) provided tax credit and real estate financial services for the team, which also includes Ridge Point Non-Profit Housing Corporation, a nonprofit organization founded in 1968 to develop and operate low- and moderate-income housing and community facilities in Bayview–Hunters Point.
The public partners were primarily the San Francisco Housing Authority (SFHA) and the Office of Community Investment and Infrastructure, as the successor to the San Francisco Redevelopment Agency (SFRA), which had been dismantled along with other redevelopment agencies as part of the statewide austerity measures taken during the Great Recession. The SFHA operated Hunters View, with rents set according to resident income based on 30 percent of monthly adjusted income, with a $25 minimum rent. The San Francisco Mayor’s Office of Housing (MOH) and the California Department of Housing and Community Development (HCD) facilitated the public side of the project, along with the U.S. Department of Housing and Urban Development and Enterprise Community Investment, which straddles the public and private sectors.
As is the case with many affordable housing projects, the tapestry of financing requires contributions from the philanthropic community. In this case, the Ford Foundation, Fannie Mae, Stewardship Council, Bay Tree Fund, and Local Initiatives Support Corporation contributed to the project, as did Enterprise Community Partners and the San Francisco Foundation.
The financing structure included a $30 million infill infrastructure grant from the HCD for the overall project, $12 million of which was used for Phase I, as well as $6 million in federal stimulus funding and $2 million in other funds through the SFHA, $14 million from the SFRA, $18 million from the MOH, and about $30 million in low-income housing tax credit equity using both the 4 percent and 9 percent tax credits. The Federal Home Loan Bank’s affordable housing program was also tapped.
The John Stewart Company plans to find developers to undertake two market-rate blocks (blocks 2 and 3) that are ready to be built on as soon as developers perceive adequate private demand for units on the view-rich site. These blocks are strategically located at the northern portion of the site, just east of the extension of Fairfax Avenue and north of blocks 4, 5, and 6 that were just redeveloped in Phase I. Directly across from the new Promontory Park and adjacent to the bluff, they have views of Sutro Tower and downtown San Francisco, as well as across San Francisco Bay to the Bay Bridge and Yerba Buena Island.
One need not enter the former public housing site to enter these blocks, giving the eventual private builder a more favorable environment in which to pioneer new, market-rate units. Jack Gardner, president and chief executive officer of John Stewart, says he thinks that homebuilding contractor/developers can control costs enough to bring the units to market in the range of $400,000 to $500,000 per unit, which he considers an affordable entry point for San Francisco, especially for new units with views, parks, and parking.
Gardner estimates that such developers could afford to pay more than $50,000 per unit in entitled and serviced land costs, which he says is only half what other developers are paying, even for rental land, in other parts of San Francisco. While he expects stacked-flat condominiums to be the form developers prefer, townhouses at lower density stepping up the slopes may also be feasible as a first market-rate project. Gardner points out that mixed-income housing has been successfully integrated into projects in Seattle; Chicago; Portland, Oregon; and other locations. John Stewart’s similar project in the North Beach section of San Francisco is anchored by a Trader Joe’s grocery store and has blended well into its neighborhood.
Though Hunters View is a more challenging site because of its history, Gardner believes mixed-income housing is feasible here, too. Others fear that, because of the Great Recession, the inability to build market-rate units simultaneously with subsidized units may doom the integration of mixed-income housing.
Market-rate housing cross-subsidizes the affordable housing to some extent, Gardner notes. While a $17.5 million land-cost contribution for the 350 market-rate units initially forecast may seem minor in a project this size, that contribution pays for infrastructure costs and generates a city tax base that can grow over time. And the units help protect the value of public investment in infrastructure and development of all the affordable and public housing units.
Key to the project is a return to the city’s characteristic urban design principles.
Just over half of the project is planned for households with incomes from less than 10 percent of area median income (AMI) to 120 percent of AMI. Of the 800 units entitled, the 741 currently planned include 267 public housing units, 83 rental units targeted at households making up to 50 percent of AMI, 59 for-sale units for households making up to 80 percent of AMI, 17 Habitat for Humanity units, and 315 market-rate, for-sale units. Most of the market-rate units are to be built in later phases as the plan is realized.
The mixed-income nature of the project on a formerly troubled site represents a new solution to a festering urban problem. As stated by Gardner, “This is the new model for the revitalization of public housing.”
Though the mixed-income, grid-based plan might be emulated elsewhere, at a gross cost per unit of more than $607,000 for 741 units (including the infrastructure cost of about $70 million, or about $95,000 per unit) or $563,000 per unit if 800 units are built, it may be hard to find funding for the subsidies required. Solomon says that extra costs were incurred because of the topography, which includes slopes of up to 20 percent. Gardner concurs, also noting the extra costs incurred to retain all the residents on site during construction, along with the costs associated with making the affordable units indistinguishable from the market-rate ones and providing larger units with up to five bedrooms.
Nevertheless, the increased spending to overcome decades of neglect and reintegrate Hunters View with the city to create a sustainable mixed-income community may be justified by the economic, social, and environmental returns—as well as the creation of human capital through lives that are helped to become more productive.
Hunters View may serve as a model of how to incorporate a large swath of isolated property in a troubled area into the rest of a city by first distilling the essence of urban design principles unique to that city and then extending them to the site to reweave the urban fabric. Even a place that was once as isolated and grim as Hunters View can be transformed into a living, functioning part of the larger grid of a humanly scaled city.
William P. Macht is a professor of urban planning and development at the Center for Real Estate at Portland State University in Oregon and a development consultant.
(Comments about projects profiled in this column, as well as proposals for future profiles, should be directed to the author at [email protected].)