Framing the Future of Downtown San Francisco

How we can bring more jobs into the region's most transit-rich employment center

Urbanist Article

Just over 20 years ago, San Francisco approved its first-ever comprehensive plan for downtown. Responding to the growth boom of the 1970s and early 1980s, the Downtown Plan presented a vision that strengthened our historic job core, limited new high-rise development to the north and west, protected neighboring historic areas such as Chinatown, Jackson Square and Union Square, and focused new office growth south into SOMA.

The vision encouraged housing around downtown and encouraged transit ridership through limits to parking downtown. This vision of the Downtown Plan largely has been fulfilled.

Today, San Francisco's downtown is at a cusp of another change, which requires careful thinking and planning to make sure that we create an appropriate downtown for our economic and social needs.

The driving tension in our downtown is between whether we will maintain our strong office core or we will become a downtown characterized by more mixed-use residential and entertainment uses. These tensions are best expressed as the conflict between the vision of maintaining a "Central Business District" versus moving more toward the creation of a larger "Central Social District." Today, we already have the beginning of a social district that extends from the San Francisco Museum of Modern Art past the Westfield shopping center and through Union Square. New buildings such as the Four Seasons on Market Street (in a site that was originally proposed for office use) are indicative of this emerging Central Social District.

While both the CBD and CSD can coexist within a large downtown, the tension emerges when different uses begin to compete for the same space. This is happening in San Francisco, particularly along Mission and Howard streets.

SPUR is taking a close look at our historic and evolving downtown to help determine what "the future of downtown" should be like. With this paper we both launch that project and frame some of the key question we must ask, including:

  • Will downtown's new residents work in San Francisco, or will our downtown become a bedroom community for Silicon Valley?
  • Is there enough land for more office buildings to support continued job growth?
  • Can downtown become a viable neighborhood if many of its residents are working elsewhere during the day?
  • What is the impact on our job base if we make residential development a higher priority than office uses?
  • Can our downtown truly compete with suburban office centers in the Bay Area as a location for knowledge industries?
  • Is the current growth of technology firms in San Francisco indicative of a major new force in our downtown, or a temporary trend?
  • What is the role of other downtowns - such as Oakland's - in housing our workforce or in providing jobs for our residents?
  • What policy tools best shape residential demand or business location decisions?

This article presents both the Central Business District and Central Social District models as distinct and competing visions for the next phase of downtown. We argue that the CBD model is essentially the current vision as expressed by the Downtown Plan of the 1980s, while the CSD is a new mode based on a renewed interest in downtown living and a belief that suburbs will continue to be more competitive for business growth than central cities. Despite their attractiveness, both models end up failing as complete solutions to the needs of our city. These failings must be resolved as San Francisco plans and prepares its new downtown.

Downtown San Francisco today: the Central Business District
The Central Business District is the traditional high-rise downtown office district found in the center of most major cities in North America. The CBD model maintains an employment core, with housing, entertainment and tourism uses located in neighborhoods around the edges of the CBD. It requires a regional transit system to bring a wide range of workers from the suburban areas into the downtown core.

The primary benefits of the model:

  • Labor market: Access to wide range of knowledge-sector and support-services jobs for workers throughout the region. Further, the presence of large corporate employers provides many middle income jobs that do not require college education.
  • Environmental: Limits to parking and centralized transit allow a high percentage of workers to commute without their car. High density promotes walkability within downtown.
  • Economic: High concentration and proximity allows greater interaction among businesses within the core and facilitates frequent face-to-face communication



Yet the same success factors are also failings. By focusing on being a nearly exclusive office district, the CBD crowds out other uses and places unnecessary limits on its transportation infrastructure based on a one-way commute. Further, recent economic growth has been more likely to occur in suburban office parks than urban downtowns.

First, by maintaining a nearly exclusive office core, the CBD typically shuts down at 6 p.m. Although surrounding neighborhoods in the broader "downtown" have activity, the CBD model inevitably results in an empty core in the evenings. This means that downtown has limited services for the few residents who do live there, or for workers who stay late. And as a single-use office district, it is particularly vulnerable to swings in the real estate market, as the emptiness can seem even more profound during economic downturns.

Second, the one-way commute pattern makes inefficient use of our transit infrastructure. BART trains enter downtown full and return to the East Bay nearly empty. And with limited capacity in the Transbay Tube (during the dot-com boom, BART reached full capacity and many commuters shifted to AC Transit), continued job growth in the downtown without comparable nearby housing growth leads to more commute trips for a system lacking much room for expansion. This leads many to claim that achieving a greater jobs/housing balance is a desirable goal for downtown San Francisco.


Third, the CBD vision of the office center was in fact a replacement for the loss of an earlier, more mixed-use downtown (think of the downtown department stores of the 1950s and 1960s). The CBD responded to a concentration of corporate services in key global regions, and benefited cities such as San Francisco, New York and Boston. But some argue that today's urban centers can no longer compete with more affordable suburban environments, which suggests that cities must find an alternative to remain successful.

Despite these challenges of the model, San Francisco still retains a strong CBD, with among the highest commercial rents in the region. Yet as jobs continue to grow at a faster rate in the suburban areas and our share of the region's office market and jobs continues to decline, we must reconsider our competitive advantage as a city. Further, the desire for a thriving center at all times of day befits a global city such as ours. Many thus consider the Central Social District a viable goal and alternative to the current CBD model.

A possible alternative: the Central Social District
The Central Social District - a phrase attributed to the Urban Land Institute's William Hudnut, who learned it from a business person describing the transformation of downtown Chicago - is a vision of a downtown with increasing numbers of residents as well as new restaurants, shopping and entertainment opportunities. While office uses would remain, the goal of the CSD is to create a mixed-use, livable, 24-hour downtown neighborhood. The CSD is associated with cities such as Vancouver, Paris and Barcelona, where the center is a place of interaction, culture and living.

The CSD idea is predicated on two notions. First: Central cities cannot compete with suburban locations for jobs and businesses. Future job growth will continually favor suburban areas at the expense of traditional CBDs. This suggests that cities need to focus on their core advantage in historic buildings, entertainment and a dense, walkable core, and also expand activities that range from shopping to theater to music. Second: Reducing the jobs/housing imbalance through adding housing in the urban core is a way to combat urban sprawl.

This second notion assumes that San Francisco has more than its share of jobs but lacks sufficient housing to support its workforce. Reducing the imbalance emphasizes building more housing for our workforce while being less concerned about expanding the already large job base.

The primary advantages of the CSD:

  • Reinforces a core competitive strength of center cities such as San Francisco - as a destination for experience and entertainment. Given the high cost burden of San Francisco (rents, taxes, labor costs) and the limited political support for economic growth, it is likely that we will continue to see a greater share of the region's jobs continue to locate in the suburban counties of the Bay Area.
  • Responds to the continued demand for residential production in San Francisco. The CSD then becomes a strategy for capturing a greater share of the overall population growth (and locating those residents downtown).
  • Provides housing nearer to work and gives downtown more life after the offices close.
  • Recognizes that downtown is one of the few districts citywide where high-density housing is politically acceptable. If we wish to add significant units to our housing stock, building high-rise downtown neighborhoods is perhaps the most likely approach.


While the Central Social District model is appealing, there are significant downsides to it, particularly for San Francisco. These include limits to economic growth, potentially decreased job quality, reduced densities, increased rents and more cars.

First, economic growth in the CSD model may be diminished as the remaining sites for office buildings become used for new residential, retail or other non-office uses. Already, much of the areas in and around the Transbay Terminal are proposed for residential development. Given that downtown's growth is constrained to the north, east and west, the only areas for growth of the office district are into SOMA and up Market Street. Further, as millions of square feet of office space convert to residential, there is a possibility we could actually reduce the total number of jobs within our downtown.

As previously noted, one of the main advantages of a concentrated office downtown is access to a range of jobs for workers from through the region. If this job concentration is reduced, through downtown increasingly becoming a social district, the new jobs will become further spread throughout the region. This means that these jobs won't be as easily available to the broad range of the workforce who can currently arrive in downtown San Francisco by way of existing transit infrastructure. This trend thus reduces the competitive advantage of our downtown as a center of commerce and harms the ability of local residents to access good jobs. There also is a potentially negative fiscal impact in the conversion of office space to residential use, given the importance of the revenue from San Francisco's payroll tax and the City's lack of an income tax.

Second, as we increasingly convert to a Central Social District, the quality of jobs also decreases. Average wages in the "experience" industries (accommodations, performing arts, retail, restaurants) most associated with a CSD are significantly less than in the knowledge industries (finance, IT, professional services, insurance) that cluster in a CSD. Nearly 70 percent of the experience-sector workers are in jobs that pay less than $40,000 per year (and have either high school or some college education). For workers with a bachelor's degree, only 3 percent of experience sector jobs pay greater than $70,000. In contrast, about 35 percent of growing knowledge sector workers with bachelor's degrees earn more than $70,000 per year. If we prioritize an economic development pattern with significantly lower earnings per worker, we produce demands for other needs such as affordable housing.

Third, as we increase the residential component of our downtown, the density may decrease. This is because residential dwellings typically are lower density per square foot than offices. Average office workers take up about 250 square feet while small apartments (for one person) often are greater than 600 square feet. For example, a 300,000-square-foot building might support more than 1,200 office jobs but only 500 residents. The residential building may support only minimal jobs in property management and security.

Fourth, as office is converted to residential, rents on remaining space will rise and have a disproportionately negative impact on startups and nonprofit organizations. This trend occurs because residential conversions tend to happen to older office buildings, which are historically Class B space (Class B buildings are typically well-maintained but were constructed before 1965). These buildings offer space at rents almost $11 less per square foot than Class A buildings in San Francisco and thus are more likely to house space for nonprofits and startups. If that lost space is ever added back to the overall supply, it is likely to be as Class A office space at higher rents. Residential conversion typically results in rising rents for remaining office buildings if demand remains constant.

The conversion trend is growing. While more than 700,000 square feet of office space already has been converted from office to residential or hotel uses, another 1 million square feet are in construction and more than 3 million are approved or proposed. If much of this space is ultimately converted from office to other uses, it will be equivalent to the loss of more than 10 Transamerica pyramids.

Finally, car usage in a CSD may increase. As employment concentration declines, many of the new residents will out-commute to jobs throughout the region. Since residential construction is currently providing more parking spaces per capita than office uses, this is likely to continue. The increase in driving is counter-intuitive to the notion of living downtown. However, given that many jobs throughout the region are in places that do not permit easy transit access, many of the new downtown office workers will drive from their high-rise homes to the jobs in the suburban regions. This trend is already happening in Chicago and New York, particularly for residents whose jobs are outside of the city. Today, commuters are much more likely to drive to jobs that are more than a 10-minute walk from a BART station. If residential uses emerge on potential office sites closer to Market Street, new office spaces may be pushed to areas that are less accessible by transit, further increasing the demand for driving to these jobs.


There are no doubt drawbacks to both a pure Central Business District or Central Social District model of a downtown. Taken to their logical conclusions, both visions fail to meet the needs of our dynamic and changing city. The way we resolve this ongoing tension between CBD and CSD is as of yet unclear.

Further, many of the factors that influence our downtown are external. Demand for downtown living is driven by demographic changes and social forces that shape people's preferences for where they want to live. While individual firm location is still most influenced by the distance to the chief executive officer's home, the scope and size of our economic base results more from emerging technologies, changing global investment patterns and competitive pressures from other metropolitan regions.

Given these factors, our ability to shape our downtown is inherently limited. Zoning, tax policy and investment in the public realm are three of our most powerful tools. We must also consider changes in other regional centers. Downtown Oakland could become a larger job center, thus strengthening the "reverse commute." East Bay and Peninsula communities could increase the production of housing near BART and Caltrain stops and thus facilitate greater transit commuting. Understanding these scenarios is part of helping determine the role of downtown San Francisco in the Bay Area.

Ultimately, the measure of success for our downtown should be based on outcomes:

  • Labor market: availability of wide range of jobs for San Franciscans and workers throughout the region.
  • Environmental: concentration of development near transit to allow residents to commute to work and get around their home community on transit/without a car.
  • Social: increase face-to-face contact among a diverse population.


Both the CBD and CSD help achieve these outcomes.

This article presents a preliminary set of questions about the future of our downtown. The real work will take place in the upcoming year. Our goal is to help inform the decision we make today to shape the downtown we need as a city.


Egon Terplan is SPUR's Economic Development and Governance Policy Director. Lisa Bell is a graduate student at the Goldman School of Publicy Policy at UC Berkeley.