This is the third in a series of SPUR policy reports on ways to increase the supply of housing in San Francisco. The first paper, published in March of 1998, proposed that San Francisco plan for a new mixed-use neighborhood in the Central Waterfront. The second paper, titled "Zoning for More Housing," proposed a set of zoning changes that will encourage greater housing production city-wide. In this third paper in the series, SPUR takes on the complicated issue of increases in housing costs associated with storing automobiles.

Introduction

Parking spaces are expensive to build, especially where land values are high. Typically, a parking space adds $30,000 to the construction cost of a new housing unit in San Francisco. If we can find a way to build less parking, we will see both lowered housing prices and more efficient use of urban land for the city as a whole.

In every American city, vehicle ownership has increased more than population over the last several decades. The dilemma of how to deal with our cars confronts all cities with greater urgency as each year goes by. In San Francisco, where the 19th century street grid puts absolute physical constraints on the ability of people to move around by car, we have made a virtue of necessity and decided to promote a Transit First policy. SPUR, in fact, first proposed the Transit First policy in 1972. It was adopted as part of the City Charter two years later. It is undoubtedly in part because of this policy that San Francisco has retained so much of its historic fabric and pedestrian friendliness. Given the conflict between maintaining our compact, Transit First city and our society's increasing dependence on the automobile, we have no choice but to work out a compromise.

There is a lot at stake on our decisions about parking, in addition to the cost of housing. If we increase the amount of urban real estate we devote to our parking needs, we push activities farther away from each other, thereby forcing people to drive for even more daily needs. Increased parking becomes a vicous circle, where the more parking we build, the more people have no choice but to drive. Access-by-proximity, the great advantage that belongs to city dwellers, depends on a compact, intimate mingling of people and land uses. We cannot simultaneously provide parking spaces for each person at each destination, and still be a city that enjoys access-by-proximity.

When we talk about the ability to get around without a car, we are getting near the heart of what makes San Francisco different from most other American cities. San Francisco's urban form is compact rather than spread-out; human-scaled rather than highway-scaled. Even after these many decades of increasing car-orientation, we are still a city where people can live their lives without cars. Consider some of the following numbers, based on 1990 census data:

1. Fewer than 40% of San Francisco residents commute to work in a single-occupancy vehicle.

2. 31% of San Francisco households have no cars and 42% have just one car.

3. In the most compact, walkable neighborhoods, auto ownership rates are below 0.5 autos per household.

Our policy and urban design decisions today will determine the way future generations of San Franciscans live. If, over time, we remake our city to accommodate cars at the expense of walkingand transit-oriented environments, people who live here will drive more. If, over time, we create more pedestrian-accessible land uses and invest in our transit system, people will drive less.

SPUR recommends reducing parking requirements in certain locations and under certain conditions. These recommendations would lower housing costs, increase housing production, and improve San Francisco's quality of life by:

• Promoting more efficient use of our limited land by allocating more space on a given lot for housing and less for automobiles

• Allowing residents to choose to forego the costs of having a parking space

• Making it easier for people to use cars when they need them, without having to own (and park) their own car

• Making the approval process for transit-friendly housing developments more predictable

Outline of the Paper

SPUR's recommendations are summarized here, followed by a discussion in the body of the paper. As a first step towards rethinking parking requirements, SPUR highlights two issues that deserve intensive study by the Planning Department:

1. Is there a way to lower minimum residential parking requirements in neighborhoods where it is convenient for people to get around without a car? We suggest serious consideration of the possibility of defining "Transit-Intensive Areas" in the Planning Code.

2. When residential developers build and market parking as a separate parcel from the dwelling unit, thereby "unbundling" the cost of parking from the costs of the home, should the parking requirements be reduced? Our knowledge of market economics tells us that when the true costs of parking are made visible in the housing transaction, we can expect a certain percentage of people to opt to forego the cost of a parking space.

Next, SPUR recommends five sets of policy and Planning Code changes that promise to either rationalize or reduce a portion of housing costs associated with parking.

1. Encourage convenient pay-per-use automobile services such as car-sharing organizations.

2. Reduce minimum parking requirements for housing that serves populations that do not have a high frequency of car ownership.

3. Allow reductions in off-street parking requirements when a new curb cut would take away as many on-street spaces as the private parking spaces it would create.

4. For historic buildings, allow exemptions from new off-street parking requirements triggered by a change in use or occupancy.

5. Adjust the Planning Code's parking space size requirements to conform to the Department of Parking and Traffic's size requirements and to allow for a greater proportion of compact parking spaces.

It should be clear that these housing policies must go hand-in-hand with appropriate transit policies, both locally and regionally. The ability of people to live without daily use of cars depends on a well-functioning transit system and on good land use-transportation coordination. In this paper, we confine ourselves to a discussion about residential parking requirements as an aspect of San Francisco's housing policy, but our goals of reduced auto dependency and lowered housing costs can only be achieved by simultaneously increasing our local and regional investments in transit and other mobility options.

Concept # 1 for Planning Department Study

It is the policy of San Francisco to cluster higher-intensity land uses such as dense residential developments near major transit lines. In areas that are well-served by transit and which are located near local retail services, can San Francisco define a set of "Transit-Intensive Areas" in the Planning Code and reduce the minimum parking requirements in these areas?

It almost goes without saying that when people find it easy to get around without a car, they drive less and tend to own fewer cars. San Francisco neighborhoods range in their auto ownership rates from under 0.5 per household to more than 1.5 per household. (See accompanying map, San Francisco Auto Ownership Rates.)

Perhaps we can find a way to pay attention to this difference, rather than treating all locations the same with respect to their parking requirements. In neighborhoods that are walkable and close to major transit lines, can we reduce, or perhaps eliminate altogether, the minimum parking requirement for new residential construction?

"Transit-Intensive Areas" are defined by two characteristics: they have commercial districts (so residents can walk for many daily errands) and they lie on major transit lines (so residents have easy transit access to destinations throughout the city and region). For the purpose of amending the Planning Code, Transit-Intensive Areas would need to be rigorously defined. As a starting place, SPUR suggests that any area that meets the following criteria be considered a Transit-Intensive Area:

1. Any area within 1,000 feet* of a Transit Center (as defined in the General Plan)

2. Any area within 1,000 feet of the overlap between a transit corridor and a commercial or mixed use district (transit corridors are herein defined as the General Plan's Transit Preferential Streets or Rail Lines; commercial and mixed use districts are defined in the Planning Code)

3. Any area within the Downtown Parking Belt (an area within downtown designated by the General Plan as having only short-term parking)

These neighborhoods are great urban locations, filled with human activity and scaled to pedestrian use. They are the kind of neighborhoods that evolved historically in San Francisco. In areas such as these, where the infrastructure is in place to provide good transit and pedestrian amenities, it makes sense to give housing developers the flexibility to provide as little parking as they think they need to market their units. This would be achieved by either significantly reducing minimum parking requirements, or perhaps setting a maximum ratio of one parking space per unit with no minimum ratio.

In order to make this work, we would need to address the issue of on-street parking,, meaning the problem of "parking spillover" where people who don't have their own off-street parking space would crowd the on-street spaces. One way of solving this would be to make the Transit-Intensive Areas into residential parking permit zones. It might even be possible to charge market rates for any non-permit holders to park on the street, and to use the monies for neighborhood improvements. If a Transit-Intensive Area were a parking permit zone, the Planning Commission would have a tool at its disposal for helping to fit new developments into existing neighborhoods: as a condition of approval, the Commission could make residents of new units that do not come with parking spaces ineligible for parking permits. A person who would choose to live in a unit created under this new provision, who would benefit from the lowered cost of housing, would simply not be allowed to park on the street for extended periods.

As San Francisco adds to its transit system over time, new neighborhoods would come under the "Transit-Intensive" designation, and the parking requirements could be adjusted accordingly. The goal would be to enhance neighborhood character and walkability, which means that the specific design solutions would need to be tailored to each unique neighborhood.

Concept # 2 for Planning Department Study

When residential developers build and market parking as a separate parcel from the dwelling unit, thereby "unbundling" the cost of parking from the costs of the home, should the parking requirements be reduced?

Some of the costs associated with automobile ownership are hidden in that they are inseparably bundled with something else. A "free" parking space which comes with a home, workplace, or shopping center is not really that. It's just bundled into the cost of the house, the employer's overhead, or the retailer's fixed operating costs.

A 1997 study for the San Francisco Planning Department (using 1996 data on housing prices) found that housing units without parking spaces were significantly more affordable and sold significantly more quickly than those with a parking space. On average, the value of an off-street parking space for a single family home was $46,391; for a condominium unit the value was $38,804. Approximately 20% more households could qualify for mortgages for units without off-street parking than can qualify for units with parking. Most importantly, for people who wonder how great the demand is for housing without parking, single family units without parking sold five days faster and condominium units without parking sold 40 days faster!

Bundling parking spaces into the cost of housing increases car ownership rates because it prevents the proper functioning of a market for parking. If we made the added cost of providing parking explicit in the transaction between developer and owner or between owner and renter, the market would determine how high the demand for parking really is for each location. This could be done by creating parking units on a separate deeded or leased parcel. Parking units would be designated in the same fashion for rental units.

Ideally, this pattern would encourage multiple developers on the same block to combine their parking units into one structure. While this would devote a piece of land to a parking garage, not always a desirable goal, it would have the benefit of using city land more efficiently and lowering the construction cost of housing by agglomerating parking into one location. As a significant side benefit, it would reduce the number of curb cuts encroaching onto the sidewalk.

Based on the expectation that residents who do not need parking will choose not to purchase it, the Planning Department should conduct a study to determine the proper reduction in parking requirements when parking is provided as a separate parcel. We stress that these would be reductions in the minimum parking requirements; more parking could still be constructed according to the developer's understanding of market conditions.

As a related measure, it might be appropriate to amend the Planning Code to allow the lease and sale of parking spaces to off-site residents. While this practice takes place "under the table" today, it should be encouraged as part of the creation of a true market for parking.

Recommendation # 1

Encourage pay-per-use car services such as car-sharing organizations.

Car-sharing is a practical alternative to private car ownership that has been pioneered in Switzerland and Germany over the last ten years. Car-sharing cooperatives provide members with pay-per-use vehicles, located throughout their city. Members book their vehicles by phone, walk to a nearby garage or lot, and drive off; they get billed by mail at the end of the month.

Car-sharing organizations reduce over-dependency on cars, while still providing access to them when needed. Economists have long pointed out that individual car ownership encourages over-reliance on the automobile because so many of the costs are fixed costs. Insurance, financing, and registration are paid up-front, and are constant, no matter how much or how little one drives. Car-sharing converts these costs into a variable time and mileage charge, so that people pay based on the amount they drive. When the fixed costs of car-ownership are made visible in each use of the vehicle, people have been shown to switch many of their trips to foot, bike, or transit, using a car only when it is truly "worth it."

The idea holds great promise as part of the solution to the parking dilemma for urban areas such as San Francisco. Here, where people have excellent access-by-proximity, there is literally not enough space for everyone to have a dedicated personal parking space at home and at work. By making more intensive use of each car (and each parking space), car-sharing cooperatives could greatly reduce the total number of vehicles that need to be stored in San Francisco.

SPUR endorses an effort now underway to establish car-sharing in San Francisco (see sidebar). There are also some important roles for the City in helping the concept take root and making sure it leads to lowered housing costs.

Provide or help obtain funding for pilot projects and for the evaluation of these projects. Based on the European (and more recently, the Canadian) experiences, car-sharing organizations can grow to be financially self-sufficient, with start-up funding from the government.

Dedicate spaces in city-owned garages for car-sharing services.

Change the Planning Code to allow developers to build fewer parking spaces if they provide access to a shared fleet of vehicles. This shared fleet could be provided through a car-sharing service, an on-site rental car agency, or a car "concierge" to an off-site location. For residential construction, reduce the current parking requirement of one space for every one unit to one space for up to every four units if the developer commits to provide access to a shared fleet of vehicles and dedicates at least one shared parking space for every 10 residential units.

As with many of the recommendations in this paper, this one depends on the ability of the City to develop appropriate enforcement mechanisms. We would suggest that the Planning Department, as a condition of project approval, could require the developer to demonstrate the continued presence of a car-sharing service on an annual basis, to ensure that some form of car-sharing is provided in perpetuity.

Amending the Planning Code in this way would not force developers to build less parking. Rather, it would give more flexibility to developers who are willing to work with the City to find creative solutions to San Francisco's parking problems.

Recommendation # 2

Reduce minimum parking requirements for housing that serves populations that do not have a high frequency of car ownership.

Some housing types primarily or exclusively serve populations which have very low utilization of private vehicles. Parking requirements increase the costs or subsidy of these units and provide little value to the residents. SPUR supports the current parking ratio of one space for each five units in developments for senior and disabled residents. For deed-restricted affordable housing outside these districts, SPUR recommends the following guidelines:

• One parking space for every 20 rooms of SRO (single resident occupancy) housing

• One parking space for every 3 units with 600 square feet or one or less bedrooms

• Two parking spaces for every 3 units of 2 or more bedrooms

• Further reductions allowed for developments in which units are permanently designated for households at 35% of median income or below by conditional use

• These changes to the Planning Code would apply throughout the entire city

Recommendation # 3

Allow reductions in off-street parking requirements when the curb cut would take away as many on-street spaces as the private parking spaces it would create.

Often in districts with lower residential densities, the creation of new off-street parking spaces takes away an equal number of on-street parking spaces. From the perspective of the neighborhood as a whole, there has been no net gain in parking, but the curb cuts have increased potential hazards for pedestrians and reduced retail continuity in commercial districts.

It can easily cost between $20,000 and $50,000 or more to create an off-street space as part of a residential remodel. Where the presently required creation of a new off-street space would be directly cancelled out by the removal of an on-street space, the property owner should have the choice not to build the parking space.

There is already a precedent for this approach in the provision in the Planning Code in the Bernal Heights Special District, which allows for the administrative reduction of parking requirements that are triggered by building enlargements if it can be shown that the imposition of the requirements would cause a loss of on-street spaces. This policy should be expanded to all residential districts, not just Bernal Heights.

In commercial districts, new curb cuts and garage entries also reduce pedestrian safety and disrupt retail continuity. To prevent this impact, off-street parking requirements for buildings on commercial streets should be able to be modified if the imposition of the full requirements would create such negative impacts. Residents living above commercial uses are less likely to need cars for shopping, so there is an additional logic to allowing for the reduction of parking in mixed use areas. In both cases, the appropriate mechanism for evaluating the reductions would be a zoning administrative modification.

Recommendation # 4

For historic buildings, allow exemptions from new off-street parking requirements triggered by a change in use or occupancy.

Owners of historic buildings face a lot of challenges. Parking requirements don't have to be one of them. San Francisco should waive parking requirements for these buildings both as a bonus to people who maintain our architectural heritage and as a way to protect historic character from the incursion of new parking garages.

The issue arises only when a change in use or occupancy would normally trigger increased parking requirements. For buildings that are listed as historic (or eligible for listing), or buildings that are contributory to landmark districts, there should be no required increases in parking for use or occupancy changes, provided the changes are certified as appropriate by the Landmarks Preservation Advisory Board.

Recommendation # 5

Adjust the Planning Code's parking space size requirements to conform to the Department of Parking and Traffic's size requirements and to allow for a greater proportion of compact parking spaces.

The Planning Code says that "regular" off-street parking spaces must have a minimum area of 160 square feet and "compact" spaces must have a minimum area of 127.5 square feet.

The Department of Parking and Traffic defines the minimum dimensions of a parking space as 8 feet by 18 feet for a standard space (equivalent to 144 square feet) and 15 feet by 7 1/2 feet for a compact space (equivalent to 112.5 square feet).

SPUR recommends changing the Planning Code to use DPT's measurements.

In addition, the Planning Code currently allows one out of every four parking spaces to be compact. In San Francisco as a whole, supporting the use of small vehicles is one of the most obvious ways to strike a reasonable compromise with the automobile. And in fact, even with the recent proliferation of large "sport utility vehicles," the average San Francisco car is not as long as cars were several decades ago. Both out of recognition of the fact of smaller cars and out of the need to encourage them, SPUR recommends changing the Planning Code to allow one out of every two spaces to be built to compact dimensions, for any structure for which two or more spaces are required.

Current guidelines allow 100% of parking spaces in the South of Market district to be compact. This is a good precedent. SPUR recommends expanding it by allowing 50% of the parking spaces in the rest of the city to be compact.

Conclusion

We cannot fit high numbers of additional cars into San Francisco. If we design our city along suburban lines, where we construct a parking space for every person, at every destination, San Francisco's walkable streetscape will be compromised and the intensity of activities-the very definition of urbanity-will decrease. It will also become increasingly difficult to get around, as traffic slows down not only cars, but also blocks the way for bikes and transit. If we want to evolve in a way that builds on our urban strengths, it is in our best interest to limit the amount of parking we construct, and simultaneously, to invest heavily in our transit, bicycle, and pedestrian infrastructure. Our relative independence from cars can be leveraged into a way to make living in San Francisco more affordable. Re-thinking our parking requirements as suggested in this paper would lower the cost of housing by:

• Making available new housing without parking, for people who would choose to forego a car in order to bring down their housing costs

• Fitting in more housing on a given lot size, by trading garage space for living space

• Supporting the emergence of pay-per-use car-sharing services that make more efficient use of each vehicle and provide residents with access to cars when needed

• Providing more certainty in the approval process for new housing developments that try to take full advantage of San Francisco's transit and pedestrian accessibility

• Through changing our parking requirements along these lines, San Francisco can build more housing while reinforcing the city's unique urban strengths so that, over time, it will grow more pedestrian-friendly, more transit-rich, more environmentally-sustainable, and more affordable.