How to Make San Francisco Neighborhoods Stop Worrying and Love More Housing
In San Francisco, the problem of unaffordable housing has reached calamitous levels and threatens the vibrancy of the city. Today, San Francisco shares with its Bay Area neighbors the dubious distinction of having the most unaffordable housing in the United States. While in most of the nation, the median house costs three times the median household income, in San Francisco, the average home costs over nine times an already higher average income.¹ Due largely to high housing costs, San Francisco and its suburbs lose over 10,000 net domestic migrants to other states every year, mainly to lower cost-of-living states such as Arizona, Nevada, and Texas.² The reason for these abnormal housing costs is a lack of new residential construction. In the period since the financial crisis, San Francisco built only one new home for every eight jobs created.³ High housing costs and population flight is now threatening new job creation.⁴
Due largely to high housing costs, San Francisco and its suburbs lose over 10,000 net domestic migrants to other states every year.
Almost all researchers attribute these high housing costs to burdensome government restrictions on building.⁵ San Francisco at one time became the national poster child for the anti-growth movement, and its intentionally byzantine development process has done much to create its housing shortage.⁶ The California state legislature, by contrast, has passed numerous law to force cities like San Francisco to allow more intensive development.⁷ Unfortunately, state attempts to encourage development have a long, lamentable, and largely ineffective history. The simple reason is that local politicians have a greater incentive to monitor development than distant state legislatures and bureaucrats. NIMBY, or “Not in My Back Yard,” neighborhood opposition, usually wins relative to distant Sacramento lawmakers.
The best way to encourage more development in San Francisco is to make San Franciscans actually want development. If we allow local neighborhoods, which admittedly bear most of the costs of development, such as congestion and crowding, to also keep most of the benefits, we could turn the political debate about development on its head. This paper proposes the framework for an act, which could be passed through a popular vote, city ordinance, or state bill, which would redirect many of financial benefits of local development to the communities that are today the locus of opposition to development. It specifically proposes to redirect impact fee revenue from new buildings, which currently goes to the city government, back to San Francisco’s 36 neighborhoods. This would help turn many NIMBYs into Yes in My Back Yard advocates, since they would stand to capture those new benefits.
The post will focus its discussion and its proposed reform on the City of San Francisco, but it is potentially applicable anywhere in the Bay Area, in California, or, as a broad framework, anywhere in the United States. It will also demonstrate how such a reform could reshape the current planning and zoning regime in San Francisco.
The Current Planning Process in San Francisco
San Francisco, like every other city in California, is mandated by state law to adopt a “general plan,” which establishes distinct zones that determine the size and use of all new buildings. Due both to NIMBYs’ peculiar influence in San Francisco dating back to the 1960s, and planners’ tendency to defer to the structure of existing neighborhoods, San Francisco’s general plan makes new development difficult. Approximately 74% of the land is zoned for no more than three-unit homes, with the majority dedicated to one- or two-unit homes. Most of the city has a maximum 40-foot height limit for all new development.⁸
Fig. 1. Adaptation of San Francisco zoning map to demonstrate allowed level of residential construction.
Unfortunately, even the restrictive general plan downplays the difficulty of building in the city. The ever-proliferating bureaucratic documents undergirding this plan muddy whatever potential clarities developers could glean from the plan itself. A typical example is the Planning Department’s East South of Market (East SoMa) neighborhood plan, drafted in 2008. The plan lays out 42 separate “objectives” the city wants to achieve through development in the area, with no ability to rank them in case they conflict, as many clearly do. (Objective 4.8 aims to “Encourage Alternatives to Car Ownership and the Reduction of Private Vehicle Trips,” while Objective 4.9 aims to “Facilitate Movement of Automobiles.”) The reader’s despair only increases when one realizes that each of these objectives also contain numberless policy subheadings exhorting urban planners to fulfill indecipherable goals, such as policy number 8.5.3., “demonstrate preservation leadership.”⁹
The very vagueness of San Francisco’s planning documents illustrates their ultimate arbitrariness. There is simply no way for planners or zoning administrators to evaluate a proposed development’s conformity to the overall “plan” based on such vacuous verbiage. Therefore, almost all “planning” regarding development in San Francisco is the result not of a preexisting map of developments, but of a concatenation of political pressures exerted on every parcel that a builder proposes to be developed.
In order to officially approve any development, even one that conforms to the general plan and to zoning requirements, San Francisco has a particularly convoluted process, one which not atypically can last five years from first proposal to final approval of the project (before actual building begins), and which often lowers the size and scope of the new development, if it gets approved at all.¹⁰ The city’s permit process mandates greater community input into planning and development than most other cities, and is replete with numerous public hearings, in which NIMBYs have substantial power. I will describe the labyrinthine outlines of this process here.
Most developments in San Francisco require some demolition of existing structures, which means a developer must send a “Dwelling Unit Removal Application” to the Planning Commission, which then requires a public hearing followed by approval of the application by the Commission. Then, before a formal building application is submitted, there is a mandated “Pre-Application Meeting,” and a requirement that neighbors and neighborhood associations be mailed invitations to voice complaints at that meeting, which complaints the Planning Commission will take consideration of in future hearings. There is also a “Preliminary Project Assessment,” conducted privately by the Planning Department (the superior agency of the Planning Commission), to alert the developer to any potential violations before plans are officially filed. Once plans are officially filed, most projects will be subject to an “Environmental Evaluation” by the Planning Department, which will then require a public hearing. According to developers, the environmental evaluation process usually takes from 12 to 24 months. Most developments, including those who have since secured residential demolition permits or those in a larger “planned unit development,” will also require a conditional use permit from the Planning Commission, which must hold a public hearing to decide whether the proposed use synchronizes with the plan and uses of the neighborhood. This process usually takes at least four months.¹¹ The Planning Department must then write a “Shadow Analysis,” paid for by the developer, with input from the Recreation and Parks Commission, to determine if the developments’ shadows affect any park.¹² If it is determined that the shadow does affect any park, another public Planning Commission public hearing and approval is required.¹³ At this stage or earlier, any member of the public can request an overall “Discretionary Review” of the project or any aspect of the project, demanding that the Planning Commission hold public hearings and accept or deny the project on purely discretionary grounds.¹⁴
Once the Planning Department’s job is complete, different departments have to review the building plan for conformity to different city and state laws and regulations, such as a Fire Department Plan Check for fire safety requirements.¹⁵ The Department of Building Inspection (DBI), itself with subsidiary offices such as Plumbing Inspection and Electrical Inspection, must make several non-public approvals of the building plan.¹⁶ Yet, before the DBI issues its final permits, any member of the public can appeal their decisions to either the Abatement Appeals Board (for decisions relating to DBI requests for abatements), the Access Appeals Commission (for disability-related building issues), or the Board of Examiners (an expert board which examines new materials used in construction as well as all DBI interpretations of the building code), each of which holds hearings on these appeals.¹⁷
Once any of these building department decisions are made final, they can be appealed to the city’s administrative Board of Appeals, which holds public hearings, and then to the Board of Supervisors, which also holds a final public hearing, before voting to uphold or dismiss the appeal. Since the implementation of district elections in 2000, this “final” political forum for appeal is particularly attuned to neighborhood complaints and the concerns of the local district supervisor.¹⁸
Even a normal project… could be subject to at least 12 separate public meeting requirements before boards, departments, and commissions of the city.
On the whole, this process means even a normal project, without any special historic or environmental impact and falling within the already existing “general plan” requirements, could be subject to at least 12 separate public meeting requirements before boards, departments, and commissions of the city. Builders complain that the single largest source of delay for their projects is constant neighbor appeals and the resultant hearings.¹⁹ At each of these hearings, publicly elected or appointed officials are held in thrall to the demands of oppositional neighbors. Many of these public hearings will in fact involve multiple hearings in front of boards and the public before a final decision is rendered. One developer said they had to engage in over 100 community meetings for a single project.²⁰
If a development involves a change in existing planning and zoning rules, a not uncommon occurrence considering their detailed nature, there are requirements for both a zoning change initiation hearing and zoning change adoption hearing at the Planning Commission, as well as hearings for another environmental assessment to evaluate the impact of the change in zoning (supposedly separate from the impact of the building itself), and potentially a historic assessment hearing (heard by the Historic Preservation Commission), and finally at least two separate votes after two separate public hearings before the Board of Supervisors, who must make a final decision on all zoning changes, all before the normal approval process described above can take place. This makes for potentially 16 separate public hearings, as well as their multiple offspring, required for a fairly ordinary development, before building can even begin.
After building begins, the Department of Building Inspection still has wide latitude to recommend changes before offering a Certificate of Formal Completion. This may help explain why, of the 58,800 housing units who went through the extensive hearing and approval process in San Francisco in the last 20 years, approximately 25% were never built or completed.²¹
The “discretionary review” of any of these applications by commissions, departments, or the Board of Supervisors, unhampered by any formal legal requirements, is particularly burdensome. As former San Francisco Planning Director Allan Jacobs explained, “Almost everything now has a discretionary review… One of the things I found out over time about discretionary review is if there is a disagreement, the force or the party that usually gets its way is the party with the most power.”²² Today, local neighbors and wealthy homeowners typically have the most political power, and they win the political battle against politically isolated, even if also wealthy, developers.
When California planners themselves were surveyed to describe the most severe constraints on housing development, the most commonly cited problem, after simple geographic and land size issues, was “Public Opposition,” which they considered a major constraint in more than 25% of projects. Planners also ranked public opposition as more disruptive than other, more-commonly discussed complaints such as “California Environmental Quality Act lawsuits,” “Land area zoned-single family,” or “Growth management policies” (all of which were cited as major problems in less than 10% of projects.) Although many commentators express frustration at public officials for sabotaging projects, the same survey reported that elected and appointed officials were more than seven times as more likely to encourage projects than to discourage them, while they said citizens opposed most projects.²³
One contemporary example of the burdens placed on builders by concerted public opposition should illustrate the problems. In 2014, builder Robert Tillman proposed to redevelop a laundromat and parking lot located one block from the 24th Street BART station into an eight story, 75-unit apartment complex.²⁴ The project technically falls in the “Eastern Neighborhood Plan,” which states that it aims to encourage transit-focused development. Yet Planning Department and Planning Commission studies and public hearings, for a project some referred to as “no-brainer,” stretched until November 2017, when a conditional use permit was finally approved. Yet local groups appealed the decision twice. In February 2018, when the project first reached the Board of Supervisors, they demanded that Tillman draft a historic resources study to determine if the decrepit laundromat was of great historical value, to jeers in the national and international press.²⁵ $28,000 later, historians concluded it wasn’t. Then, in June 2018, the Board of Supervisors unanimously voted to delay the project indefinitely in order to have Tillman fund the third in a trifecta of “shadow studies,” which this time would analyze the shadows the development would cast on an adjacent playground, which, coincidentally, would not be open at the time the shadows would be an issue. Many observers claimed that the only reason these roadblocks were put in place was local supervisor Hillary Ronen’s personal opposition to the project.²⁶ In September 2018, the Planning Commission in a public hearing reversed its previous approval and now turned down a conditional use permit for the building due to concerns about the shadows, although it was unclear if the Commission has the authority to rehear a previous decision.²⁷ The project currently remains in limbo.²⁸ The laundromat saga shows that whatever the formal requirements for planning approval, and whatever the formal plan says about encouraging development, that process is inevitably overridden by informal political concerns. As Tillman himself said in a recent op-ed, “San Francisco has rules, but does not follow them. Even a completely code complaint project is subject to ‘discretionary approval.’”²⁹ In a similar vein, the Board of Supervisors recently turned down a 157-unit Mission District development despite meeting all apparent requirements. Supervisor David Campos said he changed his mind “at the last minute” after hearing comments at the public hearing.³⁰
The importance of local political opposition is often underestimated in attempts to encourage more building. The power of neighborhood groups means that any attempts to reshape or reform San Francisco merely through reshaping the general plan will still encounter immense opposition and obstructionism. State legislative attempts to include more growth under formal plans will not prevent neighborhood opposition from stymying growth by other means.
Previous State Attempts to Encourage Housing
For over half a century, the California state government has demanded that cities design their general plans to accommodate sufficient new housing.³¹ Yet time and again, the state has realized its housing plan mandates have not created actual homes.
Beyond the requirement passed in 1963 that each city create its own “general plan,” since 1967 the state has demanded that the housing portion of such a plan include a “Regional Housing Needs Allocation,” with sufficient housing allowed in identifiable districts to accommodate a city’s population growth. The state Department of Housing and Community Development could veto such a plan if it did not allow for sufficient housing, and people could sue cities for the insufficiency of their housing plans. The law also has demanded that local zoning conformed to the approved housing plan.
In reality, the state has been reluctant to force local plans to open up land for more or denser housing. And in the housing mandate’s early years, most lawsuits under the act challenged plans that tried to spur more development, rather than those that encouraged it. From 1979 to 1982 a series of new laws limited lawsuits against cities’ housing plans, and hypothetically allowed judges to mandate subdivision map approval for necessary housing in local areas. Yet few, if any, of a series of new pro-housing lawsuits were successful in the courts or locals councils, due to continued local opposition and obstructionist permitting practices.³² To force cities to simplify the development process, the state in 1977 passed the “Permit Streamlining Act,” which required local governments to approve or disapprove development permit applications within strict time limits. The result was that projects which approached the deadline often had their permits rejected, accompanied with a request to submit them over again, starting a new permit approval time limit. In other cases, developers were reluctant to sue the city under the act, when they still required the city’s assistance in getting any projects approved.³³
In 1982 the California state legislature passed the Housing Accountability Act, which demanded that local governments automatically approve housing that “by-right” met the requirements of the general plan and existing zoning. It also mandated that all such housing be issued permits within 90 or 180 days of submission. In San Francisco, this requirement is dodged by the comprehensive nature of “discretionary review” by the Planning Commission, which can declare that even a simple, seemingly conforming proposal does not meet one of the innumerable requirements in San Francisco’s planning codes. In reality, the act only encourages San Francisco to add to the complexity of their plans to dodge state or judicial review, making almost no project “by-right” eligible for fast permits.³⁴ An attempt to strengthen the housing accountability act in 2017, SB 35, mandates that that those projects that contain a certain percentage of “affordable housing” be automatically approved when “by-right” they meet the zoning code.³⁵ Yet this requirement is little more than a restatement of the Housing Accountability Act in a particular field, and will likely have little extra effect.
Other state laws, such as the Density Bonus Law passed in 1978, amended as recently as 2017, mandate that housing with a certain percentage of affordable housing be given a “density bonus,” or increased Floor Area Ratio (FAR), of up to 35%, relative to the existing zoning code. This law too, has largely been ignored and dodged by local governments, who mandate their own separate density bonuses and continue to approve them on a discretionary basis.³⁶
In a similar vein, Governor Gavin Newsom recently proposed to withhold state transportation funds from counties that did not meet their housing goals.³⁷ Most likely, local governments will game their housing goals in order to ensure that they “meet” them in the future. As one Bay Area city councilman said when his city approved its mandated housing plan, “What I’m seeing here is an elaborate shell game, because we’re kind of lying… We have no intention of actually building the units.”³⁸ Whatever the intentions of the governor’s announcement, and of similar state actions, a continued focus on manipulating formal planning and zoning requirements, without taking into account local political opposition, will result in a continued dearth of new housing.
Aligning Community Costs and Benefits: The Proposal
If one attends a public hearing on a proposed development, one will hear monotonous complaints about road and transit congestion, school overcrowding, and burdens on public parks and infrastructure. Many advocates for more housing dismiss these complaints as overwrought, but their consistency and intensity suggest real issues are at stake. As Dartmouth economist William Fischel argues in his book The Homevoter Hypothesis, the fact that most households have the vast majority of their financial savings in one, undiversified, and easily depreciated asset, namely, their homes, makes them understandably intolerant of threats to that asset.³⁹ Research shows that any permanent changes in the local neighborhood, for good or ill, tend to be quickly reflected in the value of the homes. Even in San Francisco, a city with approximately 64% renters, the political combination of homeowners and existing landlords, who fear both competition and the congestion issues commiserate with development, presents a potentially unassailable political force against growth.
New development also creates substantial benefits. Unfortunately, those benefits accrue to the wider city and region, and not local neighborhoods. New development not only allows increased jobs and a more dynamic economy in the region, it also brings millions of dollars of tax revenue to city government, most of which does not the flow back to the local neighborhood. For instance, new development in San Francisco is subject to dozens of “impact fees,” which are supposed to offset the costs of growth, and which are demanded by the city in exchange for the approval any new construction. These include a Transportation Sustainability fee, a Child-Care Fee, an Inclusionary Housing Affordability Fee, a Street Trees In-Lieu Fee, a Water Capacity Charge, and so forth.⁴⁰ Although the city has not estimated the total costs of these fees per new housing unit, the estimate for other Bay Area cities like Oakland and Fremont range from about $30,000 to $75,000 in impact fees per multifamily unit built.⁴¹ Thus, as a rough but very conservative approximation, a 100 unit multifamily development in a San Francisco neighborhood which brings approximately $50,000 in impact fees per unit provides five million dollars in immediate fee benefits, beyond any other benefits to the city of increased future property tax, income taxes, and other charges.
This post proposes to change the political dynamic around building by providing some or all of those impact fees directly to the neighborhood. If each of San Francisco’s 36 neighborhoods had a political organization, similar to the currently weak “Community Boards” in New York, or the Advisory Neighborhood Commissions in Washington D.C., or merely a citizens’ body under the control of the local district supervisor, which could help distribute those fees, the gains to a neighborhood from growth could be vast.⁴² For instance, such a 100-unit development in a neighborhood with approximately 10,000 people, such as SoMa, could provide at least $500 to every man, woman, and child in the district. Whatever the political opposition to that development would say about congestion, if they had to explain to a four-family household why blocking this development would cost them $2,000, they would have a much harder case to make. In this new situation, much of the shouting at public meetings would be done by proponents of the development demanding such benefits, instead of their NIMBYish opponents.⁴³
The importance of the current reform proposal is that it would try to attract broad support from the local neighborhood in order to expedite the housing approval process. The neighborhood could, for instance, engage in “participatory budgeting,” whereby they vote on the areas to which they want new funds directed, such as schools, or infrastructure, reduced property or sales taxes inside the neighborhood, or, if the neighborhood agreed, tax “refunds” sent directly to local families. The goal would be the demonstrate to the local neighborhoods the real and concrete benefits of development.
Figure 2: The 36 San Francisco Neighborhoods. Source: www.city-data.com
Under this new proposal, any development permit would first go to the Planning Department, which would estimate the total impact fees. They would also estimate the fees if all in-kind mandatory provisions under the plan, such as affordable housing requirements, were instead translated into in-lieu fees, which would be provided to the neighborhood (the inclusion of such affordable housing mandates costs in this proposal could double the value of potential community fees, but substitution of affordable housing by fees would not be required in all situations). The Planning Department would then estimate how these fees would be allocated, between, say, school improvements, transit support, tax abatements, or other benefits, based on a baseline funding package approved by the district supervisor and the community in previous development meetings. In the next stage, the current “Pre-Application Meeting” would include a public decision on how to allocate this specific project’s funding, based on a neighborhood community board, overseen by the district supervisor, which would later be prorated by funding category if the final project increased or decreased in size before completion. To offset the particular problems caused by very nearby development, the law would allocate 25% of the funds to neighbors and landholders within a quarter mile of the project, who tend to be most concerned about development.
The goal of having this fee and funding decision at the beginning of the process is to get community buy-in as early as possible, with the hope that they would become stakeholders in the development. At least for now, no other parts of the development approval process need to be changed. (Although one would imagine that the current neighborhood groups, which are the biggest opponents of change, would become more active for “upzoning” and permitting reforms if this proposal was in place.) The only difference is that public hearings would involve many members of the community with a stake in furthering the development instead of sabotaging it.
Today, developers do occasionally provide grants to the community, yet current in-kind benefits deliver little that the community actually wants. Many such “community benefit agreements,” either required by the city or extracted by local political groups, involve increasing the percentage of “affordable housing” units provided by the developer to different groups.⁴⁴ Whatever the benefits of such housing, local families and homeowners are rarely interested in it, since these units are not available exclusively to the members of the community, and it address none of their actual crowding or congestion concerns. Developers sometimes also offer to fund certain parks or infrastructure improvements, or to provide money directly to local neighborhood groups, in exchange for political support.⁴⁵ Such infrastructure benefits, however, are usually minimal and ad hoc, and the funded local groups usually have minimal support in the neighborhood.⁴⁶ The closest existing comparison to my proposal is the “Community Facilities District” allowed under the Mello-Roos Community Facilities Act of 1982. Under this act, a neighborhood can vote to tax its property and thereby finance public improvements and services such as schools, streets, sewers, along with police protection, fire protection, etc. Unfortunately, the organization of such districts is also complicated, and since it requires other taxes above the already onerous current ones, and includes both old and new units, it is infrequently used in urban areas and does little to spur new development.⁴⁷
Some concrete examples of how a reformed impact fee would improve a neighborhood should help illustrate the proposal. A $5 million impact fee from a single 100-unit project could purchase, for instance, approximately 10 new large commuter buses for an area or it could plant 1000 urban trees. Fewer than five such projects could fund an entire new school at over 100,000 square feet of floor area. Or, such a project could provide up to $2,000 in tax abatements for a four-person family.⁴⁸
Today, a few preliminary attempts at providing impact fees to the neighborhoods are already in existence in San Francisco, and are showing some success. The “Eastern Neighborhoods Impact Fee” provides fees to areas around the Mission district impacted by development and has helped spur actual political demand for more.⁴⁹ Expansions of such neighborhood fees to other areas, even without a larger citywide or statewide law, or without reforms to the existing zoning plan, could be done on a piecemeal basis or as part of a citywide reform. In any case, once such local reforms were in place, politicians and reformers would be able to discern their influence and tweak the fee provisions to encourage more support for building.
Although there might be some potential lawsuits about the reform,⁵⁰ or problems about diverting a proportion of the city’s current revenue to local areas,⁵¹ on the whole, the increased benefits to neighborhoods, and increased development in the city as a whole, will allow everyone to share in the gains of growth.
This post argues that the solution to San Francisco’s housing problem is not merely another state mandate, but a change in the political and institutional structure of building. Only by allowing local neighborhoods to capture the benefits of growth, instead of just enduring the costs, will political opposition turn into a political demand for housing. The gains of growth are real and substantial. More housing allows cities to accommodate more migrants; it increases the productivity of all those already living in the city; it lowers the rents and costs-of-living; and it brings increased tax revenues to local governments. The costs are also real, and include congestion and crowding, but there are substantially outweighed by the positives.
To encourage more growth, we need to allow more people to trade its costs for benefits, and that means we need more people to share in those benefits.
If the net benefits of growth are so great, and yet growth is not happening, we know we have a failure of “trade.” Neighborhoods today are not able to “trade” the costs they bear from development for the benefits they could get from it, so they ceaselessly oppose all of it. To encourage more growth, we need to allow more people to trade its costs for benefits, and that means we need more people to share in those benefits. My proposal shows how we can spread the benefits of growth more broadly, and thus encourage more of it. Such a proposal could turn the one-time center of the NIMBY movement, San Francisco, into a city open to new housing, new people, and even new neighborhoods.