Reinventing Government in San Francisco

Reforms for a better city

Urbanist Article January 1, 2001


Currently, the city does very little ongoing maintenance of its facilities and has millions of dollars of deferred maintenance that becomes more expensive the longer it remains unaddressed. Instead, we pay for deferred maintenance with bond measures every few decades. Our tax dollars would go much farther if we took care of our public assets as part of routine, annual operations. The way to do this is to require every city department to set aside a certain percentage of its operating budget for maintenance, and to directly address the deferred maintenance issue with a long-range capital plan. In some ways, the capital plan would be the most controversial part of this proposal—and certainly the most interesting to the public. Think of it as a classic planning exercise: define the goal, develop alternatives, select the best alternative, and evaluate the results over time. [See article on page 10 for more on this idea.]


The city’s budgeting system currently consists of detailed, line-item department budgets that are approved by the Mayor’s Office and the Board of Supervisors. It is extremely difficult to change spending priorities during the year, once the budgets have been adopted. City employees in every department spend months preparing these detailed budgets. More importantly, this rigid budgeting system means that there is no incentive for departments to save money: if they find a way to do something cheaper, the following year they will lose that money from their budget and it will go back into the General Fund. This is why the Prop. E Muni reform measure made sure that savings would remain with Muni.

An alternative budgeting system, often called an Expenditure Control Budget, has been adopted by many states and municipalities around the country. It allocates to each city department a fixed percentage of the overall city budget. This percentage would be the same each year, unless the Board of Supervisors votes to change the allocation formula—which makes it clear to everyone, in a way that it is not today, that any increase in one department’s budget is offset by a decrease somewhere else. Similarly, if the city wants to increase the total amount of work being performed by local government, it would be clear that total revenues need to be increased.

Instead of micro-managing city departments, policy-making bodies would be free to focus on the big picture. Administrators would have more flexibility in how they spend their budgets. And most importantly, city departments would have an incentive to save money: in order to improve services, they would be directed to find ways to cut costs within their own departments. The Controller’s Office (and other oversight bodies) would spend more of its resources worrying about the important things: evaluating department performance and rooting out corruption.

A charter reform that gets rid of all the fixed dollar set-asides and replaces them with percentage set-asides, which can only be adjusted by the Board of Supervisors, would do a lot for good government.


Many things that save money in the short run cost a lot more money in the long run. Cutting back on building maintenance or road repairs are the classic examples. On the other hand, some things that would cost the city a lot of money up front could save a lot of money down the road. Management information systems, automated parking meter collection devices, or energyefficient buildings are all examples. The city’s current “modified accrual” budgeting system, which is the most common system for U.S. municipalities, focuses on the budget within the next fiscal year. We need an accounting system that looks at the long-term, not just one year out.


At the turn of the century, civil service systems were set up to eliminate nepotism and cronyism. Instead of patronage appointments or unregulated contracting for services, the public’s work was to be done by a set of professionals, hired on an objective evaluation of their qualifications. This idea remains sound, but today the civil service system is in need of modernization.

The city has more than 2,000 job classifications for its 28,000 employees. Promotions are based primarily on seniority. Departments cannot hire through the normal process of advertising and recruiting, but must instead hire from a list of people who have taken civil service exams. Many important positions go unfilled for months (if not years) because of the cumbersome hiring process, causing qualified applicants to accept other offers.

The civil service system needs to be turned into a modern human resource function that helps departments perform their duties, with performance goals for speed and quality of placements. The number of classifications and pay grades should be radically pared back. Job performance, rather than mere seniority, should play a larger role in promotion decisions.


If the goal is to free city employees to do their jobs and reduce the web of rules that consumes so much of their time, then a new way of ensuring accountability has to replace the old. Agency performance measurement is one of the best ways to ensure accountability. As consumers of city services, we should care about the results, not just how the results are achieved.

Several caveats are in order. First, performance measures work best for functions which are quantifiable and measurable public services. Ideal examples include Muni (passengers carried, on-time performance, trip time between destinations, the mode split of all trips taken, etc.); the Department of Public Works (average time to complete various kinds of projects; condition of roadways, life span of street trees, etc.); or SF General Hospital (patients helped, recovery rates, costs to provide services relative to other public hospitals, etc.). It is much harder to evaluate the quality of an urban design or the fairness of an opinion rendered by a judge.

Second, performance measures need to be tied to the core mission of city departments. For example, with the Department of Parking and Traffic, good performance measures would try to get at issues like trip time for pedestrians, bikes, Muni, and cars; quality of the pedestrian environment; mode split; accessibility; and other measures of the functioning of the streets. Too often, departments settle for easier, but somewhat trivial measures, such as how quickly they return phone calls to the public. Certainly public responsiveness is part of the job of a public agency, but what we really need to know as citizens is how well a department is performing its core mission.

In order to make performance measures work, several things are necessary:

• Information should be collected by an agency independent of the function being evaluated, such as the Controllers’ Office.

• Performance measures should be reported regularly: annually or at least biannually.

• Performance measures must be defined to include the costs to achieve the services.

• Comparisons with similar cities or organizations are essential for evaluation of performance.

• Incentives, such as pay bonuses, can be tied to performance measures to increase their effectiveness.

The City Controller’s Office has several staff people working full time to develop performance measures for city departments, one of the most promising developments in city government.


Millions of dollars in property taxes are left on the table each year. By more rigorously assessing property, the city will gain funding that is already owed to it. Similarly, the city could raise a great deal more each year in parking revenues to fund Muni if it enforced parking citations more systematically and took a more market-oriented approach to setting rates in cityowned garages.


In some situations, the city could provide more or better public service by introducing competition into service provision. Most of us grew up with the idea that monopolies under capitalism are inefficient and lead to abuses of power. Monopolies are able to pass the costs of internal inefficiency onto their customers in the form of higher costs or worse service. All of these things can be true about monopoly in the public sector as well.

Contracting out, as a solution to the problem of monopoly, is very different from privatization. Privatization means that government no longer pays for or manages the service, but leaves it to the private market to provide it. Under contracting out, by contrast, we still pay for public services with tax dollars, still allocating money to departments in the normal way, and still having city policy makers decide what the departments should do. But then, when it comes to finally spending the money to perform the services, the city contracts with for-profit or non-profit organizations to actually do the work. It puts city employees in the role of manager, rather than service provider— watching out for the public interest, making sure work is performed at the highest standards, and supervising the work of the service providers.

In most of city government, it doesn’t make sense to contract out work. The city does all sorts of things that are either inherently monopolistic or not “services” at all—enforcing laws, resolving conflicts, etc. But for a limited number of public services, contracting out could be part of the solution for government effectiveness.

One of the most successful areas of contracting out is in affordable housing. The old style was to have the government build and manage public housing projects itself. Today, non-profit organizations such as Mission Housing, Asian Neighborhood Design, and Bridge Housing provide most of the city’s affordable housing, although it is still paid for with tax dollars.

In some jurisdictions, public agencies and private contractors each get to bid on public contracts. Interestingly, experiences have shown that at least half the time, public agencies are able to win based on service and costeffectiveness.

Effective management of the contracting-out process is difficult. City staff need to have the capacity to develop requests for proposals, evaluate competing proposals, write contracts, supervise contractors, and evaluate the results, free from pressure from elected officials to bypass professional norms of selection and evaluation for the benefit of favored contractors. Moving from a monopoly (do it ourselves) approach to a competitive (hire outside firms to do it) approach means shifting to a more managerial, higherlevel governance model for many city departments. It is a given, in San Francisco, that any move towards increased competition in service provision would have to guarantee job security of current employees; this is a long-term idea that can only be used for certain situations.

Prop. J, passed in 1976, directed the city to contract out services whenever it could be demonstrated that contracting out would save the city money. While it has been implemented to some degree, more remains to be done. Now that contracts require benefits and a “living wage,” we can make the contracting decision purely on the basis of what’s best for the city. An independent, objective analysis should be carried out to identify other opportunities. What has been contracted out successfully in other cities? What have the results been, intended and unintended?


The whole point of contracting out public services is to help increase the quantity or quality of the public services we buy with our taxes. For it to work, the contracts must be awarded objectively to the individuals or organizations that are most capable of performing the work. If companies seeking city contracts are able to donate money to politicians or the political causes of those politicians, then the objectivity of the contracting decision may lack integrity if elected officials influence the decisions of professional staff. Certainly it creates an appearance of impropriety. Questions will be raised about whether the best organization was awarded the contract. An even worse scenario is the private firm that is able to translate its political power into a monopoly status, creating the worst of all worlds: a profit-motivated firm, operating without competition, feeding at the public trough.

There is a straightforward solution, which can be established through the legislative process or written as a condition of individual contracts: prohibit any organization that receives a substantial city contract from influencing the outcome of any local election. This was at least part of the intent of Proposition J, which the voters approved on the November 2000 ballot, but it is unclear at this time whether Prop. J will work as intended (or whether it will be upheld by the courts).


Public employees know better than anyone where the current systems break down. We have certainly found this to be true in our work on Muni reform. In order to make changes to our city agencies, public employees need to be invited in to be part of the solution. Every department has opportunities to streamline its work processes, find better ways of doing its work, and increase the effectiveness of its work, but most of these opportunities will not be discovered unless a formal structure is set up to facilitate the process.


Energy-efficient buildings, also known as “high performance” or “green” buildings, are being adopted all over the world. By designing for maximum use of daylight, highly efficient appliances and fixtures, natural ventilation and cooling, water conservation, and other green building techniques, buildings save money over their lifetimes. The public sector, because of its ability to think about a long time horizon, has the opportunity to invest in energy-saving techniques that have longer, but still substantial payback times. San Francisco has a green building program, but it is tiny—really just a pilot program—and behind schedule. Other opportunities are being lost because of the PUC’s recent decision to disband its energy conservation division. We should consider an “eco-bond” to retrofit municipal buildings on a much broader scale.


The Municipal Fiscal Advisory Committee (MFAC) exists to provide pro bono management consulting services to the city. Currently, MFAC is a very underutilized resource. Virtually every department in the city has projects that could benefit from an outside perspective and technical expertise. This kind of collaboration could allow the city to learn from large organizations in the private sector— and as similar public/private partnerships in other cities have shown (such as Joint Venture Silicon Valley’s program of lending executives to public schools for a year), the private sector might learn a lot as well.SPUR logo


About the Authors: 

Gabriel Metcalf is deputy director of SPUR.

Get The Urbanist

Join SPUR and get our magazine in your mailbox

Become a member