SPUR Endorses the Public/Private Delivery of Wi-Fi in San Francisco

SPUR Report January 22, 2007

In 2004, Mayor Gavin Newsom declared that increasing digital inclusion was one of his signature policy goals. As a result, he created TechConnect, an initiative to explore how to deliver free wireless Internet throughout San Francisco. In 2005, the City issued a request for information and comment about the development and delivery of a free citywide WiFi network. After a lengthy public process, the City selected a team of EarthLink and Google. The EarthLink/Google team then began negotiations with the City over an agreement to develop and maintain a WiFi network across the entire city.

WiFi is a broadband network that provides wireless access to the Internet to anyone within its zone. Unlike other Internet delivery options such as fiber optic cables (or the earlier "dialup"), it does not require a fixed infrastructure or digging up streets. Instead, the WiFi network would be a series of transponders mounted on City-owned and PG&E-owned telephone poles throughout the city with each transponder delivering wireless access to a limited surrounding area. The proposed network would cover the entire city.

The proposal is an agreement between the City and County of San Francisco and EarthLink, an Internet service provider. Key aspects to the proposal:

  • The system will cost $15 million-$20 million to build and operate (according to EarthLink estimates).
  • The cost to city taxpayers will be zero.
  • EarthLink will pay for the maintenance and operation of the network through the collection of user fees (advertising and subscriptions).
  • The system will provide basic broadband Internet for free and higher-speed Internet access with a modified system for approximately $22 per month.
  • There will be a limited number of subscriptions at a reduced rate for low-income households.
  • Revenue to the City includes an ongoing 5 percent of gross revenues received in the operation of the network (funded by use fees), as well as an initial pre-payment of $600,000 for access to the public right-of-way.
  • The City will use the revenue it collects to help expand access to the Internet for residents of San Francisco.

Once approved by the Board of Supervisors, there would be an initial pilot project in a five-square-mile area (less than 10 percent of the city) within two to three months. The City would then review the results of this pilot project and issue further approval. The roll-out will take approximately nine to 15 months. If the Board of Supervisors and SFPUC approve this proposal over the next month, most of the city could be covered by the end of the year. If not, this project would be delayed until 2008. If the proposal were approved, San Francisco would become the first major city in the country to offer free universal wireless Internet access.

Despite the benefits, this proposal is not universally supported.

The Alternative: a Municipally Owned and Operated WiFi System

Some members of the Board of Supervisors have declared that instead of partnering with EarthLink and Google, San Francisco should develop and manage its own wireless Internet system.

Municipally owned and operated WiFi networks do exist within the United States. However, they most often exist in smaller communities and sometimes where electric power is already delivered through a municipal utility.

The proponents of the municipal WiFi cite a variety of reasons they are concerned about the public private partnership with EarthLink and Google. Some of the key arguments and responses:

  • Argument 1: The City and County of San Francisco should operate the WiFi network, not a private company. Efficiencies should be captured by the public not as profit for a corporation. Municipal WiFi has worked in other cities such as Burlington, Vt.

Response to Argument 1: The municipal model requires the City to not only maintain the operations, but also to pay for and build the basic infrastructure. Given that telecom is a complex and rapidly changing industry and its infrastructure demands are quite high, the City would not be able to capture any of the economies of scale and expertise as exhibited by the proposed vendors. EarthLink is in the business of building and maintaining Internet systems, and has capacity to do so throughout the country. The risks are too high that the City would not capture efficiencies and might not deliver a "profit" to the public.

  • Argument 2: In signing this agreement, the City is selling off its radio airwaves to a private company (since WIFi will use radio frequency to transmit).

Response to Argument 2: There is no transfer of ownership or control of the radio spectrum in this deal. The project is using unlicensed radio frequency spectrum, which by law nobody controls. The only entity that may sell radio spectrum is the Federal Communications Commission.

  • Argument 3: Municipal WiFi offers greater privacy than a system controlled by private companies. Google is motivated by the ability to capture data on the searches of the highly web literate San Francisco population.

Response to Argument 3: Any user in the San Francisco model would be able to turn off the location-based tracking feature of the WiFi network, which provides information to a user based on the user’s location. This mechanism allows a user to "surf the web" without sharing directly their location with the internet service provider. SPUR encourages the City to mandate that the default position for the system not be capturing the user location information. It should also be noted that the privacy concerns about gathering data on users is just as much a concern among a government (municipal) service provider as a private firm that runs the WiFi system. In fact, civil liberty advocates have raised objections in other cities to municipal ownership of networks because it gives governments direct access to user data.

  • Argument 4: The City is giving up too much control in this deal.

Response to Argument 4: Under the agreement, the City maintains strong control over the outcomes. The agreement includes strict timelines over a four-year agreement (while a 10-year agreement is the industry norm to develop a WiFi network). At the conclusion of the four years, the City and/or Earthlink could choose to opt out of extending the agreement.

  • Argument 5: The City is not getting enough revenue out of this deal. Given the huge marketing benefits to both Google and EarthLink, we should be able to secure a better deal.

Response to Argument 5: Despite these claims that we should get more, the City is getting 5 percent of revenue, a pole-use fee and citywide, free service. There is always a trade-off in any negotiation, and it appears that San Francisco is getting a better deal out of this than it might get if it were to build and operate the network itself. Further, other proposed models nationwide (such as Philadelphia and Minneapolis) do not include free WiFi service.

In general, many of the claims of benefits of the municipal WiFi model are based on a misunderstanding of the complexity of establishing such as system.

The Case for the Proposed Public/Private Partnership WiFi Network

The above claims are largely built around the notion that what we have is not good enough and that the City could do better. Even if that were true, this issue requires weighing whether interests of the residents of San Francisco are best served by securing the deal we currently have on the table, or waiting another two years (or more) to try to secure something that might possibly be better - although it is debatable whether a better deal would be forthcoming or whether the City could successfully finance and build a WiFi system. Further, the proposed EarthLink/Google network is not exclusive and thus allows any additional system to be included within it.

The following are some of the key arguments in favor of the EarthLink/Google proposal:

  • The proposed network will not cost taxpayer money. There will be no additional cost to San Franciscans to build or maintain the WiFi network. This is distinct from a municipal WiFi system, which would require 100 percent taxpayer financing through tax increases or debt financing. Given the huge number of other pressing public needs, it is not prudent to use scarce public resources for WiFi when we have an alternative that is free.
  • Operating a telecom network is complex and difficult, and should be given to an entity with expertise. Running a WiFi network requires billing, customer service, ongoing maintenance, responsiveness to changing technology, etc. Capital costs are small relative to operational costs. Earthlink has been in this business for a while and knows this well. The City and the SFPUC have no experience in building and maintaining WiFi.
  • The providers will keep SF at the cutting edge in innovation in WiFi. Google and EarthLink are at the forefront of the innovation in deploying innovative services. As innovations occur in both delivery of service and maintenance of the wireless system, these companies will be able to deploy these changes in San Francisco. A municipal WiFi system would not have access to this knowledge. There is a major risk that a City owned-and operated WiFi system would become obsolete by the time it is completed. Unlike the delivery of water, telecommunications is a rapidly changing and evolving technology that is best left to firms that are at the forefront of these innovations.
  • The proposal guarantees free access to WiFi anywhere in the city. While the municipal model might provide free access, it is not guaranteed, given the potential (and uncertain) costs of building and maintaining such a system. EarthLink and Google are required by contract to provide this service for free to any computer with wireless capability throughout the city. There are potentially 200,000 residents who lack Internet service in their homes.
  • The proposed model has built-in opportunities for new competition, which should lead to better prices and offers the greatest range of choice for consumers. Due to the "open access provision" of the agreement, other companies are still able provide additional value-added services to customers. This means that Earthlink/Google have to allow other companies to use the WiFi network to sell to consumers. For example, DirectTV might choose to offer TV service to WiFi customers and work with EarthLink to develop a bundled package to offer both TV and high-speed wireless. Consumers would still be free to choose existing services that bundle TV and cable (such as Comcast and AT&T). The WiFi system provides an alternative to the existing offerings.
  • The model includes the chance to capture location-based information, which is a huge potential benefit to small businesses in San Francisco. This means that a user can opt-in to the location-based system, which would tell the network where the user is located and thus provide specific information on activities and businesses nearby. For example, the customer could access all of the "Google local" information without inputting their location. This would allow Google to provide targeted results of searches and advertisements, which would cater to the smaller neighborhood businesses. But the user also could opt out of this feature and elect not to share this information with EarthLink/Google.
  • The proposed network will strengthen "digital inclusion." By providing citywide Internet access and using revenues to help get more residents online, the proposed network will help reduce the digital divide in San Francisco. Increasingly, access to jobs and engagement with government necessitates access to the Internet. We need to do more as a city to wire our classrooms and neighborhoods and the proposed plan would get us closer to that goal.
  • WiFi is not an exclusive technology. During the life of the contract, the City could explore other alternatives, such as continuing to put in fiber. Because the proposed network costs zero to the taxpayers and residents of the City, it does not use resources that could otherwise be spent on exploring these other technologies. Meanwhile, other service providers are implementing other access standards to the Internet at high bandwidth (speed) and lower prices (including AT&T, Comcast, satellite, etc.)


The EarthLink/Google agreement would implement a WiFi system that is free, relatively high speed and offers the City approximately $2 million in revenue sharing. The concept before the City offers an excellent opportunity to establish San Francisco as a national leader in this area. It will reduce inequality in the city and provide an added boost to our attractiveness as a tourist destination.

While it is always prudent to weigh the costs and benefits of any major decision (as proposed by the current Board of Supervisors resolution), SPUR believes that these issues have been thoroughly discussed throughout the prior 18 months. On the whole, the deal before San Francisco is a good one - for the users, the taxpayers and the City. If the City chooses to go the route of municipalization, not only will it delay the implementation of a WiFi network by years, but also there is no guarantee of success.

Given the potential benefits of the proposed EarthLink/Google system and the likely negatives of the municipal route, SPUR requests the Board of Supervisors and Public Utilities Commission to adopt the proposed agreement with EarthLink and Google. END

About the Authors: 

Gabriel Metcalf

Get Email Updates

Get SPUR news and events delivered straight to your email inbox.

Sign up now