Localizing the Internet: Five Ways Public Ownership Solves the U.S. Broadband Problem.
The term "public-private partnership" is widely used to describe a bewildering variety of municipal broadband projects, projects as different as Philadelphia, where a private company will own and operate the network, and Saint Louis Park, where the city will own a fiber and wireless network and contract with a private company to manage and provide services over the wireless portion of the network.
The Status Quo: The dominant business model for telecommunications networks in the United States is a network owned and operated by a private, forprofit company that is also the only or primary provider of monthly subscription services. This is true of your local phone and cable companies. They own the infrastructure, and you as a customer have no choice in who delivers the service.
Cities have little regulatory authority over these networks. (As explained above, these networks are subject to few regulations at any level of government.) For example, they do not have the authority to require phone companies to expand their DSL coverage, nor can they include provisions related to equitable or affordable Internet access in their cable franchise agreements.
Franchise Model: A privately owned and operated, for-profit network that does not have the city as a major customer. The city grants the private company use of public assets for some period of time, and the company compensates the city for use of those assets. Cities typically work with a company that asks for a franchise and do not issue a request for proposals (RFP), although some have done so as a way of soliciting competing offers. One of the first wireless franchise agreements was in Anaheim, California. Earthlink will pay the city a fee for use of the public assets needed to support a Wi-Fi network. The city will not be an anchor tenant on Earthlink's network, because it is deploying a city-owned Wi-Fi system for municipal use. The franchise agreement does not include any requirements beyond the network providing a certain level of speed, coverage and reliability.
This model poses few risks, but also few benefits. It requires no public investment and little public involvement of any kind. The benefits are modest amounts of revenue from pole attachment fees, and the possibility of additional competition. The city has little influence over the network coverage quality of service, or the prices charged. Franchise models do nothing to overcome the digital divide between higher and lower income households.
Anchor Tenant Model: A privately owned network, with the city agreeing to become the anchor tenant by agreeing to buy a minimum annual level of services. The city grants the private company use of public assets (or assists in negotiating access from private entities), and also agrees to be a major customer of the network (an anchor tenant). In exchange, the city is compensated for use of public assets. The agreement contains a public benefits section that may include a share of revenue or limited free access to the network.
One of the first anchor tenant models was in Minneapolis. As explained above, under the terms of the contract, the City will pay the private owner of the network a minimum of $1.25 million annually for services over the 10-year life of the contract. The company will give five percent of net revenues to a digital
inclusion fund managed by an outside foundation, and provide free access in selected parks and community technology centers.
The largest benefit of this model, in the eyes of many elected officials, is that the city does not have to finance construction of the network and assumes no responsibility for its ongoing operation. The city gains a new competing network to its incumbent phone and cable companies, and receives funds for public benefit projects.
This model, however, does have substantial risks. Since the city will rely on the network for its own internal communications and revenue for public projects, it cannot allow the network or the company that owns it to fail, even when its intervention contradicts the public interest. Consider the recent case involving Massport (Boston-Logan Airport).
Massport entered into an agreement with a private company that would provide for-fee wireless Internet access throughout the airport and share a portion of its revenues with the airport. After the for-fee service was introduced, Massport tried to prevent airlines from offering their own free wireless Internet access in the airport. The conflict ended up at the FCC, which eventually ruled in favor of the airlines, on the grounds that landlords cannot prevent tenants from using legal technologies of their choosing.
Cities also face the possibility that state or federal legislation will preempt their authority to enforce these agreements at some future date, as has happened with cable franchise agreements.
The Dollars and Sense of Public Ownership
Every city that is seriously exploring a citywide broadband network should do a detailed economic and financial analysis. This will serve it well even if it should end up choosing a privately owned system because it will allow it to negotiate with the private company from an informed perspective. The analyses can use different assumptions.
Some of the issues involved are:
- Who will manage the network? This may be the entity that owns the network, or management may be contracted out.
- Will the network be for profit or not-for-profit?
- Will the owner of the network sell retail services only, wholesale access only, or a combination of the two?
- Will the city be a major customer?
- Will ongoing operations be supported by monthly subscriber fees, advertising revenue, sponsorships, municipal uses, or a combination of these?
A complete analysis requires that the city examine different ownership structures. A number of companies are offering to build networks at no up-front cost to the city. City officials should understand that although seemingly attractive for its convenience, such a model may not offer the city and its households and businesses the best long term benefits.
A financial analysis includes several key items.
Capital expenditures Capital expenditures include wireless hardware and software, backhaul (the connection from wireless access points to the larger local network, which in turn connects to the global Internet network), network engineering and deployment. It also includes core network equipment (i.e. servers and routers). The city's existing assets -- streetlights, electric poles, optical fiber connecting public buildings, etc. -- can significantly affect the cost of a network.
Costs depend on the technology. Wi-Fi hot spots, like those found in cafes or homes, are inexpensive. Ongoing costs may be as much as ten times the capital investment, however, since each hot spot must be connected to a wired connection in the existing last-mile infrastructure.
More typical is the use of Wi-Fi mesh that reduces the number of wired connections in the network by allowing information to hop from one access point to another before reaching a wired connection. Wi-Fi mesh networks for municipal use only (public safety, meter reading, mobile municipal workforce) can be deployed for $100,000 or less per square mile. Residential
service networks, typically designed to reach 90 to 95 percent of homes and businesses, can cost upwards of $200,000 per square mile.
Fiber to the home is the most expensive alternative, but it is also the longest-lived and the only "future proof" option. Estimates range from $600 to $3000 per home, depending on existing infrastructure and building density.
Operating expenditures For municipal use only wireless networks, the rule of thumb is that operating expenditures are about 15 percent of capital expenditure annually. This includes 24-hour network operations, pole attachment fees and electricity, monthly equipment maintenance and software upgrades, and Internet bandwidth. For combination wireless networks, operating expenditures are about 30 percent of capital expenditure for a retail network, 15 to 20 percent for a wholesale network. The added retail costs include customer service, billing and marketing as appropriate for retail or wholesale customers.
Localizing the Internet
For fiber to the home, annual operating costs will be around 5 percent of capital expenditure, though this may be slightly higher for smaller cities.
More detailed breakdowns vary by location. For example, average pole attachment fees are in the range of $36 annually in California, but $86 annually in Louisiana. Wi-Fi Access points with a single radio may draw $20 worth of power annually, while multi-radio deployments combined with high-powered wireless backhaul can draw five times more.
Wireless hardware maintenance will be in the range of 7 to 10 percent of equipment costs annually (though this may be higher for some backhaul components). Internet bandwidth consumption will depend on the number of subscribers and the average bandwidth use per subscriber, generally assumed to be 250 kbps to 500 kbps per user on average, and 1 Mbps per business on average.
Revenue. Monthly subscriptions are one of two major sources of revenue. Monthly rates depend on whether the network is wholesale only or retail. In a wholesale network, the city would be responsible for maintaining the network (or contracting for management) and relationships with companies that sell retail services. In a retail network, the city would be responsible for retail service and support, as well as all marketing and advertising. Gross wholesale revenue will typically be about onequarter to one-third of gross retail revenue.
The wholesale rate that can sustain the network will depend not only capital expenditures and projected subscription rates, but also the division of responsibilities between the wholesaler and retailer(s). Fiber to the premises can generate much higher revenues than wireless, because the networks can easilysupport television.
The other major revenue category is municipal use. Many cities currently budget for mobile computing, most often subscribing to cellular data services that are both slow (half the speed of a typical DSL or T-1 connection) and expensive ($60 per month). Within the city, the Wi-Fi network replaces these subscriptions, directly saving the city hundreds if not thousands of dollars each month. Other direct savings may come through replacing leased lines to public buildings with fiber or high-speed wireless connections that provide faster speeds at a lower price, or replacing local-use cellular phones with Wi-Fi phones. Cities that have invested in fiber connecting public buildings typically have a five to eight year payback relative to the expense of leased lines.
Advertising may be a source of revenue for wireless networks, but it would be unwise at this point for a municipality to count on that as anything other than an added benefit of perhaps one or two dollars per user, per month. The most challenging aspect of the evaluation will be to estimate second order effects. Some can be evaluated directly. For example, if the city has a choice between hiring a new building inspector or using wireless to improve the efficiency with the same number of inspectors, the salary of the inspector not hired can be credited as an avoided cost. But there is also a wide array
of machine-to-machine communications (automated meter reading, wireless parking meters, traffic monitoring, etc.) that may improve provision of municipal services but do not directly reduce the city's expenditures.
If the city is planning to purchase these as communications services from a private network owner on a per unit basis, the value of the cost savings must be directly determined. Many of these are zero marginal cost applications, which is to say there is no additional cost beyond that of the hardware, that are essentially free to the city if it owns the network. There are other, less tangible but very important benefits the city should take into account, including economic development, reducing the digital divide, and increasing municipal efficiency and service levels.
The city should also take into account the citywide impact of reduced rates due to competition. Sometimes cities see this as a disadvantage. They worry that incumbents will reduce their rates below those of the city owned network. In no case of which we are aware, did this result in a city network losing substantial amounts of money. Moreover, the city, by the nature of its mission and charter, should have a broader balance sheet. A drop in prices by incumbents by $10 per month translates into millions, perhaps tens of millions of dollars in collective savings to city households and businesses. That not only enriches their individual balance sheets, but keeping at least part of these savings will enrich municipal coffers as well.
-----
