It's hard to dislike the idea of free municipal wireless Internet access. Imagine your town as an oversized Internet cafe, with invisible packets floating everywhere as free as the air we breathe. That fanciful vision inspired many cities to announce the creation of free wireless networks in recent years. This summer, reality hit—one city after another has either canceled deployments or offered a product that's hardly up to the hype. In Houston, Chicago, St. Louis, and evenSan Francisco, once-promising projects are in trouble. What happened—was the idea all wrong?
Not quite. The basic idea of offering Internet access as a public service is sound. The problem is that cities haven't thought of the Internet as a form of public infrastructure that—like subway lines, sewers, or roads—must be paid for. Instead, cities have labored under the illusion that, somehow, everything could be built easily and for free by private parties. That illusion has run straight into the ancient economics of infrastructure and natural monopoly. The bottom line: City dwellers won't be able to get high-quality wireless Internet access for free. If they want it, collectively, they'll have to pay for it.
For the last 20 years or so, the thorniest economic issue in the telecommunications world has been the "last mile." Physically, the last mile consists of the wires that run from your home or business to the local phone or cable company. It's pricey and uses old technology, but almost everything depends on it and a few giant companies—like AT&T and Comcast—control it. The last mile is a bottleneck: The price and speed of the whole Internet depends on it. When people talk about the United States lagging behind the world in broadband speed and access, they're talking about the last-mile problem.
In the late 1990s and early 2000s, dozens of companies were launched that had new ideas for busting through the last mile and getting the Internet into homes. I remember going to industry trade shows where grown men demonstrated robots designed to crawl through city sewers and deliver a fiber-optic cable to your toilet. (That firm, CityNet, received $375 million in funding and actually wired the sewers of Albuquerque, N.M.) Others proposed to fire laser beams at homes to bring Internet access through the proverbial bathroom window. And in 2004, the New York Times wrote that "Internet service over power lines is probably a year or more away from becoming widely available." Oh, really? While both the FCC and paid industry analysts have continually predicted an "explosion" in broadband over power lines, its current market share is approximately 0.008 percent.
Each of these ventures proved a dismal failure—with the exception of satellite service in rural areas, no competitor to DSL and cable has gotten far in the United States. The startups have run into the oldest problem in the regulated industries book: the barriers to entry created by sunk costs. The phone and cable companies have already recovered the initial billions they've spent over decades, making it possible to set prices at levels that cannot be matched. The competitors brought weak products that were not substantially different. Against that kind of competition, the newcomers never stood a chance.
So much for the market solution—how about government? The failures of other ventures made municipal Wi-Fi seem an ideal alternative. After all, cities provide their citizens with water and garbage pickup—why not the Internet, too? A few important points seemed to distinguish muni Wi-Fi from the toilet robots. First, wireless skipped the whole issue of feeding wires into people's homes, the stumbling block for so many ventures. And Wi-Fi routers, if not perfect, are a proven and cheap technology. They work great on college campuses. Even my mother has a Wi-Fi router.
In 2004, when Philadelphia announced it was considering deploying the first major citywide Wi-Fi system, many assumed it would be free, or near-free, just like when you get Internet access from a generous neighbor. But that kind of system, of course, would cost real public money. The city would have to pay for the deployment with no hope of return.
By 2005, it became clear that major cities didn't really want to build out Wi-Fi networks as public works projects. Instead, places like Philadelphia and San Francisco announced "private/public" partnerships. That meant giving a private company the right to build a wireless network and try to make money off of it. Often, this simply meant giving a company like Earthlink the rights to install Wi-Fi devices on street lamps and charge citizens for access. The cities then washed their hands of the issue of success or failure.
The result, as this summer has made clear, has been telecom's Bay of Pigs—a project the government wanted to happen but left to underqualified private parties to deliver. Firms like Earthlink promised too much, and the cities have stood by and watched as the firms trying to build Wi-Fi systems have twisted and died on the beachhead. This summer, Earthlink fired half of its staff, including the head of the municipal Wi-Fi division. Major projects in Chicago and San Francisco have been stopped cold, and Houston has fined Earthlink for falling behind deadlines.
Some observers blame these failures on Wi-Fi's technical limits. Wi-Fi does have serious limitations, but wireless Internet technology has worked well even on large college campuses. The deeper problem is economics. When municipal Wi-Fi became a private service, it fell into the same economic trap as the toilet robots. Private municipal wireless networks have to compete against competitors with better infrastructure who paid off their capital investments years ago.
Setting up a large wireless network isn't as expensive as installing wires into people's homes, but it still costs a lot of money. Not billions, but still millions. To recover costs, the private "partner" has to charge for service. But if the customer already has a cable or telephone connection to his home, why switch to wireless unless it is dramatically cheaper or better? In typical configurations, municipal wireless connections are slower, not dramatically cheaper, and by their nature less reliable than existing Internet services. Those facts have put muni Wi-Fi in the same deathtrap that drowned every other company that peddled a new Net access scheme.
Today, the limited success stories come from towns that have actually treated Wi-Fi as a public calling. St. Cloud, Fla., a town of 28,000, has an entirely free wireless network. The network has its problems, such as dead spots, but also claims a 77 percent use rate among its citizens. Cities like St. Cloud understand the concept of a public service: something that's free, or near-free, like the local swimming pool. Most cities have been too busy dreaming of free pipes to notice that their approach is hopelessly flawed.
The lesson here is an old one about the function of government. When it comes to communications, the United States relies on a privateer system: We depend on private companies to perform public callings. That works up to a point, but private industry will build only so much. Real public infrastructure costs real public money. We already know that, in the real world, if you're not willing to invest in infrastructure, you get what we have: crumbling airports, collapsing bridges, and broken levees. Why did we think that the wireless Internet would be any different?