The Need for Speed

roadband Internet in the United States is a disaster. Its appalling. Its embarrassing. Its preposterous. Compared to the rest of the world, our connections are slow, balky, and expensive. If you divide speed by cost, Australias Internet access is three times better than ours; Frances is nine times better; and Japans thumps us twenty-one times over. Were catching fish with our hands, while they are out in trawlers. And the reason doesnt have to do with anything intrinsically American. Its not, for example, that the country is too rural. Broadband stinks even in Chicago. The problem is almost entirely a failure of policy and of imagination.

Our tortoise-like downloading isnt just a problem because it means YouTube videos buffer and stall: it means were missing out on a whole generation of useful technology. Doctors in a Tokyo hospital routinely examine tissue samples from patients in rural areas, using high-definition video and remote-controlled microscopes. In South Korea, it is so easy to share files and upload video that a whole mini industry of video tutors has flourished. According to a report published recently by the Information Technology and Innovation Foundation, the teachers who provide the best test-prep lectures have become national celebrities.

Americans should be envious, but also worried. The countries that deploy ubiquitous high-speed broadband first are the countries that will develop the tools to use it. America built the worlds first computers, and then along came Microsoft. America pioneered the Internet, and along came Google. Its hard, however, to imagine that the technologies of the future will be hatched here. In 2001, the first year that the OECD recorded data, the United States ranked fourth in the world in a metric measuring broadband cost, speed, and penetration. Now we rank fifteenth, and are dropping steadily.

Americas problem starts with one fact: we have a total lack of competition. In most places, you can only get broadband through your cable company or through the DSL your phone company sells. Both industries have recently gone through merger madness, meaning that there is not only minimal competition between industries, theres minimal competition within industries. Youre basically locked into one of two bad options. The government tried to solve this problem in the 1996 Telecom Act, which required the incumbent monopolists to lease their infrastructure to competitors. But the law wasnt enforced, and, eventually, the incumbents succeeded in neutering it.

That lack of competition is partly caused by a structural problem. Its easy to lay fiber-optic lines across the country. These wires are the Internets backbone, and weve got plenty of them. But the fiber stops at nodes in cities or towns and doesnt actually go into peoples homes. To make that journey, its got to slog across slow copper wires. The path taken by an e-mail sent from Washington, D.C., to San Francisco isnt that different from the path taken by a traveler. The beginning and the end are really slow: you stuff bags into taxis, get stuck at stoplights, and have to wait for the shuttle at the parking lot. Then you zip over the middle three thousand miles in a jet airplane. The e-mail does the same kind of thing. It crawls out from your home, speeds across the nation at about the speed of light, and then trudges across the so-called “last mile.”

How do you make that last mile faster? The easiest way would be for the companies that have the rights and access to your home to run the same fiber networks into the house that they use to send the information across the country. To stretch the metaphor, this would be the equivalent of sending a Harrier jet to pick you up at your house and drop you off on the steps of your final destination. But the cable and phone companies dont have much incentive to do this. If youre making good money off Slim Jims, why bother trying to sell sirloin?

Optimists believe that the Obama administration can reverse the problem with the 2009 stimulus package, using the rules and regulation it attaches to the $7.2 billion targeted for providing broadband in rural areas. But while this money will create great public benefit, it wont help the United States catch South Korea. Right now, only about half of American households have broadbandmostly of the really slow variety. If 70 percent of households had broadband, that would certainly be a useful improvement. Still, connecting a slightly less embarrassing percentage of the population to the Internet at embarrassingly slow speeds isnt exactly solving the problem.

What the country really needs is a massive investment to build out the fiber lines that go to peoples homes. There are at least four ways to do this: subsidize the current providers; try again to inject competition; let users buy the fiber directly; or start a completely new company. People close to the Obama administration are thinking about all of these, and we can hope that one strategy, or a combination of strategies, will appear in February 2010 in the Federal Communication Commissions critical paper on national broadband strategy.

Massively subsidizing the existing providers is the Asian model. Japan, for example, allowed its national provider, NTT, to completely write off the expenses incurred by building its network. South Korea did something similar. Its not clear, though, that this solution is transferable to the United States. Do you really want to try to solve a problem by giving billions of dollars to the companies that created itand which have records of squandering past subsidies, and powerful incentives to squash innovative solutions? Even today, incumbent telecom lobbyists are busy trying to crush plans for municipal WiFi and city-run fiber-to-the-home experiments.

Creating competition is, in general, the way the Europeans have built their networks. And while the time for mandating line sharing (as the 1996 Telecom Act did) has passed, there are some great ideas that can be taken from the other side of the Atlantic. For example, the nonprofit group Free Press has recommended that the Obama administration adopt an idea from Switzerland and offer tax credits to companies that build multiple lines of fiber into homes. If you lay in one strand of fiber, on which youll sell your service, you dont get a credit. But if you lay in two or three, which competitors can rent, you get a huge deduction.

The third idea, floated by Columbia professor Tim Wu (who also is the chairman of Free Press), is to let customers own their own fiber connections instead of leasing them from the telecom companies. Cities or communities would decide to invest in a fiber network, which some central group would manage and maintain. Individuals, subsidized by the government, would then buy their own lines and decide what services theyd want sent over them.

The last, and probably the most likely to succeed, option is for the government to just jump into the market itself, creating a model something like the postal service. This is like what Australia has recently announced. There, a new government-controlled company is going to spend $31 billion connecting homes with fiber. It could start with fiber strung into schools and community centers, then government buildings, then public housing projects, and, eventually, every home in the country. Private companies would compete, just as UPS and FedEx compete with the postal service.