The really long tail
Last month, construction was completed on a pilot project that ran fiber optic cables to 400 homes in Ottawa. Stringing fiber optic cables isn't a big deal by itself—Verizon has been running fiber to millions of homes in the US—but the Ottawa project comes with a twist: rather than providing Internet access for a monthly fee, the company plans to sell the fiber strands outright to individual homeowners.
This isn't how we're used to doing telecommunications infrastructure. Traditionally, a "last mile" copper loop, coax cable, or fiber strand has been owned by an incumbent telephone or cable company, and the customer has paid a monthly fee for telecom services. But, if the Ottawa experiment is a success, that could change.
In the future, it could become commonplace for homes to come with "tails." These customer-owned, fiber-optic connections would link them to a network peering point. Without the expense of rolling out last mile infrastructure to every home, many more ISPs could afford to serve a given neighborhood by running wiring to the peering point, leading to more competition and lower prices. Perhaps best of all, the growth of customer-owned fiber could make debates over "open access" and network neutrality moot, as robust telecom competition should prevent the worst of the monopolistic behavior exhibited by telco and cable incumbents.
It's a tantalizing prospect, but a lot of practical difficulties will have to be overcome to make it a reality. Incumbent telecom firms may not like the prospect of increased competition, and they've never been shy about using regulatory barriers to strangle potential competitors in red tape. Meanwhile, promoters of the concept face a kind of chicken-and-egg problem: a customer-owned fiber strand is only useful if there is robust competition among ISPs at the other end, but companies aren't going to enter the residential ISP market until there's a critical mass of customer-owned fiber.
Most importantly, it's not yet clear how the economics will work. The Ottawa trial suggests that a fiber connection can cost less than $3000 per household, but the exact cost depends heavily on site geography and the rate of customer uptake. It will take several years for the concept to prove itself. And, if it does, it will take a number of additional years for it to be widely adopted.
Confronting the "last mile" problem
In the quarter century since the breakup of Ma Bell, policymakers have struggled with the "last mile" problem. Competition is good for consumers and for innovation, but rolling out redundant copper, fiber, or coax lines is wasteful—especially in dense urban areas where it requires ripping up busy streets. In the late 1990s, the FCC attempted to deal with this problem with a strategy called "open access," which forced incumbent telephone companies to "share" their infrastructure at rates dictated by the federal government.
The effort wasn't exactly a rousing success. Competitors accused the Baby Bells of foot-dragging and obstructionism. The Baby Bells, in turn, complained bitterly that the prices set by the FCC were too low to recover their costs. They also noted that they had little incentive to invest in infrastructure upgrades because any new infrastructure would have had to be shared with competitors.
After years of litigation and regulatory uncertainty, in 2005 the Supreme Court confirmed that cable incumbents were not subject to open-access rules. Soon afterwards, the FCC threw in the towel on open access for the Baby Bells as well. The result was the now-familiar duopoly: most American consumers can now get broadband service from their local phone company or their local cable company, but that's it.
The new arrangement is an undeniable improvement over the old Bell monopoly, but it still has plenty of problems. In his landmark paper that coined the phrase "network neutrality," Tim Wu positioned network neutrality regulations as a better strategy for achieving the goals of open access regulation. As it became clear that the open access model was failing, some activists shifted their focus to demanding the passage of "network neutrality" regulation.
But network neutrality regulation has proven no less controversial. Critics, including me, warn of the dangers of putting the Internet under the control of the political process. Just as uncertainties in the 1996 Telecom Act led to almost a decade of costly litigation, so too could the vagueness of network neutrality regulation could lead to a decade of litigation and the investment-stifling uncertainty that comes with it. Moreover, there are plenty of historical examples of regulatory schemes with unintended consequences. Most famously, the Interstate Commerce Commission, the nation's first modern regulatory agency, did more to protect the railroads from competition than to protect consumers from the railroads. There is a real danger that incumbent broadband providers could find ways to manipulate network neutrality regulations to their advantage, just as the railroads did a century ago.
The result has been a broadband policy stalemate. Congress considered, but failed to pass, network neutrality regulation in 2006. Since taking control of Congress, the Democratic majority has shown no interest in reviving the proposal. Proponents of regulation have touted pledges from every Democratic Senate challenger to support regulation, but the prospects for action in the next Congress remain cloudy. It wouldn't be the first time that the telcos have derailed legislation they didn't like.
Customer-owned fiber
Customer-owned fiber may offer a way out of this regulatory morass. It's hard to believe today but, as Google's Derek Slater points out, there was a time when the phone company owned the entire telephone network, including the wires inside our homes and the phone on our desks. Shifting the demarcation point to the outside of our homes created a vibrant market for customer premises equipment: not just telephones, but modems, fax machines, answering machines, and other specialized gear. With customer-owned fiber, the demarcation point would be shifted even further from the customer. That would once again mean more responsibility for the customer, but with offsetting benefits that could flow from greater competition.
Slater is working with Tim Wu on a forthcoming paper that discusses the customer-owned fiber model. Under their proposal, each customer would pay the up-front costs of stringing fiber to her own home. The fiber would terminate at a carrier-neutral colocation facility, where a variety of ISPs would have a presence. Because customers, rather than any given ISP, would own the "last mile" to the customer's house, many more ISPs could afford to compete on an equal footing for customers' business.
reader comments