Forgetting Covad: Bloomberg's Misbegotten Net Neutrality Counterproposal

I've been meaning to write a much more comprehensive reaction to the incredibly uninformed, misleading, and blinkered treatment that the normally reasonable Reason has provided lately on the subject of net neutrality. Their series "Don't Tread On My Internet" is full of the same sorts of half-truths and evasions that has polluted their coverage of this subject in the past, pretending that the competitive landscape is anything other than what it is (i.e. dominated by government-granted monopolies or cartels), ignoring (or minimizing) the Netflix vs. Verizon standoff while shouting simultaneously that net neutrality is "A Solution That Won't Work to a Problem That Simply Doesn't Exist" (the words of FCC commissioner Ajit Roy), and clanking on the actual history of the net and pretending that "net neutrality" was some arcane legal concept invented by an attorney rather than a legal description of the de facto operating principles of the Internet.

What I have been hoping for from them, and has thus far been conspicuously missing, is a legitimate framework for increasing competition among ISPs. Such a framework would have, as a starting point, elimination (or at least, dramatic reduction of) barriers to competition. For terrestrial carriers, that would mean loosening access to physical rights-of-way and easements; for wireless carriers, permitting of ultrawideband devices (say). Instead, we commonly see the sorts of proposals that came across my desk today from Bloomberg: the substitution of one kind of regulation for another, and then confusing it with a market. It's a confused piece, which is not much of a surprise given its tendentious apparent origins:

Most Americans get broadband Internet access from their local cable company. The competition from DSL—digital subscriber line service, provided by telephone companies over copper wires—is fading. And satellite Internet, which suffers from crippling speed issues and usage caps, is relevant only in rural areas with no cable service. The lack of competition has allowed providers like Comcast and Verizon to rake in billions of dollars more from consumers vs. what they could charge in a competitive market.

Net neutrality won’t fix that. The high prices and the service horror stories—the most recent of which is Comcast’s refusal to cancel service for a man whose house burned down—will continue because net neutrality is not ambitious enough.

I am unaware of anyone making the claim that net neutrality will fix Comcast's epic, horrible customer service. Moving on:

The one crucial step the FCC could take to get us [real competition] would be to mandate local loop unbundling. That is a policy that would force the cable companies to lease access, for a price determined by the FCC, to what’s known as the last-mile connection: the copper and fiber-optic cable, switches, and local offices that connect the main arteries of the Internet to individual homes and buildings.

We’d get competition in one stroke, as new entrants would vie to provide broadband access. That competition would lower prices, improve service, and also ease the concerns about discrimination that provoked the FCC’s net neutrality mandate in the first place. In a competitive market, if Comcast throttles Netflix, then people who love Netflix can pick up the phone and arrange service with one of Comcast’s competitors. Competition would discipline Comcast’s behavior more reliably than any net neutrality regulation.

It's surprising to read anyone make such a proposal today seriously, because it completely forgets what happened to Covad, Northpoint Communications, Rhythms NetConnections, and a host of other, smaller players who were in the DSL reselling business when that was briefly a Thing at the end of the 1990's and into the early 2000's. The idea was that Covad et al. would sell ADSL services at prices below the incumbent local telcos using the latter's outside plant, i.e. physical network. This made for a bad deal for customers — who now found themselves with two entities needed to make a connection work. Worse, each entity was competing with the other for some part of the transaction. The incumbent telcos also sold ADSL services, which meant they had an incentive to screw over the "competitive local exchange carrier" (CLEC), and their customers.

This meant that the "competition" couldn't in fact last very long, and it was only two years between IPO and bankruptcy for Covad. Of course, much of the problem came from debt incurred due to prior capital expenditure building their own network, but the CLECs wilted like spring flowers in the summer heat. It also didn't help that Covad wasn't actually an ISP, merely a middleman for physical hardware. Actual ISPs reselling services disappeared, leaving Covad customers hanging. The advantages of having the ISP also own the physical connection to customers became obvious, and Covad and its customers quickly became victims. It's amazingly hard to believe someone is now, only a dozen years after Covad's bankruptcy, again making such a proposal.