IP Transition: The One Percent Problem

The House Communications and the Internet Subcommittee is holding hearings Wednesday on the IP Transition (which I’ll be attending in person, BTW) so it’s timely to discuss the issues again.

The hardest thing to do in Washington is to unwind a body of regulation with a long history, and the telephone has been around forever in technology terms. In 1876, Alexander Graham Bell picked up a telephone and said: “Mr. Watson–come here–I want to see you.” In 1913, The Justice Department made an out-of-court settlement with AT&T that allowed it to operate as a government-sanctioned monopoly as long as it provided universal service within a defined territory and allowed other telephone systems to interconnect with its network to serve others. By 1984, technology changes made this arrangement untenable and the Justice Department’s anti-monopoly suit against AT&T broke the company up into a number of parts. After 12 years in which the nation’s telecom business was run out of a judge’s chambers, Congress finally passed the Telecommunications Act of 1996, imposing new obligations meant to promote competition for local and long-distance telephone service. By 2010, the Telecom Act was irrelevant as telephone users had began to end their Plain Old Telephone Service (POTS) subscriptions in favor of mobile telephony and broadband. Today POTS is not economically viable.

Mobile and broadband networks using Internet Protocol are more functionally rich than POTS, but their ascendance is threatened by the desire of regulators to preserve certain features of the regulatory framework articulated by the Kingsbury Commitment and the subsequent corpus of telecommunications law.

Public Knowledge in particular wants to impose five general principles on broadband IP networks:

  1. Universal Service
  2. Mandatory Interconnection
  3. Traditional Consumer Protections
  4. Single-network Reliability Mandates
  5. Vague Public Safety Mandates

The FCC has chartered an “IP Transition Task Force” to identify its own set of goals, mandates, and practices, and AT&T and others have proposed limited trials to identify issues that need to be resolved before a wider phase-out of POTS can commence.

What are the key issues and what is the appropriate course to follow in order to retire POTS and stimulate the full deployment and adoption of IP broadband networks? Here’s a list.

  1. Is the broadband IP infrastructure ready to replace POTS as the universal medium of electronic communication today and for the foreseeable future? If not, why not?
  2. The United States is ahead of many other countries in this process, so we will need to maintain some form of connectivity with the rest of the world as we wait for other nations to catch up. To what extent is it necessary to mimic POTS in order to maintain international connectivity, and when can we assume that IP-based communication services such as Skype, VoIP, email, and text messaging are sufficient? What is the definition of ”adequate connectivity”?
  3. To what extent does the cost of universal service fall on carriers and other service providers, and to what extent is it simply a tax that is rightfully collected from and dispensed by the treasury in order to subsidize broadband services for rural and poor people?
  4. The global IP network operates with a minimum of regulations and no interconnection mandate, and we know mandates distort market dynamics. What justification, if any, is there for introducing IP interconnection mandates? What justification, if any, is there for mandating interconnection between services such as VoIP and Skype?
  5. Are consumers adequately protected by broadband and mobile network protections regarding service transparency and billing?
  6. How do we establish reliability goals in a multiple provider, multiple mode marketplace? Why do so many “public interest” advocates pretend that the modern broadband marketplace works the same way as the old monopoly?
  7. What responsibilities does Public Safety have for connecting to IP networks at an appropriate level? Should there be an “Advanced 911” service permitting citizens to contact law enforcement by VoIP, Skype, text, email, HD voice, and video call? If not, why not?

The most pressing issue is universal service. The notion of a universal communication service spanning the nation comes down from the Constitution’s Postal Clause, a self-funding entitlement that was meant to produce revenue for the federal government. The Kingsbury Commitment extended universal service from the realm of paper and goods to electric communication, and shifted its funding to a private sector actor.

Universal service for IP does not depend on monopolies, so it’s no longer efficient to tax the elements of the IP ecosystem for its support. Many of the benefits of universal IP fall well outside the IP industry in any case: the social value of communication networks consists largely of the externalities created for those who are simply users of the network. Hence, it’s appropriate to return the responsibility for universal service to the treasury.

In most instances, citizens are able to meet their needs for IP services by dealing directly with private firms. Two exceptions to this rule arise:  1) low-income individuals, including children, who lack the means to pay market prices for services and the requisite hardware and software; and 2) moderate-income individuals living in remote areas where the costs of providing service exceed even their ability to pay.

For the first group, a straightforward subsidy program is the best solution, with real eligibility verification and service levels high enough to meet basic educational, employment, health care, and public safety needs.  The UN regards broadband affordability as five percent of income, which would be as much as $100 month for the 46 million Americans at or below the official poverty line ($23,021 for a family of four.) Subsidies in this category in the US are on a sliding scale tied to income, with current eligibility at 125% of the poverty line.

For the second group, the situation is more complicated. The two major mobile networks plan to reach 98% of the population in the near future, and satellite reaches most of the rest. Satellite has adequate performance, but limited capacity in terms of concurrent users, roughly half a million to a million connections. This will serve 0.8% of the population, leaving a gap of as much as 1.2% or four million people if LTE service contours were identical, which they probably won’t be. Hence, the 1.2% probably comes down to one percent. Thus, there will be rural service gap of 3.6 million people after LTE networks are fully deployed and satellite is fully utilized. This is small in comparison with the poverty gap, but still significant as rural residents are much, much more expensive to reach than urban ones.

A number of options and policy choices exist for filling the gap, such as:

  • Who qualifies for subsidies?
  • How are subsidies dispensed?
  • How can subsidies be reduced as technology and markets reduce the cost of broadband networks?

In general, the way forward is to incentivize the formation of private firms operating with diminishing levels of public subsidy. One important lever for rural broadband is to tie First Net, the public safety network, to public networks in areas where there is no general LTE service in the plans. Another is reverse auctions for subsidized rural operators, and yet another is low interest loans and outright grants for infrastructure projects such as additional satellites and fiber runs.

The current Connect America subsidy fund is $4.5 billion/year, which should probably be the upper limit for broadband subsidies.

Taking the state of current networks and public plans for network expansion into account, we find that the universal service gap is on the order of one percent of the population. That is to say that the combination of existing networks and planned expansion of LTE is likely to leave one percent of Americans outside the broadband coverage area. How to extend service to this one percent is the broadband universal service problem.

I hope lawmakers will ask panelists to share their visions about subsidizing this one percent. I would be inclined to raise the money to launch one more satellite and also to make First Net facilities available for commercial use; but I would prefer to run First Net over commercial facilities across the country in any case. There’s no argument for having government operate broadband networks in the US when there are so many firms with vastly superior expertise.