The Progress & Freedom Foundation Blog

Bennett's Going Mobile paper is an important contribution to the growing net neutrality literature. It elaborates on his excellent earlier paper, "Designed for Change: End-to-End Arguments, Internet Innovation, and the Net Neutrality Debate," in describing how the Internet was "designed for change," and giving reasons why heavy-handed net neutrality mandates are particularly inappropriate for the mobile Internet.

The Mobile Internet, like the Internet as a Whole, Remains a Work in Progress - It Needs to Continue to Change

Going Mobile traces the historical development of the fixed-line Internet and mobile telephone networks, and identifies points of friction where these two very separate worlds collide in today's nascent mobile Internet market.

The question raised is this: who is in the best place to resolve these frictions, an administrative agency like the FCC, by way of notice-and-comment proceedings, or network engineers, operators, and applications developers working cooperatively through standards setting or other dispute resolution bodies?

Bennett correctly suggests a "co-regulatory" approach that leaves resolution of network engineering and operational problems to the network engineers, operators and content, applications, and services developers. We should heed Bennett's admonition to let the Internet experiment continue under a "light-touch" regulatory framework.

It is unclear exactly what problem the proposed net neutrality rules are going to solve.
Over 90% of the country lives in areas that have a choice of 4 or more mobile phone providers; the mobile wireless market is effectively competition by any reasonable measure.
Bennett demonstrates that the net neutrality advocates' fear of one provider acting as a terminating access monopoly doesn't fit the wireless scenario. Mobile wireless networking offers many options for connecting to the Internet, including some "pay as you go options." Consumers therefore have alternatives if they find one provider "blocking" or slowing particular content or applications.
The FCC recognizes in its Net Neutrality NRPM that wireless networks will require a different approach than wireline due to spectrum capacity constraints, need for more intensive network management to handle mobility, roaming, and in light of different market structure and product/service provisioning.
CTIA recently filed an excellent analysis with the FCC demonstrating that the predictions of wireless net neutrality advocates that the wireless Internet market would permit consumer harms to occur in the handset, services, and applications markets have been demonstrably incorrect.
The record before the FCC remains devoid of evidence of actual anti-competitive or anti-consumer behavior on the part of mobile Internet service providers that cannot be remedied through after-the-fact enforcement of existing antitrust or consumer protection laws.
To the contrary, what we see is phenomenal growth and dynamism in a sector undergoing a rapid transformation from voice-only to data everything.
This would appear to be a particularly bad time to transition from a "light touch" to "heavy handed" regulatory approach.

Internet Engineers Fear a Lack of Change While Net Neutrality Advocates Fear Change

Another important point that Bennett raises is while Internet engineers fear a lack of change, net neutrality advocates fear change. The "best efforts" Internet built on end-to-end design concepts "just works" today, according to Bennett, and is under great strain as more and more users join and more and more bandwidth intensive applications are developed; it must change just to keep up. This is true in both the fixed-line and mobile Internet markets, but is a more pressing issue for the mobile operator due to capacity and spectrum constraints.

Going Mobile demonstrates that the "stupid network" concept of one treatment for all types of Internet traffic is neither practical nor necessary and that it fails to provide a sound basis for legal requirements.
Current Internet standards don't fully support the range of capabilities required by mobile users so the wireless engineering community is working on the necessary supplements—this is a work in progress.
According to Bennett:"It would be unfortunate if well-meaning regulators add to the Internet's general ossification by forbidding necessary and constructive changes while the question of the Internet's future architecture remains unsolved."

Bennett's Helpful Policy Suggestions on "Reasonable Network Management" and "Disclosure"

Going Mobile suggests that the way forward should entail a careful and diligent examination of historical rules and precedents with an eye to creating a new framework that will enable the next generation of networking to flourish. This is absolutely correct.

Any network transparency and disclosure rule should be written on a technology-neutral basis.
Tying requirements concerning "reasonable network management" to a disclosure requirement can help avoid nettlesome problems like requirements that are too general to be meaningful, or too specific to permit adequate operator flexibility for rapid response to conditions in the field.
As Bennett suggests, a practice can be considered presumptively reasonable if it "does what it says it does" without violating basic user rights of access to Internet content, applications and services and is adequately disclosed by the network operator.
Bennett also recommends that the FCC form a "Technical Advisory Group" to develop a consensus set of disclosure requirements.

I agree with the general approach of government collaborating with industry to develop workable requirements.
I question whether this function should be housed at the FCC, as opposed to the Federal Trade Commission. The FTC has broader, industry-wide jurisdiction and a good deal of experience developing and enforcing consumer disclosure requirements.

We now depart "Planet Engineering" and arrive on "Planet Law."

The FCC Proposes to Regulate "the Internet"

In comments filed in the Net Neutrality rulemaking, I demonstrated that from a legal perspective, the FCC's proposed rules would constitute regulation of "the Internet." This is true whether one uses the definition of "the Internet" contained in either section 230 of the Communications Act, the FCC's Net Neutrality NPRM, or an earlier FCC definition of the Internet.

Bennett provides a useful explanation of what "the Internet" is from an engineering perspective: it is a virtual network that each individual network "joins" as a "member network" rather than a discrete physical network that is "accessed." That is, a "network of networks" interconnected through the use of common protocols. This view is reflected in all relevant definition of "the Internet."

In Bennett's words: "An Internet is a virtual network (or a "meta-network") that works with networks as they are, imposing a minimal set of requirements."

The Internet does not exist apart from the Internet service provider networks.
It is not a "place" or a "thing."
It is a set of agreements.
The "end-to-end" principle means that there is no part of the path between one Internet network-connected system and another that is not part of "the Internet."

This of course is consistent with the earlier meme that the Internet existed in "cyberspace."
Like many, Bennett also describes today's Internet as an "ecosystem" that is weakly managed in terms of quality of service.
Larry Strickling, Assistant Secretary of Commerce for Communications and Information, took issue with the "ecosystem" metaphor in a speech delivered on February 24th at the Media Institute, entitled "The Internet: Evolving Responsibility for Preserving a First Amendment Miracle."

According to Strickling: "It is more accurate to describe the Internet as an agglomeration of human actors-it's a large and growing social organization."
Strickling's speech calls for the initiation of an "Internet Policy 3.0." He describes this as a collaborative, rather than regulatory, approach whereby Internet stakeholders and government officials to come together to address unresolved tensions on the Internet. This sounds like Richard Bennett's suggested "co-regulatory" model.
Internet Policy 3.0 is intended to replace the earlier "hands off" approach of the U.S. government toward the developing Internet, in recognition of the central and vital role the Internet has come to play in societies around the world.
Strickling took pains to make clear that this initiative was not aimed at instituting a heavy-handed, slow moving, backward-looking regulatory approach to rising tensions on the Internet.

Whether one visualizes "the Internet" as "cyberspace," a "virtual internetwork," an "ecosystem," or an "agglomeration of human actors," the rules that the FCC proposes to adopt in its Net Neutrality NPRM will regulate that thing in the traditional manner of FCC regulation: prescriptive rules, enforcement actions, and petitions for reconsideration or waiver, with years of litigation to follow.
And therein lay the rub for the FCC: "Internet Policy 2.0" as reflected in Section 230(b)(2) of the Communications Act continues to state that it is the policy of the United States that "the Internet" and "other interactive computer services" "remain unfettered by Federal or State regulation."

"Interactive computer services" as defined in section 230(f)(2) include "information services" and the FCC has defined broadband Internet access service provided via cable modem, wireline, powerline, and wireless networks to be information services.
No matter which way you slice this, the Act appears to preclude FCC regulation of "the Internet" and "broadband Internet access services."

This brings me to the question of the FCC's legal authority to adopt its proposed net neutrality rules.

The Self-Perpetuating Regulatory Machine

Glen O. Robinson, Former FCC Commissioner and Professor of Law Emeritus at UVA School of Law, gave a fabulous lecture on February 18th at the George Mason University School of Law entitled, "Regulating Communications: Stories from the First Hundred Years." Professor Robinson discussed the tendency of regulation to persist and spread, long after the reasons for its origins have ceased to exist and often without explicit legislative direction. Among the topics he covered was the FCC's implicit or "ancillary jurisdiction."

Professor Robinson characterized the FCC's creation of the doctrine of "ancillary jurisdiction" as "regulatory parthenogenesis" whereby regulation can reproduce without outside help in the form of legislative determination.
The creation of cable regulation 20 years before the addition to the Communications Act of Title VI, governing cable communications, is the original and primary example of this phenomenon.
Professor Robinson observed that the FCC is taking advantage of vagueness about what "the Internet" is so that it can define it in way that it can regulate it. In other words, if the Internet cannot be regulated by the FCC, then by definition what the FCC proposes to regulate cannot be the Internet.
In Professor Robinson's view an agency that can self-define its mission like this can no longer be thought of as an arm of Congress. Rather, it is a machine that, once established, goes on by itself, free of legislative limitations.
The question, according to Professor Robinson, is whether it is consistent with Congressional intent for a regulatory agency to act as if legislative limitations are simply irrelevant.
The U.S. Court of Appeals for the District of Columbia Circuit may soon be providing us an answer to the Professor's question in its review of the FCC's Comcast-BitTorrent decision.

The Jurisdictional Dilemma

Bennett's suggestions for building a better net neutrality mouse-trap leave to one side the question of who is to be doing the building, and under what body of law.

The FCC's jurisdiction to act on the proposals put forth in the Net Neutrality rulemaking is not self-evident.
The FCC's theory that it may adopt net neutrality rules under its "ancillary jurisdiction" is exceedingly weak, according to most observers and the comments filed in the response to the Net Neutrality NPRM.
Nor does it appear that there is any other jurisdictional foundation for regulation of the provision of Internet services available to the FCC under the Act.
Suggestions that the FCC simply "reclassify" broadband Internet access as Title II "telecommunications services" have a certain surface intellectual appeal. But it is far from evident that the courts would accept such an Emily Latella "Never Mind" moment from the agency.
Not to mention the view from noted investment analysts that such a regulatory about-face would, according to Stifel Nicholas' Rebecca Arbogast, "totally freak people out" (as reported Feb. 23rd by Communications Daily - subscription required) and, according to Bernstein Research's Craig Moffett, send investors "running for the hills."
Scaring off the investors necessary to fund infrastructure upgrades and deployment would completely undermine the efforts to develop a National Broadband Plan to extend broadband Internet deployment to all the people of America.

In Closing

Bennett's suggestions for what sensible rules that take economic and engineering concerns into account should include are sound, but the fundamental question remains: must these be mandated as enforceable rules by a sector-specific regulatory body or may we continue to rely on market forces, backed up by general antitrust and consumer protection laws, to protect consumer interests? I remain unconvinced that a priori FCC rules are the way to go.

Regulators should turn away from creating unnecessary regulatory constraints on a market that is functioning reasonably well today, and focus instead enabling the further growth and development of Internet services, fixed and mobile, by removing regulatory impediments.

We are already hearing about some very promising steps in this direction from the FCC's National Broadband Plan task force. With any luck, the signals from the FCC's right brain will penetrate to its left brain and signal the net neutrality rulemaking team to "stand down."