Monday, September 21, 2009


Making Network Neutrality Sustainable, Revisited

Today FCC Chairman Genichowski announced that the FCC's Network Neutrality Proceeding is entering the rule-making stage. This is a historic milestone, worth celebrating, but the milestone is on a road with hairpin turns. If you look directly above us, you can see we're in almost exactly the same place we used to be when the pro-competition provisions of the 1996 Telecom Act were intact and the distinction between telecommunications service and information service was meaningful, but now we are a lot lower.

Today the Internet connection provider sector renews its resistance. As long as there are no major structural changes in this sector, we can expect the resistance to last decades. We can expect it to spread from astroturf think-tanks to courts, to legislatures and state PUCs, and to Congress. We can expect it to enlist law enforcement, Hollywood, mainstream-media news, international bodies and established technology companies. We can expect today's gained ground to erode under our feet. We can expect to keep fighting a long time. Or we can use today's momentum to change the core of the entities driving the fight so it's simply not in their interests to continue it. The choice is ours.

I wrote an essay about this day sixteen months ago. It was early. Below I reproduce my essay in its entirety in the hope that now the timing is better.

by David S. Isenberg, May 29, 2007

Executive Summary: Network Neutrality
as currently conceived requires changes
in carrier behavior that are contrary to
their corporate culture and business model,
so we can expect their active opposition
even after Network Neutrality becomes law.
If carrier resistance prevails, the
Internet stands to lose its key success
factor. The Network Neutrality movement
can learn from history; the demise of
Unbundled Network Elements (UNEs) and the
ensuing collapse of telephone and Internet
competition provides a parallel.
The solution is strategy that is more
ambitious and more patient, that addresses
industry structure rather than carrier

Network Neutrality Movement vs. Carriers

I'm proud to be part of the Network Neutrality movement, which
raised the prohibition of, "any service that privileges,
degrades or prioritizes any packet . . . based on its source,
ownership or destination," from an unknown issue in 2005 to a
cause célèbre in 2006. It achieved this victory despite a
press blackout so complete that Project Censored named Network
Neutrality its #1 most under-reported story of 2006! The
Network Neutrality movement is leading a struggle for the
Internet's essence; the Internet would not be the everyday
necessity it is today, or hold promise for tomorrow, if it
were not neutral.

At the same time, I've grown concerned that Network Neutrality
rules and regulations based on constraining carrier behavior
are not sustainable as long as the carriers -- the telephone,
cable and mobile companies -- whose behavior these rules would
constrain, continue to operate according to their legacy
business model. And I've seen signs that some of the Network
Neutrality movement's leaders don't seem to take account of
how the carriers' vertically integrated business model and
special-purpose networks have shaped carrier culture. Just as
understanding the cultures of Iraq might have guided the U.S.
to a different course there, so might understanding the legacy
that motivates telephone, cable and cellular companies help us
make a neutral Internet sustainable.

The task is urgent, because as I write carriers are trialing a
new infrastructure called Internet Multimedia Subsystem (IMS)
that will embed discrimination in their entire Internet access
infrastructure. When IMS is deployed, it will effectively
prevent the return to a neutral Internet.

However, before I launch into this exploration of carrier
culture, carrier business models and how we can make the
Internet's neutrality stable and lasting, let me clearly
emphasize two things, lest my message be distorted by Network
Neutrality's opponents:
1. Network Neutrality as currently conceived is a good thing
and an important step forward.
2. The leaders of the Network Neutrality movement are heroes
who have devoted their careers to the creation of good
technology policy and who made miracles in 2006.

How Carriers Understand the Internet Threat

In the waning hours of 2006, during the FCC'S negotiations on
AT&T's merger with BellSouth, Network Neutrality advocates
fought hard and won several very important concessions.
However, in post-negotiation discussions, they adopted a
"talking point" to the effect that Network Neutrality would
not hurt the giant merged telephone company's business
interests. One of the movement's negotiators said, "The
conditions placed on this merger will show irrefutably that
Network Neutrality and phone company profits are not mutually
exclusive." Another said, "The fact that AT&T reported nearly
$2 billion in profits, up 17% from a year ago, double-digit
growth in earnings per share, growth in residential lines
should put to rest any concerns that Network Neutrality
requirements will harm AT&T's growth now or in the future."

At best, the talking point is inaccurate, because we have not
yet seen systematic Internet Discrimination or its effects on
carrier profits, but I think it points to deeper
misunderstanding. Carriers don't spend $1.5 million a week as
they did in 2006 lobbying against Network Neutrality unless
they believe they will be harmed by it! I think the carriers'
belief is correct; Network Neutrality rules strong enough to
keep the Internet neutral will indeed weaken their business.
(I don't think that's a bad thing provided we can figure out
other ways to provide Internet access.) I've been saying for
a decade that the Internet is incompatible with telephone
companies in their current form
Then I warned (with David Weinberger) that there's an
untenable paradox when carriers that are built on a legacy of
special-purpose networks sell plain, neutral Internet
connectivity . In 2002, many of my
colleagues and I wrote to the FCC urging that it should avoid
propping up incumbent carriers and let them fail fast so new,
more Internet compatible operating models might emerge The conclusion of
this work for carriers is that the neutral, stupid, end-to-end
Internet is such disruptive technology that they must denature
it or face the risk that it could weaken them and ultimately
put them out of business.

A decade ago the big telephone companies were complacent about
the Internet. Now they see Internet applications beginning to
have revenue impact on their core businesses. Skype, for
example, is an Internet telephony application that is capable
of better voice quality than telephony, with useful features
impossible for a conventional telephone company to deploy. It
isn't tied to the telephone company's network and it can run
on any Internet connection. In a similar manner, video
applications such as Vuze provide disruptive Internet
alternatives to conventional cable-based video services, and
wi-fi appliances using Voice over Internet Protocol promise to
disrupt the mobile telephony sector.

In the early 2000s carriers began to understand the threat.
Carrier executives started speaking publicly about it several
years before Ed Whitacre's famous complaint about how popular
Internet applications are using "his pipes" for free. In
2003, for example, AT&T CEO Dave Dorman complained, "Email is
a feature that nobody pays for," and called for the
restoration of "network resident" applications. Coincidentally, 2003
marked the first of three carrier milestones that rolled back
their obligations to provide a neutral Internet. These were
the FCC triennial order of 2003 (which lightened key public
obligations on the installers of local access fiber), the
Supreme Court's Brand X decision in 2005 (which lightened many
public obligations of cable owners) and the subsequent FCC DSL
order (which lightened the public obligations of DSL

Viewed against these milestones, Network Neutrality is a come-
from-behind tactical reaction that only arose after the legacy
of common carrier obligations had been hollowed out, after
critical distinctions between infrastructure and information,
carriage and content, and basic and enhanced services had been
defined into fragmentary meaninglessness, and after the
competition envisioned as better than government regulation by
the Telecom Act of 1996 had devolved to a grunch of giants.

Accordingly, we need more than legal policy if a neutral
Internet is to endure. We must address the non-neutrality of
the carriers' technological infrastructure, core business
model and, indeed, their self-concept.

The Carrier Business Model

Carriers are slow to act, but once they do, they're
relentless. Their next step, the introduction of Internet
Discrimination, is likely to take a decade, maybe two. It is
an economic imperative to them. Discrimination is built into
the special-purpose networks that are the foundation of their
business model.

For 130 years, if you wanted telephony on a telephone network,
you used the telephone company's telephony application. There
were no alternatives. Application discrimination was
automatic. Now, in contrast, on a neutral Internet connection
you can run Skype or Vonage or Gizmo or CallVantage or dozens
of other Internet telephony applications. But on Verizon's
conventional telephone network you can only run Verizon
telephony. These facts may seem obvious, but they're
important because tying the application to the underlying
network is the cornerstone of the carrier business model.

In other words, until the Internet arrived, carriers have
always sold the application and used application revenues to
operate the underlying special-purpose network. So, for
example, a cable company's core business is selling video
entertainment it chooses rather than connectivity, via its
cable, to anything, including other video entertainment! The
Internet breaks the special-purpose network based carrier
business model.

Carrier executives are now scared. In private, when I talk
about Network Neutrality with them, they talk about capital
expenditure, incomplete amortization, the loss of traditional
customers and the growing strength of application-based

It is extremely difficult for established companies to adopt a
new business model. It is not clear how companies build
successful new business models in the first place, but Eric
Beinhocker in The Origins of Wealth suggests that successful
models may come more from trial and error than from insight
and intent. Clayton Christiansen's Innovator's Dilemma
describes how businesses actively suppress innovation; at
budget time when there's a decision between improving an
established product or developing a young, risky, marginally
profitable one, it's a no-brainer. In addition, Robert
Jackall, in his study corporate culture published as Moral
Mazes, observes that under Management by Objective, bottom-up
innovation causes pain for one's boss, which, in turn, reduces
one's promotability with predictable effects on innovation.
In all cases, the larger the change, the more likely that
change will be suppressed. A new business model based on a
nondiscriminatory Internet would be difficult and risky at

Carriers see themselves as providers of telephony, video
entertainment and mobile telephony. These applications have
shaped their corporate culture, their way of doing business
and their physical infrastructure. Carriers see Internet
access as a new, supplemental business. They see their road to
profitability paved by Internet Discrimination, because
Internet Discrimination casts the Internet in terms that are
congruent with their historical, established business model.
So the carriers are intent on rolling back the legal
prohibitions against Internet Discrimination.

In addition, they're developing and testing a new network
architecture that tracks packets across the network and
enables differential packet-by- packet treatment and charging.
It is called Internet Multimedia Subsystem, or IMS. IMS is to
be the technological realization of the carriers' plan to cast
the Internet in terms consistent with their legacy business
model. Indeed, IMS will only have value to them if Internet
Discrimination is legal.

The Lesson of Unbundled Network Elements

Network Neutrality advocates should learn from history about
how the carriers work. Take, for example, their persistent
campaign to neutralize the idea of Unbundled Network Elements
(UNEs). UNEs were created under the Telecom Act of 1996 to
enable new competition. Specifically, the problem UNEs were
created to solve was that a new Competitive Local Exchange
Company (CLEC) or facilities-based Internet Service Provider
(ISP) that wanted its own network would need a massive chunk
of capital, then a period of network construction, before
seeing revenue dollar #1. So UNE rules were introduced
whereby incumbent telcos (ILECs) would make elements of their
network (elements such as local loop, switching, etc.)
available to new CLECs at prices that would allow these new
companies to offer services and earn revenues from them.

The theory was that new CLECs would build their own physical
network facilities gradually as business revenues grew. The
ILECs owed their success to their privileged role as a
monopoly with guaranteed profits because they provided a
public good, rather than to technological superiority or
competitive prowess. So the framers of the 1996 Act saw UNEs
as a reasonable way to re-distribute that public good to
introduce competition.

The ILECs saw UNEs differently. UNEs were against their
interests. UNEs enabled their competitors. Thus the ILECs
framed UNE's as an unfair taking of their private property.
And they behaved accordingly.

The ILEC influence on initial UNE rules was so heavy that even
AT&T, then a long-distance-only company, was not able to
launch a viable UNE- based local telephony business. The
conditions under which the ILECs were to offer UNEs (known in
the trade as "necessary and impair") were sufficiently
ambiguous as to be subject to endless litigation. An ILEC
could simply out- lawyer, out-appeal and out-wait new
entrants. Hundreds of small CLECs (here CLEC includes
facilities-based Internet Service Providers, or ISPs) sprang
up between 1996 and 2000 planning to use UNEs to offer network
services and grow. Virtually all of them went out of business
over the following few years as the entire UNE concept was
worn away by a constant trickle of seemingly minor technical
FCC and court decisions.

The ILECs survived even as they continued to complain that
they were selling "their" network elements, "below cost."
They had other fiscal troubles due to (a) the rapid adoption
of dial-up Internet access, (b) the equally rapid abandonment
of dial-up Internet access as customers switched to cable and
then DSL too, (c) a parallel adoption and abandonment of fax
machines, and (d) the rapid shift to mobile phones. The ILECs
were left battered but standing. The CLECs were wiped out.
In the end, some two trillion dollars in market capitalization
was destroyed.

UNEs were not the only cause of the CLECs' demise, to be
sure. Overspending, irrational exuberance, bad growth
projections, ILEC-friendly regulators, incompetent management
and even criminal behavior played a role. But the demise of
UNEs was a major and under-recognized strategic means of

In 2003, the FCC essentially eliminated UNE rules for
broadband networks. The competition envisioned by the Telecom
Act of 1996 was dead.

When Network Neutrality Dies

There is a clear parallel between UNEs and Network
Neutrality. Like Network Neutrality, UNEs were envisioned as
a fair, public-spirited means of ensuring competition. Both
ideas are actively opposed by the telcos because they are
contrary to their business interests. In other words, just as
the telcos saw UNEs as using "their" infrastructure to enable
their competitors, so do telcos and cablecos see Network
Neutrality as enabling application providers to offer "their"
applications. Like UNEs, Network Neutrality is, at inception,
already a weak compromise, and like UNEs, we can be sure that
the telcos will exploit every ambiguity, litigate every "and,"
"but" and comma, in every Network Neutrality rule and
regulation, and will not rest until Network Neutrality has
been rendered totally ineffective.

Then, just as the demise of UNEs spurred the collapse of the
entire CLEC sector, so would the collapse of Network
Neutrality gut the now- vibrant Internet applications sector.
If Network Neutrality collapses -- and history teaches us that
policy alone is not a strong enough bulwark against carriers
defending their legacy -- our carrier will stand between us
and our Internet searches, us and our private correspondence,
us and our medical information, us and our travel plans, us
and our financial transactions. When Network Neutrality goes,
eBay, Amazon, Yahoo and Google will need to fight for their
lives, and a thousand lesser- known apps and services, will be
captured, neutered, destroyed or forced into some inaccessible
corner. The walls enclosing quasi-public services like
MySpace and FaceBook will grow higher. My blog and yours will
be shoved into a "free speech zone" in some barbed-wire corner
of the Internet.

The above scenario may not play out exactly like this, but the
vector of carrier opposition to Network Neutrality is
obvious. We can expect the carriers' push against Network
Neutrality -- even after rules and regulations go into effect!
-- will be relentless.

Making Network Neutrality Sustainable

If it is to succeed, the pro-Network Neutrality campaign must
be as persistent and forward- looking as the carriers'.

I am skeptical about the long-term viability of simply
prohibiting Internet Discrimination. The occurrence of
discrimination might be hard to establish, and carriers might
see penalties as just a cost of doing business. More likely,
exigencies will arise -- terrorism, copyright violations, et
cetera -- that are manipulated to make broad-daylight Internet
Discrimination seem acceptable and moot even the strongest ex
ante rules and deterring penalties. So whether or not we
succeed in making Internet Discrimination illegal, we should
also take initiatives like the following:

1) We should put the concept of structural separation back on
the table! If 1.6 million save-the-Internet petitioners can
understand Network Neutrality enough to realize it applies to
them, they can understand the idea that NETWORK OPERATORS MUST
CARRY. This is a bright line. It will be obvious if carriers
cross it or obfuscate it. But instituting it will be a long-
term, come-from-behind strategic effort. It should begin now.

2) We should expand the coalition of Internet customers to
*all* users of the Internet. As Internet customers, Boeing
and GE and Monsanto, and the AFL-CIO and AARP and United
Health Care, share more interests with citizen Internet users
and Internet companies than they do with carriers. This too
must be a long-term persistent effort.

3) We should clearly frame the current telco industry
structure as monopolistic. After the mergers of MCI, AT&T and
BellSouth, US telecom competition is all but dead. The only
thing worse than a monopoly is an unregulated monopoly. Even
worse is a monopoly that sees its business threatened by
freedom, innovation, competition and technological progress
afforded by an open, neutral Internet.

4) The Network Neutrality movement should frame its advocacy
in Congress, at the FCC and in the States in terms of a
national telecommunications policy to unify what now might
seem to be independent projects, including advocacy of faster
access at lower prices, community and municipal Internet
access networks, progressive CALEA, 911 and universal service
policies that are not weighted against new competitors,
explicit and clear terms of service, regulations that permit
using any device on mobile telephone networks, and the
harmonization of U.S. spectrum policy with technological

Dilemma: The Internet Connectivity Providers Are the Anti-

The largest providers of today's Internet infrastructure are
also the strongest opponents of Network Neutrality. If their
profit stream diminishes, which it must if the Internet is to
remain neutral, stupid and open, then we weaken the
infrastructure for that which we value. This is not a new
thought, see The Paradox of the Best Network What is new is that the opposition of
the telcos and cablecos has now crystallized in a full-on
assault on the Internet's neutrality. Their end game is a
corporatized Internet that stifles freedom, democracy and
innovation incidental to reifying the telco-cableco business
model. Ultimately, the vision of the Network Neutrality
movement must encompass more than the circumscription of
certain carrier behaviors; it must be structural.

We must resolve to persist until today's dinosaurs evolve into
birds. That is, we must face the fact that if the Internet is
to survive as a neutral network, sooner or later we will need
Internet access without carriers as we know them today. So we
need to decide whether we keep the neutral Internet or we keep
today's carriers, because we won't be able to have both.

Technorati Tags:

I'm not sure if you've seen this yet, but Comcast's executive vice president, David Cohen, responded to the FCC chairman today in Comcast's blog
What's interesting here is that you could substitute "existing health insurance carriers" for "incumbent local carriers" and "affordable, universal health care" for "network neutrality" and it would work in the health care domain. For example, the last paragraph:

We must resolve to persist until today's dinosaurs evolve into birds. That is, we must face the fact that if affordable, universal health care is to exist as a human right, sooner or later we will need health care without the existing health insurance carriers as we know them today. So we need to decide whether we keep alive the idea of universal health care or we keep today's existing health insurance carriers, because we won't be able to have both.
I'm an ISP. I have no interest in the content I carry. However, onerous "network neutrality" regulations would make my network financially unsustainable.

David's argument, above, uses an old rhetorical trick which Richard Stallman also used in the preamble to the GPL. It goes roughly like this:

* First, we say that people have an inalienable right of some kind or another -- whether it is to use software or to send packets through the Internet.

* Second, employ a "pivot word," such as "freedom" or "neutrality," to that right.

* Finally, we begin to apply the same word to inanimate objects (such as software code or data packets), effectively "endowing" those (now anthropomorphized) objects with the "right" that was earlier claimed for human beings.

The net result, in the case of David's document, is to endow packets with a "right" to transit, and a "right" not to be subject to discrimination, throttling, or blocking.

The fact is that packets don't have rights; people do. And no one has an innate right to infinite amounts of any costly resource, such as Internet bandwidth, nor to a given priority level for his or her traffic on the network. Both of these things are salable commodities, and it is reasonable to expect people to pay for both.

There's no "bandwidth fairy." Bandwidth isn't free, networks cost money to build, and those of us who sweat blood building them deserve to be paid for our efforts.
Post a Comment

This page is powered by Blogger. Isn't yours?