This paper was presented at the Association for Education in Journalism and
Mass Communication in San Francisco August 2006.
I am not the author. If you have questions about this paper,
please contact the author directly.
If you have questions about the archives, email rakyat [ at ]
eparker.org. For an explanation of the subject line, send email to
[log in to unmask] with just the four words, "get help info aejmc," in the
body (drop the "").
[ Bill Herman, U. of Pennsylvania ]
Opening Bottlenecks: On Behalf of Mandated Network Neutrality
Among the most contentious issues in communication policymaking in
2006 is the topic of "network neutrality," the principle that
internet service providers should generally treat all data and
applications equally. This paper defends network neutrality
regulation in the face of withering criticism by legal scholar
Christopher Yoo. This paper rebuts Yoo's claims of improved network
management and economic welfare, and it presents the case for
regulations requiring network neutrality on First Amendment and other grounds.
How do you think they're going to get to customers? Through a
broadband pipe. Cable companies have them. We have them. Now what
they would like to do is use my pipes free, but I ain't going to let
them do that because we have spent this capital and we have to have a
return on it. So there's going to have to be some mechanism for these
people who use these pipes to pay for the portion they're using. Why
should they be allowed to use my pipes? The internet can't be free in
that sense, because we and the cable companies [sic] have made an
investment and for a Google or Yahoo! (YHOO ) or Vonage or anybody to
expect to use these pipes [for] free is nuts!1
In an interview last November, SBC Telecommunications CEO Edward
Whitacre is honest about his company's market position. As half of
the broadband duopoly,2 he confesses his industry's disproportionate
market power and his intention to seek monopoly rents. SBC spokesman
Michael Balmoris insisted the company will not block consumer access
to popular websites, but Whitacre's words—uttered the same week SBC
won regulatory approval to buy AT&T—are still frightening for many.3
Other telecommunications executives have since stated either their
intention or desire to charge online content providers for the right
to reach customers at the fastest speeds.4 Coming from a firm that
faces little competition as it delivers access to the internet, a
medium of increasing importance in interpersonal and mass
communication, the threat of digital discrimination represents a
danger to important national values. Congress would be misguided to
sit idly by and allow broadband firms to begin demanding fees from
content providers; that business model would erode the potential for
internet communication to continue to foster innovation.
Nonetheless, some in Congress not only oppose such legislation, but
seek to strip the FCC of any authority to create a generalized
principle of network neutrality. A draft bill released on March 27,
2006 contains a network neutrality clause5 that would specifically
eliminate the Commission's rulemaking authority on the matter.6
Democrats,7 public interest groups,8 and online content providers9
are miffed. The bill's author, House Commerce Committee Chairman Joe
Barton, blew off this criticism as he was bragging that the bill
would pass this term and daring reporters to bet against him.10 If
this bill becomes law,11 the model that fostered breathtaking online
innovation in the U.S. may become an historical artifact.
Broadband provision is an important national goal, and broadband
service providers (BSPs) deserve the right to make a profit. Yet the
noncompetitive foreseeable future of the industry is cause for
concern. This lack of competitiveness raises a host of problems in
the development of broadband-dependent content and applications.
Broadband providers could unduly restrict what types of data may pass
over their networks, and they are already threatening to charge
content providers for the right to reach end users. Whether firms are
restricting certain types of applications or blocking certain
websites, their unique market power gives them the ability to unduly
shape the direction of broadband use into the future.
In this paper, I argue on behalf of legislation mandating
network neutrality, requiring BSPs to permit all legal,
nondestructive uses of their internet service on the same terms. As
Timothy Wu and Lawrence Lessig explain:
Stated more technically, the criteria of discrimination that
cause concern tend to
be those of the shared network, or Internet, such as
discrimination based on IP address,
domain name, cookie information, TCP port, and others. Hence,
the general principle
can be stated as follows: absent evidence of harm to the local
network or the interests of
other users, Broadband Carriers should not discriminate in how
they treat traffic on their
broadband network on the basis of internetwork criteria.12
As part of this principle, BSPs would be permitted to prevent
destructive transmissions, preserve network stability. BSPs could
continue to charge varying end user prices based on neutral measures
of bandwidth such as maximum bandwidth and total amount of uploads
and downloads, but not for the right to use certain sites or
applications. Additionally, BSPs could not demand fees from internet
content providers for the right to reach broadband customers.
While many proclaim the value of a neutral internet platform
and the legal requirements to ensure it,13 few scholars defend BSPs'
rights to discriminate among nondestructive data. Professor
Christopher Yoo is the most vocal such author. In a forthcoming
article,14 Yoo argues BSPs should be permitted to restrict users in
any way they see fit, though he contends that restrictions will
primarily be intended to manage network congestion.15 Elsewhere, he
argues that a diverse set of specialized BSP networks would be
preferable to a redundant set of general-purpose networks, and that
this anticipated positive development is hindered by a neutrality
regime.16 Unless one subscribes to a Lochneresque view of private
property rights17 or buys into the mistaken notion that BSPs should
exert some sort of editorial control over the internet,18 these two
arguments—congestion and network diversity—are two of the strongest
arguments against a neutrality regime. In arguing for that regime, I
will respond accordingly and take on Yoo throughout as the main voice
In Section II, I present a generalized description and defense
of the networking principles that under gird the calls for network
neutrality. In Section III, I discuss some of the past and likely
future instances of broadband providers placing undue restrictions on
subscribers' network uses. Fourth, I argue that the present level of
competition is insufficient to insure neutral networks. In Section V,
I demonstrate that ad hoc regulation is inadequate to the task of
stopping even the grossest anticompetitive acts of network
discrimination. Section VI briefly details a regulatory option that
could better preserve neutrality into the future. In Section VII, I
rebut the alleged disadvantages, including the argument that a
neutrality mandate would leave network administrators with too few
tools to deal with network congestion. I conclude with a brief
overview and a glimpse of how this debate fits into the broader
discussion of internet policy.
II. In Praise of Neutral Networks
Computer networks can be designed either to discriminate between
applications and data or to faithfully transmit all data regardless
of content. While a neutral network is not necessarily desirable in
every type of network architecture, society generally and internet
users specifically should and generally do prefer a neutral network.
In the first half of this section, I discuss the importance of a
neutral network in encouraging and rewarding valuable, unpredictable
online innovation. Second, I detail the role that neutrality serves
in preserving important First Amendment values such as free speech
and freedom of the press.
A. A Stable Platform for Innovation
We have clear examples of both types of network
architectures—intelligent networks designed to carry specific types
of information and nondiscriminatory, stupid networks designed to
carry any information users send. An excellent example of the former
model is the "smart" network administered by AT&T through most of the
last century. "[A]t every layer in the AT&T distributional chain, the
AT&T network had been optimized for voice telephony. But this
optimization meant that any effort to change a layer in the AT&T
distributional chain would disable other layers. … so change became
impossibly difficult."19 In contrast, those who built the internet
organized it on the latter model.20 The network is "stupid,"
faithfully carrying all data and placing the intelligence at the ends
of the network.21 While "smart" networks predestine certain uses,
stupid—or neutral—networks liberate "large amounts of innovative
energy."22 Neutral networking protocols have unleashed the explosive
growth of unforeseeable, symbiotic online innovation in the recent
past. From email to the World Wide Web to wikis to peer-to-peer
networking, the radical innovations in networking applications have
been built upon neutral internet protocols.23 It is therefore
unsurprising that an informal survey of computer technologist
websites reveals that the community is nearly unanimous in supporting
Over two decades ago, Jerome Saltzer, David Reed, and David Clark
authored a clearly articulated case for neutral networking,25 which
is still "amongst the most influential of all communication protocol
design guides."26 Network engineers still defend this design. "Stupid
Networks have three basic advantages over Intelligent
Networks–abundant infrastructure; underspecification; and a universal
way of dealing with underlying network details, thanks to IP
(internet Protocol), which was designed as an 'internetworking'
protocol."27 Infrastructure is cheaper to add, accelerating expansion
and creating abundance. Further, underspecified network architectures
and a standard internetworking protocol empower innovation:
If I have a Stupid Network and I get an idea for a communications
application, I just write it. Then I send it to my buddy, and my
buddy can install it, too. If we both like it, we can send it to more
people. If people really like it, then maybe we can charge for it -
or even start our own company. Yahoo!28
Perhaps the most significant development on the internet was the
World Wide Web, a user-friendly graphic user interface (GUI) and
effective means for computers running different operating systems to
communicate with each other. The creator of the World Wide Web, Tim
Berners-Lee, developed the Web to perpetuate a neutral network built
on end-to-end principles.29 As neutral and therefore uncontrolled
platforms, both the internet generally and the web specifically have
spawned a dazzling rate and range of innovation.
Threats to network neutrality could reduce the level and
variety of online innovation. Consider the worst-case scenario: a
system where all innovation is channeled through—and therefore must
meet the interests of—a major telecommunications firm. AT&T formerly
prohibited the attachment of all unapproved external devices to the
phone system. The effect "was to channel innovation through Bell
Labs. Progress would be as Bell Labs determined it."30 Broadband
providers are unlikely to attempt to recreate the internet on this
model. Yet even modest rollbacks of the end-to-end principle can
greatly erode the creative power of the internet. "Whatever other
closed and proprietary networks there might be, polluting the
internet with these systems of control is a certain way to undermine
the innovation it inspires."31
Consider the additional value of guaranteed neutrality from
the standpoint of innovators and the investment capitalists who fund
them. Wu and Lessig plead for a neutrality regime in order to
guarantee a stable, predictable platform on which innovators can
bank. "Their funding depends on the existence of a stable,
addressable market for their products. Such developers would benefit
the most from knowing that they can rely on a [consistent] broadband
Just as the electrical grid gives innovators a stable, consistent
system on which one can count in developing applications,33 a neutral
broadband network permits innovators to plan based on stable
expectations. This leads to greater investment in cutting-edge
applications and thus more innovation. Even minor interruptions in
the norm of neutrality, however, cause market uncertainty, leaving
investors to wonder which applications or sites will be targeted
next. This undermines the perceived future value for networking
innovations and threatens to reduce investment in research and
development and therefore reduce innovation itself.
Yoo insists BSPs will allow innovation because they are in a perfect
place to capitalize on the value of any useful progress. 34 Under
this model, however, broadband providers have a direct incentive to
allow only those innovations on which they can capitalize.35 Yoo
forgets that mandated nondiscrimination was the policy bedrock on
which the internet revolution was built:
Absent policy-mandated openness, the Regional Bell Operating
Companies (RBOCs) and monopoly franchise CATV networks would
certainly have explored only the paths of direct benefit to them. It
is doubtful that without such policy-mandated openness the internet
Revolution would have occurred. Indeed, many of the most successful
paths challenged the very core of the phone monopoly business as well
as the industry's technology and business assumptions. For example,
the internet is largely distance price insensitive, … [which] forced
profound change for the traditional telephone companies.36
If the then-current policy had permitted telephone companies to
manage network congestion by blocking or surcharging dialup access
numbers,37 the internet as we know it may not have come to pass and
certainly would not be nearly as revolutionary. This disadvantage
also applies, if not as starkly, to a scenario under which telephone
companies would have adopted Whitacre's policies of charging internet
companies more than customary telephone interconnection fees. Baby
bells would have had every incentive to choke off ISPs at rates that
gave the bells near-monopoly control of the ISP market, allowing them
to charge excessive prices and/or deliver lower quality service.
Despite the gains in online access and creativity that have already
come to pass, the lack of a neutrality mandate could still today
erode the potential for future innovation.
Yoo dismisses the applicability of precedents from the era of
telephone monopolies, even while acknowledging that they helped spur
useful innovations. As he sees it, broadband competition is too stiff
for discrimination to occur in today's market.38 The classic,
decades-old telephone attachments cases, such as Hush-a-Phone39 and
Carterfone,40 are indeed from a different technological era.
Nonetheless, perhaps no regulatory lesson from today's
telecommunications order rings louder than the resounding success of
government-mandated common carriage in spurring online innovation
over the past 15 years.41 If we are to turn our backs on this
successful strategy, we should do so only in the face of compelling
evidence not merely that the market is different, but that things are
so different as to require the exact opposite of what has worked in
the recent past. Especially in light of these past successes, Yoo's
description of "vibrant" broadband competition borders on laughable.
In almost every zip code in the US, the broadband market is highly
concentrated and certainly on no path toward meaningful competition.
This point merits further discussion, which I provide below.42
B. An Open Channel for Communication
Most of the debates over network neutrality revolve around
innovation, a neutral network is also socially valuable in that it
does not discriminate based on the moral, political, or aesthetic
value of content. A neutral network is free not only to technological
innovation but also to controversial media content that would never
be aired on older media platforms such as television and radio. When
permitted, telecommunications companies have an incentive to restrict
certain speech based exclusively on the claim that offensive content
is bad for business.43 Preserving a neutral network is therefore a
clear means of furthering First Amendment values.
The First Amendment stands for more than prohibiting
government censorship. First Amendment values are best upheld by
ensuring media diversity—not merely content diversity, but a
diversity of stakeholders who have editorial control over that
content.44 This is especially true in an era when gigantic firms with
large shares of media markets can dictate the contents of our
information ecosystem.45 For decades, the Court has held that the
health of our democracy demands "the widest possible dissemination of
information from diverse and antagonistic sources."46 While this is
subject to other First Amendment values such as editorial
discretion,47 it is upheld as the guiding principle in the current
case law regulating cable television.48 Cable companies certainly
have a reasonable claim to editorial discretion, yet they are forced
to carry certain programs in the name of a healthy local news sector
and greater net diversity of news outlets. The value of diversity is
even clearer in the case of BSPs, who disavow any editorial control
over the internet. Wu and Lessig also dismiss the idea of BSPs as
editors.49 This claim can be extended into an even clearer argument
for a neutrality regime. At least one First Amendment scholar
believes the Constitution requires state intervention when
state-created telecommunications monopolies obstruct the speech of
In contrast, Yoo insists that First Amendment values are best
upheld by permitting broadband providers to act as editors of the
internet.51 This elides the utter lack of either a general
expectation or industry-wide practice of editorial discretion on the
part of ISPs—not to mention the clause tucked into the Communications
Decency Act specifically stating that ISPs are not editors.52 It is
more useful to view each content creator or end user as her own
editor of the internet, subject to other non-ISP exceptions such as
workplace norms and content-filtering software. Especially in light
of the value that the court places on editorial diversity, the First
Amendment claim of editorial control for broadband providers is
On the uncensored internet, "just about anybody could own a
digital printing press ... and have worldwide distribution."53
Internet communicators can bypass the inherently narrower editorial
control of old media, a development embraced by authors of all
political stripes.54 Among other positive outcomes, this opens the
political game to outsiders, creating several political outcomes in
which government officials who expected little public resistance were
suddenly restrained by popular campaigns.55
The point is not that we should allow unfettered online
communication merely because that communication can and does permit
those with relatively less power to shape political outcomes; this is
just one of the most obvious positive results of a suddenly much more
equitable spread of communication power. Cast in more general terms,
we as a society should guarantee that every online communicator
serves as his or her own uncensored editor because that best upholds
the democratic values of free speech and freedom of the press. Now
that we have a communication system with the technical capacity to
support millions of independent media outlets, we should guarantee
that the editorial control over that system stays as widely
diversified as possible. A broadband provider should no more be able
to stop a customer's email or blog post due to its political content
than a telephone company should be permitted to dictate the content
of their customers' conversations.56 The guarantee that these speech
acts be legally unconstrained "is a fundamental aspect of individual
Yoo disagrees. He argues that the media diversity should, at
least generally,58 take a back seat to economic efficiency. Further,
he accuses neutrality proponents of failing to help policymakers
decide when we have "enough" diversity and the state should begin
permitting the benefits of concentration such as economic
efficiency.59 For example, he acknowledges:
There is noting [sic] incoherent about imposing regulation to promote
values other than economic welfare. The problems ... are more
practical than conceptual. Unless protecting the widest possible
diversity of sources is a virtue in and of itself that trumps all
other values, such a theory must provide a basis for quantifying the
noneconomic benefits and for determining whether those benefits
justify the economic costs.60
This is a straw-man representation of neutrality advocates
specifically and those who support diversified media ownership
generally. It is an artificially high burden of proof to expect them
to defend media diversity in the face of all other values. Yoo has
not demonstrated much risk to other constitutional values that are
considered comparable to First Amendment values. His concern is for
admittedly minor gains in economic efficiency, which is best achieved
under a neutrality regime.61 As far as the courts are concerned,
efficiency weighs little compared to a genuine First Amendment claim.
Second, the cited authors provide more than incoherent arguments on
behalf of the belief that a diversity of voices is a more important
value. For instance, in the article that Yoo cites, 62 Benkler
references an earlier article63 in which he argues:
Justice Breyer recognized that [cable] regulation "extracts a serious
First Amendment price." But, he wrote, that price can be justified by
the "'basic tenet of [our] national communications policy, namely,
that the widest possible dissemination of information from diverse
and antagonistic sources is essential to the welfare of the public.'"
That policy is not an economic policy, but rather "seeks to
facilitate the public discussion and informed deliberation, which, as
Justice Brandeis pointed out many years ago, democratic government
presupposes and the First Amendment seeks to achieve."64
When weighing the First Amendment value of increased diversity
against even other First Amendment values, communication diversity is
the core value guiding communication policy and therefore wins in
this highly analogous case. For neutrality proponents and for the
Turner court, economic efficiency is an even less important value.65
This is certainly true for other well-reasoned communication law and
policy scholars who would gladly trade economic efficiency in favor
of constitutional values such as a diverse information ecosystem66 or
privacy.67 Benkler and the Turner court may be unpersuasive to Yoo,
but many have argued quite coherently that economic efficiency is not
our country's core value.
Perhaps Yoo finds those who believe in the primacy of a
democratically diversified media system to be incoherent because he
stubbornly refuses to speak their language. For instance, as he
alleges elsewhere, "by valuing speech for its contributions to
democracy, these theories adopt a consequentialist approach that is
at odds with the autonomy-centered vision that has long dominated
free speech theory."68 Yet even the very footnote in which he makes
this claim cites an article by Baker that contends that literally any
incremental diversity is better due to the inherently more democratic
diversification of editorial power:
For many people (and most theories), true democracy implies as wide
as practical a dispersal of power within public discourse. Dispersal
of ownership also may promote the availability and consumption of
diverse content—but no theorist of whom I am aware believes that this
will always be true. But democratic values mean that it makes a huge
difference whether any lack of a particular type of diversity is
imposed by a few powerful actors or reflects the independent
judgments of many different people, for example, owners, with the
ultimate power to determine content. The key goal, the key value,
served by ownership dispersal is that it directly embodies a fairer,
more democratic allocation of communicative power.69
As far as Baker is concerned, promoting maximally democratic control
over the media is part of autonomy-enhancing democracy rather than a
cog in some "consequentialist" belief; any diversification of
communication power promotes procedural democracy. "Without more, and
regardless of empirical investigations or controversial economic
analyses, this value judgment provides a proper reason to oppose any
media merger or to favor any policy designed to increase the number
of separate owners of media entities."70 Interpretations vary, of
course, but that is not unique to the value of procedural democracy.
The pseudo-objectivity that Yoo applies to the economic question of
the BSP market's competitiveness, critiqued below, illustrates that
both core values in this debate suffer from the same problem.
Epithets of incoherence aside, Yoo is really accusing
neutrality proponents of failing to explain why their values outweigh
his. On this count, Baker provides quite solid justifications for his
reasonable policy stance: promote maximum media ownership wherever
possible until and unless other considerations prove overwhelming.71
On the other hand, Yoo himself fails quite ironically "to engage in
even a minimally adequate normative or policy analysis of the
issue."72 Yoo's rhetorical move is a clever trick, inverting the
burden of proof that he should face. Considering the almost
incomparable value of the First Amendment in the US legal canon, and
the current case law that defines that value as requiring diversity
of opinion, Yoo should be proving why economic efficiency outweighs
communications diversity in general or in this particular policy
debate. As an even less supportable debate trick, he expects those
who support diversity to prove their value claims on his terms—in a
quantitative form that translates these values into a form that can
be weighed in his economic calculus.73 Yet he offers no such calculus.
Unless Yoo believes that the democratic value of diversified
communication power could not possibly trump economic efficiency, he
has also failed to provide a coherent means of deciding which values
should win under which circumstances. He derides his opponents for
making such decisions based on an approach that "has remained
decidedly ad hoc,"74 but not even two accurate, valid systems for
measuring these abstract values75 separately could (or should)
determine which is more important under which circumstances. A human
intermediary would still have to decide upon the exchange rate
between the two currencies. Yoo's demand for a quantifiable a priori
means of resolving interminable value debates is therefore misguided
at best. After all, "it is harder to get agreement about which things
are ugly or which actions evil than about which things are
rectangular."76 If Yoo expects media policy scholars to solve the
problem of moral objectivity and create objective justifications for
First Amendment principles, he is asking them to solve a
philosophical problem deemed insoluble by some of the greatest
American philosophers of the last century.77 There may be no
quantifiable or even objective reason why Comcast should not be
granted editorial discretion over their customers' online speech, but
there are still plenty of coherent reasons.
III. Bottlenecks and Roadblocks: Actual and Potential Threats of Discrimination
There are several well-documented past and current instances of BSPs
preventing their users from making nondestructive uses of their
connections; augmenting them is a reasonable fear of content
discrimination. In perhaps the only such empirical work to date, Tim
Wu "surveyed the network designs (to the extent that the information
was available) and usage restrictions in subscriber agreements and
incorporated acceptable use policies"78 of the nation's 10 largest
cable modem and 6 largest DSL service providers as of 2002. Technical
and policy constraints have changed since then, but one thing
remains: threatened and actual discrimination are still endemic. In
the first subsection, I summarize and offer brief updates of what Wu
found in 2002. Next, I detail the continuing discrimination against
Voice over Internet Protocol, or VoIP. Third, I examine how the
threat to block specific applications such as VoIP and peer-to-peer
networking distorts the market for online innovation. Finally, I
consider content-specific threats to neutrality that may erode
customers' right to serve as their own editors.
A. Broadband Discrimination in 2002
After studying the contractual and technical constraints of the 16
largest broadband providers, Wu paints a picture of widespread
discrimination in 2002, both contractual and architectural.79 In the
part of his argument that is most relevant today, he notes three80
types of common, controversial restrictions that are embedded in
cable operators and one third of DSL operators restrict users from
acting as servers or in other substantive ways from acting as
providers of content. "This restriction has the greatest potential
significance because it affects the broadest class of
applications—those where the end-user shares content, as opposed to
simply downloading content."82 This reinforces a model of the
internet that looks like a million-channel television, favoring
one-to-many communication over many-to-many communication, eroding
the internet's power to facilitate democratic communication. After
surveying many BSPs' terms of service, this appears to be at least as
applicable today as in 2002.83
Even more commonly, BSPs often restrict commercial uses of broadband
internet connections.84 "The broadest and most controversial of such
restrictions barred home users from using 'Virtual Private Network'
(VPN) services, which are used by telecommuters to connect to their
work network through a secure connection."85 This restriction on
commercial activities is generally enacted in an effort to price
discriminate; if customers use their home connections for
business-class activities, BSPs would like to charge them
business-class prices.86 Yet this restriction, to the extent that it
is enforced, discourages a new employment model that permits
increased productivity for those who work at home,87 not to mention
the increased job satisfaction from the ability to telecommute.88
Again, this is endemic today,89 though none of the current cited
terms of service specifically ban VPNs, which are now much more
commonly used.90 Further, some BSPs still do not prohibit commercial
applications that meet the remainder of the acceptable terms of service.91
A third common contractual restriction in 2002 was the restriction
of home networking. Wu breaks this into wired92 and wireless93
restrictions, but they are similar and I will treat them jointly.
Because a user owns two computers and wants to network them to the
same broadband connection, some BSPs hypothesize that the connection
is more valuable to the subscriber and seek to charge accordingly.
While Wu's study does not systematically examine enforcement, he
notes that the then-current AT&T contract described unauthorized home
networking as theft of service and threatened to invoke criminal
punishment.94 This restriction is largely extinct.95
B. Voice over Internet Protocol (VoIP)
Since Wu's article, perhaps the most anticompetitive discrimination
has been BSPs' blocking of VoIP traffic. VoIP allows one to make and
receive phone calls over a broadband connection without paying
interstate long distance fees. Vonage, for instance, offers a VoIP
package that includes free long distance to the US and Canada for
$24.95 per month.96 For BSPs in the voice telephony business, this is
a clear cause for concern. Even Yoo objects to this type of
discrimination. "Another anticompetitive problem [that] can arise in
a convergent world is when a broadband provider bars access to an
Internet application that competes directly with its core business.
Consider Madison River Communication's attempt to protect its long
distance telephone business by blocking its DSL customers from using
VoIP."97 In the ensuing case,98 the FCC cited Madison for failing to
fulfill its duties of common carriage.99 Madison settled the case for
$15,000 and promised to stop blocking VoIP traffic on its
networks.100 While Madison appears to be holding to its end of the
bargain, other telephone companies appear to be preventing or
discouraging VoIP use on their networks. Vonage insists that two
other BSPs were still blocking their calls,101 a point to which I
C. Threatened Innovation
Until the last few years, BSPs relied on simple port blocking
to degrade or restrict disfavored applications; today, network
managers have much more sophisticated tools at their disposal. "Since
sophisticated, packet-level network-management tools allow
administrators to determine the types of traffic flowing across their
networks, it's possible for network operators to 'block' or otherwise
degrade the service for specific types of traffic."103 Blocking VoIP
is just one such threat. In another, several BSP executives have
publicly threatened to block the ports over which peer-to-peer
networks run.104 The threat to block peer-to-peer traffic is just one
example of the general problem of threatened innovation. As soon a
new application increases the value of network resources (e.g., VPNs)
or disproportionately draws upon those resources (e.g.,
peer-to-peer), BSPs may intervene either to seek rents or to minimize
their own expenses. They can demand fees from end users in relation
to the perceived value of the new technology or block
bandwidth-hogging tools in lieu of upgrading their networks. This
systematically favors the status quo, reducing the competition
between applications that led to innovations such as VoIP and
peer-to-peer networking in the first place.
Yoo denies that application suppression will likely lead to
decreases in welfare. He insists instead that network owners are in
an ideal position to capture all of the marginal value of increases
in the worth of their networks.105 Yet, even if BSPs allow all
innovations to pass and seek only to capture any increased value in
the network,106 this eliminates the profitability of future
innovations, destroying the economic incentive to innovate. Even if
BSPs can perfectly predict the extent to which these innovations will
eventually increase the value of their networks, the attempt to
capture all of the positive value of online progress constitutes a
"socially perverse"107 tax on innovation with unforeseeable and
unacceptable deadweight losses. By urging regulators to permit
discrimination, Yoo turns his back on the very policies that led to
the internet's success. The threat to peer-to-peer is merely
emblematic of what, if left unchecked, will be a looming thundercloud
over the head of generations of online innovations to come.
D. Restrictions on Content
In a final violation of network neutrality, broadband providers
explicitly reserve the right to censor the content uploaded or
downloaded by their customers. This policy statement by Cox
Communications is typical. "Cox reserves the right to refuse to post
or to remove any information or materials from the Service, in whole
or in part, that it, in Cox's sole discretion, deems to be offensive,
indecent, or otherwise objectionable."108 AT&T takes it up a notch,
reserving the right to block any content for any reason. "AT&T and
its designees shall have the right (but not the obligation) to
monitor any and all traffic routed though the Service, and in their
sole discretion to refuse, block, move or remove any Content that is
available via the Service."109 Further, in July 2005, "Telus,
Canada's second largest telecommunications company, actively blocked
access to Voices for Change, a website supporting the
Telecommunications Workers Union."110 While such censorship is
fortunately sparse, it remains the disheartening case that ISPs face
no civil liability for even willful acts of censorship.111
In the not-too-distant past, ISPs exercised fairly censorious
powers over private online speech; they regulated the content of
forums,112 private chat, and email.113 The fact that BSPs universally
reserve the right to exercise that authority over any type of online
communication carried over their pipes is unsettling, whether they
exercise that right frequently or rarely. "The system of freedom of
expression requires institutional arrangements that promote rather
than impede people's opportunities to communicate. Censorship,
whether by governmental, private, or structural forces, is
presumptively objectionable."114 Further, infrequent exercise of this
power does not disprove the essential point. As Baker observed in
relation to the abuses of media concentration generally, "[a]lthough
this power may seldom or never be exercised, no democracy should risk
the danger."115 Even competition in the market is insufficient to
guarantee that last mile providers not engage in censorship. "The
owner of the second wire is often likely to engage in the same
censorship, for the same reasons, as the owner of the first wire."116
In that light, the mere threat of BSP censorship of constitutionally
protected speech is simply unacceptable.
IV. Current Broadband Competition Guarantees Little
Yoo insists that competition in the broadband market is adequate to
prevent anticompetitive discrimination on the part of broadband
providers.117 FCC Chairman Kevin J. Martin supports the belief that
the last-mile broadband market i, though he reserves the right to
mandate neutrality should broadband providers begin placing
restrictions on users for reasons other than network management.118
Both further insist that even greater competition is just around the
corner due to technologies such as wireless (wifi) and Broadband
Power Line (BPL) transmission, even though Yoo's best-case future
scenario features at most three wireline broadband providers for most
US households.119 Any effort to label the broadband market as
competitive is wildly optimistic, to say the least. In most of the
country, one or two providers dominate the market and therefore enjoy
substantive market power over price and quality of service. While new
technologies are expected to dent this system of regional duopolies,
the era of truly vibrant competition is many years ahead under the
best scenario if it is to come at all—a condition that is hardly guaranteed.
In this section, I first demonstrate that the broadband market is
far from competitive and explain how the system of regional duopolies
discredits Yoo's primary mechanism—consumer choice—for restraining
monopolistic behavior. Second, I argue that the only free market
mechanism that could preserve a generalized norm of neutrality is the
competitive pressure of regional broadband competition at the
consumer level—and not, as Yoo suggests, the quest by website and
application developers for national market share. Third, I co-opt
Yoo's cable television analogy, as it provides an excellent policy
precedent for regulation to preserve content diversity.
A. Reigning Duopolies Gaining Speed
The first and only two vehicles for home broadband to enjoy
widespread adoption are coaxial cable, which was first deployed to
carry television signals, and Digital Subscriber Line (DSL) service,
carried over telephone lines. These vehicles currently serve almost
the entire broadband market. "Today, cable and DSL providers control
almost 98 percent of the residential and small-business broadband
market."120 Over one quarter of consumers have just one choice—cable
(23%) or DSL (5%).121 Even in well-populated markets with both
services available, most residential customers currently or will soon
face just one choice for each type of service. "In many markets,
consumers face a duopoly, forced to choose between a single cable
provider and single DSL provider—many of which bundle broadband with
television or telephone service for a pricier package."122
As measured by objective economic standards, nearly every regional
broadband market is very highly concentrated. In measuring market
concentration, the Justice Department and the Federal Trade
Commission use the Herfindahl-Hirschmann Index (HHI). To obtain the
HHI, square each firm's percentage market share and sum the squares.
For instance, consider a very optimistic scenario where four
broadband firms in a region each have one quarter of the market. By
taking the square of each firm's market share (that is, 25 squared,
or 625) and adding them all up (625 + 625 + 625 + 625), one obtains
an HHI score of 2500. Note that this is the lowest possible HHI for
four firms. If two had 40% market share each and the others had 10%,
the HHI would be 3400. An HHI between 1000 and 1800 indicates
moderate market concentration; a market over 1800 is highly
concentrated.123 The broadband market in a typical region is over
5000, explained by the FCC:
If we assume that a typical residential (and small business) market
consists of the ILEC provider, one cable provider, and one other
non-ILEC, and assume that the national figures can be used to
represent a typical local market, the HHI is approximately 5200. If
we don't allow for an additional non-ILEC and again assuming that the
national numbers of ILEC/RBOC and cable non-ILEC can be used to
calculate market shares representative of a typical local broadband
market, the HHI ranges between approximately 5500 and 5800.124
This is three times the level of competition required for a market
to be considered highly concentrated.125 "Measures of typical local
broadband markets, moreover, understate the problem because they
ignore the fact that in some local markets there is no competition at
all or, where it does exist, it is only available to some of the
customers within the market."126 If there are any, there are
certainly no more than a handful of residential broadband markets
that are truly competitive. Making policy decisions based on the
vigor of competition is therefore foolish.
The tepid competition in the broadband market will soon be
even weaker. It is technically possible for cable and
telecommunications firms to allow other BSPs to offer service over
the same set of wires. As part of the common carrier regulatory
legacy of telephony provision, telecommunications firms that sell DSL
had been required to provide access to competing BSPs. Cable
companies, in contrast, were classified as providing "information
services" and were therefore free to block competitors from using
their lines. An independent BSP, Brand X internet Services,
challenged this classification in federal court in an effort to
secure access to customers via cable lines.
Overruling the Ninth Circuit, the Supreme Court ruled that the
FCC was within its statutory rights to classify cable as an
information service and therefore exclude cable companies from common
carriage regulation.127 Within weeks, the Commission then ruled that
DSL was also an information service; less than one year from today,
current common carrier regulations requiring access to these lines
will expire and telecommunications firms such as Verizon and SBC will
be free to exclude competitors.128 "Now that these rules have been
abandoned, consumers in even the largest markets will be restricted
to two choices — the local cable provider or the local DSL provider.
This duopoly ensures higher prices, slower connection speeds and
poorer customer service."129 Considering that unregulated cable BSPs
have historically imposed more restrictions on consumers' use of
broadband connections,130 this deregulation also escalates the
likelihood that DSL operators will engage in similar discrimination.
B. Regional Market Concentration Matters
In developing the case against mandated neutrality, one of
Yoo's main arguments is that competition between broadband providers
removes the need for a neutrality mandate. He portrays a broadband
market "in which competition among providers checks anticompetitive
conduct."131 But to draw this conclusion, he relies on the premise
that concentration in the broadband market should be based on
national market share:
[A]pplication and content providers care about the total number of
users they can reach. So long as their total potential customer base
is sufficiently large, it does not really matter whether they are
able to reach users in any particular city. This point is well
illustrated by a series of recent decisions regarding the market for
cable television programming. As the FCC and the D.C. Circuit
recognized, a television programmer's viability does not depend on
its ability to reach viewers in any particular cities, but rather on
the total number of viewers it is able to reach nationwide. … This in
turn implies that the relevant geographic market is a national one,
not a local one.132
Yet earlier in the same article, Yoo acknowledges that it is customer
choice is required to drop anticompetitive behavior. "As long as
consumers have the option of switching to alternative broadband
providers, any attempt to use exclusivity to harm competition will
prove futile, since any frustrated end user will simply reallocate
their purchases to another provider."133 Yoo makes this argument
despite the clear evidence that in many areas, consumers have one or
two choices for home broadband service. The cable-DSL duopoly shows
no signs of abiding; their share of the broadband market has grown
from 94.5% in 1999 to 97.5% in 2004.134 This leaves customers with
little recourse even in light of egregious customer service,135 let
alone BSP efforts to block specific applications or websites.
Whitacre's threat to extort access fees from profitable web
businesses is also a reasonable disproof of Yoo's belief that
competition will restrain this sort of rent-seeking behavior.
High-value websites are increasingly dependant on broadband service
from regional cable and telecommunications monopolies. If
superimposed into different contexts, Whitacre's threat is laughable.
After allowing access to the general internet, no dialup provider
could have made this threat credibly. Ten full years ago, when dialup
was king and the ISP business was therefore easily entered and
fiercely competitive, America Online (AOL) was forced to allow
consumers out of their walled garden and offer general access to the
internet. AOL could no more have demanded a cut of the profits from
those whose websites were reached by their customers than Texaco
could demand a cut of the automobile industry. Like gasoline,
internet access is a homogenous commodity and if one provider blocked
access to Google (or Toyotas), customers would simply go elsewhere.
But Whitacre himself notes that he controls one of just two major
routes to broadband access in his territory.136 In those
circumstances, the threat has real teeth.
Yoo claims that it does not really matter to companies like
Yahoo! and Amazon, but those companies have vocally proclaimed that
it matters a great deal to them and the future of the internet. Along
with several other major online firms and technology trade
associations, they have formed the Coalition of Broadband Users and
Innovators explicitly and solely to lobby for network neutrality.137
For the startup VoIP companies who have fought bitterly for access
each and every time they are port blocked by broadband firms guarding
their telephony share, every user in every market is important.138 It
is further mistaken to lump all internet content into one national
market. Several prominent regional websites exist within the
boundaries of any given regional bell or cable company; giving those
broadband providers the power to choke off some of the most lucrative
customers would kill or cripple most of these sites. Most daily
newspaper websites, for instance, are of little interest to a broader
national audience and could easily lose a substantial portion of
their most lucrative audiences at Whitacre's whim.
Finally, Yoo underestimates the destructive threat that losing
even a minority of the national audience represents for application
and website developers. The computer industry is rife with network
externalities, or changes "in the benefit, or surplus, that an agent
derives from a good when the number of other agents consuming the
same kind of good changes."139 In other words, the computer industry
is filled with applications (e.g., Microsoft Office, Adobe Photoshop)
and networking systems (e.g., eBay, AOL Instant Messenger) that
become more valuable to users as more users join.140 This creates
successions of "serial" monopolies in each application or service
type.141 Once enough users decide to use such an application or
service, it enjoys near-monopoly status for years and new competitors
face a steep uphill climb, substantially undermining Yoo's claim that
the market for applications and content is of no competitive
concern.142 Even if SBC or Verizon controls only a substantive
fraction of the national broadband audience, this may be enough to
decide who does—or does not—enjoy short-term success as the serial
monopolist of the day. In this context, exclusivity arrangements are
particularly likely to have anticompetitive implications and should
therefore be avoided on economic grounds alone. The FCC recognized as
much in the AOL-Time Warner, and this economic theory "seems well
within the confines of antitrust in the new economy."143
C. The Cable Television Precedent
As part of Yoo's argument for measuring concentration based on
the national broadband market, he draws an analogy to cable
television. The statutory and economic reasoning that actually
underlies that argument, however, suggests greater rather than weaker
government protection of diversity. Citing Time Warner v. FCC,144 he
contends that a network's ability to reach a substantive national
audience is all that matters. Yet applying this case to a rebuttal of
network neutrality is misguided. First, as the Time Warner court
notes, communication policy has long been sensitive to the needs to
ensure media diversity:
Statutory authority flows plainly from the instruction that the
Commission's regulations "ensure that no cable operator or group of
cable operators can unfairly impede, either because of the size of
any individual operator or because of joint actions of operators of
sufficient size, the flow of video programming from the video
programmer to the consumer."145
Yoo leans on the FCC ruling in this case to argue that, so long as
content providers can reach a sizable national audience, local acts
of discrimination should be unproblematic.146 Yet the statute147 on
which the court relies comes to almost exactly the opposite
conclusion, demanding that no cable system provide preferential
treatment to networks in which the cable system has a stake.
Specifically, it requires that the FCC "ensure that cable operators
affiliated with video programmers do not favor such programmers in
determining carriage on their cable systems or do not unreasonably
restrict the flow of the video programming of such programmers to
other video distributors."148 Yoo is defending a broadband policy
that takes the exact opposite stance, permitting network owners to
discriminate in favor of affiliated content. This reverses the
precedent that common carriers of media content must provide
nondiscriminatory access for multiple diverse sources. Here as in
other cases, the federal government can be, should be, and is even
more vigilant against anticompetitive exclusion than in non-media industries.
V. Ad Hoc Regulation Is Inadequate
In light of admittedly problematic discrimination, as in the Madison
River case, Yoo suggests targeted FCC regulations to punish the worst
instances of discrimination.149 Others might argue that antitrust
enforcement would provide an adequate remedy, but both Yoo150 and
James B. Speta, who strongly supports neutrality,151 conclude that
antitrust regulation is inadequate. Any network neutrality regulation
should go through the FCC. The Commission can regulate either in an
ad hoc fashion or by enforcing a generalized regime of neutrality,
especially one backed new legislation. In this section, I argue ad
hoc regulation is inadequate.
Despite Yoo's enthusiasm for the Madison River case, the best and
example of the inadequacy of ad hoc regulation actually stems from
the settlement. First, consider that the implicated companies did not
fit the stereotype of VoIP port blocker. Madison River is a
telecommunications company, so blocking VoIP traffic preserves their
long distance telephone business. The two newly implicated companies,
however, are recent entrants into the voice telephony game, to say
the least. The first company accused by Vonage is Clearwire, a
company that sells long-distance wireless broadband (WiMAX152) in a
handful of states. Clearwire reached an exclusivity agreement this
March with Bell Canada to provide internet telephony over its
networks.153 The other is an unnamed cable company, which was
allegedly still successfully interfering with VoIP traffic a month
after the Madison settlement.154 Cable companies are increasingly
becoming players in the VoIP market,155 giving them an incentive to
degrade or cut off VoIP service from their competitors. In other
words, a BSP does not need to be a traditional phone company to have
an incentive to block VoIP traffic; the desire to be the only VoIP
provider on their broadband networks is incentive enough.
These companies clearly fail to meet Yoo's test for targeted
intervention, in which "a broadband provider bars access to an
Internet application that competes directly with its core
business."156 Rather, these incumbent BSPs seek to extend their
market power in the broadband business to capture potential rents in
profitable adjacent markets. Even the potential for such rent seeking
is a deterrent to the investment in and development of innovative
online applications.157 The continued discrimination against VoIP
traffic by companies that are not themselves telephone companies
shows the potential for such rent seeking in markets that are new to
a given BSP; the list of such markets will only grow.
Second, consider the utter failure of the Madison River
settlement to deter these BSPs from obstructing Vonage's calls for
competitive reasons. At least one online commentator believed that
Clearwire's certification program was an excuse to continue to block
or degrade voice traffic from competitors,158 a credible claim since
Vonage appears to have gotten its voice data through by hiding it
from Clearwire.159 Vonage has not brought complaint to the FCC over
these two latest incidents, though this may be due to a fear of
setting an unfavorable precedent. "Since Clearwire is not a
traditional telephone service provider, it is unclear what, if any,
legal recourse Vonage might have. In fact, Clearwire's terms of
service claim that its service is 'not a telephone service,' and as
such may limit users' 'rights of redress before federal, state or
local telecommunications regulatory agencies.'"160 Unlike telephone
companies such as Madison River, WiMAX and cable companies fall into
the relatively unregulated category of "information services"
providers and are therefore not subject to common carrier
regulations.161 There is certainly no guarantee that the FCC will
force them to carry competitors' voice traffic; in light of the Brand
X case, it even seems fairly unlikely. These BSPs are acting
accordingly, seeking to extent their market power into adjacent markets.
This example alone demonstrates at least the continued
potential for discrimination, which serves as a deterrent to
investment in online innovation even if actual discrimination remains
rare. Without a generalized norm of a stable platform for innovation,
provided so well by the electric grid, for instance,162 planning and
investment is less rational. The diminished potential for online
innovations that improve our collective welfare is an excellent
example of a market failure that warrants statutory and regulatory
intervention. Since unpredictability is a key element of that
failure, a principled regulatory stance is a key part of the solution.
VI. Mandating Neutrality
Wu and Lessig propose a straightforward neutrality mandate to be
adopted by the FCC. They believe the Commission has clear legal and
constitutional authority to do so.163 The regulation, which I believe
would be best implemented via new legislation, would mandate
neutrality except in a specific set of circumstances:
Broadband Users have the right reasonably to use their Internet
connection in ways which are privately beneficial without being
publicly detrimental. Accordingly, Broadband Operators shall impose
no restrictions on the use of an Internet connection except as necessary to:
(1) Comply with any legal duty created by federal, state or local
laws, or as necessary to comply with any executive order, warrant,
legal injunction, subpoena, or other duly authorized governmental directive;
(2) Prevent physical harm to the local Broadband Network caused by
any network attachment or network usage;
(3) Prevent Broadband users from interfering with other Broadband or
Internet Users' use of their Internet connections, including but not
limited to neutral limits on bandwidth usage, limits on mass
transmission of unsolicited email, and limits on the distribution of
computer viruses, worms, and limits on denial-of-service—or other
attacks on others;
(4) Ensure the quality of the Broadband service, by eliminating
delay, jitter or other technical aberrations;
(5) Prevent violations of the security of the Broadband network,
including all efforts to gain unauthorized access to computers on the
Broadband network or Internet;
(6) Serve any other purpose specifically authorized by the Federal
Communications Commission, based on a weighing of the specific costs
and benefit of the restriction.164
The increased predictability from a generalized norm of neutrality
would greatly facilitate online innovation, and BSPs would not even
be tempted to censor speech with which they disagree. Under this
mandate, network administrators would still be permitted to prevent
harmful activity, comply with legal duties, and neutrally manage
bandwidth. Considering the deep, abiding constitutional and economic
values that flourish under a neutrality regime, it would take a firm
belief in substantive disadvantages to dissuade most from agreeing
that this regime would be a good idea—at least in a market as
concentrated as the broadband market.
VII. Rebutting Alleged Disadvantages
A. Network Congestion
Recall that BSPs' supposed inability to effectively manage network
congestion is the first of Yoo's two major objections to a neutrality
mandate. He acknowledges that managing bandwidth congestion is
ideally done via usage sensitive pricing, but he concludes, "the case
for usage-sensitive pricing becomes somewhat less compelling once
transaction costs are taken into account."165 As recently as ten
years ago, transaction costs represented a major portion of a
telephone company's expenses, leading to generalized flat-rate
pricing.166 Yoo assumes the same is true of metering bandwidth,
concluding that it may be more efficient for BSPs to manage bandwidth
by discriminating based on application type.
This parallel is riddled with holes. First, he never
quantifies the expense of bandwidth metering, an ironic failure from
a scholar who expects his opponents to quantify a trade-off in
political values. He cites no networking literature to hint that
metering is cumbersome. Considering that BSPs already log their
users' web activities, this claim requires empirical support. Second,
even if the analogy with telephone metering holds, telephone
companies continue to bill by the minute for long distance and to
offer plans with cost-per-call rather than flat-rate local calling.
Third, recall that mandated neutrality for telephone users remained a
good idea despite the "unfair" network burden created by dialup
internet users. If Yoo's reasoning had ruled the day fifteen years
ago, telephone companies would have been permitted to reduce network
congestion by discriminating against dialup ISP numbers, either
preventing residential internet use entirely or seeking additional
rents from internet-using customers.167
Yoo does, however, acknowledge that metering costs may not be
prohibitive; if they are not, then network restrictions are
unwarranted. "This is not to say that all exclusivity arrangements on
the Internet are innocent. Indeed, under my approach such
restrictions would not be justified when the transaction costs of
metering bandwidth usage are relatively low."168 Both Yoo and
neutrality proponents believe that a metered regime is preferable to
one that throttles or surcharges specific applications. Payments
should ideally reflect objective measures of bandwidth, based on
total bandwidth use and/or maximum down/upload speeds. Maximum speed
is a good substitute for total use—much more so than
application-specific port blocking. Application-specific blocks can
be creatively engineered around, as demonstrated by Vonage in their
dealings with Clearwire. In contrast, instant bandwidth capacity is
an effective means of price discrimination. "[S]ervice providers can
keep endlessly upgrading their customers' connections, and use
increasing speeds as a market segmentation device. The significance
of the low utilization of data networks is that what matters to users
is not getting lots of bits, but getting a moderate amount of bits
quickly, in other words low transaction latency."169
Network owners already can and do price discriminate based on
maximum speed and/or total per-month usage. Instant bandwidth is the
more common basis for price discrimination and is becoming nearly
ubiquitous. Verizon offers DSL service at two tiers of connection
speeds,170 as does AT&T.171 Comcast offers two separate speeds of
service via cable modem.172 Verizon has also begun to deploy fiber
optic networking in limited areas, featuring three tiers of download
speeds at prices starting at $34.95 per month173 for a total of five
As Odlyzko argues, networks are generally underutilized;174
the problem is therefore not total bandwidth use but congestion
during online rush hours. Yet even if total bandwidth use matters
greatly, then network congestion can be and is managed along those
lines as well. Several BSPs, especially cable companies, enforce caps
on the total bandwidth usage per billing period. Cox Communications,
for instance, provides three tiers of service that distinguish users
based on instant bandwidth and total per-month usage.175 While
Comcast is less explicit up-front with their customers, they also
enforce caps on total bandwidth used by their customers. The trouble
of monitoring total bandwidth cannot be beyond the budgets of many
BSPs; University of Connecticut students who live in the residence
halls are subject to caps of five gigabytes of total per-week
bandwidth usage on their residential T1 lines.176 These are
profoundly captive "customers" whose service fees are built into
their boarding charges, so the school could block specific
applications such as peer-to-peer applications with little economic
loss, yet they find it perfectly feasible to enforce reasonable
network usage via a per-week bandwidth cap. Therefore, since "the
transaction costs of metering bandwidth usage are relatively low,"177
Yoo's own reasoning leads us to conclude that BSP-imposed limits on
specific applications are unwarranted.
B. Network Diversity
Yoo insists that the internet of the future may be more innovative
if networking resources are divided into a set of separate functions.
He acknowledges that the norm of the neutral network has caused the
exponential innovation of the recent past, but insists that changes
in the internet require rethinking neutrality. "There can be no
question that the Internet's meteoric success invites treating the
status quo as the relevant baseline and to place the burden on those
who would deviate from it. In recent years, however, the environment
in which the Internet operates has changed radically."178 In light of
these changes, Yoo anticipates a plethora of network designs among
last-mile providers, each optimized to a different niche market:
Indeed, it is conceivable that network diversity might make it
possible for three different last-mile networks to coexist: one
optimized for traditional Internet applications such as e-mail and
website access, another incorporating security features to facilitate
e-commerce and to guard against viruses and other hostile aspects of
Internet life, and a third that prioritizes packets in the manner
needed to facilitate time-sensitive applications such as streaming
media and VoIP. 179
Yet Yoo references little if any technical literature to support this
vision of special purpose last-mile networks.180 Quite the contrary,
one of his sources, the Blumethal and Clark piece that describes the
internet's recent changes,181 sounds a call to preserve neutrality,
not to create multiple, special-purpose networks. Here is the very
last sentence of their article: "We argue that the open, general
nature of the Net, which derived from the end-to-end arguments, is a
valuable characteristic that encourages innovation, and that this
flexibility should be preserved."182
While supporting neither the legal nor the technical regime
described by Yoo, those who support end-to-end networking generally
do acknowledge that deviations from that principle can also be
useful. "[F]rom the beginning, the end-to-end arguments revolved
around requirements that could be implemented correctly at the
end-points; if implementation inside the network is the only way to
accomplish the requirement, then an end-to-end argument isn't
appropriate in the first place."183 For instance, the authors note
that locally cached, two-stage delivery via intermediate servers is
particularly useful for streaming media content.184 Yet the potential
benefits of deviations from the end-to-end principle seminally
developed by Saltzer, Reed, and Clark do not disprove the value of
the neutrality regime proposed by Wu and Lessig. The proposed rules
would prevent BSPs from obstructing nondestructive communications; it
certainly would not prevent them from adding additional, useful
functionality such as intermediate caching. The text of the ban
itself is clear enough on this point, but Wu and Lessig's fourth
exception is even clearer. It specifically grants network providers
the power to "[e]nsure the quality of the Broadband service, by
eliminating delay, jitter or other technical aberrations."185 Within
this power, BSPs would certainly be permitted to introduce minor
impurities into the end-to-end architecture so long as they do not
degrade the general availability of a neutral communication platform.
If BSPs want to introduce tools like intermediate caching, they
certainly may do so, as long as the tools are open to all senders
Blumenthal and Clark believe BSPs, in seeking vertically
integrated business model, are looming threats to online innovation.
"The concern here, however, is that investment in closed islands of
enhanced service, combined with investment in content servers within
each island, decreases the motivation for investment in the
alternative of open end-to-end services. Once started down one path
of investment, the alternative may be harder to achieve."186 This
sincere fear rebuts Yoo's reasoning nicely; online innovation will
not be fostered by, but rather slowed by any attempt by BSPs to
create and market competing packages of "closed islands" of services.
Yoo also insists that, regardless of their ultimate shape, more BSP
networks will grow out of inefficient restrictions placed on network
traffic by incumbents.187 Yet this analysis is flawed for at least
three reasons. First, he incorrectly assumes that new BSPs will enter
and succeed due to unique packages of proprietary content and
applications. Yet as Odlyzko explains:
[T]here is far more money in providing basic connectivity, [which]
people have always valued far more, and have been prepared to pay
more for. (The far greater revenues of cellular carriers in the U.S.
than of cable TV providers is just one example…) But while content
delivery does lend itself to a closed network, connectivity does not.188
This is related to the point about network externalities; users are
far more likely to use a service or application that connects them to
their fellow users.
Second, Yoo incorrectly believes that network neutrality
requires regulated pricing.189 But this argument conflates network
neutrality with mandated interconnection of the type that was
rejected by the FCC in the Brand X decision. Note that the proposed
neutrality regime excludes mandated interconnection.190 The FCC would
not estimate the fair market price that Google would pay for the
right to be used by SBC's customers. Whitacre's wishes
notwithstanding, this price would be zero, because SBC would be
forbidden from obstructing or degrading access. In that sense, this
policy would prevent the imposition of prices where none are paid
now, but BSPs new and old would continue to charge their paying
customers whatever the market will bear.
Third, Yoo elides a substantial first-mover advantage. If a firm
enters a market first, serving as a monopoly, a second firm faces a
substantive disadvantage in entering that market.191 Because there is
far greater money in connectivity, that will always provide the
greatest incentive for new market entrants. Yet the first one or two
firms in a regional broadband market will always already have a vast
majority of the market locked up, even at inefficient prices. New
entrants can erode those profits, but they can rarely afford to
charge low enough prices to achieve a market share comparable to that
of the current monopolist.192 This is even more problematic in
industries such as the broadband market that involve substantial sunk
costs. Verizon has already laid the cables and must only maintain
them; a new BSP faces substantial build-out costs, and Verizon can
likely afford to match or beat their prices. This built-in
disincentive to new market entry erodes the potential for new market
entrants to discipline inefficient monopolistic practices. Decades of
bipartisan FCC policymaking recognized this:
Indeed, under both Republican and Democratic Administrations, the FCC
respected the efficiency and possible inevitability of natural
monopoly in the market of physical, fixed wire links to households.
... The FCC's goal has routinely been not to insist that competitors
always bypass bottlenecks, such as by building redundant local
access, but instead that bottlenecks be shared where that would be a
means to the end of competition in services offered to end users.193
Yoo's prediction, an immediate future populated by a diverse array of
broadband networks featuring highly customized features and content,
defies both history and economic logic.
VII. Concluding Bits
The principle of generalized network neutrality is responsible for
the internet revolution, and to allow BSPs to erode that principle in
the name of better profit margins is shortsighted and
anti-democratic. The continued and varied forms of discrimination are
noteworthy and regrettable. By prohibiting customers from serving
content, BSPs increase the cost of participating in online discourse,
a tax on speech. Through prohibiting commercial applications such as
VPNs (or the next innovative equivalent), BSPs drag down overall
economic growth in a partially effective attempt at price
discrimination. If we permit exclusivity arrangements with products
such as VoIP, we are explicitly allowing competition-destroying
maneuvers by incumbent utility monopolists. Through unpredictable ad
hoc bans on innovative applications such as peer-to-peer networking,
BSPs reduce the economic incentives to create powerful new networking
technologies. Finally, even if seldom exercised, the mere possibility
of content censorship on behalf of BSPs is stomach-churning for any
fan of free speech.
In the face of such actual and potential discrimination, it would be
wonderful if consumers could switch providers in a competitive
market; unfortunately, the broadband market is characterized by
regional duopolies, a problem that is going to get worse over the
next several years. Ad hoc regulation by the FCC is already failing
to discourage blatantly anticompetitive discrimination; unless the
Commission shifts policy, innovators and consumers will not be
guaranteed predictable, consistent access to broadband lines. As
outlined above, however, the Commission could provide a reasonable
guarantee that BSPs will not interfere with nondestructive
communication. This policy will not prevent BSPs from successfully
managing their networks; it provides reasonable and explicit
exceptions for preventing destructive uses, and it does not preclude
useful deviations from a pure end-to-end design such as local
caching. In an era where BSPs ranging in size from major
telecommunications firms to state universities already monitor their
users' total per-week or per-month bandwidth usage, BSPs should be
expected to impose bandwidth limits neutrally rather than picking
technological winners and losers. Further, we should set aside the
unsubstantiated hope that a diverse array of specialized BSPs will
begin to challenge incumbent duopolies with highly differentiated products.
By almost all accounts, broadband service is a commodity
market characterized by natural monopolies that serve as bottlenecks.
Decades of bipartisan regulatory tradition have forced these
bottlenecks to provide access on a nondiscriminatory basis, doing so
for purposes both economic and democratic. Among other unexpected
advantages, this tradition brought us the widespread adoption of the
internet, widely hailed as an unprecedented source of uncontrolled
innovation and uncensored speech. If Congress and the FCC do nothing
to preserve net neutrality—or, even worse, if Representative Barton's
bill becomes law—the future of the internet may be channeled through
the short-term interests of a few powerful broadband companies. In
contrast, if strong regulation forces broadband companies to leave
their bottlenecks open to all data, regardless of application or
content, the unexpected innovations in applications and content will
continue to astound us for years to come.
1 Edward Whitacre, CEO, SBC Telecommunications, interview with Roger
O. Crockett, At SBC, It's All About "Scale and Scope", BUSINESSWEEK
ONLINE, ¶ 25-26 (Nov. 7, 2005), at
2 While no one company provides broadband nationally, most customers
are effectively forced to choose among two broadband providers—a
telephone company that offers Digital Subscriber Line (DSL) service
and a cable company that offers cable modem service. This
concentration grants them economic power characteristic in
noncompetitive markets. See infra, Section IV.A.
3 Arshad Mohammed, SBC Head Ignites Access Debate, WASHINGTON POST,
D01 (Nov. 4, 2005), available at
4 FreePress, How Real Is the Threat?, at
http://www.freepress.net/netfreedom/=threat (last visited March 31, 2006).
5 H.R. ____, 109th Cong. §2 (2006), Title II, available at
6 Id. § 715(b)(1).
7 Anne Broache, Democrats Attack New Bill Over Net Neutrality, ZDNet
News, at http://news.zdnet.com/2100-9588_22-6056156.html (Mar. 30, 2006).
8 Martin H. Bosworth, Net Neutrality Gets Short Shrift In Congress,
9 Roy Mark, 'Clear and Present Danger' for Telecom Reform Bill,
InternetNews.com, at http://www.internetnews.com/bus-news/article.php/3595576.
10 John Eggerton, Telecom Bill? Bet on It, Says Barton, Broadcasting
& Cable, at
11 Both the House Judiciary Committee and the Senate may play a role
in preventing this outcome. See David Hatch, Senate Telecom Bill to
be Broader than House Counterpart, National Journal's Insider Update,
(March 29, 2006).
12 Ex parte Letter of Timothy Wu and Lawrence Lessig, Inquiry
Concerning High-Speed Access to the internet Over Cable and Other
Facilities, Declaratory Ruling and Notice of Proposed Rulemaking, 17
F.C.C.R. 4798, at 14 (2002) (Aug. 22, 2003) (CS Docket No. 02-52),
available at http://faculty.virginia.edu/timwu/wu_lessig_fcc.pdf.
13 The array of authorities cited infra represent but a small portion
of those available.
14 Christopher S. Yoo, Network Neutrality and the Economics of
Congestion, 94 GEORGETOWN L.J. (forthcoming June 2006) [hereinafter
"Yoo, Congestion"], at
15 Id. at 6-34.
16 Christopher S. Yoo, Beyond Network Neutrality, 19 Harv. J.L. &
Tech. 1 (2005) [hereinafter, "Yoo, Beyond"], available at
17 See Julie E. Cohen, Lochner in Cyberspace: The New Economic
Orthodoxy of "Rights Management", 97 MICH. L REV. 462, 463 (1998).
18 Id. at 58-61.
19 LAWRENCE LESSIG, THE FUTURE OF IDEAS, 38.
20 Id. at 39.
21 See David S. Isenberg, The Dawn of the Stupid Network, ¶11 (June
6, 1998), at: http://www.isen.com/papers/Dawnstupid.html.
22 Id. at ¶ 20.
23 See LESSIG, supra note 19, at 41.
24 See, e.g., Susan Crawford, Network Neutrality v. Platform
Competition (Oct. 30, 2005), at:
Carlo Longino, Verizon Wireless: Scrap Network Neutrality (Nov. 10,
2005), at: http://techdirt.com/news/wireless/article/6123; SBC CEO
Slammed for Comments (Nov 4, 2005), at:
http://www.dslreports.com/shownews/69175 (featuring comments
expressing particular offense at Whitacre's proposal and defending
the norm of neutrality).
25 Jerome H. Saltzer, David P. Reed, & David D. Clark, End-to-End
Arguments in System Design, 2 ACM TRANSACTIONS IN COMPUTER SYSTEMS
277 (1984), available at
26 Jean-Patrick Gelas, References: About "End-to-End" Arguments, ¶ 2
(Feb. 2004), at http://www.cs.utk.edu/~gelas/references.html.
27 Id. at ¶ 11.
28 Id. at ¶ 20.
29 TIM BERNERS-LEE, WEAVING THE WEB, 99 (1999). Berners-Lee writes:
Whether inspired by free-market desires or humanistic ideals, we all
felt that control was the wrong perspective. … Technically, if there
was any centralized point of control, it would rapidly become a
bottleneck that restricted the Web's growth, and the Web would never
scale up. Its being "out of control" was very important.
30 LESSIG, supra note 19, at 30.
31 Id. at 156.
32 Id. at 4.
33 Id. at 3.
34 Yoo, Congestion, supra note 14, at 39-40.
35 If BSPs begin charging intermediary fees, they will have an
incentive to disfavor nonmarket communication behind which there is
no sender willing to pay for delivery. To borrow from Benkler's
analysis of the cost that strong copyright protection creates for
information inputs, major commercial content creators ("Mickeys")
would be most able to pay intermediary fees, while individual and
group creators who seek no direct market remuneration ("scholarly
lawyers" and "Joe Einsteins") would be least able. This would
directly favor commercial over noncommercial content. See Yochai
Benkler, Free As the Air to Common Use: First Amendment Constraints
on Enclosure of the Public Domain, 74 N.Y.U. L. Rev. 354, 401-412 (1999).
36 Francois Bar, et. al, Defending the internet Revolution in the
Broadband Era: When Doing Nothing is Doing Harm, E-conomy Working
Paper 12, at 9 (August 1999), at
37 See Andrew Odlyzko, Pricing and Architecture of the Internet:
Historical Perspectives from Telecommunications and Transportation
(Aug. 29, 2004), 24, at
http://www.dtc.umn.edu/~odlyzko/doc/pricing.architecture.pdf. He argues:
[F]lat rates for local calling played a key role in the rise of the
Internet, by promoting much faster spread of this technology in the
U.S. than in other countries. (This, as well as the FCC decisions
about keeping Internet calls free from access charges, should surely
be added to the list of "the 10 key choices that were critical to the
Net's success," that were compiled by Scott Bradner .)
38 Yoo, Congestion, supra note 14, at 9-10.
39 Hush-a-Phone Corp. v. United States, 238 F.2d 266 (D.C. Cir. 1956).
40 Use of the Carterfone Device in Message Toll Telephone Service, 13
F.C.C.2d 420 (1968).
41 Bar, supra note 36, at 6-10.
42 See infra, Section IV.
43 C. Edwin Baker, Merging Phone and Cable, 17 HASTINGS
COMMUNICATIONS AND ENTERTAINMENT LAW JOURNAL 97, 123 (1994)
[hereinafter "Baker, Merging"].
44 See C. Edwin Baker, Media Structure, Ownership Policy, and the
First Amendment, 78 S. CAL. L. REV. 733, 734-739 (2005) [hereinafter,
"Baker, Ownership Policy"]. Baker supports media diversity in the
name of the democratic value of diffusing editorial power rather than
the mistaken belief that more diverse ownership will inherently
create more diverse content.
45 See, e.g. ROBERT W. MCCHESNEY, THE PROBLEM OF THE MEDIA: U.S.
COMMUNICATION POLITICS IN THE TWENTY-FIRST CENTURY (2004).
46 Associated Press v. United States, 326 U.S. 1, 20 (1945).
47 Miami Herald Publishing Co. v. Tornillo, 418 U.S. 241 (1974).
48 Turner Broad. Sys., Inc. v. FCC, 520 U.S. 180, 189 (1997)
[hereinafter Turner II].
49 Wu & Lessig, supra note 12, at 9. "Primarily, it is the ends—the
user of the Internet or a remote speaker—who decide on the content of
transmission, not the broadband operator. The only influence the
operator has over the content of what it carries is through the act
of restricting usage or blocking content." Id.
50 Baker, Merging, supra note 43, at 124, n. 107.
51 Yoo, Congestion, supra note 14, at 58-61.
52 47 U.S.C. § 230(c)(1) "No provider or user of an interactive
computer service shall be treated as the publisher or speaker of any
information provided by another information content provider." Id.
The same paragraph does exempt ISPs from civil claims resulting from
good-faith efforts to obstruct objectionable materials. Id. §
230(c)(2)(A). Especially when read in light of the immediately
preceding clause, however, this content-specific protection from
liability is clearly not to be confused with a recognition of
editorial rights in general.
53 DAN GILMOR, WE THE MEDIA: GRASSROOTS JOURNALISM, BY THE PEOPLE,
FOR THE PEOPLE 13 (2004).
54 See, e.g., JOE TRIPPI, THE REVOLUTION WILL NOT BE TELEVISED:
DEMOCRACY, THE INTERNET, AND THE OVERTHROW OF EVERYTHING (2004), HUGH
HEWITT, BLOG: UNDERSTANDING THE INFORMATION REFORMATION THAT'S
CHANGING YOUR WORLD (2005) (arguing that the internet uniquely
permits the dissemination of their preferred brand of left- or
right-wing views, respectively, and therefore facilitates widespread
55 See, e.g., BRUCE BIMBER, INFORMATION AND AMERICAN DEMOCRACY:
TECHNOLOGY IN THE EVOLUTION OF POLITICAL POWER 2-4 (2003) (describing
how libertarians used online communication to reverse FDIC policy),
and MCCHESNEY, supra note 45, at 280 (discussing how multiple groups
of online activists helped reverse FCC policy). The role of
information in breaking up "iron triangle" political favoritism is
well documented. See, e.g., FRANK R. BAUMGARTNER & BRYAN D. JONES,
AGENDAS AND INSTABILITY IN AMERICAN POLITICS (1993).
56 See Baker, Merging, supra note 43, at 100 (1994).
58 I make this caveat on Yoo's behalf; he details no exceptions.
59 Yoo, Beyond, supra note 16, at 53-57.
60 Id. at 54.
61 See infra, Sections IIA, IV.
62 Yochai Benkler, From Consumers to Users: Shifting the Deeper
Structures of Regulation Toward Sustainable Commons and User Access,
52 FED. COMM. L.J., 561, 565-68, 578 (2000).
63 Yochai Benkler, Free as the Air to Common Use: First Amendment
Constraints on Enclosure of the Public Domain, 74 N.Y.U. L. REV. 354, (1999).
64 Id. at 376-377 (citing Turner II, 520 U.S. 180, 226 (1997)
(Breyer, J., concurring in part)).
65 Additionally, note that economic efficiency is not necessarily the
product of unconstrained market behavior; especially in the case of
economically atypical products such as media content, a great degree
of regulation is often required to maximize efficiency. See C. EDWIN
BAKER, MEDIA, MARKETS, AND DEMOCRACY, 20 (2002) [hereinafter, "BAKER, MEDIA"].
66 See, e.g. MCCHESNEY, supra note 45, at 236; Mark Cooper, Open
Access to the Broadband Internet: Technical and Economic
Discrimination in Closed, Proprietary Systems, 71 U. COLO. L. REV.
1011, 1020 (2000). See also, generally, NEIL W. NETANEL, COPYRIGHT
AND A DEMOCRATIC CIVIL SOCIETY (JSD) (arguing that copyright law
should be crafted to maximize the health of debate in civil society,
drawing a contrast between himself and those who seek to maximize
copyright's economic efficiency).
67 OSCAR H. GANDY, JR., THE PANOPTIC SORT: A POLITICAL ECONOMY OF
PERSONAL INFORMATION (1993).
68 Christopher S. Yoo, Architectural Censorship and the FCC, 78 S.
Cal. L. Rev. 669, 675 n. 17 (2005).
69 Baker, Ownership Policy, supra note 44, at 734-735.
70 Id. at 735.
71 Id. at 734-741.
72 Id. at 741.
73 Yoo, Beyond, supra note 16, at 54, as quoted above.
Disappointingly, Yoo relies on economic analysis based in antitrust
debates that happened in other industries. He argues, for instance,
that, "Over time, courts and commentators began to recognize that
because many industries are subject to economies of scale, preserving
small producers has a price." Id. at 55. Yet this begs the question
of whether, as a society, we should or do value diversity of control
in media at a higher level than diversity in other industries—or
whether communication should actually be entirely commodified. On
this last point, see Baker, Ownership Policy, supra note 44, at 742-747.
Further, adding the production of media content to the list of
economic activities that enjoy economies of scale is more than mere
understatement; it elides properly economic reasons that justify
media exceptionalism and challenge the applicability of general
economic regulatory strategies. For almost every type of media
product, the "first copy" costs of developing and marketing something
to reproduce and distribute greatly overwhelms the costs of
reproduction and distribution. Unlike almost every other type of
product imaginable, media products as a rule feature marginal costs
that are almost always lower than average costs. GILLIAN DOYLE,
UNDERSTANDING MEDIA ECONOMICS, 13-14 (2002). This public good
characteristic of media leads to underproduction of "some media
content that an audience wants—content whose value as measured by
willingness to pay is greater than its cost." BAKER, MEDIA, at 20. It
can also lead to ruinous competition. Id. at 30-31. It is careless
for Yoo to apply rebuttals to populist antitrust reasoning without
discussing these fundamental economic differences between media
products and most other products. Further, considering the
disproportionately high degree of externalities in the media
industry, id. at 10-11, drawing on precedents primarily reached in
other industries is arguably a substantial straw-manning of those who
support media regulation that exceeds the antitrust regulation
appropriate in other sectors.
74 Yoo, Beyond, supra note 16, at 55.
75 Even antitrust suits that consider only the economic efficiency
end of Yoo's proposed two-value equation are notoriously
unpredictable. See, e.g., James B. Speta, FCC Authority to Regulate
the Internet: Creating It and Limiting It, 35 LOY. U. CHI. L.J. 15,
76 RICHARD RORTY, PHILOSOPHY AND SOCIAL HOPE, 51 (2000).
77 Rorty, id., for instance, approvingly describes John Dewey's
defense of democracy:
Dewey offered neither the conservative's philosophical justifications
by reference to eternal values nor the radical's justification by
reference to decreasing alienation. He did not try to justify
democracy at all. He saw democracy not as founded upon the nature of
man or reason or reality but as a promising experiment engaged in by
a particular herd of a particular species of animal. … Dewey's
conservative critics denounced him for fuzziness, for not giving us a
criterion of growth. But Dewey rightly saw that any such criterion
would cut the future down to the size of the present.
Id. at 119-120.
78 Tim Wu, Network Neutrality, Broadband Discrimination, 2 J.
TELECOMM. & HIGH TECH. L. 141, 156-57 (2003).
79 Id. at 143. "[O]perators indeed had implemented significant
contractual and architectural limits on certain classes of
applications. Operators showed an unfortunate tendency to want to ban
new or emerging applications or network attachments, like WiFi
devices or Virtual Private Networks, perhaps out of suspicion or an
(often futile) interest in price-discrimination." Id.
80 As described below, Wu breaks restrictions on wired and wireless
home networks into two classes of restrictions; I consider them
jointly here, thus collapsing his four categories into three.
81 Id. at 158. "The following pages provide further details on the
language of the most controversial restrictions: (1) providing
information to the public or operating a server, (2) commercial uses,
(3) Home Networking, and (4) WiFi network operation." Id.
82 Id. at 159.
83 See, e.g., "Terms and Conditions," § 8b (May, 2005), at
[hereinafter, "AT&T Terms"]; "About Cox: Policies and Agreements," §
6 (Feb. 22, 2005), at http://www.cox.com/policy/default.asp
[hereinafter, "Cox Terms"]; "Comcast High-Speed Internet Acceptable
Use Policy," § A.xiv (April, 2004), at
http://www.comcast.net/terms/use.jsp [hereinafter, "Comcast AUP"].
84 Wu, supra note 78, at 160.
86 Id. at 153.
88 Yoo and others might object that this is mere wealth transfer,
with users upgrading to commercial-class connections according to
their perceived value. Yet as discussed below in Section IV, most
broadband providers are in monopoly or duopoly markets. This suggests
that pricing for broadband generally and business-class connections
specifically will be fixed at Cournot equilibrium prices, which are
higher than competitive equilibrium prices and therefore feature
inefficient underproduction. See ROBERT S. PINDYCK & DANIEL
RUBINFELD, MICROECONOMICS (6th Ed.) 431-433 (2005). Simply, BSPs
deliberately set prices beyond the reach of many customers in an
effort to maximize profits in a noncompetitive market.
89 See, e.g., AT&T Terms, supra note 83, § 8b; Cox Terms, supra note
83, § 5; Comcast AUP, supra note 83, §A.ix.
90 This lack of explicit VPN blocks does not mean that the practice
has died. See "NetBIOS Blocked from UCI's Network" (July 23, 2004),
at http://www.nacs.uci.edu/security/netbios.html (describing the
mechanisms by which VPN administrators can circumvent ISPs' port blocks).
91 See, e.g., "Terms of Service" (2005), at
http://www2.verizon.net/policies/tos.asp? [hereinafter "Verizon
Terms"] (regrettably requiring one to login as a Verizon customer).
92 Wu, supra note 78, at 161-162.
93 Id. at 162.
94 Id. at 161.
95 See, e.g., AT&T Terms, supra note 83; Cox Terms, supra note 83;
Comcast AUP, supra note 83; Verizon Terms, supra note 91.
96 Vonage, Vonage: Leading the Internet Phone Revolution (2006), at
97 Yoo, Congestion, supra note 14, at 49. (2006)
98 Madison River Communications, LLC, Order, 20 F.C.C.R. 4295 (2005).
99 Id. (citing 47 U.S.C. § 201(b)).
100 Id. at 4297.
101 Ben Charney, "Vonage Says Its Calls Are Still Being Blocked"
(Mar. 21, 2005), at
102 See infra, Section V.
103 Paul Kapustka, "Clearwire May Block VoIP Competitors," ¶ 14
(March 25, 2005), at
104 E.g., Cynthia Brumfeld, "BellSouth: We Might Want to Block
Ports," ¶ 1-3 (Oct. 25, 2005), at
105 Yoo, Congestion, supra note 14, at 43.
106 This is a dubious claim, considering their historical willingness
to suppress innovations such as VoIP that challenge their current
107 Baker, Ownership Policy, supra note 44, at 748. Baker highlights
the difference between welfare-based economics, which seeks to
optimize total social value, and enterprise-based economics. To wit:
Of course, the enterprise's economist might be sensitive to some of
these [broader welfare values] for instrumental, but sometimes
socially perverse, reasons. The economist might check for newly
created opportunities to externalize costs cheaply or identify
someone from whom to collect (internalize) some of the enterprise's
otherwise positive externalities. Neither of these, however, and
certainly not the first, should be treated as welfare enhancing or
efficient even though beneficial to the firm.
108 Cox Terms, supra note 75, ¶ 6.
109 "Welcome to AT&T DSL Service—Terms and Conditions," § 10 (May
2005), at http://www.att.net/general-info/terms-dsl-data.html#worldnetiquette.
110 Michael Geist, Telecommunications Policy Review Submission, 5
(August 2005), at
111 See 47 U.S.C. § 230 (2005).
112 Mike Taylor, "Conversations with Fred" (Nov. 6, 1990), at
113 "Home: Censorship: Online Services,"
(last visited Dec. 6, 2005).
114 Baker, Merging, supra note 43, at 122.
115 Baker, Ownership Policy, supra note 44, at 735.
116 Baker, Merging, supra note 43, at 123.
117 Yoo, Congestion, supra note 14, at 61.
118 FCC Chairman Kevin J. Martin, Comments on Commission Policy
Statement 1 (Aug. 5, 2005), available at
[hereinafter Martin Comments on Policy Statement].
120 S. Derek Turner, Broadband Reality Check: The FCC Ignores
America's Digital Divide, 3 (Aug. 2005), available at:
121 Id. at 15.
122 Id. at 15.
123 See United States Department of Justice, The Herfindahl-Hirschman
Index, at http://www.usdoj.gov:80/atr/hhi.htm (last visited Dec. 8, 2005).
124 Amendment of Parts 1, 21, 73, 74 and 101 of the Commission's
Rules to Facilitate the Provision of Fixed and Mobile Broadband
Access, Notice of Proposed Rulemaking, 18 F.C.C.R. 6722, para. 124 (2003).
125 This makes Chairman Martin's claim of a competitive market
126 Harvey Reiter, The Contrasting Policies of the FCC and FERC
Regarding the Importance of Open Transmission Networks in Downstream
Competitive Markets, 57 FED. COMM. L.J. 243, 292 (2005).
127 National Cable & Telecommunications Assn. v. Brand X Internet
Services, 345 F.3d 1120 (9th Cir. 2003), rev'd, No. 04-277 (U.S. Jun.
128 Appropriate Framework for Broadband Access to the internet over
Wireline Facilities, Report and Order and Notice of Proposed
Rulemaking, FCC 05-150, slip op. (rel. Sept. 23, 2005) [hereinafter
DSL Ruling], available at
See also Marilyn Geewax, Bells Win Ruling on DSL Service, ATLANTA
JOURNAL-CONSTITUTION, Aug. 6, 2005, at 1F.
129 Turner, supra note 120, at 17
130 Wu, supra note 78, at 157. (2003)
131 Id. at 8.
132 Id. at 45.
133 Id. at 11.
134 Turner, supra note 120, at 12.
135 See id. at 17. I personally suffered three weeks without a dial
tone or DSL this summer, despite dozens of hours speaking with
Verizon customer service representatives and supervisors. An informal
survey of friends, colleagues, and consumer-review websites led me to
believe that I would be as likely to suffer comparably bad service at
the hands of any of Verizon's competitors—even in the relatively
competitive Philadelphia market. I would have preferred to have
switched providers entirely; my next best option to punish them for
poor service was to choose their cheapest DSL plan.
136 Again, cable BSPs are not and soon DSL BSPs will not be required
to interconnect with other would-be broadband providers as common
carriers. See DSL Ruling, supra note 128. This lack of common carrier
regulation permits infrastructure providers to price other BSPs out
137 See Press Release, Coalition of Broadband Users and Innovators,
Broadband Group Urges FCC to Ensure Consumer Freedom on the internet
(Nov. 18, 2002), at
138 See Charney, supra note 101.
139 S. J. Liebowitz & Stephen E. Margolis, Network Externalities
(Effects), ¶1, at
http://www.utdallas.edu/~liebowit/palgrave/network.html (last visited
March 31, 2006).
140 Gerald R. Faulhaber, Access and Network Effects in the "New
Economy": AOL-Time Warner (2000), in JOHN E. KWOKA, JR., & LAWRENCE
J. WHITE, THE ANTITRUST REVOLUTION: ECONOMICS, COMPETITION, AND
POLICY, 4TH ED. (2004), at 453-475.
141 Id. at 472.
142 See Yoo, Beyond, supra note 16, at 16-17.
143 Id. at 473.
144 Time Warner Entm't Co. v. FCC, 240 F.3d 1126 (D.C. Cir. 2001).
145 Id. at 1131, citing 47 U.S.C. 533(f)(2)(A) (emphasis added by the court).
146 Yoo, Congestion, supra note 14, at 45. "As the FCC and the D.C. Circuit
recognized, a television programmer's viability does not depend on
its ability to reach viewers in
any particular cities, but rather on the total number of viewers it
is able to reach nationwide." Id.
147 47 U.S.C. 533(f).
148 47 U.S.C. 533(f)(2)(B).
149 Yoo, Congestion, supra note 14, at 50.
150 Yoo, Beyond, supra note 16, at 69-70.
151 Speta, supra note 75, at 17-21.
152 See generally "About the WiMAX Forum" (2005), at
153 Bernard Simon, "Canadian Telecoms Rivals Agree Wireless Venture,"
FINANCIAL TIMES, 27 (Sep. 19, 2005).
154 Charney, supra note 101.
155 Marguerite Reardon, "Cable Goes for the Quadruple Play" (Nov. 7,
156 Yoo, Congestion, supra note 14, at 49. (2006)
157 Wu & Lessig, supra note 12, at 3-5. The authors argue:
A network that is as neutral as possible is predictable: all
applications are treated alike. Since the Commission wants to
maximize the incentives to invest in broadband applications, it
should act now to eliminate the unpredictability created by potential
future restrictions on network usage. The value of network neutrality
can be seen clearly in another context: the nation's electric
system. Because it remains neutral, the electricity network has
served as an important platform for innovation.
Id. at 3.
158 Carlo, "Clearwire To VoIP Providers: Get Certified Or, Oops, You
Might Get Blocked" (Sept. 21, 2005), at
159 Mike, "Getting Around Blocks By Playing Packet Hide and Seek"
(Apr. 22, 2005), at http://techdirt.com/articles/20050422/0946236_F.shtml.
160 Kapustka, supra note 103, ¶ 13.
161 See Inquiry Concerning High-Speed Access to the Internet Over
Cable and Other Facilities, Declaratory Ruling and Notice of
Proposed Rulemaking, 17 F.C.C.R. 4798, 4819-39 ¶¶ 33-71 (2002); Nat'l
Cable & Telecomm. Ass'n v. Brand X Internet Servs., 125 S. Ct. 2688 (2005).
162 Wu & Lessig, supra note 12, at 3.
163 Id. at 10. "Whether restricting use of the network amounts to
speech doesn't ultimately matter, for the Commission's authority is
secure in either case. The bans on discrimination that at the center
of any neutrality regime are a textbook case of a content neutral
regulation of conduct, supported by substantial government interests." Id.
164 Id. at 13.
165 Yoo, Congestion, supra note 14, at 7.
166 Id. at 27-29.
167 Yoo would object to this characterization, at least in part; he
insists that the era of healthy broadband competition is upon us and
we therefore need not bother with legal precedents set in the era of
Regional Bell Operating Company monopolies. Recall from above,
however, that local broadband markets are anything but competitive;
Yoo's hypothetical objection would therefore need to explain why a
near-total broadband duopoly is sufficiently different from total
monopoly to guarantee that the next "unfair," revolutionary use of
networking resources is permitted to thrive without threat of discrimination.
168 Yoo, Congestion, supra note 14, at 49 (2006).
169 Odlyzko, supra note 37, at 28, specifically lectures BSPs for
mistakenly seeking to create vertically integrated streaming media
centers when ever-faster broadband pipes serve the clearest route to
finely detailed price discrimination.
170 Verizon, Packages and Prices, at
(last updated 2006).
171 AT&T, Internet Services, at
http://www.usa.att.com/dsl/plans/index.jsp (last updated 2006).
172 Comcast, Savings and Services at My Address, at
http://comcast.com/Buyflow/default.ashx?PromoID=20439 (last updated 2005).
173 Verizon, Verizon FiOS: Packages and Prices, at
(last updated 2006).
174 Odlyzko, supra note 37, at 28.
175 Cox Communications, Limitations of Service, at
http://www.cox.com/policy/limitations.asp (last updated 2006).
176 University of Connecticut, Bandwidth Usage, at
http://www.security.uconn.edu/guides/bandwidth.html (last updated May
177 Yoo, Congestion, supra note 14, at 49.
178 Yoo, Beyond, supra note 16, at 21.
179 Id. at 31.
180 Having consulted innumerable online resources and several current
or former IT professionals in preparation for this paper, I concluded
that the vast majority of those with the technical skills to
develop—or even implement—the next great online innovation support a
generalized internet protocol.
181 Marjory S. Blumenthal & David D. Clark, Rethinking the Design of
the Internet: The End-to-End Arguments vs. the Brave New World, 1 ACM
TRANSACTIONS ON INTERNET TECH. 70 (2001) (cited in Yoo, Beyond, supra
note 16, at 21, n. 59).
182 Id. at 99.
183 Id. at 80.
184 Id. at 83.
185 Wu & Lessig, supra note 12, at 13.
186 Blumenthal & Clark, supra note 181, at 73.
187 Yoo, Beyond, supra note 16, at 48-53.
188 Odlyzko, supra note 37, at 28. While Odlyzko explains why open
networks will tend to win, note that the reasoning above, infra
Section I and II, demonstrates why the exceptions are both common
enough and, even when rare, bad enough to warrant intervention.
189 Yoo, Beyond, supra note 16, at 38.
190 Several of the authors cited here also support mandatory
interconnection for all ISP comers, imposed on last mile
infrastructure owners such as telecommunications and cable firms.
While that debate is also worthwhile, it is a separate debate.
191 ROBERT S. PINDYCK & DANIEL RUBINFELD, MICROECONOMICS (6th Ed.)
192 Id. at 448.
193 Reed Hundt, The Ineluctable Modality of Broadband, 21 YALE J. ON
REG. 239, 249 (2004).