alone. It creates value by enabling us to
do things we could not do otherwise, or
to do things more efficiently. Applications are the tools that enable us to realize this value. Through the applications
and content it offers, the Internet has
enabled us to become more productive
in our professional and private lives, to
interact with relatives, friends, and complete strangers, to get to know them,
communicate, or work with them, focusing on anything we like, to educate
ourselves using a wide variety of sources, and to participate in social, cultural,
and democratic discourse. In the process, it has spurred economic growth,
improved democratic discourse, and
created a decentralized environment
for social and cultural interaction in
which anyone can participate.
ISPs ability to discriminate changes
this. In a world without network neutrality rules, ISPs determine which applications and content can become
successful, distorting competition in
the markets for applications and content. As we have seen, who network
providers decide to support and who
they decide to exclude may be motivated by a number of factors that are not
necessarily aligned with users’ preferences, leading to applications that users would not have chosen and forcing
users to engage in an Internet usage
that does not create the value it could.
If I am working on an open source project that uses Bit Torrent to distribute its
source code, and the network provider
chooses to single out BitTorrent to
manage bandwidth on its network, I am
unable to use the application that best
meets my needs and use the Internet in
the way that is most valuable to me. If
I am interested in content that my network provider has chosen to restrict,
my ability to educate myself, contribute to a discussion on this subject, and
make informed decisions is impeded.
Instead, ISPs gain the power to shape
public discourse based on their own interests and idiosyncratic content policies. In addition, the risk of being cut
off from access to users at any time and
at the sole discretion of the network
provider reduces independent innovators’ incentive to innovate and their
ability to secure funding. Throughout
the history of the Internet, successful
applications such as email, the Web,
search engines, or social networks have
may have an
incentive to block
that threatens the
or does not comply
with the network
been developed by independents, not
network providers. By threatening the
supply of all those exciting new applications that have not even been thought
of yet, discrimination by network providers reduces the Internet’s ability to
create even more value in the future.
But do we really need regulation?
That competition will solve any problems, should they exist, is a common
argument against network neutrality. If
AT&T blocks an application that its customers want to use, the arguments goes,
customers will switch to another provider that lets them use the application.
This argument comes in many flavors: Some, like many European regulators, use it to argue that the problems
that network neutrality is designed
to address are caused by the concentrated market structure in the U.S.,
but are not relevant to European countries that, due to open access regulation, have more competition among
ISPs than the U.S. Others, particularly
in the U.S., argue that governments
should focus on increasing competi-
e In the U.S., most residents have a choice between at most two providers, the local telephony company and the local cable modem provider (residents in 34% of ZIP codes in the U. S.
have only one or zero cable modem or ADSL
provider who serves at least one subscriber living within the ZIP code). These providers are
not required to (and generally do not) let independent ISPs offer Internet services over their
infrastructure. By contrast, Europeans often
have the choice between cable and DSL services, and can choose among a number of ISPs
offering their services over the DSL network.
tion among ISPs instead of enacting
network neutrality rules.
These arguments neglect a number
of factors that make competition less
effective in disciplining discriminatory conduct than one might expect.
First, if all network providers block the
same application, there is no provider
to switch to. For example, in many
countries all mobile network providers
block Internet telephony applications
to protect their revenue from mobile
voice services, leaving customers who
would like to use Internet telephony
over their wireless Internet connection
with no network provider to turn to.
Second, customers do not have an
incentive to switch if they do not realize the network provider interferes
with their preferred application. If
network providers secretly slow down
packets or use methods that are difficult to detect, their customers may attribute an application’s or Web site’s
bad performance to bad design, and
happily switch to the network provider’s supposedly superior offering.
Third, finding another ISP and
making the switch requires significant
time, effort, and money, reducing customers’ willingness to switch. All this
suggests that while increasing competition is good for other reasons, it is
not a substitute for a robust network
Finally, some argue that allowing
network providers to discriminate
against applications and content is
necessary to foster broadband deployment. This argument concedes that
network providers have an incentive to
discriminate to increase their profits.
By removing the ability to discriminate, network neutrality rules reduce
network providers’ profits.
Fewer profits may mean less money
to deploy broadband networks. I am
not convinced that network neutrality
rules would reduce ne twork providers’
profit enough to push deployment incentives beyond the socially efficient
level, or that network providers would
really use the additional profits to deploy more networks instead of using
the money to please their shareholders.
f After all, network providers would still be able
to offer their own applications or content, but
they would not be able to give them an advantage over competing products.
feBRuaRY2009 | vol. 52 | No. 2 | CommunICatIons of the aCm