The Federal Communication Commission’s National Broadband Plan - Adoption, Usage, and Global Impact of Broadband Technologies - page 68

for email and basic web browsing at the low end,
with higher speeds recommended for customers
accessing high-bandwidth content such as video.
The ability to discriminate against (all) high-
bandwidth content (regardless of the content's
provider) can (if designed properly) be used to
further induce customers accessing this content to
upgrade their service. This increased willingness
to pay for higher-speed access increases the abil-
ity of providers to pass the cost of infrastructure
upgrades on to customers.
The ability to discriminate would also allow
ISPs to charge content providers - which in some
circumstances would increase the incentive for
infrastructure upgrades. Any ability to charge
content providers for transmission allows a new
revenue source. However, economic theory would
only predict that this would translate into a desire
on the part of the ISP to upgrade the network
if the upgrade would sufficiently increase the
revenue that could be extracted from content
providers. This is most likely to be the case if
the most desirable content to consumers (and
thus the most profitable to content providers) is
the highest-bandwidth content. While this seems
plausible on the surface, it's not clear whether
this will ultimately be the case or not. In general,
consumers seem fairly unwilling to pay for any
content, whether low bandwidth (such as news,
reference and search) or high (such as video or
voice), and few providers have been able to use
advertising to translate free content into a profit-
able business model.
Neutrality proponents discount these argu-
ments. They (correctly) point out that ISPs are not
predicted to spend any extra revenue on upgrades
just because the money is available. As described
above, economic theory would argue that the
ability to earn extra revenue from discriminating
would only encourage upgrades under particular
circumstances. Absent this ability (where Internet
access is largely commoditized) ISPs still have the
primary incentive to upgrade because customers

will still shop for speed and reliability in order to
access the content they desire.

4. Should potential net neutrality rules only
pertain to the infrastructure layer of the
Internet, or should they apply at any layer of
the Internet stack where a bottleneck exists?

Stakeholders on both sides of the issue have
demonstrated a convenient ability to compart-
mentalize their arguments. If content providers
object to ISPs having the ability to discriminate
against particular content, it seems convenient
(in the least), if not entirely disingenuous, for the
content providers to not face the same require-
ment (Szoka and Thierer, 2009). Google Voice,
for example, has recently become the target of
complaints from AT&T, after acknowledging that
the service blocks particular numbers (Schatz,
2009a). Many content sites discriminate - Google
and other search engines, for example, do not
present results based upon a blind algorithm, but
place “sponsored” links near the top.
While clearly embarrassed over the apparent
inconsistency, advocates for neutrality point out
that the barriers to entry are several orders of mag-
nitude higher at the ISP level than at the content
level. It is far easier for a user to switch search
engines than for a user to change ISPs. Once again,
the argument centers on how competitive Internet
service provision is today and how the level of
concentration might change in the near future.

5. What role, if any, should managed IP services
play?

Some suggest that a two-tiered Internet holds
the solution to the neutrality debate. Under such
a system, access providers, such as AT&T, would
allow neutral access to a portion of its network,
while reserving a portion of the infrastructure to
preferred content. This would increase the incen-
tive for the providers to invest in next-generation
networks, as some providers and users would be