Home » Internet

Institute for Policy Integrity responds

By 25 January 2010 6 Comments

Inimai Chettiar and Scott Holladay from the Institute for Policy Integrity New York University School of Law have responded to my critique of their report with the following rebuttal posted in its entirety.  Note that Digital Society does not agree with the following views, but we’re going to publish it as is and respond to it with our own rebuttal posted here.


In a recent post, George Ou offers some criticisms of a report that we released earlier this month.  Despite his claim that our report was based on a “flawed analysis”, we believe that the differences between our positions are smaller than it might seem.

In an earlier post on January 14th, Ou laments that network neutrality proponents “suspend economic reason” when advocating open Internet laws.   Our report does exactly the opposite by discussion some of the potential implications of net neutrality for the overall economic health of the Internet.  Our argument in Free to Invest: The Economic Benefits of Preserving Net Neutrality, boils down to five points:

  1. The existence of “positive externalities” from the Internet—benefits to Internet users and others that are not captured by ISPs or content providers—will lead to systematic underinvestment in the Internet.
  2. Because of these positive externalities, there is a role for government to step in and try to correct this underinvestment and restore the market to equilibrium.
  3. ISPs implementing price discrimination when charging content providers for access to ISPs’ networks will transfer wealth from content providers to broadband companies.
  4. Most of this wealth transfer will reward ISPs for past investment, rather than incentivize them to invest in new infrastructure.  Targeted government support—which is relatively easier to do for infrastructure compared to content—can help ISPs expand the network where needed.
  5. While some content providers may benefit from arrangements offering content providers better access to ISPs’ subscribers, others will lose out.  Because ISPs will face perverse incentives when setting up these “prioritization” schemes, there is legitimate concern that these schemes will reduce the value of the Internet as a whole.

Some pieces of our argument are less controversial—many would agree that an ISP threatening to block traffic from content providers unless those providers pay charges to the ISP is a bad thing. (This is what we refer to as “pure price discrimination” in our analysis.)   This practice would result in large wealth transfers from content providers to ISPs, and may run afoul of existing law.  (The new FCC rules would remove any ambiguity on this question and clearly ban this practice.)

There is greater disagreement about ISPs offering different connectivity options to content providers for a price.  In his post, Ou lays out a variety of potential content delivery models—such as “paid peering”—that might be prohibited under network neutrality regulations. In our report, these and other arrangements for better connectivity at a price are analyzed in the section on prioritization.

Paid peering is an arrangement in which content providers pay for a dedicated line from their local servers to the ISP’s physical point of presence. This allows the content provider to bypass the Internet backbone and move their content over a “private pipe” directly to the access point to the Internet Service Providers last-mile network.

Even if some think that paid peering would be “cheaper” than other forms of prioritization, paid peering still raises the same economic issues described in our analysis of general prioritization schemes.  First, paid prioritization allows existing market leaders in content to purchase better access to end-users; while some content providers will win out in this situation, others will be losers, ending up with the worse access to users.  Because these arrangements will benefit some while harming others, we find in the report that “it is impossible a priori to determine how the market surplus would shift.”

However, we give an important reason for concern:  ISPs will face perverse incentives.  There are a number of ways in which ISPs could pursue various prioritization schemes (like paid peering) that would increase their revenue while decreasing the value of the Internet to society as a whole. Further, the edge of the network generates a tremendous amount of wealth and ISPs are ill-positioned to consider the impact of their decisions on the economy and the Internet as whole.

The question is not whether ISPs might, in a perfect world, be able to use prioritization to increase the value of the Internet.  The question is, given their actual economic incentives, whether they will use prioritization to increase the value for everyone, or simply to increase revenue for themselves.  Because ISP incentives are not aligned with the market as a whole, there is good reason to be worried that, in the real world, failing to adopt net neutrality will lead to pricing schemes that might be good for ISPs, but will reduce the quality and value of the Internet for the rest of us.


The Institute for Policy Integrity at New York University School of Law is a non-partisan think-tank that works with advocacy organizations and governments to use economics and law to protect the environment, public health, and consumers.  Working at the national and local levels in the United States and across the globe, Policy Integrity projects bring economics to bear on issues like climate change, women’s health, and net neutrality.

Inimai Chettiar is a Legal Fellow and Scott Holladay is an Economics Fellow at the Institute for Policy Integrity.

6 Comments »

  • Digital Society » Blog Archive » Net Neutrality proponents demand a primitive Internet said:

    [...] and Scott Holladay from the Institute for Policy Integrity New York University School of Law have responded to my critique of their report “Free to Invest: The Economic Benefits of Preserving Net [...]

  • Brett Glass said:

    These arguments are specious. See, for example, Christopher Yoo’s paper and presentation at

    http://www.silicon-flatirons.org/documents/conferences/2009.02.08-210/Yoo.pdf

    and

    http://engineeringonline.colorado.edu/tegrity/siliconflatirons/siliconflatirons09s09.wmv

    Also see my filed comments regarding the NPRM at

    http://www.brettglass.com/nprmcomment.pdf

    –Brett Glass

    P.S. — Is this group at NYU funded by Google?

  • Nick Brown said:

    Read your NPRM comments. Very straight forward, logical information about a sector of the industry that seems to be being pushed to the back burner while decisions are being made.

  • Brett Glass said:

    Thank you! (By the way, it looks as if a quotation mark was omitted in my posting above, so the link came out wrong. Moderator, can you fix?)

    I fear that WISPs and small, independent ISPs in general are being marginalized in the “network neutrality” debate for two reasons.

    Firstly, the proponents of “network neutrality” regulation don’t want our existence to be recognized because their arguments are largely founded on the notion that there is no competition. The inconvenient truth that competition actually exists — and would be harmed by the regulation! — is, for them, something to be swept under the rug.

    Secondly, the large carriers, which oppose “network neutrality” regulation, don’t want the many small competitors out there to grow and eat away at their market share. So, they don’t want to acknowledge our existence for fear that the government will adopt policies that foster more competition!

    Fortunately, the study which CITI submitted to the FCC does cover WISPs.

    For an overview of WISPs, what we are, and what we do, see

    cyber.law.harvard.edu/events/luncheon/2009/12/glass

  • Richard Bennett said:

    This is simply amazing. They argue that it’s better for the Internet for applications and content providers to pay higher fees for inferior service than to permit ISPs to sell paid peering. They confuse paid peering, a service that combines elements of transit and access, with a pure access tax, ignoring the value of its transit component.

    In their paper, they claimed networks interconnect at Network Access Points, a long-decommissioned feature of NSFNET privatization that has been replaced by Internet Exchange Points, carrier-neutral exchange points. So what we have here is an economic analysis that may be very fine in some abstract universe, but which doesn’t pertain to the facts regarding Internet transit and peering today (or tomorrow, for that matter.)

Leave your response!

Add your comment below, or trackback from your own site. You can also subscribe to these comments via RSS.

Be nice. Keep it clean. Stay on topic. No spam.

You can use these tags:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

This is a Gravatar-enabled weblog. To get your own globally-recognized-avatar, please register at Gravatar.