Susan Crawford's Five Good Question

Susan Crawford, a law Professor at Cardozo and a Board Member of ICANN supportive of Net Neutrality, asks and answers five good questions about Network Neutrality. Chris Yoo, a law professor at Vanderbilt and opposed to Net Neutrality, gives his answers (along with Susan’s) here. Harold Feld, not a law professor anywhere, gives his answers below.

From Susan Crawford’s blog:

Here are five frequently-asked questions about net neutrality. Your challenge: answer each in 150 words or less.

Who am I to decline such a challenge? (I admit, keeping it to 150 words will be tough.)

1. What does net neutrality actually mean? Is it a meaningful protection for the web, or, as some say, a romanticized ideal that’s getting in the way of progress?

(This is really two questions).

For me: Net neutrality means that a broadband access provider (usually refrred to as “broadband access”) cannot leverage its market power to get money for nothing (or, as economists say, “rent seeking”). In practical terms, it means a broadband access provider cannot — without subscriber approval — offer itself or specific third parties better service than it offers anyone else.

Mind, I think offering to let a subscriber pay extra to designate a premium source or service is o.k., or services like Comcast’s “power boost” that prioritize traffic of a particular type when network capacity is available. What I oppose is Whitacre Tiering or any other exercise of market power by access providers.

As to whether this is some “romantacized ideal”: No. If I can get technicaly economic for a minute – this is a run of the mill economic regulation to prevent vertically integrated companies from leveraging their market power. To translate into English- Broadband access providers enjoy a situtation with little or no competition. The difficulty of switching services (“switching cost”) creates a phenomena called “lock in” which makes it very hard for people to switch from one provider to another even where they do have a choice. And, in economic terms, a choice between two providers isn’t a terribly good choice (we call it a “duopoly”).

Without net neutrality, it is an economic certainty that cable cos and telcos will try to use their market power as broadband gateways to the internet to maximize revenue. This is Econ 101. Companies will do whatever they can to maximize revenue. If that has bad consequences for the internet and society at large,too bad.

2. The cable and telephone companies argue that they need additional revenue to build ‘the internet of the future’ and so the Googles and Amazons of the world (who will benefit from that new internet) need to pay their fair share. Is that a legitimate argument?

This argument is bogus. No one pushes bits for free. Everyone already gets paid to build pipes, and has incentive to build bigger and better pipes.
Besides, the internet is not just the “last mile” between the subscriber and the backbone. The idea that cable and telcos should build “the internet of the future” should scare the bejezus out of us. We all built the various independent networks and applications that make the internet what it is today. Handing it back to the phone company is a step backwards for the Internet (which, back in the days when folks capitalized it as a proper noun, meant “the network of networks”).

3. Net-neutrality’s supporters are concerned that if you give the cable and telephone companies latitude to control who travels through their pipes (and at what speed), it puts those gatekeepers in a position to favor their own products and services over their competitors’. The fear is that innovation will suffer. Is that a concern you share?

Net neutrality has been the law for the entire history of the internet. I haven’t noticed a lack of innovation.
On the other hand, handing control to a central, proprietary gatekeeper is identical to the way in which radio and television broadcasting, cable, and traditional phone services were run. None of these systems have been especially noteworthy for rapid innovation — until forced to compete with the services available via the Internet.

4. Why do you think this issue has taken off with such a fervor in recent months?

Behold the power of public advocacy, and it’s limits! This issue languished until it became clear that telcos could get a law passed gutting existing rules and preventing the FCC from creating rules even if problems did emerge. Then Free Press, Moveon, and the rest of the public interest world decidd to take the case to the people using the internet to route around the disinterest of the manstream media. Further, as I have observed before, the interactive nature of the internet allows people to engage in real conversation and exploration of an issue beyond duelling soud bites.

But NN is also instructive of the limits of the blogosphere and internet advocacy. Most Americans still have no clue what’s happening, despite editorials in mainstream papers. To the extent they do, the radio and television advertising by the Bells has done a good job muddling the issue in people’s minds.

5. Could you sketch out what groundrules you’d like to see govern the internet of the future?

Ideally, I’d like to see the rules we used to have for dial up — anyone who wants to connect to customers can get access on a wholesale basis to the underlying network. We used to call the “open access” along with “common carriage” (for the wholesale part) and interconnection (so people who built out networks could connect to other people’s networks).

But I’m told that’s not politically feasible, so I will take a world where a broadband access provider (a) must disclose to users how it treats traffic, and (b) cannot overide user preferences. The operator can certainly set limits on bandwidth so that no one user can choke the whole network, and can assign priority to certain types of traffic that are sensative to lag. But if I, as a user, prefer one source of content or one application to another, the broadband access provider should not be allowed to interefere.

Stay tuned . . .

4 Comments

  1. Two points re Q5:

    1. What is the argument that says common carriage not feasible? I’m not prepared to give up this important principle!

    2. There is an important technical consideration in allowing the pipe-holders to regulate the traffic: the current protocols do not allow the APPLICATON to designate a priority or other classification request. Without this, the pipe-holders must GUESS at what kind of application they are seeing when they look at packets. This goes against an important principle of good design (c.f. Reed’s end-to-end argument). The result is that we will have really crumby software, products and services. Determining packet content by looking, can, by definition, ONLY be done for existing and very common applications, or perhaps for SUSTAINING innovations if they are common enough. It is a tautology that this cannot be done for disruptive innovation in the Clay Christensen sense. Is that what we want? Hell no!

  2. 1. I meant politically feasible, not technically feasible. No one wants to fight for it.

    2. Agree with you re: applications. However, there is no reason we ISPs cannot create such applications. Comcast’s “power boost” is an example of such an offering.

  3. I’m not following re 2. Powerboost purports to be content-agnostic. That’s ok. My claim is that it would be stiffling for ISPs to differentiate among kinds or uses of traffic. For example, all the really interesting world-changing/disruptive/different-value-chain Internet innovation is in various forms of overlay network that depend on a content-agnostic network.

  4. I think we agree, then. I want networks to remain content agnostic, but do not object to giving end users the ability to designate certain types of traffic or certain applications as having priority. While this is not done at present, I believe it could be.

    My understanding of power boost was that it was agniostic as to source but did offer higher speeds for certain types of traffic, like video. This may have merely been advertising literature rather than an actual technical capacity.

Comments are closed