Author: john

I’d like to take a moment to look at neutrality through the lens of how the Internet was designed to work under TCP/IP protocol. The short version of how the Internet works is that devices connected to the Net all have an address, and send information to other devices including the home and desired address. Routers along the network, which are aware of other nearby machines and routers, pass the information along until it reaches the desired destination. Part of the address (the subnet) tells the router generally which direction to send the information to get to another subnet on the Internet (where the destination machine resides). Since the network is redundant, and the routers only know which nearby networks are live, things are quite reliable. This simple design is why the network might be called “dumb,” and this model (putting the intelligence in applications instead of routing) is what has made the Internet a success—and removing neutrality in some ways could put this in jeopardy. I start with this argument because, again, I feel it is most important and largely lost in the overall debate. To take away neutrality would destroy this design.

Turning to market arguments, I apologize for not being able to find hard evidence of the limited choice for broadband service (although one article in the Washington Post argues that for many “geography still dictates their choice of Internet Service Providers” even in a large city like Washington D.C., “On the Web Where You Live,” May 13, 2005). Even assuming that there is consumer choice in broadband service, I still believe there are a number of market reasons to promote neutrality. An understanding of how the Internet works might show that the changes that are being proposed might happen in higher layers of networking (specifically, the transport layer). It’s conceivable that a non-neutral network would operate just as transparently to the user (for example, if a content provider pays for better access, the user may have no indication of this because the fact would be hidden by the network). Essentially, this is creating a separate “market” for content provider access to high-speed lines. This has great potential to lock out future innovators of Internet technologies, and possibly limit one’s ability to be a personal content provider (such as a popular video blogger). Additionally, again taking from an understanding of the existing architecture and building on the first argument, I would say that markets do a terrible job of creating what made the Internet successful: standards. As anyone who has to deal with multiple digital video formats may know, incompatibilities between the various types and systems can be a true pain. The TCP/IP non-neutral standard has created a commons where headaches like this just don’t happen. While the providers seem to be on the same page for moving away from neutrality (this oligopoly is another argument against the effectiveness of the market), there is no indication that they could agree on a common standard for a non-neutral Internet.

I’m not arguing against the market; rather that it may not be entirely effective in this case. Another historical view of network regulation (“Rules From Truth,” Bar and Sandvig, 2000) offers further insight into prior reasons for regulating communication networks. They point out that network externalities and “increasing returns” often increase the power of the network owners and move towards a natural monopoly, and argue that we have traditionally considered access to communication a fundamental right. They conclude that maintaining the “ability to design” is a crucial part to our previous communication policy, which has driven innovation because of the openness and transparency of the network. They explain this drive well: “Network users invented new patterns of communication and implemented them on a network infrastructure where their ability to design was promoted by open TCP/IP standards and open access to some of the telephone network’s building blocks.”

If I didn’t say it before, thanks again for the comments. I apologize if this has started to sound more like a paper than a blog, but this is turning into an enjoyable debate. As Stanley Fish said of his time traveling with Dinesh D’Souza debating the politics of education, “Neither of us, I think, changed the other’s mind on the issues we debated, but it is fair to say that both of us sharpened our arguments in the course of the … [debates].”

Sorry about the delay in replying to recent comments, it’s been a busy few days (and I never honestly expected to have any readers…guess that’s the magic of trackback). While networks and economics aren’t really my specialty, I appreciate the chance to have a good discussion about such an important issue. I can see that there’s a need to clarify some of my thinking, so here are the main reasons why I think network neutrality is something to value (from least to most important).

• Network economics: I’m still searching for where I heard recently that, while broadband availability is on the rise, most only have a choice of one service (more research on this fact is desperately needed for an informed debate). If this is in fact the case, it strengthens the argument that broadband ISPs are acting as a monopoly or oligopoly, or that there isn’t really enough competition to create market effects. This isn’t the most important point, however, because I would argue that decisions can and should be made for reasons other than just what the market may or may not provide.
• The commons and progress: Historical examples offer one of these reasons, for they show many instances of where the government has actively taken part in creating a commons on which both individuals and commerce can benefit.
  • The early postal system was a large motivating factor in the creation of national roadways, created physical locations where people would meet and discuss issues of the day, and was a source of pride for early Americans (I realize this is a controversial claim, so I’ll provide a cite: Spreading the News, Richard R. John).
  • Early copyright required individuals to register their works and deposit them with the Library of Congress, which quickly created one of the nation’s greatest libraries at low cost to creators.
  • The interstate system is open so that WalMart, Mom n’ Pop’s freight, and the public as a whole can benefit from the system.
  • Finally, in the early telephone monopoly, regulation drove network development and spread out the costs of doing so.
  • In each case, with the exception of the job of managing the commons, there has been no barrier of entry to anyone who wishes to use the network to their benefit. I would argue that removing neutrality would create such barriers (since start-ups may not be able to afford access). In my mind, the success of a commons need not always be measured monetarily, but rather in terms of social benefit (for more on all of this, a good book is Paul Starr’s The Creation of the Media).

• Preserving the end-to-end: This is where I think the true value of the above examples of commons lie. The idea behind end-to-end is that whatever is getting transmitted (be it cars, packets, or envelopes) gets delivered without regard for content or sender. The architecture places all of the “intelligence” at the ends, or in the applications. This content and sender neutral design allows any application designer to have a theoretically equal chance as any other because the network works the same for all of them. Neutral design left low barriers of entry to Internet business, while a tiered design would break this principle and make it more difficult for smaller Internet start-ups. Further, getting more out of the resources available has driven business and innovation further (video compression being one good example). The end-to-end principle is what makes the Internet, the Internet. Network neutrality and redundancy is what makes it so flexible, reliable, and resistant to physical attack. From my perspective, this debate tangentially is addressing the question of who is going to pay for network development. I believe sacrificing one of the crucial features of the Internet is an indirect way to solve this problem. (For more on the benefits of end-to-end design, see Lemley & Lessig’s “The end of end-to-end.”)

To me, maintaining this openness is the ultimate form of non-regulation, it allows users of the network to do whatever they choose within the bounds of the law. By regulating the commons provider, the online market is enabled to flourish. I would argue the debate shouldn’t be about limiting access, but rather about maintaining the integrity of the network architecture. Perhaps the most important thing at this point, if it’s not too late, would be to gather the necessary data about the problem to rise the debate above theory and rhetoric.
Cheers,
-john