I encourage readers to visit Doc Searls post "Broadband vs. Internet" for a discussion about things that matter regarding the future of Internet access for most Americans.

The Internet is no more capable than the infrastructures that carry it. Here in the U.S. most of the infrastructures that carry the Internet to our homes are owned by telephone and cable companies. Those companies are not only in a position to limit use of the Internet for purposes other than those they favor, but to reduce the Net itself to something less, called “broadband.” In fact, they’ve been working hard on both.

There is a difference between the Internet and "broadband." Broadband is a connection that is always on and tends to be somewhat faster than the dial-up speeds of 56kbps. Broadband could connect you to anything... could be the Internet or to an AOL like service where some company decides what you can see, who you can talk to, and the rules for doing anything.

The Internet is something different. It is anarchic, in the textbook definitional sense of being leaderless. It is a commons. As Doc says,

The Internet’s protocols are NEA:

• Nobody owns them.*
• Everybody can use them, and
• Anybody can improve them.

Because no one owns it, few promote it or defend. Sure, major companies promote their connections to it (and when you connect to it, you are part of it) but they are promoting the broadband connection. And the biggest ones (Comcast, AT&T, Verizon, Time Warner Cable, etc) will do anything to increase the profits they make by being one of the few means of connecting to the Internet -- including charging much more and limiting what people can do over their connection, etc.

This is one reason the connections from major corporations are so heavily tilted toward download speeds -- they want consumers to consume content. Just about every community network built in the last 3-4 years offers symmetrical connections by contrast.

Last I heard, the fastest cable offering in the upstream direction was 12Mbps. Cox, our cable provider in Santa Barbara, gives us about 25Mbps down, but only 4Mbps up. Last time I talked to them (in June 2009), their plan was to deliver up to 100Mbps down eventually, but still only about 5Mbps up. That’s competitive as long as all you want is “content delivery.” But what about when you want to live “in the cloud,” and all your data is elsewhere? In the long run you’ll need a lot more upstream as well as downstream capacity for that. Internet service optimized for media delivery (where TV especially wants to go) won’t cut it. But then, most people aren’t looking at that. They’re looking at TV on their iPads over broadband, and thinking that’s way cool enough.

So here we are, smack up against what John Perry Barlow warned us about in Death From Above, way back in early 1995. There he wrote, “The cable companies and Baby Bells have a model for developing the next phase of telecom infrastructure which, were it applied to the design of physical superhighways, would have us building them with about five thousand lanes in one direction and one lane in the other.”

This is where Bob Frankson comes in, reminding us that the big cable and phone companies are good at billing, not connecting. Their methods and procedures are optimized to maximize their revenues, not to maximize the benefits of the Internet or anything else. Back to Doc:

The division is between what communications wonks crudely characterize as “net-heads” and “bell-heads.” Think of conflict as one betwee any and only. Net-heads want the Net to support anything. Bell-heads want communications systems optimized only for the businesses they prefer — namely, their own — and to avoid even talking about the Internet. (Bell-heads have never been comfortable with the Net, because it was not made to bill. TV and telephony are easy to bill, and so is “content” in general. Thanks to Apple’s and Google’s pioneering work —mostly in league with the operators — so now are apps.)

Community Broadband generally sides with the net-heads. The focus tends to be on what is best for the community as a whole, rather than what is good for a single company or industry. After all, if a community had a choice between one business providing 5,000 jobs and 500 businesses each providing 10 jobs, they would be crazy to opt for the single employer.

Verizon had been the only major company investing in next-generation networks with its FiOS deployment. It is done expanding -- unless you live in one of the wealth neighborhoods or suburbs that got it, you won't. AT&T has even ceased its pathetic U-Verse upgrades, even as they spend millions to prevent communities from building their own, much better networks. Communities that want to be relevant in 10 years take notice -- and take a good hard look at building a locally owned network that responds to the needs of the community.

Doc Searls photos used under creative commons license, courtesy of Flickr's Irisheyes

So Comcast and Level 3 are in a peering dispute following the Netflix partnership with Level 3 to distribute their streaming movie service. Studies suggest Netflix movie streaming has become a significant chunk of Internet traffic, particularly at peak times.

A quick primer on peering: the Internet is comprised of a bunch of networks that exchange traffic. Sometimes one has to pay another network for transit and sometimes (commonly with big carriers like Comcast and Level 3) networks have an agreement to exchange traffic without charging (one reason: the costs of monitoring the amount of traffic can be greater than the prices that would be charged). (Update: Read the Ars Technica story for a longer explanation of peering and this conflict.)

Comcast claims that Level 3 is sending Comcast 5x as much traffic as Comcast sends to Level 3 and therefore wants to charge Level 3 for access to Comcast customers. Of course, as Comcast only offers radically asymmetrical services to subscribers, one wonders how Level 3 could be 1:1 with Comcast…

At Public Knowledge, Harold Feld ties the dispute to network neutrality:

On its face, this is the sort of toll booth between residential subscribers and the content of their choice that a Net Neutrality rule is supposed to prohibit.  In addition, this is exactly the sort of anticompetitive harm that opponents of Comcast’s merger with NBC-Universal have warned would happen — that Comcast would leverage its network to harm distribution of competitive video services, while raising prices on its own customers.

Susan Crawford wrote a lengthier piece about Comcast, Netflix, network neutrality, set-top boxes and NBC that is well worth reading (as is just about anything she writes).

However, for the purposes of this post, we will assume the 5x traffic imbalance is true (and unique and that Comcast has no ulterior motive for charging Level 3 (and its partners like Netflix) a fee for anti-competitive reasons (like its own TV Everywhere service). I want to explore the world in which Comcast has pure motives to explain why even in that world, policy should address the market power of Comcast.

Comcast wants to charge Level 3 for access to their customers, which means that content distributed by Netflix has a disadvantage relative to content distributed by Comcast. Even before buying NBC (does anyone really expect the Obama Administration to halt this terrible merger?), Comcast owned content creators. In fact, it has long used its market power as a massive cable distributor to acquire a stake in channels -- as detailed here and here.

As a Comcast customer, it becomes harder and harder for me to choose content not owned by Comcast. Even if Comcast does not act anti-competitively, the content it owns is simply easier for me to find and access. This creates a barrier for new content providers (and encourages Netflix to give Comcast a stake in the company).

These inevitable barriers to entry, even when Comcast is not abusing its power exemplify the problems of massive scale for a company that owns both content and the (increasingly sole) means of transmitting it.

A far better arrangement is structural separation, where the network owner has no stake in the content transferred. This observation informs our preference for community owned networks -- ideally open access networks with a multitude of independent service providers. Even if Comcast behaves itself, it has too much control over the future of content -- from web sites to television programming.