Kill Network Neutrality, Get Slower Networks

Sat, November 19, 2011 | Posted by christopher

If you want to predict the future, it helps to understand the incentives that guide action. Unsurprisingly, if a corporation has the option of being more profitable by investing less, it will do so. This is the smart conclusion of Bill Snyder at InfoWorld:

To understand their logic, consider this thought experiment: Imagine that you own a freeway -- say, Highway 101 through Silicon Valley -- and you had the power to pluck a car from a traffic jam with a helicopter and deposit it on a clear stretch of the road. Naturally, drivers who could afford the service would be happy to sign up.

"That highway is like the Internet, and the individual cars are the packets of data. The ISP is essentially the gatekeeper that controls the flow of cars on the highway. If the ISP is allowed to snatch any car from the back of a very long line and put it in front of everybody else when the driver of the car pays a priority delivery fee, would the ISP have an incentive to keep the road congested or to expand the road capacity?" they wrote.

The answer is pretty obvious: If you can make more money by keeping your network congested, why would you invest money to make it less crowded?

He was riffing on a paper, "The Debate on Net Neutrality: A Policy Perspective" by H Kenneth Cheng, Subhajyoti Bandyopadhyay, and Hong Guo.

I think many of us view this as a "well, duh" paper, but it is good to see a rigorous academic paper verifying our gut instincts.

There is a very real danger to letting a few massive corporations control access to the Internet, which is one major reason we see so many communities building their own networks. They want to ensure everyone has fast, reliable, and affordable access to the Open Internet.

Comments

Smart... and wrong

Submitted by Martin Geddes (not verified) on December 9, 2011 - 6:25pm.

More capacity is a very expensive, and potentially uneconomic, way of managing the statistical multiplexing together of multiple traffic flows with varying sensitivity to loss and delay. The outcomes we desire are (1) better customer experience, and (2) affordability. If these can be delivered via traffic management rather than capacity expansion, that's a good thing! Unless, of course, you know of an unbounded source of funds to cope with an unbounded increase in capacity to avoid ever getting contention. How much of your city would you like to concrete over to avoid ever having a jam?

Not necessarily Wrong

Submitted by christopher on December 9, 2011 - 6:41pm.

Martin, I respect your position but I think you are coming at this from a different, more technical context.  You may be correct in the situation where the network owner is not simply trying to maximize its revenue by discriminating between content to maximize its returns from a natural monopoly investment.

However, in the U.S. the network neutrality debate centers around effectively unregulated companies that have minimal competition and will discriminate in order to reduce their investment in the network while increasing their profits.

For most of us, the discussion around network neutrality is not strictly about building dumb pipes but rather Comcast and AT&T should be allowed to charge subscribers extra for downloading 10MB from netflix (a competitor to some of Comcast's services) as opposed to downloading it from Apple or Microsoft.  

If service providers want to treat all streaming video or VOIP separately from how they treat other packets, that is a different discussion. One of the problems with the term "network neutrality" is that it is too ambiguous -- referring to multiple arguments among various sub-fields in broadband.  The paper referenced in my post is examining the incentives on a monopoly, not the best network management practices. 

Termnation fees (reprised)

Submitted by Martin Geddes (not verified) on December 9, 2011 - 7:04pm.

Where the broadband operator acquires customers and then charges a monopoly rental fee to reach them, we have a regime which is functionally equivalent to that for termination of voice/messaging today. There has been a great deal of debate over whether such fees are economically efficient, and how to regulate them, particularly when users have little choice and cannot "multi-home" as they can, for example, with search engines.

The unintended consequence of regulation here is to lock-in the service-provider model (possibly with a duopoloy to "catch the customer"), and lock-out self-service community networking models. The paradoxical intervention I would suggest is to ensure the rules are set so as to enable the broadbandit robber barons to rob -- and to maximise the differential with commmunity user-owned networks. Stop trying to fix the system, and break it instead!

Bearing in mind that nearly all the bit delivered by cable are already reserved for "non-neutral" own-service use, the end effect of neutrality regulation is to starve the amount of capacity given over to "Internet" and maximise incentives to keep traffic on private sub-networks that enjoy the very business model you describe and most fear. 

Our preference

Submitted by christopher on December 9, 2011 - 7:11pm.

Broadly, our preference is that at least one network be structurally responsive to the community, probably through the ballot box or a coop arrangement.  It isn't a panacea, but at least people have the capacity to reverse dumb decisions if they so choose.  When the network is run by an entity maximizing social benefit instead of a narrow vision of profit, I think the problem of unreasonable discrimation is largely resolved.  

Unfortunately, this approach is difficult in the US due to the power of a few massive incumbent operators.  

I don't see how regulations that broadly disallow providers from discriminating in the ways I described above preclude community self-provisioning.