For some people, the special feature of the Internet, the one most worth preserving, isn't that it's a medium which is particularly free of government control, on which people can say what they want. It's the concrete detail that it uses a certain delivery model, known as "net neutrality." So far it's the norm almost everywhere, aside from censorship attempts, nannyware, spam filtering, and such. With net neutrality, a packet is a packet, and all get the same level of service regardless of protocol or source.
There are obvious benefits to net neutrality. It encourages innovation in protocols, and it avoids snafus when management based on protocol or content tries to get too clever. It helps to make sure that everyone on the Internet has a voice. But many people consider it so important and so fragile that they are inviting the federal government to stick its regulatory camel's nose under the tent flap.
There are plausible reasons to deviate from net neutrality. Some types of traffic, such as streaming video, consume much more bandwidth than others, such as text. It could make sense to make the high-bandwidth protocols a premium service or to give them lower priority. Competition allows for multiple possibilities. But the advocates of mandated neutrality insist on regulation to prohibit this. Down this road, though, once the government has a precedent, comes the long chain of power grabbing through statements of the form, "We already regulate X, so it's inconsistent not to regulate Y." If packet neutrality should be mandated, then why not a fairness doctrine, for instance?
Once the process starts, we can be sure of one thing: regulation will be skewed, as it always is, to favor those who have the most influence with the regulators. This means those who have the strongest economic interest in favorable rules, together with the greatest influence. This means companies whose size exceeds their ethics. It always works this way, and the advocates of regulation in "the public interest" always complain bitterly when they get what they ask for.
On The Technology Liberation Front, Ryan Radia discusses the distorting effects which demand for net-neutrality regulation may be creating. Comcast may, according to rumor, institute a maximum data throughput per month for users, with overage fees. (He mistakenly refers to these as "bandwidth" limits; limiting bandwidth, as opposed to aggregate data transfer, is easy to do, but providers keep increasing it because it attracts customers.)
The “Save the Internet” brigade’s insistence on neutrality and transparency has left Comcast with little choice but to resort to a metered solution to network congestion. Of course, I’m pleased that Sandvine and the “invisi-caps” will soon be history, and I look forward to consuming 249.99 GB each month on my Comcast connection.
But what about non-neutral solutions to last-mile congestion? To be sure, Sandvine was far from perfect, but who knows what innovative network management technologies will go undeveloped because of the stigma, and threat of regulation, against traffic discrimination?
Widespread attitudes would attach stigma to non-neutral approaches, regardless of the regulatory climate, but it convincing demonstrations of economic benefit, including a speedy connection for ordinary purposes, could outweigh it. Sneakiness, such as Comcast's recent attempt to limit some types of packets by improper use of protocols, would carry more stigma and thus be unattractive. When regulation is threatened, companies may (and Comcast actually did) find sneakiness preferable to attracting the wrath of lobbyists and legislators. Or more precisely, since a corporation is a person only in the legal sense, someone in Comcast decided there was less risk to his job by taking a sneaky action that only geeks would get upset about than by making a policy decision that a Congressthing might make hay from.
Let the net neutrality advocates argue that their way is the best, and let them tell users they're best off avoiding any ISP's that implement non-neutral limits. By doing this, they'll stimulate the debate and discourage foolish approaches to limitation. But they should put down the club of threatened regulation and make their point by non-coercive means.