Traditional values are under attack. The old ways are in decline, people insist on more freedom of choice than their grandparents had. The received wisdom about the very structure and organization of our world has been taken apart by science, and the only way to put it back together is to use the language of science to prop it up. So obscure terms and concepts enter public discourse with new meanings, and a alternative history is created. Fear runs rampant, men of science are assaulted for their moral bankruptcy, and an alternative science is injected into the schools.
While this paragraph clearly fits the “intelligent design” attack on the biological sciences in general and evolution by natural selection in particular, it also applies to “network neutrality”, the idea that only the government can be trusted to dictate policies embedded in Internet access networks. One of the most stark presentations of this viewpoint was written by Jeff Chester for the left’s holiest organ, The Nation magazine:
Absent net neutrality and other safeguards…[b]roadband connections would be governed by ever-vigilant network software engaged in “traffic policing” to insure each user couldn’t exceed the “granted resources” supervised by “admission control” technologies. Mechanisms are being put in place so our monopoly providers can “differentiate charging in real time for a wide range of applications and events.” Among the services that can form the basis of new revenues, notes Alcatel, is online content related to “community, forums, Internet access, information, news, find your way (navigation), marketing push, and health monitoring.”
Missing from the current legislative debate on communications is how the plans of cable and phone companies threaten civic participation, the free flow of information and meaningful competition. (ed: emphasis added)
Note the use of the terms in scare quotes, traffic policing, granted resources, and admission control. These are technical terms that come from the world of network engineering, and we can sure Chester doesn’t use them because he wishes to illuminate their importance in the engineering context. He doesn’t bother to define or explain, but takes it as given that any such words can only be destructive to “civic participation.”
The juxtaposition of network engineering language with social policy language is deceptive and inane. It’s like arguing that an electrical grid that provides alternating current to the home is responsible for politicians who flip-flop between positions depending on what audience they’re addressing. There’s a superficial similarity, indeed, but that’s where the connection ends.
We use admission control and policing on WiFi networks with something called WiFi multi-media (WMM) so that telephone calls and live video streams can happily coexist with web surfing on the same wireless network. Wireless networks don’t have unlimited bandwidth, so we have to use some finesse to provide a satisfactory experience to as many people as possible over a common network channel. Network engineering doesn’t do this in order to stifle democracy and curb “the free flow of information;” on the contrary, enabling as many people as possible to use the network as they wish to use it has the effect of enhancing free information flow. When we use “admission control” and its related priority system on a WiFi network, we can handle four times as many phone conversations as could without them. Doesn’t a phone call to a politician count as “civic participation” any more?
Have we now come to a state where the aesthetics of engineering language guide public policy more than the effects of engineering practice? I hope not, but we’re getting closer.
Chester admits that he wants to government to own Internet access networks, or failing that, that it should control them:
That means we would become owners of the “last mile” of fiber wire, the key link to the emerging broadband world. For about $17 a month, over ten years, the high-speed connections coming to our homes would be ours–not in perpetual hock to phone or cable monopolists.
Chester says, in essence, that the government is more likely to control the Internet in such a way that the public can engage in criticism of government actions and policies than anyone else. I see no empirical evidence that would encourage me to accept this article of faith, and plenty that argue for its rejection.
In our recent experience, we got to see head-to-head competition between a highly-regulated DSL service and a minimally-regulated Internet access system using cable TV systems, and the less-regulated option emerged as the clear technical winner.
Chester uses the language of science to urge us to ignore experimental data, and that’s as fundamentally unscientific and irrational as it gets.
For extra bonus points, his Center for Democracy and Technology minion Mike Godwin urges network neutrality regulation on the premise that the New York city taxi system is a transportation utopia. That’s just silly, and Matt Sherman explains why.
One can easily understand the origin of Chester’s suspicion:
And he’s got a point.
Of course, missing from the debate as well is the fact that QoS technologies are being deployed in Korea and Japan as well.
Anyway, I guess this has been a better gig for you lately than Iraq. It seems that question’s going to be settled in 2006.
Chester’s “cost per megabit” is misleading on several levels. If we apply this metric to American DSL and American cable modem we come away with an appreciation of the role that regulation plays in fixing prices. You get lots more megabits per dollar on cable, with no unsightly regulations.
And he ignores the fact that Korea Telecom blocks access to all VoIP except its own, for-fee service. This anti-neutral bundling is a key part of the economics of Korean broadband.
Why didn’t Chester mention that?
It seems to me that one of the key ideas behind network neutrality is that NO ONE can be trusted to dicate these policies in a competitive and pro-innovation way [innovation in the sense of the entire Internet ecosystem not just the access network]. It just so happens that the government has a monopoly over the power necessary to enforce that idea.
Actually, Dan, there’s another way: we can let the market enforce the idea of competitiveness. That’s what it does best.
I’m all for market competitiveness, but when the two market entrants are both government-licensed monopolists, you have a case for
“not really a market now, is it?” I don’t count satellite, because of latency issues.
For me, the magic number is four – four more or less comparable alternatives, and the possibility for collusion goes way, way down. At that point, Net Neutrality is just plain stupid. But as it is now, I think they have a point…
Manual trackback to my blog post where I disagree with you partly
The key problem I have is that you misunderstand the Cringely/Frankston proposal, although it is possible that you correctly state Chester’s understanding of that proposal
I don’t say the government should own the last mile. The Nation article I wrote refers to individuals, families, or communities owning the fiber. As for referring to “admission control” and other networking terms, they appear in documents (linked in the piece) where phone and cable companies are urged to use such mechanisms to increase revenues per users. These white papers make it clear that competitive content online will face network distribution obstacles as a way to protect our newly minted broadband monopolists from competition. Please remember, it was the cable and phone lobby that got the FCC to eliminate the U.S. Internet’s basic “end to end” non-discriminatory characteristic (cable in 2002; telcos in 2005). They phone/cable lobby did it so they could discriminate online. See more industry white papers at:http://www.democraticmedia.org/issues/netneutrality.html
Jeff, let me congratulate you for having the courage to defend your ideas over here. It takes a real man to step into hostile territory, and I commend you for it. That being said, here we go.
A) I don’t see a meaningful difference between “communities” owning the last mile and government owning it. You’re making a distinction without a difference.
B) “Admission Control” is a perfectly legitimate networking mechanism that’s used to provide Quality of Service, with or without a nefarious monopolistic motive. Your argument goes like this:
1, Hitler is bad
2. Hitler has a mustache
3. Therefore, mustaches are bad
The Internet doesn’t have an “end-to-end, non-discriminatory characteristic.” It’s a network, the purpose of which is to move packets according to user requirements. If the users require indiscriminate transport, it must provide that. If users require discriminating transport, it must provide that as well. If users require both (for different applications) it must provide them in concert. Go read Tim Wu’s first paper on NN, he lays it out quite clearly.
The Internet is a tool, not a law unto itself. Lessig is wrong. The only regulations you should be considering are rational-basis tests.
C) You have no more insight into the motives of the telco and cable lobby than the man in the moon; you’re simply speculating and offering your speculations as fact. In any case, the job of regulators isn’t to reform the motives of business people, it’s to prevent abuse of consumers and to enhance choice. Your scheme does neither, it simply criminalizes legitimate engineering practices.
jb says: “For me, the magic number is four – four more or less comparable alternatives, and the possibility for collusion goes way, way down.”
Most Americans have 4 options today thanks to EVDO, and they may soon have six, thanks to WiMax and Muni WiFi. If most choose DSL or cable over the alternatives, that’s simply the market at work.
Jeff,
I clicked on the first artcle in your supposed “smoking gun.”
And what did I see? A presentation by people who already understand a few things that you apparently do not. Specifically, VOD for IPTV takes up a lot of bandwidth. Consider the following:
I understand that even a compressed 1080i stream takes about 25 Mbps of bandwidth.
Consider small city of 100,000 has about 40,000 houses. Most people watch television in the evening, and many people have multiple TVs. Thus, we probably have 40,000 TVs that need content.
Suppose that we offered everyone open VOD. (I.e., they could watch whatever, whenever.) Well, the city would need a 1 terabit connection to the outside world, just to serve the VOD. An apartment building with 40 units would need a gigabit link, just to serve the VOD.
Are you getting the picture? There is currently no technological way to make VOD IPTV work in an unlimited way. Any implimentation will necessarily be limited.
This implies that less regulation led to the success of cable. What evidence is there for this?
It seems to be much more likely that the fact that the cable network has more bandwidth available (built much later than the phone network) is the real reason for its success.
Dan, the cable network has better wire, that’s true.
But where I live, the cable companies had to re-wire neighborhoods to get from the old uni-directional system to the new one that carries traffic both upstream and down, and they had to replace all the splitters, filters, and headends as well. An investment on this scale doesn’t often happen unless the investor feels confident in getting a return. The unbundling requirements in the 1996 telecom act pretty well dried-up investment in the last mile by telcos. Verizon was the only one to buck the trend by installing FTTH for their FIOS system, but they’ve only got 375,000 FIOS subscribers, a piddling number.
So the question really is this: why did cable companies re-wire for broadband while telcos didn’t?
MnZ makes a good point: when veterans of the telecom regulation wars look at broadband packet networks, their first complaint is the lack of dedicated bandwidth. Sadly, they don’t realize that this is the whole point of packet-switching.
The “but s/korea/japan/ has better broadband” carnard is only an arguement that makes sense if one can ignore, among other things, the laws of physics.
Please go back and view the thread Here.
As you say, the existing cable planet was uni-directional. Without the mentioned upgrades the cable companies couldn’t even begin to offer Internet access. The phone companies already had a two way network and as a result has much less incentive to replace it.
Yes, they had a two-way network of much lower bandwidth, so why didn’t they re-wire for high-capacity broadband?
Telcos in Korea, Japan, France and the UK did, but there was something else going on. These were former government monopolies that were privatized, and had piles of cash from floating shares for the first time. What else were they supposed to do with all that money burning a hole in ther pocket?
Probably because the incentive to upgrade from a network which could not do broadband Internet access to one which could is much larger than the incentive to go to a faster two way network. Especially since the initial cable deployments were not any faster than DSL anyway.
I still do not see any evidence that the regulation aspect to which I originally responded has anything to do with cable winning.
If regulation caused innovation, this isn’t the result we would expect, is it?
Richard,
Given my point above, isn’t that exactly the reason why Google and other web-based portals and Amazon and other web-based sellers are demanding draconian Net Neutrality rules? They realize that the next generation of Internet content (e.g., IPTV, QoS VoIP) will necessarily face near and middle term capacity constraints. Google and Amazon will be software providers in a world where the hardware is the limited resource. Big content providers like Disney or Viacom will be making deals directly with the ISPs. (Why would they deal with a middleman?)
Google and Amazon’s solution…make such deals illegal. Sure, it will stunt the growth of VOD/IPTV, but it keeps them in the game.
Yes, and they also want to perch on the ISPs border and stuff content into their pipe without any flow control by the ISP, the better to control the access network. You have to marvel at these guys for disguising their obvious commercial motive as some sort of valiant rescue effort by the “people’s champion.”