Google and Verizon and net neutrality

I found this excerpt from Reihan to be persuasive:

I’ll just note that Google official line on wireless net neutrality, as stated by Richard Whitt, sounds fairly persuasive. Why insist on net neutrality regulations for wireline but not for wireless services?

First, the wireless market is more competitive than the wireline market, given that consumers typically have more than just two providers to choose from. Second, because wireless networks employ airwaves, rather than wires, and share constrained capacity among many users, these carriers need to manage their networks more actively. Third, network and device openness is now beginning to take off as a significant business model in this space.In our proposal, we agreed that the best first step is for wireless providers to be fully transparent with users about how network traffic is managed to avoid congestion, or prioritized for certain applications and content.

I believe in legally mandated net neutrality when monopoly is present, as with many cable providers, but not with wireless.  Here is my previous post on net neutrality.  Here are Robert Litan and Hal Singer, defending the Google-Verizon arrangement.


I'm less than persuaded by those points.

1. Although there are five or more carriers in many markets, they do everything possible to avoid direct competition. Two-year contracts prevent customers from moving rapidly toward better prices. The complexity of the pricing systems makes it difficult for any but very sophisticated consumers to figure true costs. Phones are made not to operate on different networks (even networks that share the same technology). True competition requires that ordinary people can see which merchant is offering the better deal and instantly switch to the provider that offers said better deal. That's very hard now and will be nearly impossible when users suddenly have to weigh the costs and benefits of networks that speed up and slow down services in ways that will never be transparent.

2. The main limit on how much data can go through the air is the same as the main limit on how much data can be sent through wires: how much carriers are willing to spend on infrastructure. Are their bandwidth limitations in the air? Yes, but if you rolled out what is currently state-of-the-art technology for compression and efficiency, we'd be years away from reaching them. In all probability, as the technology continues to improve and Congress -- realizing the importance of wireless -- devotes more and more of the spectrum to it (huge swaths for broadcast TV are already getting looked at) the only real limit on capacity will probably remain the desire of carriers to spend on infrastructure.

Oddly, if their argument about limited capacity were true, it would call for more regulation, not less. If the airwaves were so limited, then you'd expect the public would get the last word in how they were used (in the theoretical, though debatable, way that regulators are expressing the public's will) rather than private companies. You don't tend to allocate societally vital resources by letting deep pocketed companies bribe carriers to force content down the public's throat by giving it faster carriage while the content that the public wants (and expresses desire for through navigation) gets forced into the slow lane.

3. Network and device openness is not beginning to take off as a significant business model in this space. People in the business, if they squint hard and allow themselves to envision nightmares, can see a possible scenario where they'd be forced into network and device openness. It's not the laughable impossibility it seemed ten years ago, but it has only become a slightly more probable dream, not a reality. And everyone in wireless will lobby for all they are worth to prevent it from becoming a reality.

4. The notion that they will be "fully transparent with users about how network traffic is managed" is laughable. The explanation would run to 50 pages in dense technical jargon that is the tech industry's way of claiming rights for itself that no sane customer would accept while being able to claim that they are being fully open. Have any of you ever read an SLA before you click the "I agree" button on software? Of course not -- and they know you haven't because no sane person would spend two hours on 25 pages of legalese before installing Flash Player, or whatever.

Multiple "competitors" doesn't always insure competition, even when they're not allowed to openly collude a la OPEC. In cases where the different companies are not competing simply on price, it can be easy to make five flavors of the same service look like five different things -- five things that don't have to compete much on price or service. Numbers could certainly convince me otherwise, but I'm not at all convinced that wireless is what any sane person would call a "competitive industry" right now.

Does that mean I want regulation? Only to the extent that you can use it to create ground rules to force the carriers to compete on price alone: no contracts, interchanable technology and net neutrality (so the Internet service you get from one company is the same Internet service you get from another and you can really compare it on price and speed in an apples to apples way.)

Comments for this post are closed

In a few years, "wireless" net neutrality will be the net neutrality, much like wireless phone service is quickly becoming the dominant phone service for many individuals (especially in cities). How many game-changing online services were (or could have) been invented in a garage? Most if not all.
This would be impossible with no net neutrality.
For you net-neutrality skeptics: Do you really believe that we'd have an amazon, a youtube, a facebook, even a MR, if AOL were able to decide which websites loaded fastest in 1999?

Comments for this post are closed

It is worth noting that, before Google entered the wireless game, it was singing a different tune:

"Notwithstanding any technical differences between wireline and wireless networks that may justify different application of the reasonable network management exception on a case-by-case basis, the record is clear that all last-mile broadband network providers have common incentives to discriminate in the absence of an effective and enforceable rule protecting consumers and competitors."

Comments for this post are closed

There's nothing wrong with efficient pricing--charging different customers different amounts for different levels of service.

Net neutrality does not seek to restrict that in any way, and the only people who say it does are its opponents.

Net neutrality is about keeping the fundamental character of Internet access intact: You pay *your* ISP for a connection to the Internet, but you do not need to pay downstream ISPs, simply for the privilege of talking to "their" customers.

Comcast would like to charge, for instance, Skype, for access to a fast lane. There is, again, nothing wrong with *users* specifically giving Skype priority over their other traffic.

But how are Internet users to know what sorts of third-party paid prioritization deals are going on behind the scenes?

Net neutrality also does not propose to block behind-the-scenes activities that actually improve the network--such as colocation of servers or paid peering. Just paid prioritization, and the creation of artificial scarcity.

These economic issues plague network technologies. There are no good answers for intercarrier compensation on the telephone network, either. It stinks that net neutrality is necessary. But it is, because true ISP competition is a pipe dream in this political climate.

Net neutrality is second best to ISP competition, but the level of competition has to be similar to that which characterized internet access in the dial-up era. One was to get there is structural separation and line-sharing. It's because of telephone neutrality that we had dial-up ISPs--phone companies hated them in the 1990s, but were legally preventing from discriminating among phone network users. Strict libertarians characterize this sort of competition as "paper competition" that is not as efficient as allowing discrimination on any network, and allowing the wire owner to also "internalize complementary externalities." But the vibrant ISP competition that characterized the dial-up days led to low prices and no call for regulation of the ISPs themselves. I think the empirical success of line-sharing--whether it's via a nondiscriminatory basic transmission network, as in the 1990s, or via local loop unbundling in Europe and Japan today--beats theories as to why it can't work.

Comments for this post are closed

Incidentally, AT&T is on the record as saying that ISP competition in the 1990s was an artifact of regulation and not necessarily good for consumers.

I'll grant that it was the result of regulation. Common carrier regulation of the phone network, after the FCC told Bell it couldn't keep "foreign devices" off the network, is the only reason we had modems, faxes, and dial-up ISPs.

But this historical example shows that basic neutrality regulation of communications networks leads to innovation and technological progress that would be much slower to come if we "efficiently" allow facilities owners to do whatever they liked. It's amazing that people think that AT&T, Verizon, and Time Warner Cable can somehow out-innovate Silicon Valley, if only they're allowed to.

Also. It's a historical fluke even the minimal amount of wireline competition we have is. We laid down parallel wire networks only because we had to. And we're lucky that cable TV was so popular in the US. The end result is we now have two general purpose local loops to most houses. Attempts to get a "third pipe" have all failed. Wireless is more a complement, and broadband over power lines failed.

If it weren't for this historical fluke, the fact that wires to houses are about as close to a natural monopoly as you can get would be more widely recognized. Actually, wires aren't expensive to run. Holes are expensive to dig.

An alternative to unbundling, mentioned above, would be to go back in time and have general-purpose ducts laid down in streets instead of cables, to more easily allow for truly facilities-based competition. Paris has a high degree of wire-based competition because of its creepy catacombs. Overbuilding does happen in metro centers, and this is as much because of the infrastructure that allows fiber to be cheaply run, as it is because of population density.

You should see the fuss, when cities (tired of having their roads dug up) tell the telcos that they have to allow their competitors to lay cable in the same holes they're digging. Or the insane, endless battles over pole attachments.

Comments for this post are closed

Leland is right, neutrality between content providers needs to be distinguished from neutrality of applications.

The net is not just the web or file sharing. While some systems still have their own infrastructure for legacy or security reasons, more and more communication is happening over the net because it's just more practical in so many ways. You have to think about not just WWW, VoIP and Bitorrent, but about NTCIP, ie: traffic lights. And how about stock markets, airports, scada, hospitals, or ATMs? Military UAVs use the same satellites as porn providers these days. Are they all truly equal?

On our roads we distinguish between cars, buses, bicycles, pedestrians, trucks and emergency vehicles, but not between Ford and Toyota, or, as drivers, you and me. Sure, we can say the web should be neutral, but the net?

Comments for this post are closed

Leland points out that "certain applications like VoIP need prioritization." The developers of the IP protocol specification figured this out some time ago, and defined a value for the "differentiated services" which requests "expedited forwarding." (See RFC 3246.) Under network neutrality, network providers are free to honor explicit handling requests such as this one, so network neutrality doesn't prevent network providers from doing a good job of supporting VoIP traffic.

In the absence of net neutrality rules, network providers would have the option of taking another approach to supporting VoIP. Rather than honor the differentiated services field, they could come up with a way of specificly identifying VoIP traffic, and apply expedited routing to this traffic. Why is this bad? As long as no innovation occurs, it doesn't matter. But suppose someone develops an improved VoIP protocol, perhaps with better compression. Under net neutrality, this new protocol could be deployed and compete fairly with the old one. It would eventually become the dominant VoIP protocal assuming its advantages are sufficient that users find it worth while to switch. But if some network providers have set up their networks to provide special handling to packets using the old VoIP protocal, the new protocal will be at a disadvantage. Even if network providers are perfectly happy to have people switch to the new protocol, they won't necessarily expend resources to reprogram their routers to recognize the new protocol, at least not quickly. This is an example of a general principle: in the absence of net neutrality change is harder because you potentially have to sell your idea to network providers as well as users.

Comments for this post are closed

I do agree that regulation in this world of competition is the only thing that would ensure that the people availing the services will have it provided in a proper manner. Net neutrality has been just start of a time where there will be a similar cases for the authority of space over the virtual world.

Comments for this post are closed

Comments for this post are closed