Home / Blogs

Old New Telco Thinking

Ahh, so the telecom incumbents have come up with a “new” idea for the Internet—usage-based pricing. That’s right, more usage (for things like VOIP and video especially) means more costs to operate the network, so users should pay by the bit, or some similar metric. It’s all so logical!

But wait a minute. I thought what sparked the consumer Internet revolution was the fact that ISPs didn’t charge by the minute, but offered flat-rate monthly fees. And what catalyzed the boom in cellular usage here in the US was the shift from heavily usage-based pricing to the largely flat rates we see today. This “new idea” is actually the oldest one in the telecom book. Even with respect to the Internet, the debate over flat vs. usage pricing is a decade old. It was at the heart of the “modem tax” debate back when I was at the FCC. At that point, the telcos were complaining about dial-up Internet access, not broadband video, yet the argument was the same. They lost that battle, but really they (and users) won, because usage and innovation skyrocketed.

Experimenting with new pricing and business models is all well and good; there’s nothing written in stone about the current fee structure for broadband access, for example. But be wary any time you hear that “economics” dictates an “efficient” consumer pricing model. Sure, as a fundamental rule, prices in competitive markets are related to marginal cost. However, local access is still not truly a competitive free market, and cost numbers in telecom are inherently fuzzy and manipulable.

Given the option, users tend to vote with their dollars. If the pricing model doesn’t reflect the value they get from the application or service, they use it less. With the kind of rhetoric we’re hearing today we should be wary about pricing decisions by network owners choking off future growth and development of Internet-based services.

By Kevin Werbach, Professor at the Wharton School and Organizer of the Supernova Conference

Filed Under


You cannot have it both ways Anonymous Coward  –  Nov 18, 2008 10:44 PM

Usage-based pricing is a bad idea. Users hate watching the meter, and will be deterred from doing things because they know it’s running. However, broadband providers have been forced into metering by the recent FCC ruling, which slapped Comcast for imposing implicit limits on customers’ usage and preventing bandwidth hogging.

Bandwidth costs money—lots of it, in rural areas. (Rural ISPs pay as much as $325 per Mbps per month for their bandwidth.) But even if it’s less expensive in a particular area, there has to be SOME limit on how much users can consume, or some of them (the big downloaders and P2Pers) will hog it, degrading other users’ service. One way to limit it is to charge a flat rate and hold the bandwidth hogs back; that’s what Comcast did. But the FCC said, no, Comcast couldn’t do that. So, Comcast had to start imposing caps. It really had no choice. Kevin, you’re going to have a say in who gets on the FCC and what it does during this administration. Perhaps you can pick someone who doesn’t make bad policy just because it affords him a chance to slap cable providers (as was the case with Kevin Martin) but will understand that ISPs do have to meet payroll. They will be as consumer-friendly as they can (they need to compete for customers), but the FCC doesn’t allow them to be consumer-friendly the results will not be good.

Revisionism The Famous Brett Watson  –  Nov 19, 2008 3:55 AM

Why wait until late 2008 to respond to this article from early 2006?

However, broadband providers have been forced into metering by the recent FCC ruling, which slapped Comcast for imposing implicit limits on customers' usage and preventing bandwidth hogging.
Comcast was slapped for selectively discriminating against particular P2P applications. Their actions may have prevented a certain class of bandwidth hogging to some degree, but the impact was limited to the upstream channels of certain specific P2P applications. It was this "discriminatory and arbitrary practice" [FCC 08-183] which formed the crux of the matter in the Comcast case, particularly given that the practices were not disclosed to customers. (On the contrary, Comcast initially denied that they were blocking access to any application or throttling any traffic -- a misleading half-truth at best.) You're clamouring for change on the basis of a revised account of history. Nobody has been "forced into metering", as you say. The FCC ruled against unreasonable application-specific interference, not bandwidth management in general.

Yes, revisionism -- on your part. Anonymous Coward  –  Nov 20, 2008 1:37 AM

It's relevant to comment on Kevin's 2006 posting because now, in 2008, he's on the Obama transition team and will be instrumental in determining (among other things) who will chair the FCC. As for the FCC decision against Comcast: it appears that you are the one engaging in revisionist history. As was reported in Comcast's filings with the FCC just last month, Comcast was not, as you claim, "selectively discriminating against particular P2P applications." It was throttling all P2P, and with good reason; keeping P2P in check is vital to ensuring quality of service for those of us who aren't doing illegal downloading of pirated music and movies.If bandwidth hogs are allowed to congest the network, the rest of us wind up with slow service and higher prices. Comcast was doing the right thing. So, why was it slapped down? Because Kevin Martin, who wants to be a telco exec when he steps down as FCC chair, was "paying forward" to the telcos by beating on the cable companies.

Here we go again. The Famous Brett Watson  –  Nov 20, 2008 5:36 AM

As was reported in Comcast's filings with the FCC just last month, Comcast was not, as you claim, "selectively discriminating against particular P2P applications."
According to their filing with the FCC, their network management technique involved inserting TCP RSTs into the upstream transfer channels of five distinct file-sharing protocols, so I stand by my earlier characterisation. Seriously, why is it that we have this serial parade of vocal Comcast apologists who rant and rail against P2P and the FCC, but insist that any kind of application-neutral bandwidth management is impossible for obscure and inexplicable technical reasons? You haven't got to that latter stage yet, but I'm waiting expectantly on the basis of past experience.

P2P mitigation Anonymous Coward  –  Nov 22, 2008 12:49 AM

The use of RST packets is a longstanding and very effective method of administratively terminating connections. Of course, in the case of P2P, it doesn't stop the transfer; it just cuts off some of the hydra's heads. (BitTorrent and many other P2P programs effectively hack the network by opening up many connections, seizing priority over legitimate traffic.) Comcast had the right to terminate ALL P2P traffic, because it is contrary to its terms of service. But it was being nice; it only limited P2P when the network got congested. And because its equipment knew which P2P malware it was dealing with, it could do it gracefully and effectively, just as a virus checker can best deal with a virus if it has a pattern for it. As fo the "parade" you are whining about: maybe it's because there are a lot of people out there who know that Comcast was right.

PR Miasma The Famous Brett Watson  –  Nov 22, 2008 10:47 AM

Comcast had the right to terminate ALL P2P traffic, because it is contrary to its terms of service. But it was being nice; it only limited P2P when the network got congested.
That's a novel theory. By extension, I suppose, a provider could effectively prohibit all activity on its network, and then selectively allow violations of these terms just to "be nice" about it. It would then be completely free to interfere with customer traffic in any way it saw fit. No doubt there has to be a boundary (albeit a vague, subjective, and indistinct one) at which this behaviour becomes "unreasonable". Having said that, I would have far less to complain about if Comcast had openly admitted they were using TCP RSTs to break P2P connections on the basis that P2P violated the AUP. I wouldn't necessarily applaud the action, but it would have been vastly more reasonable than (a) shrouding the whole thing in secrecy until forced to confess under mounting evidence, and (b) claiming that the method used was a reasonable form of network traffic management to prevent congestion. The miasma of PR spin-doctoring in this affair has poisoned (and continues to poison) rational technological debate.
P2P ... hydra's heads ... hack the network ... seizing priority over legitimate traffic ... P2P malware ... virus
Take this, for example. I detect a certain amount of "OMG P2P IS TEH EV1L" spin in what you've said, and you're not the first to sing that tune here in a CircleID comment relating to Comcast. This serves to distract readers from the core issues of Comcast's behaviour in this matter: the lack of disclosure, the fact that TCP RST insertion is not a technically reasonable approach to congestion control, the fact that they are selectively degrading specific applications, and so on. Vilifying P2P distracts from this: it shifts the focus off Comcast as the wrongdoer; it makes Comcast's actions seem reasonable on the (false) basis that anything anti-P2P must be an intrinsic good; it paints Comcast as the white knight which is protecting all the innocent, legitimate Internet users from the selfish, evil, smelly P2P filth.
As fo the "parade" you are whining about: maybe it's because there are a lot of people out there who know that Comcast was right.
No, I don't think that's it. I have a short list of plausible explanations, and "Comcast was right" didn't make the cut. "Astroturf", on the other hand, did. If you (collectively, Comcast apologists) want to wage a propaganda war against P2P or the FCC or whatever, I'd appreciate it if you could do so without promoting utterly bogus technical doctrine in the process.

Metering is Wrong Tool Michael Roberts  –  Nov 19, 2008 9:31 PM

The carriers are barking up the wrong tree with monthly aggregate data limits.  If there are good ways to deal with pathological user behavior, then this isn’t one of them.  First, spreading the monthly limits out over the approx 2.6 million seconds in a month produces bandwidth utilization generously under 100 kilobits per second.  Can’t call that “hogging,” can we?  The real problem is peak loads, isn’t it?  So what load can a normal PC or Mac generate on a broadband connection?  Many of us have gigabit Ethernet ports on our machines, but have you ever seen that?  So, actually, in the real world, you’d hardly ever see more than a single megabit sustained upstream for more than a few seconds.  And I’ve never seen close to Comcast’s claim of “up to 12 mbps” download either.  The industry talk is all FUD, a smokescreen for desire to control user access to content.  If ISP’s want to deal with gamers and other hackers trying to louse up the system, why not say so, and put the appropriate words in the service agreements rather than making 99% of their customers mad at them?

P2P doesn't produce a single peak; it causes congestion that lasts for hours or days Anonymous Coward  –  Nov 20, 2008 1:47 AM

It happened in my neighborhood. A group of kids went on a downloading spree and everyone's access was bogged down all day, because P2P keeps running 24 x 7. It took complaints and a letter from the cable company to the kids' parents to clear up the problem. When users run P2P, they cost also cost the provider more for bandwidth than they are paying. Our cable company puts it right in its terms of service that running servers and P2P is prohibited. Blocking P2P is the right thing to do, but if providers cannot do this metering is the only way to discourage it. And it has nothing to do with access to content. Comcast was not looking at what was being transferred via P2P, though if it did I am sure it would have found that all of it was illegal. It was pushing back on P2P, and that is a good thing.

Comment Title:

  Notify me of follow-up comments

We encourage you to post comments and engage in discussions that advance this post through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can report it using the link at the end of each comment. Views expressed in the comments do not represent those of CircleID. For more information on our comment policy, see Codes of Conduct.

CircleID Newsletter The Weekly Wrap

More and more professionals are choosing to publish critical posts on CircleID from all corners of the Internet industry. If you find it hard to keep up daily, consider subscribing to our weekly digest. We will provide you a convenient summary report once a week sent directly to your inbox. It's a quick and easy read.

I make a point of reading CircleID. There is no getting around the utility of knowing what thoughtful people are thinking and saying about our industry.

Co-designer of the TCP/IP Protocols & the Architecture of the Internet



Brand Protection

Sponsored byCSC

Threat Intelligence

Sponsored byWhoisXML API

Domain Names

Sponsored byVerisign

New TLDs

Sponsored byRadix

IPv4 Markets

Sponsored byIPv4.Global


Sponsored byVerisign


Sponsored byDNIB.com