Comcast

by on April 3, 2008 · 16 comments

I have said that the threat of regulation is a credible deterrent to prevent unreasonable discrimination by broadband service providers and we don’t need a new regulatory framework with the unintended consequences which always flow from regulation.

And James Gattuso, noting that Comcast and BitTorrent were already working with one another on a solution to their network problems “long before this story broke,” correctly chided me for overlooking how public opinion is also a credible deterrent. James is right, particularly when there is a competitive market. And like it or not, the broadband market is competitive.

A “duopoly,” you say?


Allow me to quote Cornell Professor Alfred E. Kahn (formerly chairman of the New York Public Service Commission, chairman of the Civil Aeronautics Board under President Jimmy Carter and a father of airline deregulation) from testimony before the Federal Trade Commission last year:

There is no consensus among economists about the likely sufficiency of competition under duopoly—in the present instance, between landline telephone and cable companies in the offer of broadband access to the Internet—although evidence of active competition between the two, such as is actually occurring, might provide sufficient basis for deregulation, particularly in light of the aforementioned rapidity of technological change. By the same token, the presence of an actively competing wireless provider or providers—would, in the mind of most, justify—indeed demand—de- or non-regulation. Wireless voice service is one of the great success stories in telecommunications over the last few decades. I understand that the prospects for wireless data in the near future are excellent. Any analysis of future competition in Internet access must consider the possibility—or likelihood—that the cable and telephone duopoly will be joined by three or four wireless suppliers in the near future.

Of course, wireless broadband is a likelihood. It’s even a certainty. Check out the latest broadband competition report from the Federal Communications Commission: There are actually more mobile wireless high-speed lines (defined as more than 200 kbps in at least one direction) than cable modem or DSL lines (35 million mobile wireless versus 34 million cable modem and 27 million DSL lines).

I realize mobile wireless broadband access isn’t yet identical in terms of price and speed. But we all know it’s heading in that direction. The point is, if landline telephone and cable companies who offer of broadband access to the Internet abuse their market dominance they will merely accelerate the growth of mobile wireless broadband services. The correct measure of competition isn’t what we have today but what we will or could have tomorrow.

Call it what you will: The threat of regulation, public opinion, enlightened self-interest or simply the desire to be a good corporate citizen – Despite the absence of regulation Comcast is working with BitTorrent and is also now rolling out DOCSIS 3.0 beginning in the Minneapolis and St. Paul region – providing up to 50 mbps downloads and 5 mbps uploads; Comcast expects to deliver even faster speeds of up to 100 mbps to its customers over the next two years with the capability of delivering higher speeds of 160 mbps or more in the future.

  • http://techdirt.com/ Mike Masnick

    Hmm. I wouldn’t exactly use the FCC’s report as evidence, considering that it’s been so torn about by the GAO for years as being inaccurate and even the FCC admitted (right as they released this report) that it’s totally inaccurate and future reports will include an entirely different system of measuring broadband penetration.

    In fact, one FCC commish called the report “stunningly meaningless.”

    That “duopoly” isn’t even really a duopoly in many places. I live in the very heart of silicon valley, and until very recently had no options other than cable broadband. DSL wouldn’t serve me (now it does, but at a speed that is less than the new broadband definitions).

    And your claims of wireless competition are also misleading. Those wireless connections cannot be used as regular broadband lines thanks to incredibly restrictive terms of service that prevent all but the most basic usage, and many have extremely low usage caps. Not real competition.

    As for future wireless competition… considering that the major players in the space are owned by the same companies who own the DSL lines, it’s not at all convincing that DSL dominance would accelerate wireless plans. The two nationwide wireless firms that aren’t owned by telcos aren’t exactly real threats. T-Mobile doesn’t have the spectrum necessary to do real broadband and Sprint (which has a ton of spectrum) is going through a bit of an implosion right now, and its chosen technology (WiMax) has serious problems that are finally coming to light.

    So, yes, we may eventually have wireless broadband, but it’s simply incorrect to claim that it’s here now or that it’s likely to be here in the near term.

    I agree that public pressure and competition can keep companies as good citizens, but your measurement of competition is totally inaccurate.

  • http://techdirt.com/ Mike Masnick

    Hmm. I wouldn’t exactly use the FCC’s report as evidence, considering that it’s been so torn about by the GAO for years as being inaccurate and even the FCC admitted (right as they released this report) that it’s totally inaccurate and future reports will include an entirely different system of measuring broadband penetration.

    In fact, one FCC commish called the report “stunningly meaningless.”

    That “duopoly” isn’t even really a duopoly in many places. I live in the very heart of silicon valley, and until very recently had no options other than cable broadband. DSL wouldn’t serve me (now it does, but at a speed that is less than the new broadband definitions).

    And your claims of wireless competition are also misleading. Those wireless connections cannot be used as regular broadband lines thanks to incredibly restrictive terms of service that prevent all but the most basic usage, and many have extremely low usage caps. Not real competition.

    As for future wireless competition… considering that the major players in the space are owned by the same companies who own the DSL lines, it’s not at all convincing that DSL dominance would accelerate wireless plans. The two nationwide wireless firms that aren’t owned by telcos aren’t exactly real threats. T-Mobile doesn’t have the spectrum necessary to do real broadband and Sprint (which has a ton of spectrum) is going through a bit of an implosion right now, and its chosen technology (WiMax) has serious problems that are finally coming to light.

    So, yes, we may eventually have wireless broadband, but it’s simply incorrect to claim that it’s here now or that it’s likely to be here in the near term.

    I agree that public pressure and competition can keep companies as good citizens, but your measurement of competition is totally inaccurate.

  • http://www2.blogger.com/profile/14380731108416527657 Steve R.

    Saul Hansell has an article in the New York Times (March 31, 2008) Comcast’s Blurry High Definition Picture.
    ——————————————
    “Not only has Comcast been slowing down Internet users exchanging files with the BitTorrent protocol, it has been quietly reducing the quality of some high definition television networks it carries as well.”
    ——————————————
    Given continued disclosure of yet another corporate “shaping” scheme, at what point do we acknowledge that free market has not reigned in an offending company?

    (PS: I am not a Comcast user, so I have no first hand experience with them. I am simply using Comcast as a “Poster Child” to demonstrate that the free market does not immediately correct the malevolent behavior of offending companies.)

  • http://www2.blogger.com/profile/14380731108416527657 Steve R.

    Saul Hansell has an article in the New York Times (March 31, 2008) Comcast’s Blurry High Definition Picture.
    ——————————————

    “Not only has Comcast been slowing down Internet users exchanging files with the BitTorrent protocol, it has been quietly reducing the quality of some high definition television networks it carries as well.”
    ——————————————

    Given continued disclosure of yet another corporate “shaping” scheme, at what point do we acknowledge that free market has not reigned in an offending company?

    (PS: I am not a Comcast user, so I have no first hand experience with them. I am simply using Comcast as a “Poster Child” to demonstrate that the free market does not immediately correct the malevolent behavior of offending companies.)

  • http://www.techliberation.com/contributors/ryan_radia.php Ryan Radia

    Hance, you are absolutely right that competition can function fairly well even with a duopoly. Cable and DSL are fairly close substitutes for most uses, and a substantial majority of people can get both.

    And with Wi-Max just a couple months away from official launch in DC and Chi, and LTE on the not-so-distant horizon, wireless competition may come sooner than we think. Who knows what firms will do with the 700mhz band when analog NTSC stops in 2009?

    Plus, the telcos can only service 79% of U.S. homes with DSL, largely due to CO distance limitations. AT&T would love to offer broadband to such customers, if only an economically feasible option existed. The 700mhz spectrum, for all we know, could be used for broadband, albeit modest, in areas not served by DSL.

    200kbps doesn’t meet my personal definition of broadband. But 800kbps–sustained throughput, not burst speeds–is broadband, and it’s what EVDO Rev A offers. Rev A devices are now available to let PCs browse the web at these speeds. So, while most people still have 2 options, a growing number have 3. Half of Americans are now covered by Rev A.

    Also don’t forget Sprint’s new unlimited voice and data plan. No hidden caps, no usage limitations.

  • http://www2.blogger.com/profile/14380731108416527657 Steve R.

    This is the last gasp of a dying company. It is not a demonstration that this is a new consumer friendly business trend. As a former Sprint customer, I can attest to the onerous business practices of Sprint bitting the hand that feeds it.

    The market has spoken. When sprint was making money customer complaints just went into the waste basket. Customers are still beating down the doors to leave and Sprint may not survive. Now that a significant number customers have left, Sprint finally comprehends that they need customers, want to be nice, and are now offering a new plan to attract customers. To bad they didn’t comprehend that before.

    PS: I am still waiting for Sprint to refund their billing “mistakes” and to apologize for these “mistakes”.

  • http://www.techliberation.com/contributors/ryan_radia.php Ryan Radia

    Hance, you are absolutely right that competition can function fairly well even with a duopoly. Cable and DSL are fairly close substitutes for most uses, and a substantial majority of people can get both.

    And with Wi-Max just a couple months away from official launch in DC and Chi, and LTE on the not-so-distant horizon, wireless competition may come sooner than we think. Who knows what firms will do with the 700mhz band when analog NTSC stops in 2009?

    Plus, the telcos can only service 79% of U.S. homes with DSL, largely due to CO distance limitations. AT&T; would love to offer broadband to such customers, if only an economically feasible option existed. The 700mhz spectrum, for all we know, could be used for broadband, albeit modest, in areas not served by DSL.

    200kbps doesn’t meet my personal definition of broadband. But 800kbps–sustained throughput, not burst speeds–is broadband, and it’s what EVDO Rev A offers. Rev A devices are now available to let PCs browse the web at these speeds. So, while most people still have 2 options, a growing number have 3. Half of Americans are now covered by Rev A.

    Also don’t forget Sprint’s new unlimited voice and data plan. No hidden caps, no usage limitations.

  • http://www2.blogger.com/profile/14380731108416527657 Steve R.

    This is the last gasp of a dying company. It is not a demonstration that this is a new consumer friendly business trend. As a former Sprint customer, I can attest to the onerous business practices of Sprint bitting the hand that feeds it.

    The market has spoken. When sprint was making money customer complaints just went into the waste basket. Customers are still beating down the doors to leave and Sprint may not survive. Now that a significant number customers have left, Sprint finally comprehends that they need customers, want to be nice, and are now offering a new plan to attract customers. To bad they didn’t comprehend that before.

    PS: I am still waiting for Sprint to refund their billing “mistakes” and to apologize for these “mistakes”.

  • Matt S

    Steve:

    I am simply using Comcast as a “Poster Child” to demonstrate that the free market does not immediately correct the malevolent behavior of offending companies.

    Traffic shaping is only malevolent behavior from the perspective of the 1% of users taking 50% of the bandwidth. For the other 99%, shaping is beneficial. Don’t get caught up in the mock outrage.

    One might even call Comcast irresponsible if they don’t shape traffic, considering winners and losers.

    The free market simply expresses the wishes of consumers. 99% of consumers don’t give a damn about traffic shaping and don’t consider it malevolent.

  • http://www.techliberation.com/contributors/ryan_radia.php Ryan Radia

    Matt, I’m with you fundamentally, but I prefer not to shape the argument in those terms. The percentages you cite are complete guesswork, and based on the estimates I’ve read, your figures may be exaggerated in Bittorrent’s disfavor.

    Here’s what we can reasonably conclude:
    -a small but not insignificant minority of Comcast customers use Bittorrent
    -Bittorrent consumes a substantial and disproportionate amount of bandwidth
    -A sizeable portion of torrent traffic is copyright infringement.

    Many torrent users who’ve suffered on the account of Sandvine neither pirate nor consume excessive network resources. On the other hand, as you rightly point out, a large majority of Comcast customers are at this point well served by a network free from P2P-induced congestion.

    The broadband market is functioning like it should, and Comcast and other ISPs are experimenting with methods to deal with increasing demand for bandwidth.

  • Matt S

    Steve:

    I am simply using Comcast as a “Poster Child” to demonstrate that the free market does not immediately correct the malevolent behavior of offending companies.

    Traffic shaping is only malevolent behavior from the perspective of the 1% of users taking 50% of the bandwidth. For the other 99%, shaping is beneficial. Don’t get caught up in the mock outrage.

    One might even call Comcast irresponsible if they don’t shape traffic, considering winners and losers.

    The free market simply expresses the wishes of consumers. 99% of consumers don’t give a damn about traffic shaping and don’t consider it malevolent.

  • http://www.techliberation.com/contributors/ryan_radia.php Ryan Radia

    Matt, I’m with you fundamentally, but I prefer not to shape the argument in those terms. The percentages you cite are complete guesswork, and based on the estimates I’ve read, your figures may be exaggerated in Bittorrent’s disfavor.

    Here’s what we can reasonably conclude:
    -a small but not insignificant minority of Comcast customers use Bittorrent
    -Bittorrent consumes a substantial and disproportionate amount of bandwidth
    -A sizeable portion of torrent traffic is copyright infringement.

    Many torrent users who’ve suffered on the account of Sandvine neither pirate nor consume excessive network resources. On the other hand, as you rightly point out, a large majority of Comcast customers are at this point well served by a network free from P2P-induced congestion.

    The broadband market is functioning like it should, and Comcast and other ISPs are experimenting with methods to deal with increasing demand for bandwidth.

  • http://bennett.com/blog Richard Bennett

    Actually, Ryan, Matt’s numbers wildly overstate the size of the user population affected by Comcast’s current practice of resetting BitTorrent seeders when the network is highly loaded.

    The only people really affected by the current policy are those who create Torrents and do their initial seeding on the Comcast network. Once you’ve got a torrent going, Comcast customers are able to download it just fine.

    I’ve never seen the numbers on how many people create legal torrents, but my estimate is that it’s a really, really small number, like less than a 0.1% of all torrent users, who are probably about 10% of all Internet users. There’s just no point in creating a torrent unless you’ve got a really, really big file to transmit, and if you have one that you came by legally you probably aren’t dependent on a residential network for your initial seeding.

    So the math problem is: What percentage of Comcast customers create (legal) torrents and seed them from their home connection?

    And BTW, I did successfully seed TPW 34 from the Comcast network, just for fun, but it was kinda slow. But BitTorrent doesn’t care about slow, it keeps trying until you tell it to give up.

  • http://bennett.com/blog Richard Bennett

    Actually, Ryan, Matt’s numbers wildly overstate the size of the user population affected by Comcast’s current practice of resetting BitTorrent seeders when the network is highly loaded.

    The only people really affected by the current policy are those who create Torrents and do their initial seeding on the Comcast network. Once you’ve got a torrent going, Comcast customers are able to download it just fine.

    I’ve never seen the numbers on how many people create legal torrents, but my estimate is that it’s a really, really small number, like less than a 0.1% of all torrent users, who are probably about 10% of all Internet users. There’s just no point in creating a torrent unless you’ve got a really, really big file to transmit, and if you have one that you came by legally you probably aren’t dependent on a residential network for your initial seeding.

    So the math problem is: What percentage of Comcast customers create (legal) torrents and seed them from their home connection?

    And BTW, I did successfully seed TPW 34 from the Comcast network, just for fun, but it was kinda slow. But BitTorrent doesn’t care about slow, it keeps trying until you tell it to give up.

  • Ryan Radia

    Richard, I was under the impression that Comcast’s TCP resets afflict all seeders who reside in areas where Snadvine has been adctivated. i.e. once you finish downloading a file, then your torrent app switches to seeding mode instead of leeching mode, Sandvine kicks in. Correct me if i’m wrong, but if I’m not, then sandvine affects a good number of torrent users. many might unwittingly seed post-downloading, but back when I used torrents regularly I usually seeded for a while after downloading. it helps share ratio on private torrent communities, plus is a matter of giving back for what you’ve taken. if nobody seeded beyond the first seeder, torrents wouldn’t get very far. anyway, i agree that very few people seed torrents from the beginning, although my wild guess is that 0.1% of torrent users is a bit low, but still admittedly small in the grand scheme of things.

  • Ryan Radia

    Richard, I was under the impression that Comcast’s TCP resets afflict all seeders who reside in areas where Snadvine has been adctivated. i.e. once you finish downloading a file, then your torrent app switches to seeding mode instead of leeching mode, Sandvine kicks in. Correct me if i’m wrong, but if I’m not, then sandvine affects a good number of torrent users. many might unwittingly seed post-downloading, but back when I used torrents regularly I usually seeded for a while after downloading. it helps share ratio on private torrent communities, plus is a matter of giving back for what you’ve taken. if nobody seeded beyond the first seeder, torrents wouldn’t get very far. anyway, i agree that very few people seed torrents from the beginning, although my wild guess is that 0.1% of torrent users is a bit low, but still admittedly small in the grand scheme of things.

Previous post:

Next post: