The Real Net Neutrality Debate: Pricing Flexibility Versus Pricing Regulation

by on October 27, 2005 · 12 comments

[Cross-posted from the PFF blog]

As Ray noted in his essay last week on Clearwire and VoIP blocking, I have long argued that broadband service providers (BSPs) will eventually will end up price differentiating based on bandwidth usage, in part because of the futility of differentiating based on service bundling or technological applications / usage.

What I mean by this is that most attempts to discriminate against specific websites or applications are likely doomed to fail or end miserably for the carriers. First, with Net surfers getting more sophisticated with each passing day, it will be very difficult to block most activities without them finding ways around the restrictions. Second, attempting to discriminate against certain types of bits is complicated for the carriers and will likely require more effort than it’s worth. Third, even if carriers were able to discriminate against certain bits, if they went overboard it would spark an intense consumer backlash and a likely exodus to an alternative broadband provider. (And if a serious alternative backbone provider was not yet available in that region, excessive blocking / meddling by the incumbent BSP would likely serve as the best incentive for new entrants to enter the market and offer a less-restricted surfing experience.)

But critics claim that I’m full of it and some of them pointed to last week’s front-page story in the Wall Street Journal as evidence. In their article, “Phone, Cable Firms Rein in Consumers’ Internet Use,” reporters Peter Grant and Jesse Drucker claim that, “Several large telephone and cable companies are starting to make it harder for consumers to use the Internet for phone calls or swapping video files.” The story goes on to note that some BSPs “have begun to closely monitor the uses of their network with an eye toward controlling activity by users who are swapping movies, TV programs, pornography and other video files. Operators say file sharing is growing so quickly, it threatens to sharply slow down other uses.”

What are we to make of this? Are BSPs hell-bent on controlling our web-surfing experiences and even resorting to “discrimination” against certain types of uses or activities? Should that be illegal? Are there other ways of handling this problem?


Let’s begin by acknowledging that there is a potential problem here. While some intensive corporate and individual Net users may not realize it (or just don’t care to admit it), excessive web-surfing by a small handful of users can have negative externalities. Other, more casual users of broadband networks could be impacted by a handful of bandwidth hogs who eat up the bulk of system capacity. Just as lanes on a highway can get clogged up from too much traffic, the same goes for broadband networks.

In the case of broadband networks, however, a small handful of people can be responsible for putting more cars on the road than others and causing the online traffic jams. As the WSJ story reports, according to Mike Lajoie, chief technology officer of Time Warner (TW), fewer than 10 percent of TW subscribers consumer more than 75 percent of the bandwidth. And I’ve actually heard even more extreme numbers reported by other BSPs, with the ratio being more along the line of 5-10 percent of users eating up closer to 90 percent of bandwidth.

Of course, these numbers shouldn’t really be surprising. At least at this point in time, the vast majority of web surfers are basically looking to do a few routine things online in any one sitting and then getting offline and back to the real world. These casual users jump online to check e-mails, monitor a few eBay auctions or a favorite newsgroup, and perhaps look at the news on a few websites. And then there’s the other users who might best be labeled “Netizens,” since they almost literally live in cyberspace on a full-time basis. These are avid online gamers, the file-sharers, and the online video watchers, and yes, the porn-surfers. They spend endless hours uploading, downloading, sharing, and playing online. And they consume massive chunks of bits in each sitting. (I want to make it clear here that I do not have a problem with any of these activities, rather, I am merely trying to point out that a small handful of websurfers impose the largest burden on broadband networks. In fact, I am probably in this group myself!)

So, here’s the key question in this debate: How should BSP deal with these bandwidth hogs? If a small handful of Netizens really are imposing more of a burden on the system, what can and should be done about it (if anything at all)?

Currently, some BSPs deal with the problem by basically sending the most intense users a dirty letter essentially saying “Please cut it out, or else” but it is somewhat unclear what the “or else” is. Under many of the Terms of Service agreements I have read, BSPs say they will cut of your service for certain “violations” of the agreement, including excessively burdening the system with traffic. But it is my understanding that such provisions are rarely enforced. (Seriously, if anyone knows of a bandwidth hog being cut off, please send me info because I try to keep track of these things).

As the WSJ article suggests, the other strategy that some BSPs are currently considering involves tinkering with networks to potentially restrict or slow access to certain sites or activities. In other words, the BSPs start to act like more of a full-time traffic cop and try to direct flows during times of peak usage.

In my opinion, both of these strategies represent a silly way to deal with the problem of excessive bandwidth use by a small minority of users. The second strategy presents a major enforcement nightmare for the BSPs and would create much ill-will with a core part of their customer base. Moreover, if BSPs really did start cutting off service to large groups of users or, worse yet, slowing down or blocking access to certain sites or services, it would open them up to intense scrutiny from the regulatory community. Because even though we are moving away from common carrier notions in the world of communications, many still want to regulate high-speed network providers as if they were common carriers. Indeed, Net Neutrality regulation is really nothing more than common carriage law for the Information Age. It seeks to enshrine those old principles into law for the next generation of networks and services.

But if lawmakers went that route–and recent drafts telecom policy reform on Cap Hill all include some sort of Net Neutrality provision–it is important to understand that this means the continuation of price controls for communications. This is because, in practice, common carriage law is toothless without price regulation. You can have all the common carriage / non-discrimination principles on the books that you want, but unless regulators have the ability to also control prices those principles are meaningless. After all, a BSP faced with merely a behavior constraint could simply price access to the protected service or application at a different rate to discriminate against it.

This raises the most interesting issue in this entire debate: Why is it that BSPs are not currently attempting to meter broadband usage and price it to account for demand and “excessive” usage by some users? In my opinion, this would be the most efficient and least meddlesome why of dealing with this problem. Per-minute or per-bit pricing schemes could help conserve pipe space, avoid congestion, recover costs and enable BSPs to plow the savings into new capacity / innovation. Despite this, no BSP seems willing to engage in any sort of metering of the pipe. Why is that?

I think there are two reasons that BSPs have so far been unwilling to price discriminate. First, broadband operators are probably concerned that such a move would bring about unwanted regulatory attention. Second, and more importantly, cable and telco firms are keenly aware of the fact that the web-surfing public has come to view “all you can eat” buffet-style, flat-rate pricing as a virtual inalienable right. Internet guru Andrew Odlyzko, has corretly argued that “People react extremely negatively to price discrimination. They also dislike the bother of fine-grained pricing, and are willing to pay extra for simple prices, especially flat-rate ones.” And George Gilder, another famous Net guru, noted in his book Telecosm that, “Everyone wants to charge different customers differentially for different services. Everyone wants guarantees. Everyone wants to escape simple and flat pricing. Forget it.” Gilder bascially argues that simple and flat pricing is almost always preferable from a consumer perspective and, therefore, network providers should avoid more complicated pricing schemes.

For these and other reasons, BSPs probably don’t want to rock the boat too soon with more creative price discrimination schemes, but someday they may have to as bandwidth-intensive web sites start to eat up more and more pipe capacity. While simple and flat pricing seems like the sensible approach, it remains likely that some BSPs will eventually attempt to craft tiered or metered pricing schemes. While some consumers will cry foul, a number of bandwidth-intensive Internet vendors and website operators will likely be absolutely apoplectic over such a move, and some may even run to regulators seeking redress.

This raises the important question of whether or not broadband operators should have the right to price network access in this manner. And, would Net Neutrality regulations prohibit such innovative pricing schemes from being employed in the first place?

The answer remains uncertain, but clearly, if some form of network “non-discrimination” rule is on the books, some website operators and content providers may push to invoke it against a BSP that suddenly announces a new metered pricing scheme for bandwidth-intensive web offerings. It would be very unfortunate if this scenario came to pass, since such creative pricing schemes may be part of the long-run solution to relieving Internet congestion and allowing carriers to accurately assess user charges for Web activities. Supply and demand could be better calibrated under such pricing schemes and broadband operators may be better able to recoup sunk costs and make new investments in future infrastructure capacity or network services. As Odlyzko argues:

Thus even if it is not optimal from a global point of view, it might be necessary to introduce complexity in order to be able to construct and operate the telecom infrastructure, especially the residential broadband networks that are so eagerly awaited by government and industry leaders. That might mean allowing carriers to charge differently for movie downloads than for Web surfing. That, in turn, might require a new network architecture. Such a move would not be unprecedented. The key (although seldom mentioned) factor behind the push for new network architectures appears to be the incentive to price discriminate. It is an incentive that has been operating since the beginnings of commerce.

The bottom line is that it should be left to markets, not regulators, to determine what pricing schemes are utilized in the future to allocate scarce space on broadband pipes. The broadband marketplace is still in an early developmental stage, having only existed for a few years. What business model will prevail or make network activities profitable in the future? Pay-per-view? Advertising? Metered pricing schemes? Some hybrid of these and other systems? No one knows for sure, but policymakers need to allow network operators the freedom to innovate and employ creative pricing and service schemes so that market experimentation can answer that question.

This what the fight over Net Neutrality is really all about: Pricing freedom versus pricing regulation.

  • Christopher

    A better thing to do would be for the government to step in and say “One price for 20GB, 40 GB, so on”.
    That would solve the problems of bandwidth leeches of which I am admittedly one (download 20 gb of stuff a month or more).
    But if the government were to do that, I would cry foul UNLESS they lowered the prices dramatically as compared to unlimited access plans.

  • Christopher

    A better thing to do would be for the government to step in and say “One price for 20GB, 40 GB, so on”.
    That would solve the problems of bandwidth leeches of which I am admittedly one (download 20 gb of stuff a month or more).
    But if the government were to do that, I would cry foul UNLESS they lowered the prices dramatically as compared to unlimited access plans.

  • http://tregear.org Jonathan Tregear

    It’s arguable whether there really is such a thing as bandwidth scarcity in landline broadband capacity. Even if there is, it’s nothing like radio spectrum scarcity where price differentiation is somewhat more defensible.

    However, the issue is not whether metered service would be more efficient or a more rationale way to pay for bandwidth, the issue is whether price differentiation will result in different types of application bits being priced differentially. Will my file download bit cost me less than my VoIP bit? That is what price differentiation is about; it is not about metered service.

    Price diffentiation is a real threat, since BSP’s could and are attempting to use it to inflate user costs for non-BSP provided services.

  • http://tregear.org Jonathan Tregear

    It’s arguable whether there really is such a thing as bandwidth scarcity in landline broadband capacity. Even if there is, it’s nothing like radio spectrum scarcity where price differentiation is somewhat more defensible.

    However, the issue is not whether metered service would be more efficient or a more rationale way to pay for bandwidth, the issue is whether price differentiation will result in different types of application bits being priced differentially. Will my file download bit cost me less than my VoIP bit? That is what price differentiation is about; it is not about metered service.

    Price diffentiation is a real threat, since BSP’s could and are attempting to use it to inflate user costs for non-BSP provided services.

  • http://www.coises.com/ Coises

    Three decades ago there was already a model for computer resource allocation: timesharing. (I presume something like it must be used by today’s pre-emptive multitasking PC operating systems, such as any 32-bit version of Microsoft Windows.) In timesharing, the priority of a user’s tasks decreases as the user’s recent resource consumption increases and as the time since the last user action (e.g., a terminal keystroke) increases. A similar principle in TCP would say that you divide the available bandwidth among everyone who wants it, weighting recent requests higher than continuing service of older requests and weighting users who haven’t used much bandwidth recently (we’re talking seconds or minutes here, not hours or days) higher than those who have been taking a lot. The net effect would be to prioritize short, simple tasks (like getting an ordinary web page) so that their response time remains rapid while slowing higher bandwidth activities (and letting real-time high-bandwidth activities fail) during times when the system is struggling to meet demand… but slowing them only as much as is needed to maintain the chosen target response time for simple tasks.

    This should keep typical users happy and encourage high-bandwidth users to do their big tasks at times when the overall system load is low, without imposing arbitrary punitive measures or engaging in the practice of trying to determine just what users are doing (something I think should be avoided like the plague it could become).

    The next step in this business model would be to sell priority: e.g., if you pay more, the parameters used to compute your priority relative to other users would be changed so that you could take more bandwidth at busier times before being slowed (all relative to what others are paying, of course; hmmm… this could turn into bidding for priorty: perhaps a good thing, perhaps not).

  • http://www.coises.com/ Coises

    Three decades ago there was already a model for computer resource allocation: timesharing. (I presume something like it must be used by today’s pre-emptive multitasking PC operating systems, such as any 32-bit version of Microsoft Windows.) In timesharing, the priority of a user’s tasks decreases as the user’s recent resource consumption increases and as the time since the last user action (e.g., a terminal keystroke) increases. A similar principle in TCP would say that you divide the available bandwidth among everyone who wants it, weighting recent requests higher than continuing service of older requests and weighting users who haven’t used much bandwidth recently (we’re talking seconds or minutes here, not hours or days) higher than those who have been taking a lot. The net effect would be to prioritize short, simple tasks (like getting an ordinary web page) so that their response time remains rapid while slowing higher bandwidth activities (and letting real-time high-bandwidth activities fail) during times when the system is struggling to meet demand… but slowing them only as much as is needed to maintain the chosen target response time for simple tasks.

    This should keep typical users happy and encourage high-bandwidth users to do their big tasks at times when the overall system load is low, without imposing arbitrary punitive measures or engaging in the practice of trying to determine just what users are doing (something I think should be avoided like the plague it could become).

    The next step in this business model would be to sell priority: e.g., if you pay more, the parameters used to compute your priority relative to other users would be changed so that you could take more bandwidth at busier times before being slowed (all relative to what others are paying, of course; hmmm… this could turn into bidding for priorty: perhaps a good thing, perhaps not).

  • http://www.u-blog.net/syd1980 Julien Boyreau

    As said above, bandwidth scarcity is a myth in wired-based network ; Switches&TransmissionMedium Capacities grow much faster that bandwidth usage ; in a pure IP network, it’s much less costly to perform over provisionning than administrate the complexity of artificial scarcity.

    The real problem is that the IP Technical Model and the Unmettered use create a commodity (IP Transport) that has no other diffenrentiation than Price and Performance i.e exactly the same market structure as Intel in the CPU Market (just replace IP Packets by instructions and ADSLx.x by Pentium)

    After having reused Cable and Copper, the next Step to fight on performance is Copper Subsitution by Fiber thus Infrastructure Investment such as Railways, Roads…
    So the fascinating question for free-market supporters (like me). Can market freedom find incitations to finance big naked fibers infras or will public/state money be unfortunately necessary ???

    Could we find an “Broadband Intel” world-sized in the Cellular/Fixed telecom crew ??
    I am quite pessimistic…what do you think about ??

  • http://www.u-blog.net/syd1980 Julien Boyreau

    As said above, bandwidth scarcity is a myth in wired-based network ; Switches&TransmissionMedium; Capacities grow much faster that bandwidth usage ; in a pure IP network, it’s much less costly to perform over provisionning than administrate the complexity of artificial scarcity.

    The real problem is that the IP Technical Model and the Unmettered use create a commodity (IP Transport) that has no other diffenrentiation than Price and Performance i.e exactly the same market structure as Intel in the CPU Market (just replace IP Packets by instructions and ADSLx.x by Pentium)

    After having reused Cable and Copper, the next Step to fight on performance is Copper Subsitution by Fiber thus Infrastructure Investment such as Railways, Roads…
    So the fascinating question for free-market supporters (like me). Can market freedom find incitations to finance big naked fibers infras or will public/state money be unfortunately necessary ???

    Could we find an “Broadband Intel” world-sized in the Cellular/Fixed telecom crew ??
    I am quite pessimistic…what do you think about ??

  • bago

    Most ISP’s I know of just charged for overage.

  • bago

    Most ISP’s I know of just charged for overage.

  • http://www.licquia.org/ Jeff Licquia

    (Seriously, if anyone knows of a bandwidth hog being cut off, please send me info because I try to keep track of these things).

    That would be me, sorta. The effect was the same, at least.

    This was back in the dialup age. I had a relationship with my ISP, they knew I was a “high-volume customer”, and they promised to contact me if I started getting to be a pain.

    Then they were bought. My next monthly bill was for something like $1800.

    The final agreement we reached was that they would only charge me the normal “unlimited” rate for that bill, and in return I would no longer be their customer, effective immediately. I suspect that was the intent of their abnormally high bill.

    I recall looking over my contract at the time, and not seeing any exceptions to the “unlimited” language, so I expect I could have fought it and remained their customer if I had really wanted to. At the time, though, there were plenty of other options, and I was thinking I would soon get this new-fangled DSL thing anyway once it came to my neighborhood.

    Of course, all this predates P2P, streaming video, widely-available broadband, and the dominance of national ISPs.

  • http://www.licquia.org/ Jeff Licquia

    (Seriously, if anyone knows of a bandwidth hog being cut off, please send me info because I try to keep track of these things).

    That would be me, sorta. The effect was the same, at least.

    This was back in the dialup age. I had a relationship with my ISP, they knew I was a “high-volume customer”, and they promised to contact me if I started getting to be a pain.

    Then they were bought. My next monthly bill was for something like $1800.

    The final agreement we reached was that they would only charge me the normal “unlimited” rate for that bill, and in return I would no longer be their customer, effective immediately. I suspect that was the intent of their abnormally high bill.

    I recall looking over my contract at the time, and not seeing any exceptions to the “unlimited” language, so I expect I could have fought it and remained their customer if I had really wanted to. At the time, though, there were plenty of other options, and I was thinking I would soon get this new-fangled DSL thing anyway once it came to my neighborhood.

    Of course, all this predates P2P, streaming video, widely-available broadband, and the dominance of national ISPs.

Previous post:

Next post: