We need to be reasonable about broadband usage caps
After posting this article on American broadband caps in perspective, it seems we’ve attracted a few angry responses. Advocacy group Stop the Cap’s Phillip Dampier pulled out all the stops to denounce me and our executive director Jon Henke that we’re essentially selling broadband consumers out. So what’s the reason for Dampier attacks? It turns out that Dampier’s Stop the Cap campaign wants the bill HR 2902 passed which would outlaw usage caps, and they can’t afford to have someone point out that every broadband provider in the world has to have some kind of usage cap (disclosed or not), and that the US has some of the most generous usage caps in the world.
The low blow on Net Neutrality
Dampier starts out his attack with the typical straw man argument that George Ou hates Net Neutrality and is probably in favor of blocking VoIP services and ISP censorship. Next he’ll probably say that I like to club baby seals and kittens too. The problem with these kinds of attacks is that they don’t advance the debate and they’re just ripping a page from the SaveTheInternet/Free Press talking points that Net Neutrality is all about stopping Internet censorship when it has never been an issue of censorship. Free Press and other Net Neutrality advocates have continued to propagate the myth that AT&T’s network blocked Pearl Jam and Verizon blocked text messages even though they never happened. AT&T’s website simply didn’t want to carry a political message when it was supposed to be a rock concert, and Verizon simply denied an initial request for a special shortened phone number but later approved it.
The actual Net Neutrality bills e.g., Markey’s Net Neutrality bills essentially banned tiered pricing on higher priority Internet services when the Internet has never worked that way. Even staunch Net Neutrality advocates like Larry Lessig and Sir Tim Berners-Lee have stated that they oppose bans on differentiated pricing on differentiated services, but we don’t see Free Press and their allies attacking them even though Lessig and Berners-Lee have virtually identical positions on Net Neutrality as I do. I suppose one difference is that Tim Berners-Lee is willing to lend his reputation to officially endorse Net Neutrality regulations that violate his own principles while I am not.
The hypocrisy over metered Internet access and usage caps
Free Press, the EFF, Tim Wu (chair of Free Press board), Larry Lessig, and even Vuze networks (the peer-to-peer company that filed the FCC complaint against Comcast) were all PROPONENTS of metered pricing as a superior alternative to network management, at least when they were all trying to fight reasonable network management.
Tim Wu said:
“I don’t quite see [metering] as an outrage, and in fact is probably the fairest system going — though of course the psychology of knowing that you’re paying for bandwidth may change behavior”
Free Press testified to the FCC:
“More importantly, if Comcast is concerned that the collective set of users running P2P applications are affecting quality of service for other users on a cable loop… they could also charge by usage.” (p. 29)
“Indeed, in many nations, network providers do meter, and bill their customers on the basis of amount used. So the transaction costs of doing so must not be prohibitively high. Indeed, a network provider can apparently meter cheaply because, in most networks, users’ traffic to and from the Internet passes through a single gateway, the network access server.” (p. 31)
I fought hard against their efforts to PROMOTE usage caps as the superior way of managing the network. I exposed the EFF’s efforts to promote usage caps as the better way to manage the network. I testified that usage caps were the most inferior way of managing a network before the FCC against Larry Lessig who argued that usage caps were the superior solution. Yet we fast forward a year or two and Free Press’ lobbying arm Free Press Action Fund is now lobbying hard for the Stop the Cap campaign and HR 2902 to make usage caps illegal.
I pointed out that some form of usage cap was necessary SO LONG AS we don’t have intelligent network management solutions in place, but Dampier quoted me out of context by claiming George Ou says “usage caps are necessary”. So what does Mr. Dampier think about Free Press and all the others who continuously argued and testified that metered pricing and usage caps were the superior solution to intelligent network management? He thinks that’s just fine because we must have misunderstood them and besides, they’re allowed to change their mind and he’s not going to attack an ally. Yet Dampier thinks it’s ok to misrepresent my work and attack me even though I am a leading advocate of metering alternatives simply because I don’t support his particular bill. I don’t believe HR 2902 is necessary in light of the fact that we have some of the most generous usage caps in the world and in light of the fact that the market already rejected Time Warner’s more restrictive caps and forced Time Warner to back off.
Dampier’s tin foil hat theories and lies
The more we read Dampier’s comments, the more it becomes clear that his rage applies to any commercial or business entity. For example, he was angry that Time Warner’s commercial customers didn’t get a cap and claimed that businesses were the biggest consumers of bandwidth. But that makes no sense because businesses usually ban P2P file sharing usage in the work place and they consume far less bandwidth than residential customers. Moreover, commercial customers PAY MORE money to get more capacity provisioned to their network, so why shouldn’t they get more capacity? The funniest part came when I suggested the obvious solution that consumers simply need to buy the commercial grade broadband connection and they would have the same benefits as the business customers. Dampier replied with his tin foil hat theory that Time Warner doesn’t allow consumers to buy commercial grade access. How does he know this? Because Time Warner never mentioned the option so that must mean they prohibit consumers from buying commercial grade broadband.
Dampier was also dismissed my suggestion that dissatisfied Time Warner customers can exercise their rights as consumers and switch to alternatives like AT&T. To make his point, Dampier selectively quotes AT&T’s 20 GB usage cap as the only other alternative. When I challenged him on this point and pointed out that AT&T had a 6 Mbps service with a 150 GB usage cap for $35/month, Dampier resorted to lies and claimed that AT&T makes you pay $45/month for 6 Mbps service while actually delivering 1. But I know that isn’t true from experience because AT&T is very conservative about offering faster DSL unless they’re sure that it can deliver stable sync rates that match the advertised speeds. I even had to spend weeks escalating to AT&T support managers to get AT&T to sell me 3 Mbps service because I lived in a borderline area at 13,000+ feet from the CO. When called on this lie, Dampier tried to clarify that he was just including taxes in the price and he was citing one particular customer who had a problem circuit but was allowed to switch to the lower rates for the lower speeds. Somehow I have a hard time believing Dampier’s explanation and it seems more like he’s willing to say anything and propagate any theory to advance his agenda.
Are unlimited usage caps for broadband even realistic?
Phillip Dampier and Stop the Cap don’t just want to stop Time Warner from implementing their 40 GB cap. That much is obvious because Time Warner already backed off from their proposal but they’re still a convenient bogeyman to rally their ultimate cause of zero usage caps. Comcast’s 250 GB caps is simply not high on their “target list” for the moment because they just want the bill slammed through congress which will make it easy to target all ISPs.
But are zero usage caps even possible in Broadband? The reality of the Internet or any packet switching network is that it’s impossible to guarantee that users can constantly use the network. The “secret sauce” of the Internet is the fact that the packet switching network efficiently shares all available bandwidth between the few active users rather than the old circuit switching telephone network that wasted the majority of capacity on the idle circuits. Packet switching networks allows Internet based networks to dynamically expand in bandwidth allocation, but the flip side of a network that can expand is a network that must eventually contract when more users are active. Packet switching networks however do have a maximum number of bytes it can transmit a month and that budget has to be shared amongst its users. It is not an infinite capacity network where everyone can saturate their connection 24×7.
All broadband networks rely heavily on this sharing concept called “statistical multiplexing” where every bit of broadband network capacity is probably shared between 20 to 40 people. This is often portrayed as some kind of evil and greedy business model, but it’s actually a great for consumers because shared broadband is 20 to 40 times cheaper than dedicated circuits. Case in point, Comcast’s 15 Mbps DSL service costs less than $3 per Mbps per month while a business class T1 circuit costs $129 per Mbps per month. Best of all, the fact that the average broadband consumer uses their connection less than 5% of the time means that consumers can get close to 100% the performance of a dedicated circuit for less than 5% of the cost. But this only works when the average duty cycle is reasonably low, and that duty cycle has to be kept low with implicit or explicit usage caps.
So at best, we can over provision commercial grade networks with sufficient revenue and capacity that caps aren’t needed. Even the broadband services that don’t advertise caps rely on some kind of implicit usage cap mechanism. They all have terms of service informing customers that they are not allowed to abuse the network and interfere with their neighbor’s ability to use the network, and they almost always prohibit the operation of “servers”, which belong in the more expensive commercial grade broadband services. But explicit caps are more transparent, and more transparency is something that all of us at Digital Society heavily supports. So now that ISPs like Comcast and AT&T have come out and clearly disclosed the usage caps that they’ve had all along implicitly, we’re going to punish them?
Intelligently managed networks mitigate the need for usage caps
Intelligently managed networks like Comcast which employs “fair share” which deprioritize customers who use the most bandwidth behind customers who use the least bandwidth can offer the most generous usage caps. That’s because intelligent networks deal with the problem of peak consumption directly without having to knock down average consumption with more restrictive usage caps. But even with a good network management solution in place, residential broadband networks need some kind of usage cap that serves the needs of the vast majority of consumers. Those consumers with more demanding needs can choose commercial grade broadband services.
Broadband providers who don’t employ these intelligent network management schemes have to resort to far more restrictive usage caps to indirectly deal with peak bandwidth consumption by knocking down the average consumption. This is why I have consistently stated that usage caps are the inferior solution and I strongly criticized Free Press and others for promoting the myth that usage caps are the superior alternative to network management. Now Free Press want to swing 180 degrees and get us to believe that ALL usage caps should be illegal, but Free Press’ hypocrisy and unreasonable stance on network management have shown us that they have no credibility.
The current global broadband market has almost universally implemented some form usage caps and the vast majority of countries have more restrictive caps than the United States. Even Time Warner’s more restrictive usage cap is generous by global standards but the American market has rejected it and Time Warner has listened to the outcry. That is the free market in action and it has worked. What we don’t need is legislative action that swings too far to the other extreme where even sensible usage caps are outlawed.