I have some free advice for Chairman Genachowski.
Get an FCC team working on Canada’s experience with broadband caps and get enlightened. He wasn’t saying it for the first time. But telling the country’s cable operators at their annual confab he’s a fan of broadband caps is taking this old debate out for a whole new kind of spin… the operative term being spin.
The Washington Post and others reported yesterday that Genachowski put it like this:
“Business model innovation is very important. There was a point of view a couple years ago that there was only one permissible pricing model for broadband. I didn’t agree.”
Some commentators find it mighty strange hearing this kind of endorsement – especially in the middle of a brawl launched by Reed Hastings over possibly anti-competitive behavior on the part of Comcast. Which announced a few days ago it was experimenting with broadband caps in ways that may or may not turn out to be consumer-friendly.
The Americans should be looking at the Canadian experience to find out just how bad things can get when the regulator and the incumbents get into bed together over caps. It’s one goddam slippery slope – even if I do agree with Pete Nowak’s assessment that, compared to what Canadians have to put up with, the Americans can go cry us a river.
The many ways the cap argument has it all wrong
The Americans have not had the benefit of seeing a great deal of dirty incumbent laundry aired in public as part of a regulatory proceeding – like we did in 2011 in the runup to the November 2011 CRTC decision on incumbent broadband pricing. Although the then-chair refused to make that proceeding about retail as well as wholesale caps, we learned a great deal about what to believe on a whole set of consumer-related issues.
A few key words have popped out in the US press coverage in the last 24 hours. The two most important of these are congestion and cost, which help the incumbents – and apparently Genachowski – win the headline wars. As in our networks are getting congested by those pesky subscribers of ours… and somebody has to pay for all the associated network engineering costs!
So here’s my laundry list of myths, misconceptions and outright bullshit about broadband caps – or what us Canucks refer to as UBB, usage-based billing.
1 – Congestion isn’t the problem for which data caps are the solution.
One way the Canadian incumbents kept the CRTC and its customers painted into a corner was by incessant references to the problem of network congestion. Congestion is at least an implicit part of the rationale for Genachowski’s shady “business model innovation” – i.e. getting captive subscribers in a broadband duopoly to pump up ARPU.
The most thorough debunking of the congestion myth I’ve read was the paper released by Michael Geist in March of last year, at the height of Canada’s UBB debate (I’ve uploaded the pdf here for reference).
Congestion doesn’t occur everywhere across the network, nor is it the result of some users using too much capacity. As Michael explains (pp.9-10), any actual congestion will arise in one place, i.e. over the incumbent ISP’s internal network, between the last mile and the handoff to the public Internet. The incumbents pay no transit or other related cost here, because they own and control this local access network. And internal networks can be cleared of congestion in several other ways besides imposing data caps on all users, including investment in greater capacity and the reallocation of bandwidth that vertically integrated ISPs are devoting to their own services, especially IPTV.
2 – Either caps are cost-based or they aren’t. If they are, they don’t deter usage. If they aren’t, they’re just a money grab.
I’m not clear whether Genachowski’s announcement signals a desire to help the US incumbents do what ours were supposed to do: ease congestion by using caps as an economic disincentive to using the Internet. But this goal was, believe it or not, part of the CRTC’s rationale for its economic ITMPs, issued as part of its 2009 review of ISP traffic management practices (Telecom Regulatory Policy CRTC 2009-657).
What we learned eventually was the incumbents were being encouraged to charge as much as C$5 for one gigabyte of data over a sub’s cap – whereas the likely economic cost of transferring that incremental GB was probably less than 5 cents. Do the math. (I see Comcast is charging or proposing to charge a whole lot less – more like $10 for 50GB according to a recent piece in Ars.)
Michael shows in his paper that, far from trying to justify their over-use pricing on cost, the incumbents stood the whole argument on its head, telling the CRTC cost-based pricing “would not take into account the broader behavioral impacts that permit the attainment of the underlying carrier’s network management objectives” (Geist, p.6). In plain English, the Commission and its licencees agreed to impose punitive caps on Canadian subs to keep them the fuck off the Internet.
Yes! Galloping into the digital future by discouraging Canadians from using the Internet to get a post-secondary education, find a job, use government services, help the kids with their homework… Meanwhile, Bell’s CEO was telling the Boys from Bay Street on a call that their big revenue jump in residential in late 2010 came from… usage-based billing, thanks to all that video congesting Bell’s under-performing network (Geist, p.7).
Maybe Genachowski’s been talking to his former counterpart and now soulmate, the former chair of the CRTC. Konrad von Finckenstein was busy weathering the UBB storm through 2011 by telling everyone who would listen, including Parliament, that we needed the UBB régime not to keep everyone off the Net – just those evil bandwidth hogs, who were spoiling ISP service for the rest of us.
Too bad that isn’t what causes congestion. And too bad punishing heavy users is therefore misguided and ineffectual. It turns out that congestion isn’t caused sporadically by heavy users; it’s a function of peak and off-peak demand, just like traffic congestion on the 401.
The demonization of hogs, which was very popular in Canada for years, fits nicely with another piece of nonsense in this reasoning: that broadband is just another utility like water or electricity. It isn’t. Unlike water and electricity, bandwidth is not a consumable resource. And consumers take an access lease based in the first instance on capacity – i.e. bandwidth – not volume, a piece of physics completely irrelevant to utilities.
One more thing about how fatuous this policy reasoning is. I wrote a post last fall entitled “How ITMP-based data caps punish light instead of heavy users.” In it I demonstrated (or claim I demonstrated) that the CRTC framework, implemented in a jurisdiction where retail broadband isn’t regulated, was probably having consequences both unintended and undesirable. I made up the chart below to show how Rogers’ rates at the time were designed to discourage light users, who I assume are over-represented for low income households, while heavy users (presumably over-represented for high income households) were getting a big volume discount.
This approach may make good business or marketing sense. But it’s no way for the regulator to manage the Internet or make social policy; and, given Genachowski’s controversial remarks yesterday, it also sets a really bad example for the FCC.
(I’ll be back tomorrow with the rest of my list…)