Canada’s broadband troubles, a year on

The initial announcement of the CRTC’s intention to launch a new media proceeding – Broadcasting Public Notice CRTC 2008-44 – was published exactly one year ago to the day (May 15, 2008). The PN included the usual mantra about our world leadership in broadband (para 18):

“Canada is a world leader in broadband connectivity. Recent data published by the Commission indicates that high-speed residential Internet access is available to 93% of households across the country and has been adopted by more than 60% of Canadian households. Canada places first among G8 countries, despite the challenges inherent in connecting a geographically dispersed population.”

Lately I’ve been trolling through transcripts of the oral interventions at the new media hearings that began February 17. From what I can tell, Chairman von Finckenstein is quite serious about recreating a cyber-version of the broadcasting system on the Web.

Let’s look at some actual data

In his exchanges with the big ISPs on March 10, the Chair queried the supplicants closely on whether it’s technically feasible to tag and measure packets of Canadian data so that the Commission might godfather some kind of “fast lane” for Canadian content – professionally produced, TV-like content. In my 2008-11 submission, I bemoaned the Commission’s decision to exclude both UGC and interactive content, since personal content and interactivity pretty much define networked digital media. But as the table below indicates, we have a much bigger problem on our hands that makes this debate about content moot: i.e. the debate we’re not having about the net-gen distribution platforms we need to build so we can deliver next-gen content.

[table removed]

The data here are adapted from an April post by Dr Larry Press, Professor of Information Systems at California State University, entitled “Why is connectivity in Stockholm so much faster and cheaper than in US cities?” I deleted a couple of Press’s examples and added Bell and Rogers to extend his comparisons to the two nearest equivalents in Toronto. There are two other differences between the original table and my mashup. Price figures for the six locales from the original table are in US dollars, so at today’s US exchange rate, the Bell and Rogers prices would be about 15% lower. I’ve calculated these prices from the Web site promos, but factored in the fine print – meaning I didn’t use the lowest come-on price, which goes up after 12 months, assumes you’re bundling, and leaves out the modem lease and taxes. But the really important difference is that subscribers in the first six locales are getting their connections over fiber.

So my first question looking at the table is: How does Canada get to be not only first in the G8, but a world leader in broadband connectivity?… Claims echoed at the hearings by intervenors like Rogers: “Canada has an enviable record. Today, Canada’s broadband connectivity exceeds that of every other G8 country” (para 9614).

The G8 countries are Canada, France, Germany, Italy, Japan, Russia, the United Kingdom and the United States. According to the table, subscribers in Tokyo, Japan can get connected at 100 megabits, up and down, over fiber, for $61 a month. Subs in Lafayette, USA can get connected at 50 megabits, up and down, over fiber, for $58 a month – if they’re on the municipal network, not the Cox network. Press hypothesizes that the most important variable in explaining connection speed and cost is who owns the network. The point is made strikingly by the example of Lafayette, where the cable incumbent, Cox, charges a rate 2.4 times higher than what is charged for access to the municipal network – even though the municipal network offers 10 times more bandwidth in the uplink.

The secret to Canada’s leadership

So far, all this evidence runs entirely counter to the claim that we lead the world, or even the G8. It turns out, however, that there is a grain of truth buried in the latest available data on the OECD’s broadband portal.

The home page divides the research into five major sections: Penetration (actual lines); Usage; Coverage and Geography; Prices; and Services and Speeds. I’ll be coming back to some of the concepts and numbers here, but for now, let’s focus on the penetration data. Not counting the press release and criteria used, that section is sub-divided into 11 datasets – one of which (1i) is devoted to historical penetration rates in the G7. If you click here, you’ll download an Excel spreadsheet that shows Canada leading the G7 and the OECD average – in number of broadband lines per 100 inhabitants. As of Q2, 2008, Canada led on this metric with 27.9 lines (or subscriptions) per 100 inhabitants, whereas the OECD average is a mere 21.3.

What the Canada boosters are doing, however, is cherry-picking a single metric out of a total of 43 (and we haven’t even got to the data from the ITU and other international sources) – a piece of chicanery my undergraduate students spotted in five minutes. The Commission might do well to start from the beginning and read the OECD’s June 2008 Press Release. Out there in the real world, the overall score looks a little different:

“Denmark, the Netherlands, Norway, Switzerland, Iceland, Sweden, Korea and Finland lead the OECD with broadband penetration well above the OECD average, each surpassing the 30 subscribers per 100 inhabitants threshold.”

Oops, no Canada. Things get even more interesting in the next item, the Broadband Subscriber Criteria:

“All statistical country comparisons should be undertaken with caution and this advice similarly applies to broadband statistics. There is a breadth of market, regulatory and geographic factors which help determine penetration rates, prices, and speeds. Therefore, it is important that policy makers examine a wide range of broadband indicators when considering key policy decisions.

“Broadband subscription penetration rankings tell nothing of the prices that users pay, the advertised speeds of connections, or whether there are restrictive bit/data caps on those lines. Countries doing well in one measure may be weaker in another.”

Do our regulators need to be reminded that you can’t make good policy with bad information? That you can’t simply pick out the numbers that justify the status quo, while ignoring affordability, the deployment of fiber, symmetric architectures, broadband quality and network ownership models?

As I’ll try to show in future posts, Canada has a great deal to lose in the outcome to the new media proceeding.