Battle of the broadband studies (3)

If this is “affordable,” I’m giving up my citizenship.

The recent study entitled Lagging or Leading: the state of Canada’s broadband infrastructure was financed by Bell Canada, Bell Aliant, Cogeco, Rogers, SaskTel, Shaw and TELUS. It says over and over that, once you root out the incompetent studies, Canada looks great, belying those over-simpified headlines about a “crisis” in the house. It returns often to how affordable broadband is in Canada:

“Canadians have access to some of the most affordable services, while also benefiting from some of the world’s fastest connection speeds for both wireline and wireless broadband services” (p.4).

Leave aside the sophistry in disconnecting affordability and speed (we have some of the former and also some of the latter; how about affordable and fast?). And let’s not quibble about USD PPP, weighted averages or whether anybody at the Berkman Center can run a decent regression analysis. Let’s just look at a few humdrum retail prices: what subscribers in France and Canada are actually paying today for broadband on its own and in bundles.

Are Canada and France perfectly comparable? No. In fact, the Lagging authors seem to think that all comparisons are invidious comparisons. They warn, referring particularly to the OECD, that the “difficulty of conducting a 30-country international comparison in a rigorous scientific manner should not be underestimated” (p.60). The OECD broadband geniuses apparently didn’t know what the hell they were getting themselves into.

One thing the OECD folks can do, as residents of France, is get what I would call affordable broadband (all prices in Cdn dollars). Check it out:

  • Today in France: 24 Mbit/s stand-alone DSL = $24/month (before VAT)
  • Today from Bell: 16 Mbit/s DSL = $65.90/ month (before taxes)

Bell says the “best price” for their fastest service – remember the French service is 50% faster – is $51.95. But that’s in a bundle. And they don’t exactly foreground the $3.95 a month you pay for the DSL modem (which I’m guessing Bell recovers the cost of in 6-8 months).

Bit caps: a Canadian success story

But the really problematic hidden cost is the bit cap. The Bell service carries a 75-gig cap; the French service has none. The Bell service is already 2.75 times more expensive, before taking this into account. Let’s look at the price of these two services as if they were both “unlimited.” Bell charges $1 a gig for any overage, which maxes out at $30 a month.

For the sake of argument, let’s say my friend in Paris uses 100 gigs in a month and I consume the same amount in Toronto. Forget about whether this is “high” consumption or otherwise. The prices now become:

  • Today in France: 24 Mbit/s = $24/month
  • Today from Bell: 16 Mbit/s DSL = $95.90/ month

On that basis, the Bell service is exactly four times more expensive than the French service – for 2/3 the nominal bandwidth. Part of the absurdity in this scheme is that as speeds increase, bit caps become more punitive. And they look less like a way to manage traffic and more like a way to boost ARPU.

Speed – and prices – through the roof

Here’s how really fast services line up in the two countries. France has the following five 100-meg services, with uplinks ranging from 2 to 100 megs (prices again converted to C$):

  • France Telecom (100/100) = $102
  • France Telecom (100/10) = $71
  • Numericable (100/2) = $32
  • Free (100/50) = $47
  • SFR (100/50) = $47

The few very fast Canadian services are not only expensive, they’re asymmetric – and all have bit caps. Here are three Canadian examples on the cable side:

  • Shaw (100/5 with 400 GB cap) = $157
  • Shaw (25/2 with 150 GB cap)  = $96
  • Videotron (50/1 with 100 GB cap) = $80

The only Canadian offering here that matches the French services is the 100/5 Shaw service. At 5 megs in the uplink, it falls halfway between Numericable at 2 megs and France Telecom at 10 megs, whose prices are $32 and $71 respectively. That makes the Shaw service five times more expensive than Numericable. More charitably, if you average the two comparable French services, you get about $50 – still making the Shaw equivalent three times more expensive.

It’s also more expensive to phone Canada from Canada than from France

Bundles are designed to save you money. But compared to what? Certainly not compared to the standard offerings in France. Free, the French operator, charges the equivalent of $47 for the following triple-play:

  • DSL (26/1)
  • TV (200 channels)
  • unlimited phone calls to 80 countries including Canada.

How about TELUS? Here’s a comparable triple-play from them:

  • DSL (15/1) – cost $38
  • TV at lowest price in build your own package – cost $25
  • unlimited calls to US, CA + 1000 min international – cost $60

The TELUS Total: $123.

If you’re on TELUS, in other words, you pay $60 to make unlimited calls across Canada, and pay another $63 for DSL and TV. If you’re on Free, you can make unlimited calls across Canada for $47 – which includes your DSL, with nearly twice the bandwidth, and 200 TV channels. How do you like them apples?

[I just checked with my Man in Paris and he confirms that none of the French broadband providers applies a) caps; b) fine print that makes “unlimited” limited; or c) any other hidden monthly charges, like Bell’s modem lease. They do have cancellation penalties.]

Enough with the affordability bullshit

Apart from all the flaws in my methodology, I fail to see how anyone in this country can say with a straight face that we have “affordable” broadband. At least I agree with the Lagging authors on one point: we need a lot more work on the demand side. But when it comes to the reasons why 30% of us keep turning up our noses at broadband, especially very fast broadband, you gotta wonder about this kind of thinking:

“[W]e believe that more attention should be focused on the demand side of the equation, to develop a better understanding of why some households choose not to adopt, or why Canadian broadband consumers seem satisfied with 5Mbps service or less, when 50Mbps service is increasingly available. Instead of focusing so much on penetration rankings, we should concentrate on identifying and eliminating the barriers to adoption. If we succeed, better penetration will be the natural result (pp. 66-67).”

Why are Canadians shunning fast broadband? Uh, because it’s so freaking expensive?

Although affordability goes hand in hand with the elimination of bit caps, North Americans will have to suffer through further experiments with usage-based billing. As the OECD points out, bit caps are the exception not the rule in most other member countries. In 609 observations made as part of their 2008 survey, the OECD found that Canada was one of only four countries (out of 30) that featured bit caps on all the services observed. For their part, the Lagging authors think bit caps are so unimportant that they devote less than one paragraph to them (p.17). By the same token, I found exactly two references in the Lagging study to symmetric access – neither of them even related to wireline platforms (pp. 13, 23).

And a final (anti-consumer) word from our regulator

Whatever else you may think of their October decision on Internet traffic management practices, the CRTC is clearly not trying to become Canada’s answer to the Federal Trade Commission. What got all the recent attention was technical ITMPs. But I’m just as worried about the encouragement the Commission has given the incumbents regarding economic ITMPs – their obscure euphemism for bit caps and other billing practices used to dampen demand.

On October 27, Chairman von Finckenstein gave a talk in Montreal in which he discussed the ISP decision. I found one of his comments very unsettling:

“We have a strong preference for economic ITMPs, given their obvious transparency. The market can be left to manage bandwidth through the interplay of supply and demand. Users should be able to decide how much bandwidth they want or need and what they are prepared to pay for it.”

ITMPs and bit caps are “transparent”? The invisible hand of the market is managing bandwidth? I think not. Sure, users should be able to decide how much bandwidth they need – except, as I’ve noted before, not even my 4th-year Communication Studies majors know what bandwidth is, let alone how much they need. This is bad policy, which does exactly the opposite of what it claims, i.e. give consumers more decision-making power.

The insidious assumption behind this remark is that network management needs will always trump consumer welfare, and the incumbents need economic ITMPs, including bit caps, to stay in business. Our policymakers don’t need economic ITMPs to signal to consumers we can’t afford fast broadband in Canada. We figured that out on our own, thanks.