Now playing at the CRTC: your precarious future on the Internet (2)



In this post, I follow up on my comments about the first day of the CRTC’s hearing to review its framework for wholesale services in the telecom industry. Since the most significant sector to be affected is Canada’s residential broadband service, I’m summarizing evidence here that was compiled recently by the Open Technology Institute (OTI) that compares broadband in 24 cities in Europe, East Asia and the US, along with Toronto. This evidence is consistent with findings from other international studies. It shows Toronto lags far behind the broadband leaders in available speeds; in the penetration of fiberoptic platforms; in symmetric connectivity (uplink bandwidth matches downlink bandwidth); and, most seriously from a social policy perspective, in the high prices Torontonians are forced to pay. I take this evidence as a strong argument in favor of maintaining and extending the regulatory regime that ensures open access to networks for smaller, competitive ISPs – including not just legacy platforms like DSL, but also emerging fiber platforms. Unless the CRTC includes these next-generation platforms, Canada will fall even further behind in its long slide into slow and expensive broadband connectivity.


“We are now ready to take our place as the most technologically advanced nation on the planet.” –Stephen Harper, Digital Canada 150, April 2014

Last month the Open Technology Institute released the third in a series of annual studies of broadband speeds and prices in 24 cities in the US, East Asia and Europe, plus Toronto (originally 22 cities). I wrote about OTI’s first report back in November 2012 (CRTC’s 2nd pro-consumer decree: 4 reasons not to celebrate); and I had comments a year later about the second report (Broadband data for Toronto: more bad news and getting worse).

The Open Technology Institute (OTI) is an independent, non-profit advocacy group that operates under the auspices of the New America Foundation. Its mission includes the promotion of “affordable, universal, and ubiquitous communications networks through partnerships with communities, researchers, industry, and public interest groups...” One of its most valuable activities is the pursuit of rigorous empirical research about the state of broadband networks, all the data from which is made openly available to third parties.


Three factors to note at the outset. One, the OTI work makes it eminently clear that fiber is not only the platform of the future, but the platform that defines the international leaders today. The second factor goes to the old argument about population densities at the population level. Since this is a study of city-level broadband, low population density as a disincentive to deployment is moot in the OTI methodology. Third, most of the data in this study reflects the kind of “view from the living room” methodology adopted by the OTI researchers. They like to design ratings that would be meaningful to a mainstream consumer in her decision-making about household communications services. That’s consistent with the OTI’s use of advertised speeds, in contrast to other studies focused on measured speeds.

Toronto Internet: bad in 2012 and getting worse

I prepared the three-year series in the chart below to show what market forces have done for Toronto broadband as measured by top advertised download speeds. Toronto has been falling on this metric over the course of the three years in question. On the top speed city ranking for 2012, Toronto came in 9th out of the 22 cities. In 2013, when the sample was expanded to 24 cities, we came in at 15th. This year, Toronto dropped again, to 18th place out of 24 (data were collected right up to this past September).


Now let’s look at the findings in Figure 1 below (screen grab from OTI report, p.12). The most important and unsettling of these is the status of data caps in the various cities. If you look down the data cap column second from the right, what you’ll see sticking out like a sore thumb is the Toronto row featuring Rogers’ contender for top speed – along with the only explicit data cap in this whole set of 24 cities. In half of these cities, 12 in all, there is no data cap (“N/A”), which includes five of the top 10. The other nine are marked with a period – indicating the researchers didn’t have the information either way.


Next, look at the prices. Rogers is more expensive (on an absolute basis) than all but four of the other cities, and all of those are in the US, where the FCC has also decided to forbear from regulating the wholesale market for fiber (OECD data show that fiber penetration in the US is about 7.8% of all connections: see previous post for details). Notice also that the 11 top cities feature one gigabit access, and of these, seven are symmetric one gigabit (1 Gbps down and up). The best Toronto can do is the Rogers platform running at 350 Mbps symmetric.

By the way, those US cities in the top tier – Chattanooga, Kansas City KS, Kansas City MO and Lafayette – have achieved this privileged position not thanks to the incumbents but in spite of them. That’s because in roughly 20 states, legislators have been persuaded by Comcast, AT&T and their cronies to outlaw any form of muni broadband, even when the incumbents had no intention of building faster networks themselves. These four cities are good examples of how muni fiber, the local electric utility and Google have begun to transform the broadband landscape, if only a very small part of it.

Getting a good deal under $40

On Monday the Competition Bureau didn’t stop at telling the Commission it should avoid any regulation of emerging fiber platforms (which are not currently covered under the open access regime). The CB also said the Commission should get out of the business of ensuring that new entrants like TekSavvy can gain access to unbundled local loops, meaning in plain English, just scrap the wholesale mandate and service-based competition altogether, while we wait for the “facts” that will let us make the “right decisions.” (I was greatly relieved when Chairman Blais, reacting to an especially misanthropic response from the Bureau, suggested the CRTC’s role was to act more like a doctor than an undertaker.)


If you had any doubts, the data in Figure 2 above should tell you what we’re in for if the Commission actually listens to the Bureau’s advice. It reflects the best deals the researchers could find in the 24 cities for broadband under $40 (converted from local using the USD/PPP formula).


In a stunning surprise, none of our incumbents was found to be offering a decent deal in Toronto for less than $40, a distinction that goes to indie ISP TekSavvy. I’m not aware of a DSL offer from TekSavvy at the indicated speeds (30/5), but I have their 25/7 DSL, which costs me $39.99 a month with a 300 GB cap (not the 150 GB cap shown in the table). They’re both good deals.

But the OTI has provided us with more than this snapshot. We have the three-year time series again to provide historical perspective. The second chart I’ve made up (below) shows how Toronto has ranked for best deals at the $35 price point in 2012 and 2013, and then $40 in 2014 (once more, we have 22 cities in the first year, then 24 in the subsequent two.


On this metric as well, Torontonians are worse off than they were in 2012. The determining variable here is speed: OTI ranks cities in this price range according to how much bandwidth you get for your money. And once again, the best deal is not with an incumbent in either 2012 or 2013; Acanac took the honors.

Then there’s the fiber. For 2012, OTI’s data shows that 11 of the 22 cities in the “best deals” ranking were on fiber; for 2013, that number dropped to 10 out of 24 cities; and in 2014, the proportion of fiber was back up, to 15 cities out of 24. For Toronto, however, the best deals are confined to DSL or cable across the three years.

Prices for plans in the 25-50 Mbps range

The next dataset I found helpful tabulates the average price in each city for Internet plans in the 25 to 50 Mbps tier, shown in Figure 4 below. Toronto is in the mediocre middle, at 12th place with an average price of $51 (OTI did not provide this metric in its two previous studies).


While there isn’t a great deal of variation in the prices in this dataset (apart from the outlier, Mexico City: thanks Carlos!), one things deserves comment. The average Toronto price of $51 is deceptive because, unlike most of the other 23 cities, that advertised fee is subject to overages because of the use of data caps. CRTC data indicates that a mere 12% of Canadian broadband subscribers were on unlimited plans in 2013. This was actually a slight drop from the level in 2011, when it was 13% (CRTC CMR, 2014, Fig. 5.3.1, p.178).

I wonder tangentially whether this methodology reflects real-world consumer decision-making. Most of us go shopping for things we can identify and describe, some of whose features will count explicitly in our choices – e.g. a 4G phone instead of a 3G. Yet this same principle does not seem to apply to shopping for Internet connectivity. I say that on the basis of a great deal of anecdotal evidence that regular consumers, as opposed to nerds or policy wonks, have no idea what bandwidth they pay for every month. My students majoring in communication studies don’t even know what bandwidth is in the first place; and when I send them home to find out what speed they’re getting, they often come back empty-handed. This mystification strikes me as a major social policy issue for a technologically advanced economy.

How money translates into a pig in a poke, aka bandwidth 

It may or may not be generally true that consumers are unable to make ISP purchase decisions based on a characteristic of signal processing they have no clue about, namely bandwidth. In the absence of anything more than anecdotal evidence, my sense is that most people wanting connectivity start with a budget in mind, then manoeuvre around their ISP’s marketing bullshit and land on a plan they feel they can afford. All that most customers have to conjure with is some silly plan name – Unlimited Semi-Light Wideband Junior.

The other two OTI charts I’ve inserted below show why this purchase dynamic is so important in terms of consumer welfare. The first one (Figure 5) ranks the 24 cities by the average speed you get for forking over $35 to $50 a month. Check out Toronto – 21st out of the 24, better than two American locales and Mexico City…


What does this information tell us? Well it tells me that, if consumers have no clue what speed they’re paying for, they can’t possibly judge whether they’re getting value for their money. In Toronto, the chart above illustrates very clearly just how little value market forces are providing. The policy issue for the CRTC (and, yes, the Competition Bureau) is that mainstreamers are not only getting screwed, they’re more or less by definition unaware they’re getting screwed. Sure, people have an idea that the customer service provided by Rogers and Bell really sucks. Except the deal is even worse than they think.

When we move up the budget scale to anything over $50, we’re still in terrible shape. In Figure 6 below Toronto ranks 16th out of 22 cities (two cities didn’t fit the model). Because these are average speeds at any price point over $50, we can assume that speeds overall are much higher in most other cities in the sample than in Toronto. Yes, some critics will say, who needs all that speed? The incumbents will certainly say that. (I see on Rogers site they do not advertise a 350 Mbps plan; the top speed there is 250.)


But that’s just a piece of sophistry that misses the point. Consumers – not to mention innovators and other valuable members of society – will find uses for all that bandwidth. Moreover, the policymaking process of the kind the CRTC is currently engaged in necessitates taking the long view. It certainly doesn’t need the bankrupt approach of the Consumer Bureau, with its self-fulfilling, anti-consumer prophesies. Broadband policy will not thrive in a climate like Canada’s, where the federal government’s target for 2017 is 5 Mbps. Let us also remember that the CRTC’s more ambitious goal of 5 Mbps by the end of 2015 is defined as “access” to that speed, not in terms of actual subscriptions to that speed.

Let the CRTC keep two things in mind as it deliberates. One, what will matter most a few years from now is how much fiberoptic we’ve got and how affordable it is. Fiber is the end-game for all concerned.

Two, the incumbents have a licence to print money, as the man said once about getting a TV station to run. They had their chance to show how well they could do on fiber deployment without having to endure suffocating, heavy-handed regulations. In a big, dense and affluent place like Toronto, we should have seen lots more fiber deployed. But the incumbents blew it. If Toronto, and Canada, expect to be part of the hyperconnected future, it’s time for the Commission to step in and fix this badly failed market before it’s too late.