Look at the data

I noticed a number of people retweeting a chart produced by Visual Capitalist that looked at the price of 1 GB of mobile data in various countries.

Canada was listed near the top with an average price said to be $12.55 (USD), along with a caption of “Canada’s high cost for data can be explained by its high market concentration. In 2018, Bell, Rogers and TELUS collectively had over 82% of all wireless service revenues”, citing the CRTC as its source. I am not able to find a source for that figure. The CRTC’s Communications Monitoring Report actually shows a higher data point for 2018, but that isn’t really what is important. It simply isn’t true that ‘high market concentration’ is responsible for Canada’s mobile service prices.

In virtually every developed country in the world, the top three service providers hold an even higher share of the market. The trajectory of Canada’s smaller carriers is increasing; in 2019, Freedom Mobile and Videotron accounted for a third of net new subscribers. Other countries are experiencing consolidation.

In his opening remarks at the CRTC Review of Wireless Services, Dr. Robert Crandall testified “Canada continues to benefit from a competitive wireless industry, an industry that is less concentrated than the wireless industries in all but two developed countries in the world – Denmark and Sweden.”

Dr. Crandall continued, saying “In a regulatory environment that has relied heavily on platform competition, Canadian carriers have invested far more per subscriber than have European carriers, who have been traditionally much more regulated.” Indeed, if you look at the text that accompanies the pretty picture (and I know, many don’t bother with such detail), you’ll find that Visual Capitalist has a section that discusses variations in prices. “Relatively wealthy nations tend to charge more for mobile services since the population can generally afford to pay more, and the cost of operating a network is higher. This is apparent in countries like Canada or Germany.”

But let’s look at the data in a little more detail. How was this Canadian average price of US$12.55 calculated? The source information for the Visual Capitalist chart comes from cable.co.uk; there is a lot of important detail to be gleaned from its original research and data set [xlsx, 128 KB].

Apparently, cable.co.uk looked at 60 Canadian price plans on February 6, 2020. The samples ranged in price from C$2.50/GB to C$140/GB. Yes, that is right. Converted into US (cable.co.uk used an exchange rate of US$0.712 = C$1.00), and we see a range of US$1.78 to US$99.68. A remarkably wide range included in the sample. How was the sample size of 60 determined? In its description of methodology, the company says “Packages were recorded up to a maximum of 60 per country – records beyond this number have negligible impact on the average.” But, that actually isn’t true.

Plans that are in the order of $100 per GB each add $1.66 to the average, more than a 10% impact on the average for each plan at that level. Canada is one of very few countries in the sample that hit the arbitrary maximum in the sample size. In itself, isn’t that an observation on the amount of choice available to consumers? Looking at last year’s data, the maximum price plan included in the Canadian samples was $80, contributing to a lower average price point, despite a worse exchange rate. The data set shows the average at C$15.92 (US$12.02) on October 26, 2019 compared to C$17.63 (US$12.55) on February 6, 2020. Does anyone think prices actually increased in Canada between October and February?

Further, the mere availability of a plan at that $140/GB level (likely something like a $15 plan with 100MB), says nothing about its relative popularity or relative affordability. It is absurd to think that someone looking for 1 GB of data would give more than a passing glance at a $140 plan, when unlimited plans are in the marketplace from about $50. Similarly, someone mainly looking for an emergency phone, needing minimal data, might find such a plan to be a better solution than an unlimited plan.

Is it even appropriate to use straight arithmetic averages to compare countries or should there be weighting based on various factors, such as which plans consumers are actually choosing in the marketplace.

It’s very easy to look at a chart on social media, nod one’s head, and retweet or reply without bothering to look beyond the headline. It is tougher to apply a critical eye, look at the data, and determine policy based on deeper analysis.