Standing corrected: Stats Can on the Internet Use Survey

CIUS analyst Ben Veenhof provides feedback on my analysis

On Monday I posted part 3 of a 3-part series on what’s wrong with the CRTC’s broadband target. While the CRTC’s specifics – especially the 5 meg downlink, 1 meg uplink, and 2015  target date – have been grist for the pundit mill, my take is a little different. In a word, the CRTC’s regulation of retail Internet access, as well as its inability to understand how the Net works, have rendered the target meaningless.

You’ll undoubtedly want to read the posts for yourself. But it’s easy to pull out my biggest issue with the Commission’s ivory-tower approach: they’re way too stuck on the geography problem to have any time for the affordability problem. In other words, where residents happen to live in Canada should be playing a much smaller role in Internet regulation and broadband development than how much money Canadians make – or don’t make.

Although this proposition has seemed notionally obvious to me for a long time, the release of the latest CIUS on May 25 seemed to offer solid empirical grounds for making this claim. As I’ve done before, I ran my ideas past Ben Veenhof yesterday to get a serious reality check. Ben is a principal analyst responsible for the CIUS and thus knows whereof the CIUS speaks. And as I’ve said before, he and his colleagues have shown themselves very willing to help out by fielding questions and explaining nuances.

I’ve pasted in below most of Ben’s email with his reactions to the Monday post. As you’ll see, he tackles two issues, one small, one bigger. The first involves a rounding error in the online and broadband numbers; the second involves the more critical matter of just who counts as “rural” and who as “urban” for purposes of the CIUS findings. He then has another couple of details to add about the demographics of online. (Disclaimer: Ben’s participation in this exchange should in no way be construed as agreement on his part, or that of Statistics Canada, with anything I have said in this post or any other.)

Here’s Ben…

“You likely derived the figure for % of all Canadian households with a [high-]speed connection (76%) from two other figures provided in the initial Daily release: 79% of all households with a connection at home * the 96% of those connected reporting high speed.  There is some rounding error in this calculation (i.e. if one had more decimal places, one would arrive at a figure of 75% for the % of all Canadian households with a high speed connection).  No error on your part, just one of those things that can be calculated when you have more decimal places at your disposal.

“You also noted “Until told otherwise, I’m going to assume that respondents located in neither a census metropolitan area (CMA) nor a census agglomeration (CA) can be considered as ‘rural’ residents.”  CMAs are groupings of municipalities situated around an urban core, and may contain a mix of both urban and rural areas, depending on the criteria one uses to define ‘urban’.  (CMAs are defined largely by the degree of integration of communities as measured by commuting flows).  However, if defining urban areas based on some other criteria, for example a minimum population density, there may be ‘pockets’ or areas that are part of the CMA that could be considered rural if they do not meet a minimum population threshold.  For example, Uxbridge is part of the Toronto CMA and while both the urban core of Toronto and the township of Uxbridge may be densely populated, there are pockets of less densely-populated areas between the two locations which could be considered rural, if defined based on another standard.  When reporting the CIUS 2010 data from the Daily you may wish to use your own language, but I think it is good that you provided the CMA and CA definition for your readers (as well as the accompanying graphic).

“I’m glad you were able to use some of the results I had sent following the May 25 release, showing rates of access by household income quartile.  I thought I would add that in the initial release, we also noted that 20% of households that lacked access at home identified cost of service or equipment as a reason – this reason was reported by proportionally more households in the lowest income quartile (24%).”