Thursday, October 30, 2008

 

Looking deeper at OECD numbers

OECDEarlier in the week, I wrote about the the latest OECD broadband report which shows that as of June 2008, Canada continues to hold the number 10 position, based on connections per 100 people.

As you know, I like to view the reporting of OECD numbers with a grain of salt.

We used to gaze enviously at Korea as a pace setter for broadband connectivity. In Canada's heyday, we were second only to Korea in broadband penetration. In the current report, the OECD points out that "Korea’s fibre penetration alone (12.2 per 100 inhabitants) is higher than total broadband penetration in 5 OECD countries."

In preparing material for a talk next week, I noticed an interesting situation when combining a couple of the OECD's tables.

Something is strange with Korea's reporting of household data. I looked at the OECD's tables for households with computers [ Excel, 50.5 KB] and households with broadband [ Excel, 37.5 KB] and tried to combine them to look at the percentage of households with computers that have broadband internet access.

One would think that homes with computers would be the asymptote for broadband penetration - why would a household subscribe to broadband if you don't have access to a computer?

Well, apparently in Korea it doesn't work that way. In fact, 20% of their broadband enabled households have no access to computers.

What is going on there?

When you dig deeper, it turns out that Korea's household broadband data includes broadband enabled mobile phones. Since virtually every phone in Korea is a 3G phone, every household with a mobile phone is scored as having broadband access.

Sometimes, you need to scratch beneath the surface.

By the way, as a percentage of homes with computers, Canada's broadband penetration ranks 6th in the OECD, behind Korea, Iceland, Japan, Belgium and Netherlands.

Technorati Tags:
,

Comments:
I don't think that's correct. If you look at the penetration per 100 figure, Korea has 8.4 DSL, 10.5 cable, 12.2 fibre and 0 other, for a total of 31.2. That works out to 26.9%, 33.6% and 39.1% each (100% total). Mobile/wireless does not look to be included.
 
In the posting, I provided links to the two source tables which have "households" as the denominator. The penetration per 100 table isn't relevant for this calculation.
 
I think I can help out here as I've swapped some e-mails with the OECD folks who put these figures together and have written a fair number of these broadband ranking stories. Mark, you're comparing two different sets of statistics - one is subscriptions (aka penetration), the other is usage. The subscription figures, which are the ones used to determine broadband rankings, do not count 3G mobile internet. According to the OECD, the subscription numbers are very reliable and easy to compare because operators know exactly how many lines they have in operation and report them accordingly. Usage figures, or the household numbers you've referred to, are reported differently in each country and therefore harder to compare. As you pointed out, Korea reports 3G mobile users, but we don't know who - if anyone else - also does (Canada may, in fact). The usage numbers are therefore less accurate and less representative of a country's particular broadband situation. As the OECD has pointed out many times, the main point of the penetration rankings is to illustrate the value proposition of broadband for consumers and businesses. Over time, and when considered with other measures such as pricing, growth rate and speed, the penetration ranking also provides a nice snapshot of the state of competition in a given country.
 
Thanks Peter for your help. Actually, the figures that I used for this particular analysis come from household survey results - the tables have the appropriate footnotes.

I confirmed with the OECD analyst that my figures are comparing apples and apples. It was the initial commenter who mismatched the tables.

What actually skews the subscription figures is that, for many countries, it includes business connections. Further, by using "pop" as the denominator, it is biased against countries with larger than average household sizes and benefits those with smaller households.

It is why I keep saying that the raw rankings and numbers are irrelevant without analysis and context.
 
Post a Comment



<< Home

This page is powered by Blogger. Isn't yours?