Sunday, 10 July 2011

Beware of Fairfax newspapers bearing data

I took this photo on Thursday - it's from the Financial Review, and it show CO2 output per capita. Australia is the yellow bar third from the top.

I scratched my head for a few moments and wondered how they produced this result.

Then I looked in the bottom left hand corner of the graph - it says "Source OECD".

Aha - you see, the OECD only has 34 member countries, whereas there are around 215 recognised countries worldwide. By using a limited dataset, they've made Australia's CO2 output look really bad.

If you look at all countries on a per head basis, Australia is 12th.

If you look at all countries on a total output basis, Australia is 16th.

I wonder why Fairfax chose to represent the data in the way they did? Hmmm....


cav said...

So how exactly is CO2 measured?

Col. Milquetoast said...

They didn't even list all of the OECD members.

Worth mentioning is that none (I think, I only verified the top 10) of the nations omitted from the top x list were Annex 1 countries under the Kyoto Treaty so they are under no expectation for them to lower emissions. I also note that what is listed as "Korea" appears to be only South Korea (also not an Annex I country).

If they had included non-OECD nations and limited it to the same range (from the top emitter to Turkey) then it would have been a top 97 list instead of a top 30 list. That a lot of emission omission.

I don't fault them for limiting the list to prevent tediousness but limiting it by OECD status instead of emissions is odd.

It would have been interesting to see the chart adjusted in light of Qatar's 55.4 metric tons per capita (wiki says their peak was 65.8 in 1997). Apparently the wiki table doesn't count CO2 from exhaling because they list Chad, DR Congo and a few others at zero.