The conversation over Beijing pollution has been extensive, but I found something that may be worth mentioning. Today I ran across an entry in Managing the Dragon, a fairly new blog by Jack Perkowski who was immortalized in Tom Clissold’s Mr. China.

Anyway, he found a piece from China Auto Review concerning the alleged success of the experiment to move cars off the road for the Olympics, and made a very brief comment on it. As I will demonstrate, the good people at China Auto Review seem to be on the Beijing payroll, or at least unfamiliar with statistical analysis. I also discovered something suspicious on an official Chinese environmental site, but we'll get to that later. First things first.

China Auto Review is a paid site so I cannot verify, but according to Jack the article said:

August 16 was a day of slight pollution with pollutant index of 116 in Beijing, according to the municipal weather report. The following days were of the same weather conditions having small wind but heavy humidity and haze not conducive for pollutant dispersion. As anticipated, however, the pollutant indexes were considerably reduced to 91, 93, 95 and 95 for four days in a row from August 17 to 20, largely due to the presence of fewer vehicles in the city areas.

There are several problems with this, not the least of which that 116 is hardly ‘slight’ pollution (it is nearly three times the legally allowed limit in Paris and the reading, which was actually 115, was within 1 point of August’s worst day). However the most ridiculous part is the claim that these four days showed any reduction in pollution at all. Aside from August 16th, the day before the experiment started, every single day in the previous week had lower pollution than any day of the test. In fact, from the beginning of August until the 17th, the average pollution rating was 79.5, well below the 93.5 during those four days.

Furthermore, the data from final 12 days of the month strikes another blow to China Auto Review’s case. Over that the pollution level was even lower, at an average of 75. For the entire month there were only five days as bad as the average of the four day period described as having “considerably reduced pollution.”I feel like I have fairly convincingly crushed their argument, but just in case you like conspiracy theories, when I went to the SEPA website where the government publishes its pollution data, I noticed something else: August 20th, the last day of the traffic experiment, is not a 95 as claimed in China Auto review, but instead completely missing. You can get data on every day from the last five years, but the one entry missing on both the English and Chinese versions is the last day of the Beijing Olympics pollution experiment. I’m not saying someone took it down. I’m just saying.

(Below is a screen shot from the website. Note the 20th is missing. Click to enlarge)