Blue Skys over Beijing or Statistical Manipulation?

A post that may be illuminating for any economist working on Chinese data, especially environmental data.

One World...One Dream..One Gas Mask: Part 349 [Time China]

../

But a closer look at the data and changes in collection methods casts doubt on the government's sunny claims -- and raises serious questions about Beijing's commitment to a green Olympics.

../

On Jan. 1, the government announced "blue sky" days had improved to 246 last year, up from 100 in 1998. The news was widely reported inside and outside of China.

What wasn't reported, though, was a change in collection methods. The Beijing API is an average of data from selected monitoring stations. From 1998 to 2005, the same seven stations -- located in the city center -- were used to measure air quality. These stations monitored areas with different characteristics, including high traffic areas, plus residential, commercial and industrial districts. In 2006, however, just as international scrutiny on China's air quality was increasing, two stations monitoring traffic were dropped from the city API calculations, while three additional stations in less polluted areas were added.

Calculating the average daily Beijing API values for 2006 and 2007 using data from the original monitoring stations changes the outcome considerably; in fact, 38 of Beijing's 241 so-called "blue sky" days in 2006 would not have qualified as "blue sky" under the old methodology. The number is even less for 2007: 55 fewer days would have attained the "blue sky" standard, out of 246 reported "blue sky" days. That translates into fewer "blue sky" days as a whole than in 2002 (which had 203 reported "blue sky" days), immediately after Beijing was awarded the Olympics, and casts grave doubt on China's reported five straight years of continuous air quality improvement.

This is not all - even the definition of pollution has been changed. Sigh.

The government also substituted in less stringent measures of pollution. Beginning in June 2000, measurements of nitrogen dioxide were substituted into the air quality calculations in place of measurements of nitrogen oxides. The new standard for nitrogen dioxide was much less stringent than the old standard for nitrogen oxides, which were the worst pollutant (in terms of number of weeks exceeding air quality standards) before 2000. Since then, not a single day has exceeded the standard, thanks to the new, more easily attainable criteria.

.