Varun Sivaram

Energy, Security, and Climate

CFR experts examine the science and foreign policy surrounding climate change, energy, and nuclear security.

Print Print Cite Cite
Style: MLA APA Chicago Close


Making Sense of China’s Energy Intensity Statistics

by Michael Levi
April 7, 2010

Stephen Howes (via Roger Pielke) takes a look at recent Chinese statistics and calls recent Chinese claims to have cut energy intensity by 14.38% from 2005 to 2009 into question. I’ll be the last person to defend the integrity of Chinese government statistics. That said, I think there may be a way to understand the discrepancy – but I don’t think it resolves the fundamental problem.

Howes cites the 2009 Statistical Communiqué as putting 2009 energy consumption at 3.10 billion tons of coal equivalent (tce) and the 2008 Statistical Yearbook as putting 2005 energy consumption at 2.25 billion tce, for a growth of 37.97%. He then cites the 2009 Communiqué as showing Chinese GDP grew 9.6% in 2009 and 8.7% in 2008, and the 2008 Statistical Communiqué as indicating that Chinese GDP grew 13% in 2007 and 11.6% in 2006. This adds up to 50.24% growth from 2005 to 2009. Combining these figures gives an energy intensity improvement of only 8.2%. This strikes me as a completely reasonable calculation.

So what gives? One obvious possibility is the officials are just making stuff up. Another explanation, though, may be that different statistics are using different definitions (or more precisely, normalizations) of GDP.

According to the 2009 Communiqué, nominal Chinese GDP grew from 18,493.7 billion yuan in 2005 to 33,535.3 billion yuan in 2009, a jump of 81.3%, or 16.0% annually. Over the same period, if you combine the CPI changes reported in the Communiqué, you find a total CPI increase of 11.9%. If you use that to normalize GDP, you get a 4-year increase of 62%, equivalent to an annual increase of 12.8%. And if you combine that with the reported increase in energy consumption, you get an intensity reduction of about 14.2% – roughly the same as the number that China has claimed.

Of course, China itself has not reported average GDP growth statistics this large – they’ve reported the ones that Stephen Howes cites. That’s because the GDP deflator they use is not the same as the CPI increase: the first index is (appropriately) based on production, while the second is about consumption. The discrepancy is naturally large in export-intensive economies like China’s.

All of this leads to two questions: First, given that I’m not an economist (nor an expert on Chinese statistics), do my numbers correctly explain (if not justify) the Chinese intensity claims? Second, if they do, is this way of measuring things one that the world should accept? After all, when China says that it’s going to cut emissions intensity by 40-45% from 2005 to 2020, the underlying definitions matter — and using CPI rather than the standard GDP deflator seems to be defining that goal down.

Post a Comment 3 Comments

  • Posted by David Cohen-Tanugi


    Thank you for sharing this insight. This is obviously an issue of great interest, and decyphering Chinese statistics and certainly be a headache. The understanding we have from the experts at the Lawrence Berkeley National Laboratory is that more often than not, China’s official statistics turn out to be largely accurate – it’s just a question of figuring out the assumptions, coefficient, units and other details which are often not included.

    I’ve taken a further look at your very reasonable inference that China’s energy intensity numbers may have been based on a CPI-adjusted GDP. One issue leaves me wondering whether we might still have not quite uncovered the entire mystery.

    First of all, I agree with your cumulative numbers: a 62% increase in CPI-adjusted GDP over the past four years, and a 37.97% increase in primary energy consumption. However, I don’t believe it’s appropriate (or mathematically valid) to reduce this to the yearly numbers.

    CPI-adjusted GDP is 1.62 times bigger in 2009 than in 2005, and primary energy consumption is 1.38 times higher in 2009 than in 2005. Therefore, the appropriate calculation is to divide the cumulative numbers to find that China’s energy intensity in 2009 is (1.62/1.38)=0.852 times what it was in 2005. Subtracting 100% give us a cumulative reduction of 14.8% over the period, not 14.2% as you find in this post.

    It’s certainly not impossible, but at least rather improbable, that the statistic cited by Wen Jiabao would *understate* China’s progress in reducing energy intensity so far. If I’m correct with this cumulative calculation, then the reduction calculated based on a CPI-adjusted GDP is actually even stronger than what Wen officially announced. This leads me to believe that either (a) your hypothesis is incorrect and Chinese statisticians were *not* using CPI here; or (b) the picture is more complicated than we may have initially assumed.

  • Posted by Stephen Howes

    Michael and David put forward the interesting idea that the Chinese claim that energy intensity has declined by (about) 14% since 2005 is based on deflating GDP by the CPI deflator rather than the GDP deflator. Interesting but not ultimately convincing, for two reasons. First, there is no precedent for this, and it would be a very odd thing to do. Second, look at the 2009 numbers. The 2009 Communique from China’s NSB gives energy growth of 6.3%, GDP growth of 8.7%, and a fall in energy intensity of 2.2%. All internally consistent, and all done using the GDP deflator (you’d get a different answer if you used CPI).

    So I’m still looking for a resolution.

  • Posted by Ian Castles

    Michael Levi speculates that the Chinese Premier’s statement may be based on a calculation by which nominal GDP is deflated by the CPI (i.e., prices of consumer goods and services), rather than the prices of goods and services making up GDP as a whole. I doubt this, but such a procedure would in any case be illegitimate.

    In 2005, individual consumption expenditure by households in China amounted to only 38.0% of China’s GDP. The corresponding proportion for the US was 70.4%. Conversely, expenditure on gross fixed capital formation was 41.5% in China and only 19.2% in the US (all data from World Bank, 2005 International Comparison Program, Tables of Final Results, 2008, Table 11).

    Emissions per unit of expenditure on capital formation would of course be far greater than emissions per unit of expenditure on consumption in both countries. It would be entirely wrong to deflate GDP by a consumer price measure if the purpose is to measure emissions intensity.

Post a Comment

CFR seeks to foster civil and informed discussion of foreign policy issues. Opinions expressed on CFR blogs are solely those of the author or commenter, not of CFR, which takes no institutional positions. All comments must abide by CFR's guidelines and will be moderated prior to posting.

* Required