One of the great intellectual puzzles for me as an economist living in China is why so many smart people obsess over official Chinese GDP growth data. Given that there is extensive research demonstrating how Chinese GDP data is managed and even Li Keqiang in his much touted index has warned against reading too much into it, people continue to read every decimal point as an indicator of good times or impending doom. Chinese GDP data is kind of like the latest Hollywood comic book tent pole, full of special effects that leaves us entertained and wanting more.
The problem, however, lies in finding alternative measures of Chinese GDP growth. Every alternative measure of Chinese GDP growth that I am aware of, has used some variation of the Li Keqiang Index using electricity and goods traffic as a proxy for economic activity with a few additional measures thrown in for good measure.
These are good attempts to produce better data but run into two significant problems. First, they are using either fragmented or imperfect data. This may be from industry groups or other sources that have good data sources but it typically has the very real potential of omitting or missing observations. Second, if they use more comprehensive data it typically relies on some government or quasi-public data collection agency. However, this creates a battle of the wits scenario. If government statisticians are smart enough to manipulate GDP data they are smart enough to manipulate secondary data especially after Li Keqiang made his well known observation. They know that you know, so clearly they cannot leave this economic variable untouched. In other words, attempts to estimate GDP data using any public or quasi public data in China suffers from a positive bias.
Greenpeace has released a report (make sure to click through to the underlying links) suggesting that in the first quarter 2015, YOY CO2 emissions and coal consumption have fallen by 5% and 8% respectively. If true this staggering and incredibly important for our understanding of the Chinese economy. There is one important caveat. Greenpeace is utilizing Chinese government data, which as I just noted, is notoriously unreliable. As one article about the Greenpeace report notes, China has previously reported large drops in coal production only to later adjust the numbers back up enormously due to producers simply not reporting output.
While we need to proceed cautiously in interpreting these numbers, for the reasons noted, I believe we can make reasonable interpretations of this data. Let us consider a couple of scenarios and the subsequent interpretations. It seems unlikely that Chinese statistical agencies are purposefully upward biasing the fall in coal and electricity usage. By that I mean, if Greenpeace is reporting that coal use in China fell 8% based upon official data, that it actually fell 12% but is biased upwards. Furthermore, the China Electricity Council, electricity growth in the first four months of 2015 grew YOY by a total of (wait for it) 0.2%.. I don’t think even the most concerned about the Chinese economy would be predicting falls in coal and electrical consumption even larger than what is being reported. Therefore, I consider it less likely that these numbers are manipulated upwards in any meaningful way.
It seems somewhat more likely that the numbers are downward biased. However, a willful downward manipulation is not completely satisfactory explanation for a few reasons. First, given the widely recognized existence of the Li Keqiang Index even within China, downward manipulation would seem to needlessly invite speculation about the health of the Chinese economy. Second, drops in domestic production of energy producing commodities are matched and even exceeded in relative terms by imports. Given the difficulty in manipulating trade data, which two countries participate in, the general trend of the data seems matched by international data. Third, this data is broadly matched by domestic Chinese data that is strongly correlated. For instance, there has been widely reported flat or falling production of coal and electric intensive industries like steel and cement. This also is matched with similar falls in international commodity trade and prices. Fourth, given that China has told the world that its CO2 emissions and coal dependency will continue to rise until 2030, there seems little incentive for it to manipulate the data downwards now. Fifth, nor do the numbers appear to be falling due to a shift in the Chinese portfolio of energy production. The speed necessary to produce a shift of this magnitude is simply enormous and while hydropower production surged in 2014 due to higher rainfall, new electricity production remains overwhelmingly coal based. In short, the secondary data fails to support the idea that this data on coal and CO2 emissions is significantly downward biased.
What seems like the most likely scenario is that these numbers are broadly representative and accurate. I should note that doesn’t mean the data is perfect and unmanipulated, only that appears most likely to be generally representative of coal, CO2 emissions, and electricity production. In my opinion, Any significant manipulation would require actual electricity growth to come in upwards of at least 2-3%. In short, much of the data matches close enough that we can feel comfortable that any manipulation is either well coordinated across independent actors or telling a similar story.
If these numbers are generally if not perfectly accurate, this has one very important implications: real GDP growth in China is nowhere near 7%. In other words, electricity consumption was flat. Producing GDP growth of 7% with total electricity growth of 0.2% would require electricity efficiency gains never before seen in modern economic history. If this number is generally accurate, this would imply that Chinese GDP growth is in the low single digits. A plausible guestimate, based upon electricity growth between January to April 2015, would be GDP in the 1-3% range. There is simply no way you can have zero electricity growth and manage 7% GDP growth.
The low growth of estimated GDP would further explain many of the strong actions taken by Beijing to support GDP growth such as the “call it anything but stimulus” stimulus to the bailout of indebted provinces. Maybe Beijing knows something the rest of us don’t?