Associate Professor Christopher Balding of the HSBC School of Business at Peking University in Shenzhen has just published a paper arguing there are huge flaws in Chinese inflation data, with major ramifications for the official GDP figures. The full paper is here, and the post that follows originally appeared on his personal site, and summarises some of the problems with the official data from China which, as Australia’s biggest trading partner, may have implications for forecasting domestic economic conditions.
Why Does Anyone Still Believe Chinese Data?
It baffles me with everything that we know about Chinese economic and financial data that people, and even smart people who should know better, continue to believe the Communist Party Propaganda press releases put out by Beijing. The Reuters team writes:
“The slowdown in the rate of deflation was taken as further evidence of a possible stabilisation of the economy, with analysts looking to data later in the day on investment and industrial output for more signs of a bottoming out in activity.”
When you simply make economic data up, of course the numbers are always going to look good. For about 10 years, Chinese unemployment has bounced between 4.1-4.3%. Now even accepting that Chinese unemployment has remained quite low throughout this entire time period, which in itself a dubious proposition, we expect more random variation than that.
Let’s take another example from my recent working paper How Badly Flawed is Chinese Economic Data? The Opening Bid is $US1 Trillion, on how the National Bureau of Statistics in China (NBSC) calculates consumer price inflation and how they should calculate consumer price inflation (CPI).
CPI should be calculated based upon what the average consumer buys. To take a simple example, if Honda sells a million cars a year in the United States and their prices go up by 2% annually, that should count a lot more than the 10,000 Bentleys a year that go up in price by 5% annually.
Despite having the highest scoring high school maths students in the world, it is apparent that none of these technically proficient students work for the NBSC. Below is a table with the official raw data on price change in private housing with a break down by urban and rural residents where the previous year is equal to 100. This pricing data is used to build the Chinese CPI.
Looking at the raw data from the NBSC, it is apparent there is some skewing of the total change to urban residents. When you calculate the total price change with the implied weighting between the urban and rural population, the NBSC is significantly overweighting the urban population. The NBSC between 2000 and 2011 gives an 80% weighting to the urban population with a 20% weighting to the rural population. In 10 out of the 12 years, if utilise a straight 80/20 urban rural, the implied total number is within 1 one thousandth of a per cent of the official number.
However, this 80/20 urban/rural weighting is not remotely representative of the Chinese population. In 2000, China was nearly two-thirds rural and 2010 reached a 50-50 urban rural split. In other words, the NBSC was mis-weighting the rural population by nearly 50 percentage points. This results in a not insignificant accumulated difference over time. The already bogus data price data gets reduced even more in the price basket by overweighting the population with the lower price increase: urban private housing residents. If we weight based upon the actual population weight rather than the Alice in Wonderland 80/20 urban/rural weighting, this raises the price of housing by approximately ½ of one per cent annually of more than 5% cumulative difference.
While this total does not drastically alter the final Chinese CPI number, it definitely adds to it. More importantly, it shows just how absolutely fraudulent the Chinese data is and how many layers the employ to manipulate the data. It is not just manipulating the price data, which they do, but also how to weight that data and who it applies to.
Furthermore, you continue to add up all these little ways that the NBSC is fraudulently manipulating the data and you arrive at some pretty big numbers. As the saying goes, a trillion here and a trillion there, and pretty soon, you are talking some serious money.
From Balding’s World, used with permission.