Understanding Chinese Statistical Data: The Devil Is in the Details

by: Susan Weerts
On November 11, the International Monetary Fund (IMF) released its report on the recent technical assistance mission to China, concluding:

“Overall, substantial progress has been achieved by the authorities in all areas of macroeconomic statistics in recent years and STA’s TA program has contributed significantly to this progress.However, much remains to be done to address the data weaknesses, which still hamper economic analysis, policy formulation, and monitoring.”

The IMF identified several risk factors in the areas of personnel turnover, institutional and data quality. Data quality has been a persistent issue plaguing the reliability of Chinese Statistical Data. The IMF highlighted the following three areas as being of the most concern regarding data quality.
(1) Limited focus on the usability and accessibility of data.
(2) A lack of credibility of statistics due to the lack of explanation for apparent inconsistencies in certain data sets.
(3) The delay in posting of GDDS (General Data Dissemination System) metadata.
The controversy surrounding the Chinese Statistical Data has been persistent through the years. Data released earlier this year individually by China’s 31 provinces and autonomous regions showed a combined GDP of about RMB 1.4 trillion more than the figures given by the National Bureau of Statistics of China (NBSC). That is, the discrepancy between the local government and national government data was as high as 10%.
Over the years, the NBSC has gradually shifted away from the Marxian Material Product System (MPS) toward the United Nations System of National Accounts (UNSNA). However, the Marxian MPS still exerts certain influences at the provincial level. The MPS and UNSNA differ in their conceptual approach on the value-added calculation for GDP. This is particularly apparent in the areas relating to non-material production, such as housing, health care, etc.
As a result of the IMF’s report and subsequent exchanges with the NBSC, Mr. Zhilong Peng, an auditing official with the NBSC, posted a commentary on the NBSC’s website on December 2, 2009 that acknowledged some statistical imperfections still existed within the system in spite of the Bureau’s best efforts. He stated, “Over the year, the proportion of consumption in terms of the total GDP has been shrinking even with a strong GDP growth.” Consumption growth in 2008 was 8.8%, well below the 9.8% GDP growth. Some areas of deficiencies in data gathering and analysis he highlighted were:
  1. Lack of representative sample survey in the current calculation of household consumption. The current urban and rural household survey procedure was established in the 1980s. It is obsolete and suffers several deficiencies, such as small sample size, non-uniform segmentation, etc.
  1. Bulk of the Chinese consumer consumption is based on cash-transactions. However, the consumption calculation is based on accounting data.
  1. Failure to capture the true picture since the survey participants are reluctant to report any “gift” payments in relation to medical care, school, job-searches and etc.
  1. Difficulty in estimating the rental and owner-occupied housing consumption. Due to lack of rental estimation, the Bureau resorts to cost estimation method in calculating the housing consumption.
When questioned about the Chinese retail sales data in the Q/A session at one recent conference last month in New York City, Dr. Feng Xiao, CEO of Bosera Asset Management with RMB 209 billion (approximately USD 30.6 billion) AUM, replied that the Chinese retail sales data underestimated the Chinese consumption.
He used his recent interior decoration of his apartment as an example. He paid it in cash without signing the final receipts. He went further to claim that majority of the Chinese consumption went without any receipts. If Dr. Xiao’s statement is true, it raises a very important question: If the majority of the sales of goods and services are not properly recorded, how good are the Chinese companies’ financial statements?
It is well-known that the Chinese firms keep at least three sets of books, one for reporting, one for tax and one for themselves. As the NBSC surveys are based on accounting data, and given that most consumption is “under the table,” the discrepancy between the true Chinese consumption and the official retail sales data is likely to be very large. Furthermore, many investors, including nearly all the investment bankers, have a strong conviction on future Chinese economic growth, stemming from the expectation of strong Chinese consumption growth and the ability of China to transform itself into a consumer economy. For those whose conclusions are based solely on the official statistical data, the premise can be somewhat questionable.
Two revisions might be needed to forecast for Chinese economic growth.
  1. Current Chinese consumption might be more robust than the official data depicts, and therefore needs to be adjusted upward.
  1. The growth potential of the Chinese consumption is likely overestimated, given that the current consumption is likely higher, and the wage growth is lagging behind GDP growth. After taking account of all the omitted consumption expenditures from the official data, the Chinese consumers might be near their optimal consumption levels.
The constraints on the future growth can be seen:
  1. Low wage growth and uneven distribution of wealth. The percentage of wages in terms of GDP fell to 11% in 2007 from its peak 17% in 1980. On December 3, 2009, The People’s Daily, cited that the top constraint on consumption is low wage growth, especially among those with the low incomes.
  1. High Housing Prices. The “Economic Blue Book” by the Chinese Academy of Social Sciences on December 8, 2009 predicts a 9% GDP growth in 2010. However, it also indicates that 85% of households will not be able to afford to buy a house at the current price level. The ministry of Commerce estimated that total retail sales in 2009 reached RMB 5.7-6trillion and nearly half of that expenditure was on real estate. This leads to the conclusion that people are sacrificing personal consumption to invest in Real Estate.
  1. Inflation. A survey showed that among the 100,000 families that purchased a home this year in the cities of Beijing and Shanghai, 70% of them allocates more than 50% of their monthly incomes to their mortgages. Recent price hikes for water, electricity and fuel will further limit their discretionary consumption. It is estimated that these households will start to take on debt if the inflation rate reaches 3% next year.

Disclosure: no position