China GDP Data Ratings
The Quality of GDP Data in China is Improving, and Up to Date in Many Respects.
But is Still Some Way From Good Quality. Use With Caution!
China's GDP data, unlike that of many developing countries, is based on a recent Base year, and on the latest internationally agreed System of Nation Accounts (SNA). It is therefore in many respects up to date and based on sensible methods. As such it is relatively useful data, and probably of better quality than that of most developing nations. It is also of increasing utility as Government makes sensible changes to improve data quality.
Nevertheless, the data is frequently subject to criticism, notably about the unusual speed with which the data is produced, and with potential over estimation of the growth estimates over time.
Factors used to judge GDP Data Quality
Six factors are used to judge GDP data quality and to provide countries with a Data Quality Ranking (DQR) score. These are:
- Base Year used to calculate the GDP data (years out of date)
- Standard of National Accounts (SNA) applied
- Estimated Size of the Informal Economy
- A Proxy for Resources Devoted to Measuring Economic Activity
- A Proxy measure for likely Government Interference in Economic Data production
- A Proxy variable for Regulatory Hurdles faced by enterprises generally
Table 1: China: Relative Factor Values and Total DQR Score
|Country||Base Year||SNA||Informal Economy||Resources||Government Interfearence||Regulation||DQR Score||Global Rank ▾||Grade ▾|
- China ’s Base Year is 3 Years out of date
- China has adopted the latest SNA standard
- China has a large informal economy at an estimated 12% of the size of the official economy
- Our proxy measure for the level of resources in China is high at 75 out of 100
- Government Interference is rated at 39 out of 100 – a low score means more interference
- On this proxy measure China scores 74 out of 100 – a high score means less regulation
Base Year: China’s current base year is 2015, three years out of date, within the recommendation that base years be regularly updated every 5 years. Out of date base years can cause considerable distortions in the measurement of GDP particularly in rapidly growing economies so the regency of the base year is a major plus point, placing the data on a par with that of many developed countries in this respect. According to the IMF the NBS has improved the ‘exhaustiveness of the GDP estimates by activity,’ but changing data collection systems to move to a system of chaining ‘is restricted by the decentralized nature of the statistical system.’ We estimate, based on a model of the impact of base year updates across the world using data from 16 counties that if China updated its base year to 2018 or adopted chaining based, official reported GDP could rise by several percentage points.
SNA: China uses the latest version of the United Nations System of National Accounts (SNA 2008). However, it must be noted that using the most up-to-date standard does not mean that all of the recommendations in SNA 2008 have been implemented. China is still in the process of moving its data collection and analysis methods between SNA 1993 and the latest version. The 2008 SNA has already been implemented in some areas such as the calculation of financial intermediation services indirectly measured and in the capitalization of research and development, but according to the IMF the techniques for deriving volume measures of some GDP components for calculating national income by the output method “are not sound and need to be improved”. The expenditure method is used to estimate annual GDP components at constant prices on annual basis, but not on a quarterly basis and even the annual data is not published.
Government expenditure and output are essential components of national income accounting and accurate estimates of GDP. Unfortunately, the quality of some government data in China is poor according to the IMF. The annual frequency of data provided on social security and extra budgetary funds is inadequate with a long time lag between reports. China uses the approved Government Financial Statistics (GFS) reporting system for general government, but provides no breakdown of expenditure by economic type apart from interest payments. The Chinese authorities have indicated an intention to begin collecting these data and to present them on a quarterly basis, to develop accrual based measures of fiscal performance as recommended by SNA 2008 over the medium term while strengthening the compilation of cash based GFS, but no specific timeframe has been revealed. Government data also excludes the operations of the local government financing vehicles (LGFVs) which have been estimated to have outstanding debt at around US$6 trillion, but which are not recorded as public sector liabilities.
Overall China has not fully implemented the recommendations of SNA 2008.
Estimated Size of the Informal Economy: In many economies, particularly in emerging markets, GDP is potentially appreciably larger than is recorded in the official statistics. The DQR employs data for the estimated size of China’s informal economy for 2015 provided by the IMF Working Paper: WP/18/172 which estimates a level of 12.1 per cent of measured GDP.
It is difficult to account for the size of the informal economy in official GDP figures, but SNA 2008 recommends that countries try to include estimates to reconcile GDP income and expenditure estimates and to investigate illicit and illegal activities. China could do more to make better estimates of the output generated by the informal economy and we believe that the employment of resources to evaluate these activities could add several percentage points to recorded GDP.
Proxy for Resources: The quality of national income estimates depends on the statistical capacity and the resources available to national statistics offices, the National Bureau of Statistics. The score for this component of the DQR is derived from the United Nations Human Development Index (HDI) used as a proxy for assessing the availability of economic resources in national statistics offices. The HDI ranges up to a value of 100 which implies the highest level of resources and China scored 75 on the HDI scale just above the median score for the Asia-Pacific region at 70.
Proxy for Government Interference: Governments interfere with the production and dissemination of basic economic data in many ways and for a variety of reasons. The DQR uses a general measure of corruption as a proxy for Government interference .The score for this component of the DQR is derived directly from Transparency International’s Corruption Perceptions Index (CPI) which measures the level of perceived corruption in 175 countries.
The CPI is measured by a score up to a 100 with a high value implying less corruption. Corruption has been endemic in China for some time. The CPI score estimates the level of corruption in China at a high level with the country scoring 39, implying less corruption than the median value of 36 for the Asia-Pacific region, but worse than the regional mean level of 43 which is skewed by the presence of New Zealand, Hong Kong, Australia, Singapore and Japan.
Proxy for Regulation: The quality of economic information provided by the private sector to the state is negatively related to the degree of government intervention and regulation. The DQR uses the World Bank Ease of Doing Business Index (EODBI) which measures the level of regulation in 190 countries. A score of 100 indicates very little regulation and according to this measure China has a score of 74, better than the median value of 68 for the Asia-Pacific region.
Impact of Improvements in Measurement on GDP Size
China is already the largest economy in the world in terms of GDP rankings measured by the World Bank in Purchasing Power Parity (PPP) terms so an improvement in data quality would not improve its relative ranking by size.
Improvements and Problems
It is difficult to produce accurate economic data in a country the size of China, but the NBS has made significant efforts to improve data quality in recent years. Nevertheless there are a number of issues that are not reflected in China’s DRQ ranking. A number of economists have been expressing concern about the accuracy of Chinese economic statistics on GDP and economic growth for some time.
The International Monetary Fund noted in July 2018 in its latest Article IV Staff Country Report that China’s data gaps undermine policymaking, the credibility of official statistics, IMF surveillance, and G20 commitments, and need to be ‘urgently addressed.’
For example, price Index data is used in order to deflate nominal GDP estimates to produce real GDP measured in constant prices. The Consumer Price Index (CPI) is extensive covering approximately 500 areas throughout the country, including around 200 counties and around 300 cities, but the expenditure category weights derived from urban and rural households’ surveys every five years, are not published. The Producer Price Index (PPI) with a current base year of 2010 has industry definitions which are consistent with international standards, but it is not seasonally adjusted.
One leading academic specialist in Chinese economic data, Christopher Balding, has questioned the reliability of price data, and its impact on GDP estimates for some time. He suggests that the reliability of the size and growth rates in real GDP are particularly questionable. In a paper in World Economics published in 2014 Balding argued that the NBS had been persistently underestimating inflation with the effect of biasing real GDP upwards by 8 to 12 percent over the period 2000- 2011 or more than $1 trillion in PPP terms.
The possible upward bias in China’s GDP estimates has also been noted (for a different reason) in a recent report published by the Brooking Institution in May 2019 which estimated that the NBS over-reported its economic growth between 2008 and 2016 by an average of 1.7 percentage points. The difference arises because local governments are rewarded for meeting growth and investment targets and manipulate the data upwards while the NBS has not adjusted data sufficiently to reflect this bias, although the IMF has noted with approval that the NBS has decided to take over the collection of provincial GDP estimates from local bodies this year which may lead to future improvements.
China’s declared GDP, therefore may well be biased upwards in any given period, but measurement problems in one of the components of aggregate demand – investment – could inflate current GDP at the expense of sustainable real growth in national output. A problem noted by Michael Pettisfor the Carnegie Endowment argues that unlike other economic systems there are fewer constraints on the ability of the Chinese authorities, particularly at the local level, to fund non-productive activities which create little economic value, but which still raise the burden of debt. Non-productive economic activities will generate employment and GDP which will be reversed in the future once debt capacity has been reached.
Overall the World Economics Data Quality Rating (DQR) of economic data in China is in the “B” grade. Although it provides a good guide for many purposes there are a number of issues needing attention. Nevertheless as measured by the World Economics Data Quality Rating (DQR), Chinese data is ranked as of better than average quality in the Asia-Pacific region, holding 7th place out of 27 countries which translates into 46th position globally out of the 154 countries ranked.