The global financial crisis has renewed interest in reforming China’s official statistical apparatus, which is burdened by institutional and structural flaws that make it impossible to produce reliable numbers. And without reliable numbers, the central government cannot make accurate economic forecasts or prudent public policy. Although the country’s National Bureau of Statistics and National People’s Congress have begun addressing the issue, much more needs to be done to remove politics from the process.
Editor’s Note: This analysis is part of a series that explores China’s industry, finance and statistics.
Analysis
RELATED SPECIAL TOPIC PAGE
The Beijing Municipal Development and Reform Commission announced on Sept. 2 that the gross domestic product (GDP) per capita in Beijing in 2008 was $9,075 and is likely to reach $10,000 by the end of 2009. The 2008 GDP figure is a 12.5 percent increase over the previous year. It also means that the standard of living in Beijing — if not the rest of China — has reached “above the middle level,” according to commission officials (presumably comparing Beijing to other comparably sized cities in the world). But the news did little to enthuse the Chinese people, who are inherently suspicious of official statistics and aware of the frequent discrepancy between government numbers and the reality of their lives.
Indeed, unreliable statistics are nothing new in China, where the methodology of collecting, interpreting and presenting numerical data remains a mystery, not only to the outside world but also within China itself. There are many reasons for this. China is a vast country with a huge population that is difficult to measure simply because of its size and distribution. China also has traditionally used its own statistical tools and techniques in order to avoid apples-to-apples comparisons using international norms and to appear more developed than it may actually be.
Even more important, China’s bureaucratic structure, extending from Beijing down to the lowest township level, gives the politicized local governing body too much influence over the local statistical office. This creates an environment in which officials are less interested in “reality” than in climbing the bureaucratic ladder. Motivated by the need to impress their superiors, Chinese bureaucrats at most levels have the incentive to falsify or exaggerate their reporting of statistics, thus giving the central government an unclear picture of the actual situation on the ground.
And when official statistics are unreliable, it is difficult for the central government to gain a sufficient macro view to make accurate socio-economic forecasts — and thus sound policy decisions. This handicap has certainly been evident during the global financial crisis, during which Beijing has had a particularly tough time determining the country’s true economic condition and direction. The central government has had such difficulty that it has begun taking steps toward comprehensive statistical reform. This has been attempted before, however, and it is no easy task.
A History of ‘Reform’
Before the 1960s, official statistics in China, reliable or not, were considered a state secret. During the so-called “Great Leap Forward,” beginning in 1958, nearly all figures were highly distorted, as regional officials competed to produce exaggerated information to meet the central government’s unreachable goals and to secure their paths through the ranks of government. Economic growth in a given locale was often rewarded by promoting the bearers of good news. This misinformation contributed to the great Chinese famine that lasted from 1959 to 1961 and killed tens of millions of people. Beginning in 1978, to accompany the dramatic socio-economic reforms of Deng Xiaoping’s “Opening Up,” efforts were made to modernize statistical measurement and standardize the system in China. Despite these efforts, skepticism over statistical reliability would become only more pronounced.
China’s entry into the World Trade Organization (WTO) in 2001, while not contingent on its meeting WTO statistical standards, did require China’s methods to be compatible with those of its trading partners. In October 2002, Chinese Premier Zhu Rongji made an inspection visit to China’s National Bureau of Statistics (NBS) and called for the creation of a “scientific, reliable and efficient modernized statistical system.” Since then, the NBS has adopted certain methods that comply with international standards set by the U.N., but given the dependence of local NBS offices on corresponding local governments (more about this later), little progress has been made so far in improving statistical credibility.
When Ma Jiantang took over as head of the NBS in September 2008, just as the global financial crisis intensified, calls increased for China to revamp its statistical processes. The world economic situation created great uncertainty and challenges for the central government, which found it increasingly difficult to gain a big-picture view of the impact of the crisis on China. State-controlled media began to editorialize about the need for such reform, and under growing pressure from the WTO, U.N., International Energy Agency and other international entities, Beijing began to intensify its efforts.
In June 2009, the country’s top legislative body (the National People’s Congress) revised the 1983 Statistics Law to prevent the falsification of official data and to impose severe penalties on officials who “intervene in government statistical work and manipulate or fabricate data.” Smaller improvements are also under way or apparently in the offing. The NBS promises, among other things, to improve the calculation of average urban income by including employees of private enterprises, who were previously excluded from the accounting process (employees of state-owned enterprises typically earn more money).
Processes and Power
China’s official statistics are processed by two different and parallel systems: the “integrated statistical system” and the “government department statistical system.” The integrated statistical system consists of bureaus at each level, from the NBS in Beijing down to offices in the townships. Although the local bureaus have the authority to validate the numbers, they are heavily influenced by the corresponding local government offices, which have the authority to promote the employees of the bureaus and fund their operations. The other system, called the government department statistical system, consists of statistical departments set up by the ministries of the central government. For example, the Ministry of Finance provides statistics on financial matters; the General Administration of Customs provides statistics on imports and exports; the Ministry of Commerce provides statistics on foreign investment; and the Ministry of Justice provides statistics on legal matters.
Problems arise mainly due to the makeup of the integrated statistical system. Under China’s political structure, every government official in the country is appointed rather than elected (except at the county and village level and in some cities). As a result, most government officials as well as public employees are responsible only to their superiors, who determine their promotions and career paths. And the government continues to consider high economic performance the primary prerequisite for official advancement.
Statistical Myths
Criticism of Chinese statistical methods has long been concentrated on economic data, such as the GDP or consumer price index, which the government uses to make macro-economic policies and promote China’s image at home and abroad. In February 2009, a debate began regarding the reported decline of electricity consumption along with the reported growth of the GDP. Observers argued that with an expansion of both the service and industrial sectors, electricity consumption must also be increasing. In April, this debate was followed by further suspicion of China’s average urban income as reported by the NBS, which claimed the figure in 2008 was 29,229 yuan, an increase of 17.2 percent from 2007. The public suspected the reported figure was significantly inflated (many Internet users joked in their chat rooms that their wages were raised by the government and their employers were unaware of it).
Moreover, figures showing that industrial output grew in May were not consistent with electricity consumption, which declined, and the NBS later attributed the inconsistency to industrial use of energy-saving technologies. Nevertheless, the inconsistency piqued public skepticism, which was further heightened on Aug. 4 when 31 provincial and municipal governments provided their GDP estimates. The estimates added up to 15.38 trillion yuan for the first half of 2009, which is almost 10 percent higher than the 13.99 trillion yuan reported by the NBS.
The debate has also focused on employment data, amid sharply increased social unrest since 2008. Traditionally, local governments conceal social problems such as unemployment and ethnic strife from the central government. It has been reported that local statistical officials help companies “structure” their lay-offs in order for the unemployment numbers to appear lower, thereby rendering the national unemployment rate meaningless. For example, since the onset of the global financial crisis, Chinese workers in some parts of the country have been encouraged to resign by being given higher severance packages than those who are involuntarily laid off, yet those who are laid off are the only workers counted in the unemployment numbers. Unaware of the magnitude of unemployment at the local level, Beijing can be caught by surprise when simmering social instability breaks into a boil. And it is strategically important for the central government to give the people the impression that it is successfully dealing with the localized crisis at hand, lest the social instability widen.
Then there is the matter of statistical standards inconsistency. China’s GDP, for example, is measured by comparing one quarter of the year to the same quarter of the previous year (year over year). European countries, Japan and the United States use both year over year and quarter on quarter (comparing one quarter with the preceding quarter), but most countries use quarter-on-quarter comparisons for international purposes. The NBS has announced it will begin using quarter-on-quarter instead of year-over-year figures to report GDP in 2010.
The NBS also accounts for inflation in a manner contrary to international practice. Most countries use either current prices (measuring GDP at prices of the current reporting period) or constant prices (measuring GDP at prices of a base period). The NBS, on the other hand, employs a “comparable-price” method, comparing the prices of new products to those of products from a benchmark year. To employ this method, Chinese companies are given price manuals that cover only certain products. As a result, they have ample opportunity to exaggerate business performance by understating actual inflation.
Similar inconsistencies are seen in other statistical areas. For example, investments in fixed assets, such as real estate, are counted as having been made when the funds are disbursed, rather than when the money is actually spent (meaning many unsold houses are actually counted as investment). Retail sales are counted when a factory ships products to a retailer, not when the retailer sells products to a consumer. Also, unsold products that the retailer returns to the factory, which in turn ships the products to other retailers, are also counted in the sales figures, as are items procured by the government.
By now, the NBS is acutely aware of the urgent need for comprehensive statistical reform, but the NBS can only do so much. The central government is trying to balance a number of urgent statistical needs: It must have better numbers to assess the state of the economy and society so that it can make accurate forecasts and more effective policies; it must manipulate those numbers at the top in order to achieve political goals, internally and externally; and it must do all that without scaring off foreign investment, which is vital in maintaining the “Chinese economic miracle.”
Of course, not on the list is the need to address the institutional and structural reasons that the statistical process is inherently flawed in China. It is one thing for the National People’s Congress to give national laws governing statistics more teeth; it is quite another to overhaul China’s political structure to make officials more accountable. And since true political reform is not on the central government’s agenda, “Chinese numbers” will continue to be an issue in assessing and dealing with China.