As China's government struggles to keep its economic growth rates from falling further, it faces persistent questions about its growth claims in the past.
China economists and scholars have argued for years over the accuracy of official gross domestic product figures. In 2007, Premier Li Keqiang described the data as "man-made" and "for reference only," when he served as a provincial leader of the Communist Party of China (CPC), according to a memo published by WikiLeaks in 2010.
Criticisms of the figures issued by the National Bureau of Statistics (NBS) have focused on the remarkable stability of GDP growth rates through good times and bad.
No matter what the external pressures, the NBS figures have had an uncanny knack not only for meeting the government's GDP targets but also for varying from quarter to quarter by no more than 0.1-0.2 percentage points, critics say.
"Over the past 16 quarters, China's headline GDP has only varied by 0.8 percentage points in total, within the range of 6.2 percent and 7.0 percent," the Rhodium Group said in a recent research note.
"If that sounds unusually smooth, it is," the New York-based consultants said.
For well over a decade, the NBS has been accused of "smoothing" the changes in GDP growth rates to project an image of stability, usually leading to overstatements of growth that extol the benefits of CPC rule.
The Rhodium analysis points to wider variations in industrial production and surrogate indicators like electricity output, citing a variety of sources including the work of University of Pittsburgh economist and historian Thomas Rawski.
The discrepancies have been particularly glaring in official GDP growth claims for 2014-2015, when China suffered a widely-acknowledged economic slump.
Although the NBS claimed a GDP growth rate of 6.9 percent for 2015, Rawski estimated the true rate to be in the range of 4 percent or less in a Radio Free Asia interview at the time.
While the arguments about growth and the relative strength of the service sector are familiar, they may be coming under a brighter spotlight now because of combined pressures from the trade war with the United States, policy debates over the global slowdown and GDP goals that the CPC has set for the 70th anniversary of the People's Republic of China this year.
Recent indicators suggest that gaps between official and actual economic performance may be as large now as at any time in the past four years. But the conflicting signs of a slowdown make it hard to tell which set of official figures to believe.
Despite the mild decline in year-to-year GDP growth rates from 6.4 percent to 6.2 percent from the first to the second quarter, according to the NBS, industrial production growth in August fell to 4.4 percent from 4.8 percent in July, marking its slowest pace in over 17 years.
Power consumption in the first eight months rose 4.4 percent compared with an 8.5-percent increase last year, the National Energy Administration (NEA) said.
Doubts about the integrity of the official reports have spurred analysts to seek more accurate measures.
Last month, The Wall Street Journal reported that satellites monitoring China's manufacturing activity have found evidence of contraction, while the NBS continues to report steady growth.
China's official purchasing managers' index (PMI) for manufacturing in September also signaled contraction for the fifth month in a row.
Despite the dim outlook from the PMI indicators, the official English-language China Daily said on Sept. 24 that manufacturing "continued to be the bright spot of China's industrial landscape during the first six months," predicting that it would "replicate [the] impressive growth numbers" in the second half of this year.
Analysts who track alternate indicators have concluded that "China's economy isn't tanking, but it is almost certainly weaker than advertised," the Journal said.
Impact on global economy
Over the years, the NBS has pursued a series of reforms and crackdowns on local data inflation and falsification, but the repeated signs of statistical smoothing suggest that at least some of the fault lies closer to home.
By now, the arguments over NBS accuracy have become old hat, but the Rhodium analysis makes the case that official exaggeration of GDP growth is a problem that goes beyond domestic myth-making and glorification of the CPC.
Instead, it is "vitally important to the global economy, given that China is the second-largest economy in the world," it says.
Bad data leads to bad timing or wrong decisions on monetary and fiscal policies, eventually affecting investment, debt levels and risks.
Doubts about the data have "real economic costs" and may undermine the credibility of the country's leadership, the analysts say.
Rhodium's proposed solution would be for the NBS to make a public revision of the GDP figures for the past five years, with downward adjustments for 2014-2015, leading to slightly higher growth levels for the recovery in 2016-2017 and a reduction in growth for 2018.
The recommendation puts no specific numbers on the annual adjustments, but it argues that "localities" in China have already made downward revisions for their 2014-2015 data that amount to 2-3 percent of the nation's 2014 GDP.
In an email message, a co-author of the research note steered clear of concluding that a revision of that size would be warranted for the 2018 figures, which stood officially at 6.6 percent.
"The short answer is we don't have a precise estimate of the size of the GDP revision in 2018 that is necessary," said Rhodium director Logan Wright, adding that the group's data base tracks industrial output rather than GDP.
But Wright voiced confidence that "actual raw industrial output had been far more volatile than reported IVA (industrial value-added) growth or GDP, particularly in 2018."
In estimating the GDP overstatement for 2014-2015, the analysis relied on previous work to assess the effects of statistical fraud cases uncovered in mostly northern provinces in recent years.
In January 2017, for example, the governor of Liaoning province, where Li served as party secretary, admitted that fiscal revenue figures were inflated by at least 20 percent from 2011 to 2014.
From 2011 to 2016, central government investigators accused at least eight of China's 31 provincial-level governments of data falsification, Rhodium Group said.
But the recommendation for a sweeping revision seems likely to conflict with a core CPC promise to double China's 2010 GDP in a decade.
The pledge made by former President Hu Jintao in 2012 was initially estimated to imply annual growth rates of about 7 percent and could probably be met now with growth of about 6 percent.
Although Hu could not have foreseen the effects of the tariff war with the United States, the government will likely remain in control of the outcome with the power to publish whatever data it wants.
The analysts argue in effect that keeping Hu's pledge without a revision would come at the cost of credibility. But the chance that the NBS will rain on the CPC's anniversary parades seems remote.
"If these longer-term Party-created targets require rates of GDP growth over 6 percent, then it is political factors that are likely pushing reported GDP growth rates higher," said Rhodium's analysis.
"For those who would argue the costs of this adjustment are too high, ... we can only reply that the costs of continuing on the current unrealistic path ... are much, much higher," it said.