Dogged by doubts about exaggerated economic claims, China's government is moving to raise penalties for statistical fraud.
On Oct. 8, the National Bureau of Statistics (NBS) announced an extensive series of revisions to its basic data gathering and reporting law.
The proposed amendment would hold officials and their supervisors directly responsible for the data they submit and increase fines for falsification, according to state media and draft revisions on the NBS website.
In a statement cited by the official English-language China Daily, the agency appeared to acknowledge that a series of earlier reform efforts had fallen short.
"We revised the law this time because we hope to build a modern statistics system in the new era and to improve the accuracy, integrity and promptness of the statistical materials," the NBS said.
The paper cited a high incidence of fraud found by investigators in a report to the Standing Committee of the National People's Congress (NPC) last year.
Figures were "seriously fabricated" in 58 percent of the companies and 94 percent of the fixed-asset investment projects that were examined for "abnormal data" since 2017, according to statistics cited in the report.
The prevalence of data faking suggests that little has changed since a similar NPC probe in 2009 publicized cases of blatant falsification, including overstatement of some production figures by as much as a factor of 10.
Over the years, data inflation has been blamed on career-minded cadres seeking promotions based on economic performance.
The exposures led to a partial overhaul of NBS operations in 2012, requiring some 700,000 enterprises to bypass local and provincial officials by reporting directly to central data collectors in Beijing.
But these efforts were also short-circuited at the local level by officials who pressured the enterprises through their own direct channels to pump up production numbers, the 21st Century Business Herald reported at the time.
The new draft amendment attempts to deal with the problem of career advancement.
"No unit or individual may use false statistical data to defraud the honorary title, material interest, or rank of job and rank promotion," says a revision in translation from the NBS Chinese-language website.
It is unclear why the agency unveiled its proposed reform 10 days before it announced that gross domestic product growth in the third quarter fell to 6 percent, its slowest pace since 1992.
On Sunday, the NBS announced that it will send 11 inspection teams to check data from nine provinces and autonomous regions, as well as two departments of the cabinet-level State Council in the next month, the official Xinhua news agency said.
In theory at least, the amendment would raise the stakes for officials who approve or transmit fabricated reports.
"The auditing and signing personnel of statistical data shall be responsible for the authenticity, accuracy, completeness and timeliness of the statistical data reviewed and signed," it says.
The draft also gives any member of the public the right to report falsification, offering unspecified "recognition and reward."
Stiffer fines are threatened for falsification, although these may depend on enforcement.
An early version of the amendment raised the maximum fine for concealing or destroying original records from 200,000 yuan (U.S. $28,300) to 500,000 yuan (U.S. $70,700), but a later version apparently revised the penalty to "more than 200,000 yuan."
The reworking of the statistics law follows years of international skepticism over the stability of China's official GDP growth figures through good times and bad.
Critics have cited the remarkable "smoothness" of the quarterly GDP growth numbers over the years.
A report in August by the New York-based Rhodium Group noted that growth rates had only varied by 0.8 percent percentage points over the previous 16 quarterly periods, despite wider fluctuations in industrial output and other indicators.
The amendment's focus on grassroots fabrications and reporting problems does little to explain how smoothing of national numbers takes place. But at least one provision appears to touch on the practice of manipulating the numbers to conform with past reports, also known as "convergence."
"The main statistical indicators cannot be obtained through administrative records or processing of existing statistical survey data," the draft said.
The proposed revisions are likely to cast doubt on the accuracy of the most recent GDP growth figures, since the provisions have yet to take effect.
The official decline in third-quarter growth after 6.2 percent in the second quarter may be only a partial reflection of China's structural slowdown and trade war tensions with the United States.
A strict application of the revised law could present a challenge to the economic success story that has been told over the years by the government and the Communist Party of China (CPC).
A draft paper submitted to a Brookings Institution conference in March by economists at the Chinese University of Hong Kong and University of Chicago found that official GDP growth rates in 2008-2016 were overstated by 1.7 percentage points.
But total elimination of data inflation could produce an abrupt drop in growth results, damaging China's image of economic stability. A downward revision of previous quarters could make the decline seem more gradual, but the effect on confidence in the government remains to be seen.
In its latest forecast, the International Monetary Fund projected moderately larger declines in China's economic growth from 6.6 percent on 2018 to 6.1 percent this year and 5.8 percent in 2020. The rates were down from 6.2 percent this year and 6.0 percent next year from the IMF forecast in July.
Whether the NBS and the government plan to use the amendment to make a significant adjustment to economic reporting remains an open question. But the extensive details of the draft revisions suggests that the NBS has been engaged with the issue for a significant amount of time.
Derek Scissors, an Asia economist and resident scholar at the American Enterprise Institute, voiced doubt that the process will lead to greater transparency in NBS accounting.
Scissors noted repeated references in the amendment to the confidentiality of "state secrets."
"If you leak something embarrassing about the financing of a single centrally-controlled state-owned enterprise, you can be found to have violated state secrets. So, it's not credible that transparency concerning macroeconomic aggregates is actually a goal," he said.
Scissors concludes that the NBS overhaul is either an exercise in public relations or a legitimate attempt to improve data quality, but only for unpublished accounting to be used by the government.
The amendment has been posted for public comment until Nov. 9, the NBS said.