China's relatively mild slowdown in the second quarter has reignited a controversy about whether its official statistics can be trusted.
Chinese growth edged down to 7.6 per cent in the second quarter from 8.1 per cent in the first quarter, and analysts said the momentum in June, from stronger bank lending to rising investment, pointed to a rebound in the second half of the year.
But rather than delivering reassurance, the numbers instead provoked questions about whether the reality is worse than the government is letting on.
Economists with Barclays noted that a deceleration in industrial production was consistent with 7.0-7.3 per cent growth. Analysts at Capital Economics said that the true figure was probably closer to 7.0 per cent.
Sheng Laiyun, spokesman of the national bureau of statistics, took aim at critics in a news conference. "I want to tell everyone that they are wrong," he said. "Our data are independently verified and it all matches up."
Doubts about Chinese data have a fine pedigree. Li Keqiang, who is widely expected to succeed Wen Jiabao later this year as premier, confided to U.S. officials in 2007 that gross domestic product was "man made" and "for reference only", according to a diplomatic cable published by WikiLeaks.
Mr Li said it was much more useful to focus on three alternative indicators: electricity consumption, rail cargo volumes and bank lending.
That advice has come back to haunt the government because electricity production has been unusually weak this year. Electricity output growth was flat in June, whereas industrial production – which accounts for about 40 per cent of GDP and usually correlates quite well to electricity – was up 9.5 per cent.
"Power production shouldn't be at zero growth when heavy industry is growing near double digits. That doesn't make sense," said Ken Peng, an economist with BNP Paribas.
However, he was quick to add that strength in a series of other indicators that are not related to electricity production, especially bank lending and retail sales, were pointing in the right direction for growth.
"If you look at the whole picture it does make sense. Policy does seem to be having some effect," Mr. Peng said.
To arrest the slowdown, the Chinese central bank has cut interest rates twice in the past six weeks and officials have also accelerated approvals of investment projects, which are expected to provide a kick for the economy in the second half.
As for the mystery of the large divergence between electricity production and industrial output, Mr. Sheng of the statistics bureau said there was a simple reason.
"Everywhere in China is extremely focused on cutting energy use and cutting emissions, and focused on technological innovation," he said. "So there was great progress in the first half of the year in cutting energy consumption. The amount of energy consumed per unit of GDP decreased."
But his explanation failed to put the controversy to rest. Analysts said it stretched credulity that Chinese industrial groups could have effectively eliminated the need for additional electricity while continuing to grow at such a rapid clip.
Cao Heping, an economics professor at Beijing University, said the problem was one of methodology, not outright deceit. "Chinese statistics might be inaccurate, but have never been manipulated," he said.
Mark Williams of Capital Economics said China's statisticians would do well to follow more transparent processes that can be cross-checked by outsiders.
"The doubts that swirl around Chinese data are an added source of uncertainty for global investors and for anyone who cares about the state of the world economy," he said.
Additional reporting by Emma Dong