Across the dozens of economic indicators released by China’s National Bureau of Statistics this week, few demonstrated the difficulty of pinning down the state of the world’s second-biggest economy better than the steel data.
Just months ago, steel output — previously subject to an informal cap as Beijing sought to curb emissions and production — was on course to expand significantly in 2023 for the first time in two years.
But — in line with an official desire to reduce output — December production fell 15 per cent year-on-year to its weakest level since 2017, a rate of decline that meant total annual output edged up, but remained essentially flat at just over 1bn tons.
“It is fair to say that we don’t believe these numbers,” said Colin Hamilton, a London-based analyst at BMO Capital Markets, of China’s December steel and pig iron data. Even before their release, Hamilton had said “data quality issues” were “raising their head again in China” and pointed to potential “strategic under-reporting to meet official targets”.
He added that the fall “did not tie in with anything else” and that coke production, which goes into the steel industry, was only marginally down year on year in December. “We have seen year-end discrepancies before, but never to this extent”.
The steel figures, announced alongside a 5.2 per cent gross domestic product growth reading for 2023 that narrowly beat Beijing’s official target, are just one example of the intense global scrutiny that greets data releases from the world’s second-largest economy.
For years, economists have used alternative gauges — from electricity consumption to energy imports — to complement their understanding of China’s GDP data and check the picture painted by official reports.
Former premier Li Keqiang, who died last year, in 2007 reportedly admitted to a US official that, given the unreliability of some provincial data, he used alternative measures such as bank loans to assess economic activity.
Beijing’s severe tightening of control over information flows during and since the Covid-19 pandemic has deepened uncertainty over official data.
Despite headline expansion that far surpassed projected global growth of 3 per cent in 2023, Chinese policymakers are still grappling with a multiyear property slowdown, deflation and consumer caution.
“There are inconsistencies in the data set that was just released,” noted Louis Kuijs, head of Asia economics at S&P and a former World Bank economist in China. “I also worry if I see that the statistical authorities are not independent from the government, which is the case in China.”
In July, authorities stopped publishing youth unemployment data, which in June hit 21.3 per cent — its highest level since the metric was introduced in 2019 — officially because of methodological concerns. They reintroduced the data this week with a new methodology that put youth unemployment at 14.9 per cent for December.
Julian Evans-Pritchard, chief China economist at Capital Economics, said he was “inclined to take [the government’s] explanation at face value”.
There was something “odd” about the previous data because it did not appear to follow the business cycle and included full-time students, Evans-Pritchard said, but added that the government should have continued to publish the previous series as well.
Rebecca Nadin, a director at the Overseas Development Institute, a think-tank in London, said a particular difficulty now was “the challenge . . . of being able to talk to people in China and verify or validate some of the economic data”. She pointed to a national security focus that may impact economic indicators.
On the full-year GDP data, several economists highlighted China’s choices for the deflator, a broad measure of prices that is used to convert nominal growth to a real figure and relies on a high degree of statistical judgment. China’s nominal GDP was less than the 5.2 per cent growth in real terms, meaning falling prices boosted headline growth.
“[The] GDP deflator needs to deflate industrial activity, needs to deflate government services and there are a lot of assumptions that flow in so that might lead to some distortions,” said Fred Neumann, chief Asia economist at HSBC.
Kuijs said the deflator China used for industrial production appeared to track the producer price index, a gauge of factory gate prices that is heavily influenced by global commodity prices. This approach could be misleading, so when commodity prices move sharply, he adjusts Chinese GDP data accordingly.
Various investment banks and research houses supplement their approach with alternative gauges. TS Lombard publishes its own “real GDP index” and this week noted that full-year real GDP growth was “likely as low as 3.6 per cent”.
Evans-Pritchard, who in 2020 noted that China’s headline GDP was “eerily stable” compared with other major economies, uses an in-house “China activity proxy”.
“Activity . . . in Q3 in particular was a lot weaker than they were willing to acknowledge,” he said, referring to Capital Economics’ own findings.
Evans-Pritchard added that he believed nominal GDP data in general was accurate, but that assumptions used in the deflator allowed “a degree of flexibility” on the headline figure that authorities could use to “nudge things in the direction that they want to see”.
But despite the doubts, there are few prospects of any comprehensive alternative to China’s official figures. “All of these [alternatives] have had flaws in them,” said Neumann of HSBC, which does not produce an independent measure of GDP.
Despite a “huge class of people” who would disagree, Kuijs thinks China’s national statistics accounts broadly describe the picture he would expect of the economy.
It would “be hard to come up with an indicator that would crowd out the GDP data from the National Bureau of Statistics,” he said. The NBS did not respond to a request for comment.
