AT HONG KONG airport, a couple hold an animated discussion about whether to buy a $350 polo shirt from Hugo Boss. Their conversation, like many in luxury shops across the city, is in Mandarin, the language of mainland China, and not the local variant, Cantonese. The mainland provides a third of the airport’s visitors and many of its most avid shoppers.
.



In 2012, according to estimates by Jonathan Garner and Helen Qiao of Morgan Stanley, a bank, the Chinese spent over 2.3 trillion yuan ($370 billion) on domestic tourism alone. And yet China’s GDP statistics captured only a tiny part of that spending, they argue, as well as missing spending on financial services, health care and housing. As a result, official figures show private consumption languishing at around 35% of GDP.


Morgan Stanley’s “bottom-up” calculations, by contrast, imply that it has grown since 2008 to almost 46% of GDP (see chart). Mr Garner and Ms Qiao draw on company reports and industry studies to fill gaps in the official data, which, they say, undercounted consumption by $1.6 trillion in 2012, more tan Australia’s entire GDP. Their calculations echo earlier studies, which also found that official statistics undercount consumption, albeit by a smaller margin.


As well as stuff bought offshore, spending online is also undercounted, the two economists argue. On a single weekend in November, Chinese consumers spent more than $3 billion on two websites, Taobao and Tmall (both part of Alibaba, an online giant), in celebration of “singles’ day”, the bachelor’s equivalent of Valentine’s day. But official statistics have failed to keep pace with changing consumer habits, Ms Qiao argues, neglecting entire categories of e-spending.


Online gaming, for example, is largely missing. Yet it amounted to 53 billion yuan ($8.5 billion) last year, according to Morgan Stanley’s tally of revenues earned by online gaming firms.


China’s statistics have long been viewed with scepticism or worse. Some economists worry that they fail to reflect reality, others that they slavishly reflect political imperatives. In 2002 Thomas Rawski of the University of Pittsburgh complained about a “tornado of deception”. Five years later Carsten Holz, then of Princeton University, said that official statistics should be taken with a “rock of salt”. When Li Keqiang, now China’s prime minister, was party chief of Liaoning province in 2007, he called the province’s output figures “man-made” and “for reference only”.


But things are not as bad as they were. China’s National Bureau of Statistics (NBS), for example, long ago stopped relying on provincial output figures to calculate national GDP. China’s economic census in 2004 gave the national statisticians a better baseline for subsequent work. In 2006 a book published by the OECD argued emphatically that China’s national accounts are inevitably “wrong”, in that they are forced to plump for one of a range of plausible figures, but that they are not politically manipulated.


But the NBS does not make it easy for independent outsiders to cross-check their work. Sceptics instead look for inconsistencies between China’s growth figures and other indicators, such as power generation or cement output. To track Liaoning’s economy, Mr Li looked at rail cargo, bank lending and electricity consumption.


Inspired by his example, three economists at the Federal Reserve Bank of San Francisco have distilled an alternative national growth index from three similar items. (It is a more sophisticated version of our own “Keqiang ker-ching” index published in December 2010.) They discovered that, contrary to popular belief, China’s growth figures are in the “same ballpark” as Mr Li’s indicators.


Not every statistical distortion serves to flatter China. Indeed, some of the biggest remaining flaws in China’s statistics are politically awkward. The official figures may, for example, exaggerate the politically sensitive income gap between urbanites and rural folk by as much as 40%, according to Jinjun Xue of Nagoya University and Wenshu Gao of the Chinese Academy of Social Sciences.


The understatement of consumption also gives ammunition to China’s critics, who worry that its economy relies too heavily on unsustainable investment and lament the government’s failure to rebalance the economy. Mr Garner and Ms Qiao’s alternative calculations imply instead that rebalancing is under way. In estimating consumption’s growing role, they assume that the hidden spending is not captured elsewhere in the GDP figures. If, in fact, this extra spending is misrecorded as something else, then consumption’s share of GDP would be even bigger.


Not everyone doubts the purchasing power of China’s consumers. At Hong Kong airport, adverts tempting customers to splurge compete with official notices designed to put them off one kind of purchase in particular. Since March 1st anyone taking more than two tins of milk powder out of the territory faces up to two years in prison, an embargo designed to retain Hong Kong powder for local mothers, amid concerns on the mainland about the safety of Chinese milk powder. China’s statisticians are not the only ones suppressing consumption.