Another slew of Chinese data, another load of raised eyebrows. Evidence of accelerating retail sales or fixed-asset investment might be taken at face value elsewhere, but these are two of China's least reliable indicators. Seasoned China-watchers accept that not everything adds up in the data-gathering business. The total of first-half gross domestic product figures released individually by China's 31 provinces and municipalities, for example, was almost 10 per cent higher than the national figure put out by the National Bureau of Statistics. But some inputs to those sums are flakier than others.

A 33.1 per cent year-on-year rise in urban fixed-asset investment, as reported by the NBS on Wednesday, does not necessarily mean that investment rose by 33.1 per cent. As John Makin of the American Enterprise Institute think-tank points out, “investment” by central-government planners is counted on the basis of funds disbursed for projects, regardless of whether those funds have actually been spent. Similarly, retail sales figures – up 16.2 per cent – are not a useful proxy for household consumption as they basically track shipments to retailers, lumping in all kinds of government and corporate spending.

China's growth is no illusion. Data on industrial value-added, which correlates increasingly closely to energy consumption, shows an unequivocal rise in output. And as long as the data-collecting process is consistent, it is the trend that matters. But the problem with counting money transfers as investment, or mixing up institutional spending with consumer, is that there may be big lags in the impact of stimulus efforts. That is as good an explanation as any for China's continuing reluctance to tighten monetary conditions explicitly, amid mounting evidence of asset inflation: urban property prices were up almost 4 per cent in October, the biggest monthly increase in 14 months. Whatever the data may say, recovery cannot yet be taken for granted.