What would China look like if it were growing at just 2 per cent a year? That sounds like a ridiculously pessimistic question given China’s performance in the past three decades. Certainly, it has manifold problems. Indeed, its economy is already slowing. But what misadventure could possibly bring its growth rate crashing down so spectacularly?

That is the wrong question, according to an influential paper by US economists Lant Pritchett and Lawrence Summers . For them, “the single most robust and striking fact” about growth is “regression to the mean” of about 2 per cent. Only rarely in modern history, they say, have countries grown at “super-rapid” rates above 6 per cent for much more than a decade. China has managed to buck the trend since 1977 by harnessing market forces, engineering possibly the longest spell “in the history of mankind”. But what goes up, the authors tell us, must eventually come down.

They have trawled through the data and drawn two powerful conclusions. One is that there is almost no statistical basis for predicting growth from one decade to another. Extrapolation is a mug’s game – or, as they put it, “current growth has very little predictive power”. From 1967 to 1980, Brazil grew at an average annual rate of 5.2 per cent. Few would have predicted, then, that for the next 22 years per capita income would grow at precisely zero.