Think of China as a continent, not as one national economy
While rich regions' growth rates may slow, poorer provinces still have plenty of catching up to do before they approach the middle income trap
We've heard a lot in recent months about how China is in danger of falling into the middle income trap.
But there's a problem with these warnings. They treat China as if it were a single country. But in economic terms, China looks more like an entire continent.
In case you missed it, the middle income trap is shorthand for the collapse in growth rates that afflicts some emerging economies once they reach a certain level of development.
The examples cited most often are in Latin America. In the 1960s and 1970s, economists predicted great things for fast-growing economies like Brazil and Argentina, forecasting they would reach developed status within a decade or two.
They never made it. In the 1980s their growth evaporated amid the Latin American debt crisis. Today, incomes in the big Latin American economies are still less than half those enjoyed by the developed world.
Now pessimists warn that something similar could happen in China. For the last 30 years China has been able to generate high growth rates simply by playing catch-up with the rest of the world.
The combination of urbanisation, capital investment in infrastructure and technology transfers from the developed world has driven huge increases in income.
But economists warn that China is now entering the danger zone. With per capita incomes approaching half rich country levels, and with fewer productivity gains to be made from buying in foreign technology, China's competitive edge is being eroded.
The likely result they say, will be a steep fall in economic growth rates over the next few years, as China's development shifts down to a lower gear.
This makes sense up to a point. As economies get richer, their growth rates do tend to slow. Just consider Hong Kong. Since the late 1980s, the city's gross domestic product per head has doubled in real inflation-adjusted terms. But over the same interval our average annual growth rate has fallen by half, from around 8 per cent to just 4 per cent.
But applying this sort of analysis to the mainland fails to account for the sheer size and diversity of China's economy.
It's true that per capita incomes in China's wealthiest provinces - Beijing, Shanghai and Tianjin - are already well above the US$17,000 a year level (at purchasing power parity) at which growth rates typically begin to slow down. Other provinces, including Jiangsu, Zhejiang and Guangdong, are not far behind.
But in most of China's provinces incomes are still less than half the middle income level. And in some - Gansu, Yunnan, Guizhou - they are less than a quarter of the US$17,000 danger mark.
Experience elsewhere in Asia indicates that national growth rates typically reach their highest level when per capita incomes are just below US$10,000, and then decline as wealth increases.
Given that China's provinces resemble national economies in a broader region, this implies that although growth rates in the richer provinces may well tail off over the coming years, the country's poorer regions can still grow rapidly while catching up.
According to researchers at the Federal Reserve Bank of San Francisco, this internal convergence may reduce the chance that China as a whole will fall into the middle-income trap.
They forecast that growth rates in China's richer coastal provinces are likely to slow to just 5.5 per cent by 2020.
However, they expect that growth in the rest of the country will remain strong at around 7.5 per cent, only falling below 7 per cent in the second half of the following decade.
This would reverse the pattern of recent years, in which richer provinces have consistently grown faster than their poorer neighbours.
But if the San Francisco Fed's researchers are right, the convergence of China's poorer regions with its rich coast should delay the onset of slower growth across the economy as a whole, perhaps helping China to avoid the middle income trap entirely.