…or almost everything, anyway, can be learned by reading Dwight Perkins’ paper in the latest Asian Development Review. The paper is very clearly written and logical, and has the benefit of being freely available online (it is published by the ADB), unlike most academic publications. Dwight is probably my favorite development economist and it is always worth hearing his thoughts about China. While it will contain few surprises for those who have been closely following developments in China over the last few years, I still find it useful to go back to basics periodically.

The paper’s case is simple: China is a country with a very high share of investment in GDP, which is reaching saturation for investment demand in some sectors. This means that both the returns on investment in those sectors is falling (affecting productivity) and the growth of total investment is slowing (affecting capital accumulation). Therefore, growth slows. One of the more interesting parts of the paper is how Dwight thinks through the difficult question of future demand for various types of investment, particularly in infrastructure.

The same issue of the journal also contains a paper by Justin Lin making the case that China’s growth does not have to slow down substantially. This looks like a more formal presentation of the case Justin has been making recently that China can sustain 8% growth, which I’ve discussed on this blog previously. It seems worthwhile to try to compare the two papers, though this involves wading through tables of growth-accounting assumptions. Interestingly, the difference between the two does not seem to be relative optimism or pessimism about “reform” and its effect on future productivity gains. Justin says 2.1% TFP growth over the next two decades is enough to sustain 8% GDP growth, while Dwight says 2.1% TFP growth will only produce 6% GDP growth.

This means that the key difference between a slower-growth and higher-growth scenario is the pace of future investment spending. Which of course makes perfect sense since investment spending has been the biggest driver of China’s growth so far. Justin’s case for a higher rate of investment growth in the future is not very clearly spelled out, but he does mention “continued large-scale infrastructure investment” and the need for new investment caused by changing technology. To be fair, it’s clear that neither author has, surprise surprise, solved the puzzle of how to forecast investment demand over the long term. But Justin’s paper, despite the nod to changing technology, seems to rely more on extrapolating the recent levels of certain indicators into the future, while Dwight does more thinking about how things might change in the future compared to today. It should be pretty clear which approach I think is better.