Author: Paul Hubbard, ANU
Reading the latest Chinese growth projections to 2050 brings to mind Karl Marx’s aphorism that history repeats itself first as tragedy, second as farce. One of the co-authors, a Yale economics professor, told the Financial Times the ‘main point of our findings is that, contrary to common misconceptions, productivity growth under Mao, particularly in the non-agricultural sector, was actually pretty good’. But this is no argument to return to Mao-era policies, which would be a tragedy for the Chinese economy. Moreover, using Mao-era productivity trends to project Chinese GDP up to 70 years later, leads to farcical conclusions.
This is a shame, because the first 55 pages of the National Bureau of Economic Research (NBER) working paper provide a forensic account of Chinese multi-sector productivity growth from 1953 to 2012. Only in the admittedly ‘speculative’ section are these historical trends extrapolated out to the 2050s. Based on reform-era (post-1978) productivity trends, they project China’s average growth until 2024 at 7.8 per cent, 5.2 per cent during 2024–36 and then 3.6 per cent during 2036–50. A reversion to Mao-era (beginning after the Great Leap Forward) trends would see 5 per cent until 2024, 4.6 per cent during 2024–36 and then 3.9 per cent.
There are three big problems with these projections.
First, the Mao-era economy may grow relatively faster after 2036, but it would also be one-third smaller than the reform-era scenario. The authors assume that China’s total factor productivity growth rates automatically converge upon US trend after reaching 70 per cent of US productivity levels. Because the reform-era economy grows faster, it hits this slow-down threshold earlier while the Mao-era scenario is still picking the proverbial ‘low hanging fruit’. Mao-era productivity growth might then seem pretty good, but reform-era growth is still better.
Second, despite the authors’ care with historical statistics, they completely neglect the introduction of the one-child policy at the beginning of the reform era. They assume labour force growth of 0.5 per cent annually, based on the ‘lower end’ of China’s population growth until 2012. But the United Nations’ World Population Prospects projects China’s workforce to contract by 0.4 per cent annually until 2025, and by around 1 per cent annually thereafter. Correcting this would knock a few percentage points off either scenario.
The third, and fatal, flaw to the projections is the underlying model. The authors assume that regardless of the policy regime chosen, Chinese workers eventually become as productive as American workers. The choice of historical scenario becomes a simple choice of which parameter to use to extrapolate future growth. Under these assumptions, institutional choices simply don’t matter. This is exactly the kind of naïve ‘Asiaphoria’ that Lant Pritchett and Larry Summers warned of in 2014 in their own NBER paper.
More careful projections, such as those by the Australian Treasury, estimate productivity potential using a rough measure of existing institutional quality. In this model, labour productivity only reaches US levels if the economy is underpinned by US-quality institutions. A country with less developed economic institutions can still experience some rapid growth, but only until it reaches its lower level of potential as determined by its institutions.
It was because China was still so desperately poor in the 1970s that the authors were able to find ‘pretty good’ productivity growth in the Mao era. This is because Mao-era institutions provided superior economic conditions than what had prevailed during the preceding civil war and Japanese invasion. But while per capita income — closely tied to labour productivity — in China grew before 1978, it was still less than double the level of the Ming dynasty some six centuries earlier.
After 1978, the experimental introduction of private property, open markets and the political acceptance of entrepreneurs gradually expanded China’s productive potential. Since reform and opening up began almost 40 years ago, Chinese per capita income has doubled roughly every eight years. It now surpasses the Soviet Union at its peak. And using the Global Competitiveness Index as a proxy for institutional quality, China still has good institutions relative to its current productivity levels. This leaves some room for more catch-up growth.
The Chinese edition of the Financial Times reported the growth projections under the headline: ‘In the long run, Mao-style policies could bring China higher growth’. This is a farcical policy implication. At best, a reversion to Mao-style policies might reduce income inequality, but only because Chinese would be more equally poor.
Paul Hubbard is a Sir Roland Wilson PhD Scholar at the Crawford School of Public Policy, The Australian National University, a visiting scholar at the National School of Development, Peking University. He is on leave from the Australian Treasury. These are his personal views and do not reflect those of the Treasury.