Thank you very much, St Louis Federal Reserve. Thank you very much.
A recent paper from the St Louis Fed by Xin Wang and Yi Wen has made the empirical case for my qualitative description of China’s economic model. China’s government investment in infrastructure is achieving a fiscal multiplier of 2. This suggests that China’s Keynesian policy of strong investment in infrastructure is responsible for the great majority of the economy’s strong growth.
The paper finds the following:
1. The fiscal multiplier in China is indeed significantly larger than 1 so that real GDP can always rise by more than the increase in government purchases. In this scenario, the added government spending also stimulates private consumption and investment even if workers are just building roads to nowhere and homes for nobody.
2. However, such a large multiplier is not necessarily a free lunch, as government spending itself may also be an aggravating source of the boom-bust cycle in China. Consequently, the benefit of the multiplier may be largely offset by the cost of the subsequent boom-bust cycles, especially when government purchases are financed by credit expansion and money creation.
The conclusions can be split into two parts: first, the work on the multiplier that relies on empirical evidence and second, the assertions that there is no free lunch because the benefit comes from building roads to nowhere and homes for nobody. In particular, the paper does not make clear the costs of the boom-bust cycles, relative to the manifest benefits of a fiscal multiplier of 2.
Both these latter conclusions smack of political expediency rather than sound analysis. Seemingly, praising the Chinese model and Keynesianism is beyond the scope of the Federal Reserve. Here are my thoughts on the latter part of the analysis.
The fiscal multiplier as an indication of productivity
The idea that a fiscal multiplier of 2 is the result of consumption and investment from building useless infrastructure is incredibly flawed. On the consumption side, China has a low propensity to consume. This is true of both consumers and firms. It seems, therefore, unlikely that the multiplier is a function of increased spending from incomes. It is also unlikely that private sector investment would occur alongside poor investments, such as roads to nowhere and buildings nobody lives in.
Instead, the multiplier will be a function of the improvement in productivity created by the infrastructure China is building. This has been a constant theme at this blog. China is seeing dramatic evidence of productivity improvement in rising wages and stable-ish prices. Productivity is in the rising wages of internal areas and the ability to consume more. It’s lowering transport costs as a share of GDP, raising wages and improving living standards. All of this creates more sustainable levels of economic activity. The data suggest China remains highly unproductive relative to developed nations. Current estimates put China’s capital stock per capita at around 10-15% of the US.
Another way of looking at productivity and the success of the fiscal multiplier is to consider China’s tax take. The increased economic activity from infrastructure is creating growth in tax from externalities which is outstripping the rate of nominal growth, as the chart below highlights.
It’s certainly not clear that with a fiscal multiplier from infrastructure investment of 2 that China should stop investing. Indeed, I am comfortable arguing that China is investing optimally and has some way to go.
The cost of inflation?
The secondary argument against taking such a high multiplier seriously is the impact of inflation and tightening cycles on the economy. Inflation is shown in the study to be correlated to government spending which is fair enough. The government is driving the economy and so cycles will be dependent upon the changes in its spending. But is inflation high enough to outweigh the benefits of government action? I would argue no. In addition, the study’s focus on the entire period of opening up from 1980 is flawed in that it is the present that is of most value. We should distinguish between the period to 2000 and from 2000. it is from 2000 that growth has been strongest and inflation weakest.
Between 1980 and 2000, GDP growth in China averaged 9.8% and the CPI averaged 7%. Since 2000, GDP has averaged 10.4% and inflation has averaged 2.4%. The period since 2000 has been remarkable. Growth has been stronger and prices more stable: the cost the paper identifies has been a lot lower than implied, in the last ten years.
This is particularly the case when taking a comparative approach. The chart below shows growth, inflation and economy size for a number of economies from 1990 to 2010. China stands out. As importantly, the figures for Brazil and Russia are only for the period from 2000 because pre-2000 inflation would be greater than 50%. China’s achieving something no other economy has come close to achieving.
Conclusion
This research from the St Louis Fed is crucial to understanding China. China has created the incentives (political not economic) required to get the infrastructure a developing economy needs. Yes, it creates waste in the form of some roads to nowhere and cities that noone lives in but from a national portfolio perspective (where diversification allows for some failure) it is achieving a significant positive return. The criticism of the model contained within the paper is simply superficial and suggests political motives. The bears have long needed to demonstrate that productivity growth is inconsequential in China. This would seem to make that case near impossible.
