Nova workboard

a blog from young economists at Nova SBE

The Balassa-Samuelson effect for China

The Balassa-Samuelson effect was first theorized independently by Bela Balassa and Paul Samuelson in 1964. This theory explains that the real exchange rate appreciates due to an increase in the productivity of the tradable goods sector. The mechanism works as follows: since the price of tradable goods must be equal among countries (Law of One Price), when there is an increase in the productivity in the tradables sector, this will result in an increase in the wage rate of this sector (from the profit maximization condition). Assuming (perfect) mobility between the two sectors, this implies that the wage rate must be equal across sectors, so the wage level will also increase in the non-tradables sector. As a consequence, this will lead to an increase in the price of these goods. Hence, there will be an appreciation of the real exchange rate.

Thus, in this case, the appreciation of the real exchange rate reflects an increase in the productivity. A result from this theory is that inflation is naturally part of the growing process of a country that experiences large productivity gains. What does this imply about China?


China is recognized as a fast-growing economy, so we should observe a real exchange rate appreciation of the Renminbi. However, this phenomenon is not observed, reason why China is accused of artificially undervalue its currency. Therefore, an article of “The Economist” claims that China should “welcome” inflation, because its price level is actually very low for its high economic growth when compared to other fast-growing economies, as the graph shows:


  (Source: The Economist)

However, empirical evidence of the Balassa-Samuelson effect on China is ambiguous. Indeed, according to some studies surveyed by Dunaway and Li (2005), the real exchange rate of the Renminbi has not appreciated in line with what Balassa-Samuelson effect predicts. The undervaluation estimated by these studies varies between 5% and 49%, which is quite a large interval. Indeed, the author points to the fact that the effect may be overstating the undervaluation. First, the assumption of full employment or a very high level of employment is not verified in China. This is a necessary assumption for the theory to hold since it ensures that wages increase when productivity in the tradables sector rises. Second, due to estimation specifications, changes in productivity must be strongly linked with the changes in the CPI/ PPI ratio for the theory to hold. The authors claim that this link might be weak for the Chinese case because several prices considered in the CPI are under the control of the government or are mismeasured. Thus, the link between wages and prices in China does not work as predicted by the Balassa-Samuelson effect.


Therefore, in fact, there is no empirical evidence of a strong Balassa-Samuelson effect for China. Nevertheless, it still seems that Renminbi is undervalued as Paul Krugman puts it:

“(…) you can disregard people who offer calculations suggesting that by some criterion – say, Balassa-Samuelson adjusted purchasing power parity – the renminbi isn’t undervalued. We know that the renminbi is grossly undervalued, not through questionable estimates that can be endlessly debated, but on a PPE (proof of the pudding is in the eating) basis: the current value of the renminbi is consistent with massive artificial capital export, and that’s that.”

Sofia Amaral, #538


Miguel Lebre de Freitas, Notes “Políticas Macroeconómicas”

Dunaway, Steven and Xiangming Li (2005), “Estimating China’s “Equilibrium” Real Exchange Rate”, IMF Working Paper.



Author: studentnovasbe

Master student in Nova Sbe

Comments are closed.