Structural Change and the Dynamics of Real Exchange Rate

In this paper, we first examine empirically the Balassa-Samuelson effect in the presence of structural change. Using cross-country data on per capita income, price level and employment shares by sector, we find that a country’s tradable sector’s share of employment is a more significant predictor of the country’s price level than its per capita income, suggesting that the widely used empirical Balassa-Samuelson model is often mis-specified and biased in the face of structural change. We then re-examine the theoretical relationship between price level and per capita income in a model with structural change and show that the positive relationship between price level and per capita income predicted by the standard model is not a robust result in a multi-sector world. In contrast, the relationship between a country’s price level and the tradable sector’s share of employment is much more robust. Finally, we extend our theoretical analysis into a dynamic trade model with unbalanced trade to quantitatively examine how structural changes and policy distortions may have influenced the dynamics of the China-US real exchange rate. We show that much of the slow increase in the value of the Chinese currency over the last decade can be accounted for by the significant expansion of the tradable sector in China.