This paper explores the robustness of the Balassa-Samuelson (BS) hypothesis. We analyze a panel of OECD countries from 1970 to 2008 and compare three different datasets on sectoral productivity, including a newly constructed database on total factor productivity. Overall, our DOLS estimation results do not support the BS hypothesis. For the last two decades, we find a very robust negative relationship between the productivity in the tradable sector and the equilibrium real exchange rate, in contrast to BS. Earlier supportive findings depend strongly on the choice of the dataset. Except for the terms of trade, the explanatory power of other variables is weak.
Real Exchange Rate Balassa-Samuelson Hypothesis Panel Data Estimation Terms of Trade