This paper addresses the role of markup variations in the transmission process of cross-sectoral productivity differential shocks and government spending shocks to the relative price of nontradables. The Balassa-Samuelson model based on frictionless goods markets predicts that a rise in the sectoral productivity ratio by 1% raises the relative price by 1% while government spending changes leave the relative price unaffected. Using panel cointegration and unit root tests applied to a panel of fifteen OECD economies, our empirical evidence does not support these implications. We find that a rise in relative productivity by 1% raises the relative price of nontradables by only 0.70% and that an increase in government spending by 1% of GDP drives up the relative price by around 1%. This paper shows that these items of evidence can be successfully explained by a two-sector open economy model in which variations in the composition of demand for nontradables give rise to endogenous changes in the markups.
