Two essays on the real exchange rate

This thesis consists of two essays that use time-series techniques and recent OECD data to examine the long-run behavior of the real exchange rate within the framework of Balassa-Samuelson model. Theses essays can be summarized as follows.;The second essay relaxes the assumption that PPP holds for traded goods and extends the Balassa-Samuelson model to allow specialization in traded goods. In this variant, the relative traded productivity can affect the real exchange rate through two channels, the Balassa-Samuelson channel which is positive and the terms of trade channel which is negative. We conclude that the introduction of specialization in the Balassa-Samuelson model (which leads to the breakdown of PPP for traded goods) can explain weak and sometime reversed relationship between the real exchange rate and the relative productivity differential.;Balassa-Samuelson model predicts that rich countries tend to have higher price levels than poor countries. The first essay tests this hypothesis with both cross-section and time series data and confirms the puzzle that although this hypothesis performs well in explaining cross-sectional data, it is relatively poor in tracking time series data (except for the case when Japan is the reference country). To explore this puzzle, this essay uses a simple model which determines both the real exchange rate and the income per capita differential in terms of traded and nontraded productivity differentials under the key assumption that Purchasing Power Parity (PPP) holds for traded goods. This model suggests that the persistence and variation of these productivity differential series are very important in explaining the long run relationship between the real exchange rate and the income per capita differential. This view is supported by test results from both simulation and real data.