Additional evidence of long-run purchasing power parity with restricted structural change

We investigate two alternative versions of Purchasing Power Parity (PPP): reversion to a constant mean in the spirit of Cassel and reversion to a constant trend in the spirit of Balassa and Samuelson, using long-span real exchange rate data for industrialized countries. We develop unit root tests that both account for structural change and maintain a long-run mean or trend. With conventional tests, previous research finds evidence of some variant of PPP for 9 of the 16 countries. With the unit root tests in the presence of restricted structural change, we find evidence of PPP for five additional countries.