A century of Purchasing Power Parity: evidence from Canada and Australia

This study empirically examines the Purchasing Power Parity hypothesis using more than a century span of annual data of Australia, Canada and Britain and a battery of unit root tests. The study finds support for the validity of the Purchasing Power Parity hypothesis in the long-run within the framework of both linear and non-linear cointegration tests. The error correction models indicate that it takes four to five years for the short-run deviations from PPP to revert back to the long-run equilibrium. The results also indicate a non-linear mean reversion behaviour in the case of Canada. Overall, the evidence of support for the PPP hypothesis is robust across specifications and testing procedures.