Tests of Purchasing Power Parity via cointegration analysis of heterogeneous panels with consumer price indices

There has been significant interest in the empirical performance of the Purchasing Power Parity (PPP) hypothesis. Initial studies were, in general, unfavorable for PPP. These results led researchers into two directions. One branch of the literature employed price data for a limited range or goods (e.g., fruits or clothing). A second branch has reevaluated the performance of PPP with more powerful methods. In this paper we combine these two branches of the literature. We use consumer price sub-indices data and recently developed panel cointegration techniques to test weak PPP. Our results are suggestive that the failure of PPP can be attributed to inclusion of non-traded goods in the overall index.