This paper studies the influence of aggregating across space when (i) testing the PPP theory or more generally pair-wise cointegration and (ii) evaluating the PPP puzzle. Our contribution is threefold: we show that aggregating foreign data and applying an ADF test may lead to erroneously reject the PPP hypothesis. We then show, on the basis of theoretical arguments as well as Monte Carlo experiments, that a sizable bias in the estimates of half-life deviations to PPP may be due to the effect of aggregation across space. We finally illustrate empirically the importance of spatial considerations when estimating the speed of price convergence among euro area countries.
aggregation across space cointegration Half-life estimates PPP price convergence