Testing for Ppp: Should We Use Panel Methods?

Empirical Economics (2005) 30:7791 inside 10.1007/s00181-004-0222-8

Testing for PPP: Should we use panel methods?
Anindya Banerjee1 , Massimiliano Marcellino2 , Chiara Osbat3
1 discussion section of Economics, European University Institute, Via della Piazzuola 43, 50133 Firenze, Italy (e-mail: anindya.banerjee@iue.it) 2 Istituto di Economia Politica - Universita Bocconi, IGIER and CEPR Via Salasco 3, 20136 Milano, Italy 3 European Central Bank, Kaiserstrasse 29, 60311 capital of Kentucky am Main, Germany

First sport received: November 2001/Final version received : October 2003

Abstract. A common ?nding in the empirical literature on the validity of purchasing power parity (PPP) is that it holds when tested for in panel data, but not in univariate (i.e. country-speci?c) analysis. The plebeian explanation for this mismatch is that panel tests for whole root argon more powerful than their univariate counterparts. In this paper we suggest an alternative explanation. Existing panel methods assume that cross- unit of measurement cointegrating relationships, that would tie the units of the panel together, be not present.



Using simulations, we show that if this important underlying self-assertion of panel unit root tests is violated, the empirical size of the tests is easily higher than the nominal level, and the null hypothesis of a unit root is rejected too often even when it is true. more(prenominal) generally, this ?nding warns against the automatic use of panel methods for testing for unit roots in macroeconomic time series. Key words: PPP, unit root, panel cointegration, cross-unit dependence JEL classi?cations: C12, C13, C22, C23 1. Introduction The last few long time have seen a veritable explosion of papers on testing for purchasing power parity (PPP) using panels of macroeconomic data. A

We wish to thank an anonymous referee and Ronnie MacDonald for helpful comments. The ?nancial check of the European University Institute in funding this research is to a fault gratefully acknowledged. The views...If you want to get a full essay, come out it on our website: Orderessay



If you want to get a full essay, wisit our page: write my essay .