برابری قدرت خرید بیش از دو قرن؟

This paper re-examines the purchasing power parity hypothesis for the dollar–sterling exchange rate using the two centuries of data from Lothian and Taylor (LT) (1996) [Real exchange rate behavior: the recent float from the perspective of the past two centuries. Journal of Political Economy 104 (3), 488–509]. Unlike LT, we conclude that the dollar–sterling RER is nonstationary, implying a rejection of the long-run PPP hypothesis. The differences in our conclusions are explained by: (1) sensitivity of ADF unit root tests to the choice of lag length, and/or (2) the presence of significant time trends in the ADF or Phillips–Perron unit root test equations.

In the last 10–15 years, a large literature has emerged on testing the long-run validity of purchasing power parity (PPP), or equivalently the stationarity of the real exchange rate (RER), using modern time-series econometrics techniques. (See Rogoff, 1996, for recent references.) Lothian and Taylor (LT) (1996) emphasize that low power in standard unit root tests, especially with short data spans, may have caused researchers to incorrectly conclude that the RER is nonstationary. They present new unit root test results for the franc–sterling and dollar–sterling RERs using annual time series spanning two centuries. With the increased test power obtained by this large data sample, they are able to reject the unit root hypothesis using both augmented Dickey–Fuller (ADF) and Phillips–Perron (PP) tests. They therefore conclude that PPP is valid in the long run for the two bilateral exchange rates considered. Although we agree with LT that the franc–sterling exchange rate is stationary, our re-examination of the dollar–sterling RER concludes that it is not stationary. The differences in our conclusions are explained by: (1) sensitivity of the ADF unit root tests to the choice of lag length, and/or (2) the presence of significant time trends in the ADF or PP unit root test equations. We argue that, with suitably long lag lengths in the ADF equations, the unit root hypothesis is not rejected. If the lag length is shortened to that considered by LT, the ADF equations have significant time trends. The time trend is also significant in the PP test equations. Either unit roots or deterministic time trends, of course, imply nonstationarity, and hence rejection of the PPP hypothesis.