This paper investigates the possibility that long-run relative purchasing power parity is
dependent upon the nature of real exchange shocks that are experienced. While existing
studies involving developed and less developed countries often find against purchasing power
parity having employed linear tests of non-stationarity or non-cointegration, we employ a
new cointegration test, recently advocated by Enders and Siklos and Enders and Dibooglu,
that tests for an asymmetric adjustment towards parity with respect to positive and negative
real exchange rate shocks. Using a sample of ten African economies with data taken from the
post-Bretton Woods floating exchange rate era, long-run purchasing power parity holds in
eight of these cases if an explicit distinction is made between positive and negative shocks.
Across the sample, we find variation in the type of asymmetry experienced and the roles
played price and nominal exchange rate adjustment.