High and low frequency variations and the cyclical behaviour of real wages

Abstract

The conflicting empirical evidence on the cyclicality of real wages may be a result of the fact that the measure of comovements employed in the literature is affected by coherence among non-business cycle variations in the data. By recursively removing components the degree to which long- and short-term cycles contaminate the existing evidence is studied. While confirming many of the results of the current literature, it is found that evidence of cyclicality is affected more by long rather than by short-term components, even though the data have been detrended by a number of different methods. It has been suggested in the literature that wages become more countercyclical as the deflators used represent less finished bundles of goods. This study shows that this result is robust to the removal of components from either end of the spectrum.