Spectral analysis of international tourism flows

This study applies the methods of univariate and bivariate spectral analysis to international tourism flows. Spectral analysis detects cycles within and between time series data sets. Its application presented in this article establishes relationships that are less readily found by more conventional time series methods. Specifically, leading cycles of dependencies are established between exchange rates and passenger flows. Not only does spectral analysis offer a fuller description of time series data, it can assist in the development of models that have historically been more commonly used in the study of tourism flows. Keywords: international tourism, time series, exchange rates, spectrum, squared coherence, phase. (C) 2000 Elsevier Science Ltd. All rights reserved.