Spectral analysis of seasonality in tourism demand

This paper aims to analyse the seasonality in New Zealand tourism demand from Australia and the USA using spectral analysis. Tourism demand is divided into four different categories depending on the tourists' visiting purposes as registered in the customs cards upon their arrivals in New Zealand. Spectral analysis based on the sample from January 1980 to December 2007 revealed that different travel purposes share similar cyclical components but their contributions to the total variation in tourism demand differ between travel purposes and markets. More importantly, the results show that analysing aggregated data is often not sufficient to reveal the underlying seasonal patterns of tourist arrivals and policy makers would benefit greatly by analysing disaggregated data rather than relying on the analysis of aggregated data alone.