A note on seasonal breaks in tourism demand time series

Abstract

In this paper, a structural break test procedure proposed in literature is employed in a tourism demand analysis context. This article is the first to apply a graphical approach using seasonal surface and seasonal contour maps, which can be important tools for seasonal pattern evolution analyses. For the purpose of empirical study, total lodging figures were selected to analyze British tourism demand in the Algarve. The seasonality of tourism demand has several effects on economic activity. Identifying seasonal breaks in tourism time series allowed us to obtain evidence of efforts to lessen seasonality and that can hopefully benefit long-run policy evaluation and planning.

The seasonality structure of demand in the portuguese hotel industry before the COVID-19 pandemic

“…Tourism seasonality in Portugal has not been thoroughly studied in terms of space, distribution of clients, quality of service offered, revenue, although some authors have studied the seasonality in Portugal or Mediterranean destinations, such as Duro (2016), Mitra (2020), Vrkljan et al (2019), Neves et al (2008), Gouveia (2014) andVergori (2017). Some of these studies have some limitations as far as the implications of the scope or method used are concerned (Gouveia, 2014;Vergori, 2017). While Gouveia (2014) studied only British tourism demand in the Algarve, this author applied a graphical approach and structural break test procedure and identified two seasonal breaks, one in 1977 and another in 1991.…”

“…Some of these studies have some limitations as far as the implications of the scope or method used are concerned (Gouveia, 2014;Vergori, 2017). While Gouveia (2014) studied only British tourism demand in the Algarve, this author applied a graphical approach and structural break test procedure and identified two seasonal breaks, one in 1977 and another in 1991. Vergori (2017) studied the forecasting process of seasonality in four countries, including Portugal where one seasonality peak was detected, but the accuracy of the SARIMA forecasts revealed to be less accurate than for other countries, such as Austria and Finland, even using data from 1990 to 2014.…”