Tourism demand for Italy and the business cycle Andrea Guizzardi, Mario Mazzocchi * Department of Statistics, University of Bologna, Via Belle Arti 41, I-40126 Bologna, Italy article info Article history: Received 1 June 2007 Accepted 28 March 2009 Keywords: Business cycle Forecast evaluation Structural time series model Tourism demand Seasonality abstract This study provides a strategy for modelling the effect of the business cycle on tourism demand under the rationale that tourism cycles are heavily influenced by lagged effects of the overall business cycle. Using quarterly data on overnight stays in Italian hotels, both domestic and inbound between 1985 and 2004, we adopt a structural time series approach to evaluate two alternative models, the first with a latent cycle component (LCC) and the second based on specific economic explanatory variables (XCV). The two models are compared in terms of explanatory power, best-fit, residual diagnostics and fore- casting ability. The results show similar performances. The policy implication is that the XCV model can be used for calibrating countercyclical interventions in tourism policy. Ó 2009 Elsevier Ltd. All rights reserved. 1. Introduction The relevance of irregular trends and cyclical patterns in tourism demand has been long recognized and research has been targeted at effective modelling and forecasting strategy. While major business cycle fluctuations strongly influence consumer demand for goods and services, such as in times of economic recession and boom, the response of tourism demand is not necessarily immediate and straightforward because of substitu- tion effects between types of destinations and lags between decision making and the actual holiday. The business cycle factor has been emphasized in some studies (see e.g. Wong, 1997), but has been implicitly accounted for in tourism demand models through explanatory variables subject to fluctuations, such as prices and disposable income. However, little attention has been devoted to modelling the long-term relationship and lag struc- ture between tourism demand and the business cycle (see e.g. Gouveia & Rodrigues, 2005). Explicit modelling of cyclical components can be effectively pursued through Harvey’s struc- tural time series approach (STS, see Harvey, 1989), which has progressively gained in popularity and exemplified by its appli- cation in Gonza ´lez and Moral (1995, 1996), Greenidge (2001), Turner and Witt (2001), Kim and Moosa (2005) and Vu and Turner (2006) among others. However, none of these papers has explicitly focused on the dynamic specification and under- standing of the cyclical component. Thus, the key question posed in this paper is whether and to what extent tourism cycles can be regarded as a direct consequence of business cycles. This would mean that cyclical movements in tourism demand may be simply explained by the delayed effect of the economic cycle. In order to address this question, we compare and assess two alternative routes to the specification of the cycle component within STS models, one based on the standard stochastic specification, and the other relating cyclical movements to the overall economic dynamics as proxied by economic explan- atory variables. If a relationship between the tourism cycle and the overall business cycle can be shown, then tourism policy could take advantage of the delay between the two cycles by adopting coun- tercyclical measures to soften the impact of adverse economic conditions. The application of this study explores the evolution and cyclical behaviour of tourism demand in Italy measured by nights spent in tourism accommodation structures. Italy is one of the world’s leading countries for tourism earnings and tourism is a key sector in the Italian economy (World Tourism and Travel Council, 2007). By separating domestic tourism from inbound tourism, we account for differences and delays in international economic cycles. Quarterly data cover the period from 1985 to 2004, which includes important cyclical events (such as the 1992–1993 recession), major events like the Football World Cup in 1990, the Roman Catholic Jubilee in 2000 and the introduction of the Euro in January 2002. The STS specification allows for a stochastic trend to capture structural changes and breaks, such as those potentially induced by the above events, and the sample is of sufficient duration to detect modifications in seasonal patterns. * Corresponding author. Tel.: þ39 051 2098225; fax: þ39 051 232153. E-mail addresses: andrea.guizzardi@unibo.it (A. Guizzardi), m.mazzocchi@ unibo.it (M. Mazzocchi). Contents lists available at ScienceDirect Tourism Management journal homepage: www.elsevier.com/locate/tourman 0261-5177/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.tourman.2009.03.017 Tourism Management 31 (2010) 367–377