Traditional analysis of tourism demand has been mainly based on the consideration of economic variables aiming at explaining the evolution of either tourists’ expenditure or overnight stays or arrivals. This study is based on a collection of both economic variables and non-economic factors surveyed since 1998–2013 for 99 Italian NUTS3 (Nomenclature of Units for Territorial Statistics) regions (provinces). It is the first study in which such a wide array of non-economic and economic variables has been investigated for a panel data at this geographical level by using a simultaneous equations model. The analysis shows that all considered variables are significant for the evolution of tourism demand and that climate, tourism supply and entrepreneurial capabilities have the largest impacts.

Most tourism destinations are affected by seasonality. Seasonal demand causes various problems for local firms and administrations hampering the efficient use of available facilities and the development of local capabilities. This paper discusses, from an econometric point of view, another important issue stemming from strong seasonality: the effect on forecasting tourist flows. This aspect of seasonality is addressed through an analysis of tourist arrivals in the Province of Lecce, southern Italy.

In questo lavoro si studiano le presenze turistiche mensili registrate in provincia di Lecce nel periodo gennaio 1988-dicembre 2008. L’utilizzo dei dati mensili consente di valutare l’impatto della componente stagionale sulla capacità di prevedere il futuro andamento della variabile ‘presenze turistiche’. In particolare, sono stati utilizzati due distinti approcci che si differenziano a seconda che l’aspetto stagionale venga esplicitamente rappresentato nel modello stocastico oppure venga eliminato dalla serie originaria e, quindi, non modellato. Nonostante la capacità previsionale dei modelli basati su dati destagionalizzati dipenda dal tipo di procedura utilizzata (TRAMO-SEATS oppure X12-ARIMA), ciò che emerge è un peggioramento delle previsioni qualora si modellino dati che presentano una forte componente stagionale.

Seasonality is a phenomenon that affects the vast majority of tourist destinations. The negative aspects of seasonality have been widely discussed from economic, social and environmental points of view. On the contrary, the unreliability of tourism demand forecasts is rarely listed among the negative effects of seasonality. This is despite the importance of the quality of forecasts for the planning of economic activities. This article evaluates the impact of different patterns of seasonality on tourism demand forecasting in the light of different volume of tourism flows. With this aim in mind, the monthly tourist overnight stays in four European countries – namely Austria, Finland, Portugal and Netherlands – are analysed for the period January 1990–December 2014. Data show both one-peak and two-peak seasonality. Results highlight that the stronger seasonality is, the less reliable forecasts are.