Complex and chaotic tourism systems: towards a quantitative approach

Purpose – Tourism systems have been considered more and more in the light of complexity and chaos theories. Most of the work done in this area has highlighted the reasons for and the issues regarding this approach. A steadily growing strand of the recent literature uses the theories to overcome the problems of a reductionist and mechanistic view that is considered unable to provide a full understanding of the structural and dynamic characteristics of tourism systems, and specifically of tourism destinations. This paper seeks to continue this approach and to provide a series of quantitative methods to assess the dynamics of non‐linear complex tourism systems. Design/methodology/approach – The time series used in the paper contains data collected from a sample of 23 large (four‐star) hotels located in Milan, Italy. For each structure daily data of occupancy, average room rate and RevPAR (revenue per available room) were recorded for the period 2006‐2009. The daily distributions of these observations are highly skewed, and therefore the median of the daily values were considered. This results in three series of 1,461 points per type (occupancy, room rate and RevPAR). Findings – The data confirm the complex nature of the destination system and its tendency towards a chaotic state. Additionally, high stability and long memory effects are detected. The outcomes and the implications of this analysis are examined. Research limitations/implications – A comparison of the values obtained leads to the conclusion that the series under study has a detectable level of non‐linearity, even if it does not reach the pure chaoticity of the Lorenz attractor. A first conclusion is that, as qualitatively assessed in many similar studies, the tourism destination is a complex system with a tendency to become chaotic. Originality/value – The picture obtained with the analyses conducted can be summarised by saying that the system under study exhibits an unequivocally complex nature. It tends towards a chaotic stage but does so at a slow pace. The stability of the system is quite high: it might be able to resist transient shocks well but, once led in one direction, its long memory characteristics tend to keep it on the resulting path.