Accurate tourism demand forecasting has an important role to play in underpinning government investment decisions concerning the development of national tourism infrastructure for future economic benefit. In this paper a contingency approach to tourism forecasting model selection is taken and the appropriateness of a multiplicative seasonal autoregressive integrated moving average (SARIMA) to modelling Greece's inbound tourism in the medium/long-term to the year 2005 is investigated. Furthermore, the findings of this paper indicate that there are circumstances where the conventional restriction and use of ARIMA for short-term forecasting only can be relaxed. Predicated upon the impact of previous Olympic Games on host nations' tourism demand, baseline shift estimates were incorporated to the SARIMA baseline time series model using simple superposition. Predicting the most likely level of inbound tourism demand in Greece, as a conse- quence of the forthcoming 2004 Olympic Games, inevitably has major implications for government policy, tourism infrastructure investment and marketing decisions at various levels of aggregation. Within this context, two candidate SARIMA models are evaluated for their forecasting ability using conventional diagnostic statistics for goodness of fit together with comparisons made with alternative establishment forecasts. Whilst the limitations of univariate time series and SARIMA non casual models are acknowledged, within a contingency approach, the strategic implications of SARIMA to tourism forecasting and tourisin marketing in data poor modelling contexts are nevertheless beneficial and insightful in the absence of more rigorous alternatives. Moreover. it systemic perspective is proposed where tourism forecasting and tourism marketing are considered to he integral components of a second order. homoscedastic. feedforward cybernetic model, i.e.. a recursive cyherfilter approach (RCA).
© 2001-2024 Fundación Dialnet · Todos los derechos reservados