Multimethod forecasts for tourism analysis

Abstract

This paper examines the use of combined forecasting methods in tourism forecasting. The use of multiple methods allows more information to be incorporated into the forecasts. The objective of the paper is to show that forecasts of tourism arrivals obtained by combining forecasting models are more accurate (therefore more useful to practicing managers) than single-method forecasts. Several combined forecasts of tourist arrivals in Florida obtained and compared to single-method forecasts gained through econometric or time-series methods. Parameter interpretation and forecasting accuracy of both combined and single-method forecasts are discussed. The conclusion of the paper offers a commentary on the usefulness of multimethod approaches to forecasting.