by Irem Önder – Univesity of Massachusetts Amherst
Overall, tourism demand forecasting is difficult due to the characteristics of the tourism industry and the unpredictability of human behavior. At the same time, it is essential for the industry for planning and scheduling as well as for the proper allocation of resources. Time series analysis is a common approach for predicting tourism demand. It leverages time series data for the purpose of identifying the behaviors, statistics, and other meaningful features of said data. As such, the purpose of this chapter is to demonstrate how to conduct time series analysis for tourism demand. The methods applied include seasonal naïve (no-change model), single exponential smoothing (SES), error trend seasonal (ETS), and combined forecasts. Berlin, the capital of Germany, was chosen as the tourist destination used to showcase these methods in the how-to section. Arrivals to Berlin from Germany, Italy, Spain, and the United Kingdom (UK) between 2005-2019 as well as overall arrivals will be used in this tutorial. Results will be evaluated based on the mean absolute percentage error (MAPE) and root mean square error (RMSE) measures, and EViews software will be used to conduct the analyses.