A meta-analysis of international tourism demand elasticities and forecasting accuracy

Abstract

Accurate analysis of tourism demand and forecasting is crucially important if tourism businesses are to develop effective marketing strategies and governments to formulate effective national/regional tourism policies. A number of methodologies, including qualitative, quantitative, and combined approaches, have been used to achieve this. Based on temporal structure and complexity, the quantitative methods can be further divided into the basic time series models, advanced time series models, static econometric models, dynamic econometric models and artificial intelligence models. However, no single method has been proved to outperform the others in all situations. Differences in data characteristics and the features of each study are possible reasons for this variation in performance. This study uses meta-analysis to examine the relationships between each of international tourism demand elasticities and the accuracy of different forecasting models, and the data characteristics and study features which may affect these outcomes. By reviewing 262 studies published during the period 1961-2011, the meta-regression analysis shows that origin, destination, time period, modelling method, data frequency, variables and their measures, and sample size all significantly influence the estimates of the demand elasticities produced by a model, and its forecasting accuracy. The interaction effects between variables are also discussed and examined. This study is the first attempt to pair forecasting models with the data characteristics and study features. The results provide suggestions for the choice of appropriate forecasting methods in different situations. Moreover, the demand elasticities at both product and destination levels are generalised by statistically integrating previous empirical estimates. This will be useful in developing effective marketing strategies across different tourism markets.