FORECASTING INTERNATIONAL TOURIST FLOWS

Previous research suggests that the ''no change'' model is relatively accurate when annual data are used to forecast international tourism demand, whereas more complex models are relatively accurate when seasonal data are used to forecast domestic tourism demand. The difference in accuracy rankings could be due to the international versus domestic dimension of the data or the difference in data frequency. This study demonstrates that the results obtained for international tourism using annual data also hold for seasonal data, so the previously obtained difference in accuracy rankings appears to be attributable to the use of international as opposed to domestic tourism demand data.