Estimation of demand models for long-distance cross-border travel

Abstract

Although long-distance cross-border travel contributes significantly to global emissions from the transport sector, transport models for this type of travel are scarce. In this study, a disaggregated travel demand forecasting model is estimated using Swedish national travel survey data 20112016 along with detailed supply data from European road, train, and ferry networks and a Worldwide air network, aiming at forecasting Swede’s long-distance travel abroad. Mode choice, destination choice and trip generation are modelled by traditional Nested Logit models and Multinomial Logit models. Results show that values of time of long-distance cross-border travel derived from the model estimation are in general higher than values of time of long-distance domestic travel. Furthermore, elasticity estimates of level-of-service attributes for train suggest that infrastructure investments in high-speed rail network may have a profound effect on demand for longdistance cross-border travel, especially for business trips.