Cross-nested logit model for the joint choice of residential location, travel mode, and departure time

This paper aims to describe the joint choice of residential location, travel mode, and departure time. First, based on random utility maximization theory, the Cross-Nested Logit model and traditional NL models are formulated respectively. House price, travel time, travel cost, and factors depicting the individual socio-economic characteristics are defined as exogenous variables, and the model choice sets are the combination of residential location subset, departure time subset, and travel mode choice subset. Second, using Beijing traffic survey data of 2005, the model parameters are estimated, and the direct and cross elasticity are calculated to analyze the change of alternatives probability brought by factors variation. Estimation results show the Cross-Nested Logit model outperforms the three kinds of NL model. It is also found by estimation results that decision makers will change first their departure times, then their travel modes, and finally their residential locations, when exogenous variables alter. Moreover, elasticity analysis results suggest that, for long-distance commuting, it is difficult to decrease car travels even if additional charges are imposed on car users. The effect on choice probability by variations in travel time of other travel mode can be considered as negligible for alternatives within 5 km commuting distance, and this effect are greatest for alternatives between 10 and 20 km commuting distance. These findings have important implications for transport demand management and residence planning.

► We propose a Cross-Nested Logit model for the joint choice of residential location, travel mode and departure time.
► The Cross-Nested Logit model outperforms traditional NL models.
► There is higher substitution between alternative residential locations than between alternative travel behaviors.
► Higher substitution exists between alternative departure time periods than between alternative travel modes.