Abstract Balancing jobs and housing (JH) can induce sustainable travel behavior, reducing dependence on private cars for commuting. Many studies have explored the possibility of either the multi-scale of place or the multi-level of people for examining the impact of JH balancing on sustainable commuting behavior. However, limited studies have focused on both the impact of the combined probability of mode choice and time to commute in a private car. Therefore, this study empirically identified the impact of JH balancing in spatial multi-scales such as neighborhoods, districts, and metropolitan cities on the combined probability of mode choice and the time taken to commute in a private car. Overall, 23,956 workers from Seoul, South Korea, were examined by applying hierarchical Tobit models. The study also demonstrated the moderating impact of JH balancing on sustainable commuting behavior based on income, education, occupation, and number of workers per household. The results confirmed that although sustainable commuting behavior decreased as the spatial scale of JH balancing increased, the effect of spatial scale was not consistent. Specifically, its effect on sustainable commuting was the greatest at the neighborhood level, although there was no distinct difference at the district and metropolitan city levels. This study also confirmed differentiation in the multi-scale impacts of JH balancing by income, education, job, and number of workers per household. Finally, the study concludes that more sustainable commuting behavior can be induced through steady investments in public transit at the metropolitan city scale, along with a supply of affordable housing near employment centers and affordable jobs near residential areas at the neighborhood level, while considering the differences at various levels of demographic and socioeconomic attributes for workers.

Highlights • Examined multi-scale moderation impacts of jobs and housing balance on commuting behavior. • Applied hierarchical Tobit models to identify the impacts of neighborhood, district, and metropolitan city scales on it • Balancing jobs and housing by multi-scale boundaries could be more efficient for sustainable commuting • The smallest spatial scale of JH balancing, which is a neighborhood unit, was the most effective in promoting sustainable commuting behavior • Personal socio-economic status such as income and education attainment should be considered in its application