Travel plans for new residential developments: Measuring self-selection effects to better understand travel behavior impacts

Travel plans can be required for new residential developments as part of the land use planning and approvals process. The aim is to manage car use of residents by providing a package of tailored initiatives and facilities that support the use of more sustainable forms of transport. However, there is a limited understanding of the extent to which such packages have influenced travel behavior or simply attracted residents who were already making travel choices to reduce car use; the latter effect is commonly referred to as residential self-selection. In this study, a survey of residents in Melbourne, Victoria, Australia, was conducted at developments with and without travel plans. Propensity score matching (PSM)-based on a set of attitudinal, preference, and demographic variables-was then used to explore how residential self-selection could be analyzed in the context of travel plans. Results suggest that residents living in developments with travel plans have lower car ownership and higher levels of walking and cycling compared with residents living in developments without travel plans. PSM was deemed to be a suitable method for analyzing residential self-selection effects, yet a sufficient sample size is required to appropriately quantify these effects.