Understanding commuting patterns and changes: Counterfactual analysis in a planning support framework

In order to contain commuting distance growth and relieve traffic burden in mega-city regions, it is essential to understand journey-to-work patterns and changes in those patterns. This research develops a planning support model that integrates increasingly available mobile phone data and conventional statistics into a theoretical urban economic framework to reveal and explain commuting changes. Base-year calibration and cross-year validation were conducted first to test the model's predictive ability. Counterfactual simulations were then applied to help local planners and policymakers understand which factors lead to differences in commuting patterns and how different policies influence various categorical zones (i.e. centre, near suburbs, sub-centres and far suburbs). The case study of Shanghai shows that jobs-housing co-location results in shorter commutes and that policymakers should be more cautious when determining workplace locations as they play a more significant role in mitigating excessive commutes and redistributing travel demand. Furthermore, land use and transport developments should be coordinated across spatial scales to achieve mutually beneficial outcomes for both the city centre and the suburbs. Coupled with empirical evidence explaining commuting changes over time, the proposed model can deliver timely and situation-cogent messages regarding the success or failure of planned policy initiatives.