Software and results for: An agent-based model for simulating the impact of social norms on active commuting interventions

Interventions to increase active commuting have been recommended as a method to increase population physical activity, but evidence is mixed. Social norms related to travel behaviour may influence the uptake of active commuting interventions but are rarely considered in the design and evaluation of interventions. In this study we develop an agent-based model that incorporates social norms related to travel behaviour and simulate the relative impact of four active commuting interventions. A synthetic population of Waltham Forest, London, UK was generated using a microsimulation approach with data from the UK Census 2011 and UK HLS datasets. An agent-based model was created using this synthetic population which modelled how the actions of peers and neighbours, subculture, habit, weather, bicycle ownership, car ownership, environmental supportiveness, and congestion affect the decision to travel between four modes: walking, cycling, driving, and taking public transport. We examined four scenarios (and a control): car-free days; increasing environmental supportiveness for walking and cycling; decreasing environmental supportiveness for driving; increasing environmental supportiveness for walking and decreasing environmental supportiveness for driving. Compared to the control scenario, the car-free days scenario increased the odds of active commuting by 81.8% (OR 1.818; 95% CrI: [1.816 to 1.820]). Increasing environmental supportiveness for walking and cycling had no effect (OR: 1.000; 95% CrI: [1.000 to 1.001]), as did decreasing environmental supportiveness for driving (OR: 1.000; 95% CrI: [0.999 to 1.001]), and increasing environmental supportiveness for walking and cycling while decreasing environmental supportiveness for driving (OR: 1.000; 95% CrI: [1.000 to 1.001]). These results provide support for car-free days as an intervention to ‘nudge’ people into active commuting. However, why built environment interventions were less effective is unclear. This release contains the agent-based model used for these results. It also contains the raw results, the Stan analysis code, and the Stan chains from the results.This research was funded by KCL–LSHTM seed funding. RG is supported by a Medical Research Council Studentship [grant number: MR/N0136638/1]. SC is funded by Health Data Research UK (HDR-UK). HDR-UK is an initiative funded by the UK Research and Innovation, Department of Health and Social Care (England) and the devolved administrations, and leading medical research charities