Using big data to identify if an off-peak fares policy within a multi-modal system can influence travellers’ boarding time preferences and improve modal integration: The case of South-East Queensland, Australia

Peak hour congestion, and peak loading are pressing issues for public transport agencies around the world. In multi-modal public transport systems, the fare system can help create an efficient system with good modal integration. An off-peak fare discount of 20% discount for travelling outside the defined peak period has been introduced in South-East Queensland but this has not been successful in persuading travellers to travel outside the narrowly defined peak hour thus not providing congestion or crowding relief nor increasing customer satisfaction on these attributes. This paper investigates the time effects and passengers' boarding time preferences of the fare policy, including the off-peak discount, using big data in the form of smart card data from automated fare collection systems. Choice models are estimated with public transport service characteristics to measure passengers' time preference to provide greater understanding and a quantification of the potential to shift passengers out of the peak. Using the outcome of the models, the paper concludes with suggestions for more sophisticated off-peak policies to congestion and peak loading.