Spatially staggered pre-peak discounts to ease subway crowding: A model-based policy study in Beijing

Abstract

High demand during peak hours leads to severe overcrowding in public transport systems, particularly in large metropolitan areas. While demand-side measures like fare discounts are widely adopted to shift travel times, existing uniform cut-off discount schemes often overlook spatial heterogeneity in demand patterns. This study proposes a spatially staggered pre-peak discount scheme to address this gap, integrating three sources of input: smartcard data, stated preference surveys, and agent-based simulations for policy evaluation. A case study is conducted based on Beijing Subway Line 5, a busy commuting line that connects large residential areas to the city center. Three discrete choice models are developed using the survey data to predict travelers’ departure time choices under pre-peak discount schemes. Twelve scenarios representing four types of pre-peak discount schemes (uniform, discontinuous, smooth and staged discount end times) are evaluated for their effectiveness in reducing peak crowding, potential side effects, and required subsidy inputs. While scenarios with intermediate discount cut-off times generally reduce peak congestion by 3%–5%, the spatially staged discount scheme is found to strike the optimal combined outcomes in terms of mitigating additional demand concentration before the cut-off time and involving moderate financial investments. The easy-to-remember scheme is thus recommended as an inspiration and potential refinement for developing future innovative fare incentives.