Exploring taste heterogeneity and substitution patterns in dockless bike-sharing parking preferences

Dockless bike-sharing (DBS) is an important sustainable urban transportation mode in many cities but faces challenges with disorderly parking management. This study aims to explore the presence of taste heterogeneity and substitution patterns in DBS users’ parking preferences and to determine how interpersonal variations, alternative-specific attributes, and socio-demographic characteristics affect parking choices. Based on stated-preference data collected in China, a mixed nested logit (Mixed NL) model is employed to account for both inter-alternative correlation and random taste heterogeneity. The results indicate that reducing the distance to parking and increasing monetary fines are more effective in discouraging disorderly parking than offering incentives for orderly parking or adjusting parking fees. Social influence also plays a critical role, as users are more likely to park disorderly when they observe others doing so. Meanwhile, the research also reveals that users are willing to pay an average of 0.8 CNY to reduce the distance to parking by 100 m, and are willing to accept on average an additional 58 m of the distance to parking in exchange for 10 min of free riding time. These insights into DBS users’ parking behaviour enhance the understanding of the effectiveness of possible policy interventions and offer a valuable reference for developing future management strategies.