Predicting owners’ willingness to share private residential parking spots

Abstract

Sharing of private residential parking spots is a new pattern of parking management in China. This pattern corresponds to the booming sharing economy and is growing very fast. It can significantly improve the utilization of parking resources and relieve parking supply pressure. Based on the real data of 1-year behavioral records of owners obtained from Ding Ding Parking (DParking), an application on smart phones, as well as various field survey data, the study analyzed the influential factors and predicted owners’ sharing willingness. Two Classification and Regression Trees (CART) were developed to answer questions pertaining to whether owners would share their parking spot and how long owners would share during peak periods of parking demand, respectively. The results showed good accuracy in both models and revealed that owners’ self-use behavior, along with owners’ private spots’ physical characteristics and rental effects of the previous month, all have significant influence on owners’ willingness to share. The influence of factors and their importance differ for the two models; thus, a detailed comparison is performed. The findings in this paper would be beneficial to the government’s parking supply policies, as well as to third parties, so as to enhance the effective distribution of parking resources.