Empirical analysis on long-distance peer-to-peer ridesharing service in Japan

Ridesharing has been attracting attention in Western countries in accordance with the rise of social and economic systems in which goods and services are shared between individuals. In accordance with the spread of ridesharing services, users' data, which concern how, when, where, and why they use these services, are being gradually collected around the world. However, there are only a few studies that deal with the empirical situation in Japan. In addition, ridesharing services in the country are not yet as popular as those in Western countries. Therefore, in this study, we aim to show the present situation of the ridesharing behavior in Japan. We conduct an empirical analysis by using actual long-distance peer-to-peer ridesharing data in the country. Firstly, we classify ridesharing drives into three classes by OD pairs: Inter-metropolitan, Low-density area, and Others. Next, we formulate a binomial probit model that explains the matching success for each drive class. The estimated model shows that the departure time and day, days to departure from registration date, page views, and driver's past experiences are important factors for successful matching. Moreover, the similarities and differences across the drive classes are discussed through the estimated model. The sensitivity of each drive attribute is explained, and ways of promoting this ridesharing service are suggested by using the estimated parameters.