Measuring transit service reliability at the route level? Exploring the relationship between reliability measures and ridership

Transit agencies are in a consistent struggle to offer an attractive service that draws a higher level of ridership. To improve the attractiveness of the service, one of the key objectives of agencies is to enhance transit service reliability. Service reliability refers to service punctuality and adherence to schedule. A considerable number of studies have focused on understanding the general factors affecting reliability. Nevertheless, it is rare to find studies that explore the association between different reliability measures and transit usage at the route level. Therefore, the aim of this study is to assess how different reliability measures relate to public transit-usage and which measures best explain variations in transit ridership. In total, 22 transit reliability measures that ranged from on-time performance (OTP) measures to service variation measures were assessed. Using land-use, socioeconomic, and detailed ridership datasets, in addition to data obtained from Winnipeg Transit’s Automated Vehicle Location (AVL) system, random coefficients mixed-effect models were estimated at the route level. The results show that, generally, deviation-based measures performed better than OTP measures in explaining transit ridership at the route level. The reliability measure of absolute deviation at terminals performed best in predicting variations in transit ridership, while controlling several influential factors. More importantly, the improvements in predication of ridership due to the use of reliability measures varied according to route’s ridership. This study offers planners and policymakers helpful insights into understanding the relationship between transit service reliability measures of choice and transit ridership at the route level of analysis.