Effects of the London Cycle Superhighways on the usage of the London Cycle Hire

This paper evaluates the effects of the London Cycle Superhighways (CS) on the usage of the London Cycle Hire. A total of 52 km CS routes and 762 cycle hire docking stations are included for a period of 5 years. Covariates such as land use, road network characteristics and socio-economic information are included in the data set. Transaction records of each docking station are obtained from Transport for London. We first estimate the effect area of the CS routes using a buffer-area based regression. Our results suggest that the effect area of the London Cycle Superhighways is 300 m from either side of the center of the CS routes. Then the propensity score matching methods are applied to select control docking stations and estimate the effects of the CS routes on the cycle hire usage. The results suggest a significant increase of 27.1% in the average ridership for docking stations within the effect area. For trips with both the origin and destination in the effect area, the increase is 73.2%. We also investigate the effects of the cycle superhighways on average travel speed and travel time of cycle hire trips. The travel speed using cycle hires is increased by 1.16 km/h (13.3%), although the effect is insignificant for trips with destinations out of the effect area. The cycle superhighways also significantly reduced the travel time by 11%. The results show no significant effects on the travel distance due to data restriction.