Fig 1.
Civil car ownership and PCDI in Beijing, Shanghai, and Wuhan from 2015 to 2018.
Table 1.
Launch dates of bike-sharing services in Beijing, Shanghai, and Wuhan from 2016 to 2018.
Table 2.
Description of variables.
Fig 2.
Second-order polynomial fit, where the 25 days before and after the entry of bike-sharing services are set as the bandwidth.
Fig 3.
Third-order polynomial fit, where the 25 days before and after the entry of bike-sharing services are set as the bandwidth.
Table 3.
Effect of bike-sharing on CDI.
Table 4.
Bandwidth sensitivity test.
Table 5.
Holiday effect.
Table 6.
Travel restrictions, extreme weather, and weekend effects.
Fig 4.
Trends of traffic restrictions for 25 days before and after the launch of new bike-sharing services.
Fig 5.
Trends of extreme weather for 25 days before and after the launch of new bike-sharing services.
Fig 6.
Trends of DNvalue for 25 days before and after the launch of new bike-sharing services.
Fig 7.
Trends of length of subway and bus lines in operation for 25 days before and after the launch of new bike-sharing services.
Fig 8.
Trends of population density for 25 days before and after the launch of new bike-sharing services.
Table 7.
Extreme value test.
Table 8.
Time heterogeneity test.
Table 9.
Heterogeneity test.
Table 10.
Further heterogeneity tests.
Table 11.
Mechanism analysis.