| The Canadian urban transit sector has been experiencing declining ridership in the 1990s, while the supply of transit services has remained relatively stable. With the declining ridership, transit systems must adjust supply, which they might adjust inefficiently. This implies the use of excessive resources in the production of transit services. Also, the declining ridership and revenues increase the emphasis on efficiency in order to avoid increased subsidization of the transit systems. The relevant issue then is “how can the transit systems reduce the amount of inputs, and at the same time maintain their level of output?” This thesis focused on addressing the question of technical efficiency in the Canadian urban transit sector from 1990 to 1998.; This thesis applies bootstrapping methods to calculate DEA scores and associated standard errors, and tests some hypotheses regarding efficiency. This thesis has demonstrated that the original bootstrap DEA estimates are consistent with those of the ordinary DEA estimates using the means test. Also, the ordinary DEA estimates and the original bootstrap DEA estimates are biased, that is, they deviate from the bias-corrected bootstrap DEA estimates. The bias-corrected bootstrap DEA estimates may not represent measures from the original efficient frontier. No other study in the transportation literature has applied bootstrap methods to analyze efficiency. Other contributions are in identifying the sources of efficiency change and making recommendations regarding them. In particular, it is found that subsidies, speed, and newer buses improve transit efficiency while peaking reduces efficiency. Finally, there is no technical change, and that scale effect may be the factor that explains efficiency change.; The main results of this study based on the bias-corrected bootstrap DEA estimates are as follows. The average technical efficiency of the Canadian transit systems is about 78%, implying an inefficiency level of 22%. Most transit systems (56%) experience increasing returns to scale (IRS), while 29% experience decreasing returns to scale (DRS). Transit systems operating less than 100 buses are subject to IRS (77%), and those operating more than 100 buses experience DRS (61%). Most of the transit systems (78%) that experience IRS have service area population between 50,000 and 150,000 persons. On the other hand, transit systems operating in larger urban centres with service area population over 400,000 persons are subject to DRS (83%). The transit systems, on average, could increase their output, by roughly one percent. Also, the transit systems could reduce the fleet size, fuel, and labour by roughly 32%, 34%, and 31%, respectively, without simultaneously reducing the output level. (Abstract shortened by UMI.)... |
