軌道運輸標竿聯盟每年度定期進行績效評比,其中臺北捷運公司於「系統可靠度」單項指標之表現被評定為全球26個會員之首;惟各項關鍵績效指標個別作為績效評比準則,則顯得不足客觀。是故,本研究欲以資料包絡分析法(data envelopment analysis, DEA)共同採納多項績效指標,以求得臺北捷運於此聯盟中之相對表現。首先,將衡量各捷運之DEA效率值,並以隨機邊界法(stochastic frontier analysis, SFA)之迴歸模型將第一階段效率值中所包含的環境變數之影響排除,以求得真正的管理面效率值,最後再以麥氏指數(Malmquist index, MI)分析各捷運公司於跨期營運效率的變動情況,找出影響生產力變動之因素為何,並提供各公司未來發展建議。本研究之投入項採用車廂數與員工數,產出項採用了客運人數與年營運收入,環境變數則是採用了補貼比例、服務城市人口密度及公車轉乘優惠。研究結果發現:臺北捷運與曼谷捷運於投入項調整前後之分析結果皆為相對有效率者,其對於投入資源之有效地運用,使得其成為標竿對象,可供其他捷運公司學習。而環境變數之衡量結果發現,服務城市人口密度對於投入項之差額有正相關之關係,該捷運服務城市的人口密度越高,則使得整體生產效率表現越好。跨期效率變動分析結果發現,蒙特婁捷運之總要素生產力呈現上升或下降的趨勢,主要是生產效率值變動之情況而導致之,雪梨捷運之總要素生產力,則是主源由於生產技術之進步或衰退等發展情況。
Being in memebership of Nova group of the Urban Railway Benchmarking Group, Taipei metro and other 25 members of Nova and CoMET have been annually estimated many key performance indexes. Taipei metro had been named of the most reliable metro system for five years.However, using single KPI as performance estimation priciple is not objective and not convincible enough. Therefore, this study would use data envelopment analysis(DEA) to estimate the relative efficiency of Taipei metro to the group that several performance indexes would be included. At first, this study would estimate the DEA efficiency scores of the metro systems, and use SFA regression model to decompose the effect of the environmental variables for the actual managerial efficiency scores. Then, this paper would analysis the change of intertemporal effciency of the metro systems by Malmquist index(MI). It would contribute to realize the factors that could affected the change of productivity, and then to provide advice to the governance. In this sudy, the two inputs are taken as the number of carriages and employees, and the two outputs are taken as the number of passangers and the annual operating revenue. Furthermore, the ratio of government subsidy, the population density of the serviced-city, and the offering of transfer discount would be taken as environmantal factors. The results indicate that Teipei and Bangkok metro are relative efficient while the the input-slacks being adjusted and non-adjusted. Because of efficiently allocation their resources, they become to being the benchmark to other corporations. The results of the evaluation of environmental effects suggest that the population density has positive correlation relationship to the input-slack. The higher the population density is, the greater performance of global productive efficiency is.