| Summary: | 碩士 === 國立成功大學 === 交通管理學系碩博士班 === 97 === Taiwanese intercity bus companies are facing the difficulty of operation due to the increase of operational costs year by year. Unfortunately, the operational efficiency of intercity bus companies have never been carefully investigated. Therefore, the main objective of this study is to examine the cost efficiency of 37 intercity bus companies operating over a five-year period in Taiwan. Several panel data stochastic frontier models were estimated by using both Cobb-Douglas and Translog cost function. The conventional models are extended in order to model the environmentally unobserved heterogeneity and the time-varying effects. The estimated coefficients and inefficiency scores are compared across six models. Finally, tobit regression and economic characteristic analysis are imposed to understand the current situation of intercity bus industry.
The findings indicate that the total cost of intercity bus company arose by the impacts of SARS in 2003. Moreover, “True random-effects model” proposed by Greene (2005) can improve the estimations regarding both heterogeneity and the time-varying nature of efficiency. Besides, the estimate efficiency is higher in the Northern region of Taiwan, the performance of service quality negatively affects the score of efficiency, the scale of bus company positively affects the efficiency scores. Among factors of input, fuel price has the most significant influence on total cost. Base on calculations, the index of economies of scale is greater than one, suggesting the presence of unexploited economies in Taiwanese intercity bus industry.
Final result suggests that all companies should develop various marketing strategy and increase their output to achieve the cost-minimising size. In addition, alternative panel models such as the “True random-effects model” proposed by Greene (2005) can be applid to explore the possible impacts of unobserved firm-specific factors and time-varying effects on efficiency estimates. Further, from government’s perspective, the adopted methodology can be also recommended to subsidiary policy making.
|
|---|
