Integrating DEA Window with Malmquist Index to Analyze Efficiency of Taiwanese Telecommunication Companies

• Main Authors: Cheng-Yu Chang, 張晟域
• Other Authors: Hsu-Hao Yang
• Format: Others
• Language: zh-TW
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/53460350132658555489

碩士 === 國立勤益技術學院 === 工業工程與管理系 === 94 === The growth rate of subscribers of telephone services in Taiwan has gradually shown signs to halt over the past few years owing to the limited market size and continuing deregulation by the government. The consequence of this halt is the declining growth rate of productivity and the prevailing merger among the companies to increase profit margins. Currently, the three largest companies,Chunghwa Telecom (CHT), Taiwan Mobil (TWM), and Far Eastone Telecommunications (FET), have captured market share up to 80 percents altogether. Despite this majority of market share, how to maintain long-term competitiveness remains a focused issue for companies. Therefore, this research analyzes the trend of efficiency and productivity change from 2001 to 2005, and discovers important factors leading to the change.To analyze efficiency and productivity, this research uses Data Envelopment Analysis (DEA). Because the data ranges from 2001 to 2005, the first part of the research uses DEA window analysis that is appropriate to observe dynamic efficiency change over the periods and to compare how one decision making unit (DMU) performs in different periods. Since there are only three companies, using window analysis can increase the number of DMUs and thus increases the discriminating power of DEA. The second part of theresearch incorporates window analysis into Malmquist productivity index to measure the technical efficiency change and shift in technology frontier over the periods. Finally, the research uses Kruskal-Wallis rank order test to verify whether the change trend coincides with the result from DEA Malmquist. According to the computational results regarding to technical efficiency, TWM performs best, followed by FET. Considering the technical efficiency can be decomposed into pure technical efficiency and scale efficiency, the results suggest that all three companies achieve good pure technical efficiency, while the CHT performs relatively poor in scale efficiency. In view of productivity change, TWM is the best, and CHT is the worst. In terms of technical efficiency change, all three companies show the trend of slightly decreasing. In terms of technological change,the results indicate the progress of technology. Using the non-parametric rank tests verifies that the productivity change not only is significantly different, but shows the trend of increasing or decreasing.