應用共同權重於麥氏生產力指數發展多週期生產力評估模式

Abstract

本研究利用資料包絡分析法(Data Envelopment Analysis, DEA)所演變而來的共同權重模式(Common Weights Analysis, CWA)，提出兩階段的分析方法，幫助績效評量者分析受評單位多週期的生產力變化。階段一利用各組織的多週期指標資訊，決定各期的指標權重，應用麥氏生產力指數(Malmquist productivity index, MPI)發展CWA的麥氏生產力指數(CWA-based MPI)來衡量各參與評比的組織每對前後兩週期間的生產力變化，並計算影響生產力變化的子評量指數(CWA-based sub-MPIs)：1. 個人競爭力變化(Self Competitiveness Shift, SCS)；2. 群體競爭力消長(Group Competitiveness Shift, GCS)。階段二以階段一為基礎，利用(Caves, Christensen, & Diewert, 1982)CCD數值轉換概念，轉換CWA-based MPI成為CWA/CCD-based MPI，使其滿足循環性檢定與時間倒數檢定，可衡量任意兩週期之間的生產力變化，使任一週期為標準衡量它期之生產力變化有穩健之分析結果，發展多週期間的生產力變化方法。最後，本研究應用我們所提出之多週期績效評量技術，評量Fortune Global 500中11家藥廠2005至2008四年間的生產力表現。

The common weights analysis (CWA), derived from data envelopment analysis (DEA), can help the performance assessor to measure the efficiencies of a group of many units of assessment (UOAs) with multiple inputs and outputs. Phase I of this research is to develop a method to measure the CWA-based Malmquist productivity index for each UOA’s productivity change between two adjacent periods. The productivity index further is decomposed into two sub-indices: Self Competitiveness Shift (SCS), and Group Competitiveness Shift (GCS). In Phase II, based on Phase I and employ the concept of (Caves, Christensen, & Diewert, 1982) we develop a CWA/CCD-based Malmquist productivity index for measuring each UOA’s productivity change between any two periods. Most importantly the productivity index satisfies Time-reversal test and Circular test so that the multilateral performance analysis would avoid the bias from choosing the fixed comparison period. One could measure the trend of productivity changes for each UOA directly. The productivity changes of Fortune Global 500 Pharmaceuticals companies between 2005 and 2008 are analyzed with our procedure.