以資料包絡法評估組織中各單位經營績效

Evaluating Operating Efficiency in Organization Using Data Envelopment Analysis

鄭雅帆(Ya-Fan Cheng)

指導教授 ： 李孟峰

國立臺北大學/商學院/統計學系/碩士(2012年)

摘要

過去企業經營的指標，多以組織的生產力及利潤等相關附加價值為衡量準則；然而在資訊時代中，企業經營的指標則以服務品質、企業形象及客戶滿意度等價值來衡量。一個企業組織之經營績效好壞都顯示於企業價值上，經營者應該如何因應外在環境變動而重新檢視調整內部資源運用便是企業組織提升經營績效的重要關鍵，而找出組織中各單位的相對效率也已儼然成為趨勢。本研究對企業組織之各單位進行績效評估，以資料包絡分析法（DEA）找出各單位相對效率，針對無效率單位進行差額變數分析找出改善方向。

本研究之實證分析為評估台灣X壽險公司中各單位之經營績效。選取各單位之業務員人數、主管教育程度、主管年資、主管年齡、主管性別與各單位所在之人口密度為投入變數；以一年內各工作月結束之平均保費收入為產出變數，進行效率分析。結果顯示，約有5.5%的單位具備競爭優勢且穩健度極高，而整體單位之平均效率值介於0.6至0.7之間，存在改善空間臻至最適效率。其中純粹技術效率平均值為0.678，低於平均規模效率值0.945，顯示總技術無效率的主因來自於純粹技術無效率。X壽險公司單位的經營無效率關鍵在於主管未能充分利用資源之管理模式失當，以致造成要素投入浪費而無法發揮應有的效益；故單位主管應考量個別單位核心競爭力以採取最有利的經營型態或策略，強化整各單位的經營體質以提高經營效率。

關鍵字：資料包絡分析法、經營績效、效率值、總技術效率、純粹技術效率、規模效率

關鍵字

資料包絡分析法 ； 經營績效 ； 效率值 ； 總技術效率 ； 純粹技術效率 ； 規模效率

並列摘要

The major evaluation criterions to the guideline of enterprise management were organization's productivity, profits and other related value added in the past. Nowadays, service quality, corporate image and customer satisfaction were the major indicators of an enterprise in the information era. The value of an enterprise was based on the management achievement. The major key of an enterprise to increase management achievement is the manager should understand how to in accordance to the change of external environment to re-inspect and re-adjust the utilization of internal resources to enhance the performance. In addition, it becomes a tendency to discover the relative efficiency of various branches in an organization. This study applied Data Envelopment Analysis (DEA) procedure to calculate the relative efficiency of branches in an organization to perform achievement evaluation. And the slack variables were analyzed to indicate the improvement direction for the inefficiency branches.

The empirical analysis of this study is to evaluate the management achievement of branches of a life insurance company in Taiwan. In this analysis, the number of sales clerks, education level, service year, age, gender of manager and the population density of the area where branch located was selected as input variables; and the average gain of monthly insurance premium within 12 months for each branch was selected as the output variable to carry out the efficiency analysis.

The results show that about 5.5% of the branches have a high degree of competitive advantage and high steadiness; and furthermore the overall average values of efficiency of branches range from 0.6 to 0.7. It shows that there is certain extent has to improve to optimum efficiency. In which the mean value of pure technical efficiency is 0.678, lower than 0.945, the average value of scale efficiency. It also indicated that the major reason of the total technical inefficiency was caused by purely technical inefficiency. The key of management inefficiency of branches of the life insurance company in this study is that some managers have not been able to make full use of resources and misconduct management model. So it results into the waste of input factor and desired benefits cannot be achieved. Accordingly, the managers should consider the core competencies of individual branches to adopt the most advantageous of management style or strategy, and strengthens the entire branch operating structure to raise the management achievement.

Keywords: Data Envelopment Analysis, Management Achievement y, Value of Efficiency, Total Technical Efficiency, Pure Technical Efficiency, Scale Efficiency.

並列關鍵字

Data Envelopment Analysis ； Operation Efficiency ； Efficiency value ； Total Technical Efficiency ； Pure Technical Efficiency ； Scale Efficiency

參考文獻

1. 王正絨（2004），「台灣地區壽險業經營績效之探討」，淡江大學保險學研究所碩士論文。

4. 李建成（2010），「我國壽險業之公司治理、風險承擔與經營績效關係之研究」，朝陽科技大學保險金融管理研究所碩士論文。

23. 鍾素霞（2005），「台灣壽險公司經營績效之研究---金控架構與獨立公司之比較」，中原大學企業管理研究所碩士論文。

1. Banker, R. D.（1984）,”Estimating Most Productive Scale Size Using Data Envelopment Analysis”, European Journal of Operational Research, 17: 35-44.

2. Banker, R. D., A. Charnes and W. W. Copper（1984）,”Some Models for Estimating Technical and Scale Inefficiencies in Data Evelopmant Analysis” , Management Science, Vol. 30, pp.1078-1092.