Setting Improvement Targets For Construction Companies : A P-DEA Approach

[[abstract]]營建公司須評估自己的經營績效，以瞭解其在營造業者間的整體競爭力，並訂定務實的改善目標。為達企業之永續經營，財務性和非財務性因素皆應納入績效評估。本研究係採用資料包絡分析(data envelopment analysis DEA)為衡量方法，對台灣地區上市營建公司進行績效評估與比較。首先從2006年至2010年服務業500大排行榜中，選取10家每年均進榜且資料齊全之營建業者為案例公司，評估各公司之相對效率。並利用務實資料包絡分析(P-DEA)模式，訂定案例公司之營運改善目標。本研究採用四個指標，財務性指標為股東權益報酬率和自有資金比率；非財務性指標為員工高學歷比率和員工平均產值率。研究驗證P-DEA模式能藉由所產生之虛擬決策單位(virtual DMU)，訂定效率前緣公司之改善目標，以幫助營造業者提升經營績效。但礙於資料限制，敏感度分析結果顯示各因子中以高學歷比率對績效的影響程度最大，其次為自有資金比率。[[abstract]]A construction company has to assess its own operation performance so as to understand its overall competitiveness relative to other firms in the industry and set practical improvement targets To attain sustainable development of the company both financial and non-financial aspects should be included in performance assessment This study has adopted data envelopment analysis (DEA) as measuring method for assessing and comparing the performance of publicly traded construction companies in Taiwan First 10 construction companies with complete data that ranked among the top-500 firms of the services industry consistently between 2006 and 2010 were selected as case firms for assessment of their relative efficiency Then the practical data envelopment analysis (P-DEA) model was used to determine the case firms’ improvement targets in operation In this study four indicators were included; the financial indicators were rate of return on shareholders’ equity and proportion of net assests in total assests while the non-financial indicators were proportion of employees with a higher degree and scaled average production per employee It has been verified that the P-DEA model through virtual decision making units (DMU) obtained from it can set improvement targets for efficiency-frontier firms so as to help construction companies enhance their operation performance However due to data constraints the results of sensitivity analyses show that of the four indicators proportion of employees with a higher degree achieved the greatest influence on performance followed by proportion of net assests in total assest