台商電子組裝業於中國大陸設廠之區位選擇

Abstract

1980年代以勞力密集產業為主的亞洲國家或經濟體如日本，台灣，香港，南韓等因國內勞工工資成本的上漲，使得競爭力遭受到嚴峻的挑戰，必須向外尋求新的生產基地。中國這個鄰近亞洲各國的人口大國，以其優異的地理位置，充沛且便宜的勞動力，又具備了潛在的龐大市場商機，吸引了世界包含了亞洲在內的各國前往投資設廠。如何選用適當的方法做投資設廠地點的評估對於企業而言至關重要。階層分析法(AHP)能夠針對多準則或多目標等問題提供決策方案。但是當AHP排序分數 (Ranking Score)非常靠近時，容易產生不可靠之決策結果；同時使用於階層分析法的數據(Samples)皆來自於專家意見(又稱: 專家問卷)，使得數據隱含專家的主觀偏好而致影響結論。我們提出混合型階層分析法 (Hybrid AHP)，建構95%的AHP排序值(Ranking Score)誤差帶之信賴區間以增加決策可靠性；同時運用統計拔靴法(Statistical Bootstrapping Approach)針對有限專家問卷進行抽樣計算(取出不放回)，儘可能消弭專家主觀偏好對結論的影響。依據我們的方法，決策原則是極大化AHP值，極小化誤差帶。為驗證混合型階層分析法，本論文以台商電子組裝業在中國大陸地區(分為北、中、南、東四區)設廠評估資料做為實證研究。研究結果顯示選擇中國東部 (江蘇，浙江，上海)作為設廠地點具有最高共識度以及最佳決策可靠性。

An upswing in labor costs and currency appreciation during the 1980s caused Asian countries and economic entities such as Hong Kong, Japan, and Taiwan to search for new manufacturing sites in order to obtain lower manufacturing costs. China, with its huge, rapidly growing market, was one of the main options for these Asian economic dynamos. How to find an appropriate method to evaluate an optimum place for a factory set up in China is essential to an enterprise. Analytical Hierarchy Processing (AHP) is a method to solve complex mul-ti-criteria decision problems. However, to make decisions only based on AHP results is not sufficiently reliable, especially when the results are too close to be precisely determined. To strengthen AHP analysis, we propose a “hybrid AHP” method. This method computes an error band (EB) of final AHP-scores using its 95% confidence interval to estimate AHP score error, and calculates EB combined with a “Bootstrap” to mitigate expert pectoral bias. Based on our methods, our decision rule is to “max-imize AHP score and minimize EB”. To verify our methods, we took Taiwanese elec-tronic assembly manufacturers selecting manufacturing sites in China as a case study. Our research found that Eastern China has a relatively higher consensus of preference for establishing manufacturing sites.