中國大陸自1978年實行改革開放以來，吸引FDI的數量和金額迅速成長，外商投資企業實際上已經成為中國大陸FDI的重要組成部份，值得注意的是，中國大陸吸引FDI的地區分佈呈現出東高西低的非均衡特徵。 因此，本研究預測中國大陸城市的FDI具有外溢效果而存在群聚現象，故本研究以2007年中國大陸281個城市之資料為研究對象，其中包含各城市外商直接投資金額、廠商規模、平均每人城市道路面積、平均每人科學教育支出、勞動者平均工資、科學園區加權權數等資料為基礎，欲探討中國大陸城市之間外商直接投資是否存在群聚現象。 本研究分別以最大概似估計法和貝氏估計法來驗證SAR、SEM、SAC Model，整體衡量此三種模型的適合度，確認出何種方法與模型最適合本研究，在三種模型驗證中，貝氏估計法都能夠和最大概似估計法一樣估計出有效的估計值，最後比較出貝氏估計法的相對優勢來驗證Lesage學者對於貝氏估計法優勢性的見解。 Ever since the reform and opening-up policies of Mainland China have been put into effect in 1978, the number and amount of FDI drawn to China increased significantly, the investments of foreign industries essentially form an important part of national economic in China. Nevertheless, as a sound economic structure within mainland China has not been established yet, the dependence on foreign investments increases in accordance with the gradually rising proportion of FDI to GNP. In view of the phenomenon, this study envisages the clustering phenomenon of FDI in Mainland China as well as the tendency of foreign investments mostly put into a certain geographic area during a particular phase. However, it is quite difficult to define these geographic areas by using the terms as province, city, or administrative district. In the course of research, therefore, the econometric model generally applied is inadequate for describing such specific characteristics. In order to enhance the understanding of FDI, for facilitating the functions of forecast and control, this study employs the Bayesian approach to spatial clustering in measuring and calculating, to verify if there is clustering phenomenon, and further explores the significance of FDI clustering phenomenon to the economic growth.
