台灣與日本的汽車產業群聚比較

Comparison of the Cluster of the Automobile Industry between Taiwan and Japan

全球汽車產業一直是經濟發展的重要推動力，隨著產業全球化和技術進步，國際間的技術合作和知識轉移日益加緊。台灣雖然不是汽車製造主要國家，但是跨國合作模式為台灣建立了汽車供應鏈。回顧過去對汽車產業的相關研究集中於單一國家或單一城市，但是對於跨國技術轉移及跨國產業群聚的相關研究是稀缺的，如何評估汽車產業群聚的經濟地理特徵，特別是識別汽車產業群聚的形成因素至關重要。這項研究的主要目的是探討台灣和日本汽車產業集聚的原因，在相關的產業群聚研究表明使用經濟地理數據具有顯著的效果。這項研究使用經濟地理數據分析台灣與日本的行政區的汽車產業群聚特徵，研究結果表明台灣汽車產業群聚的形成因素是勞動力數量，而日本汽車產業集群的形成因素是每1員工創造營收。這2個因素決定了台灣汽車產業和日本汽車產業群聚的總體差異性。K-平均值法和手肘法的模型分析在19個台灣行政區和47個日本行政區找到了最佳群聚數，歐基里德距離進一步解釋了特徵值與群聚的相關性，同一個汽車製造商在跨國產業合作的過程中，被多個本地經濟因素影響而形成不同的群聚特徵。

The global automobile industry has consistently served as a significant catalyst for economic development. With the globalization of industries and advancements in technology, international technical cooperation and knowledge transfer are becoming increasingly pronounced. Although Taiwan is not a major player in automobile manufacturing, it has established a transnational cooperation model that supports an automobile supply chain within the region. Historically, research related to the automobile industry has predominantly focused on individual countries or cities; however, studies examining transnational technology transfer and industrial clusters remain limited. It is essential to evaluate the geographical economic characteristics of automobile industry clusters, particularly identifying the factors that contribute to their formation. The primary aim of this study is to explore the factors contributing to the agglomeration of the automobile industry in Taiwan and Japan. Previous research on industrial agglomeration indicates that geographical economic data plays a significant role in this phenomenon. This study analyzes the characteristics of automobile industry clusters in the administrative regions of Taiwan and Japan by utilizing geographical economic data. The findings reveal that the key factor driving the formation of Taiwan's automobile industry cluster is the availability of labor. In contrast, the primary factor for Japan's automobile industry cluster is the revenue generated per employee. These two factors account for the overall differences between the automotive industry clusters in Taiwan and Japan. The K-means method and the elbow method were utilized to determine the optimal number of clusters in 19 Taiwanese administrative regions and 47 Japanese administrative regions. Euclidean distance further clarifies the relationship between eigenvalues and clustering. During the process of cross-border industrial cooperation, the same automobile manufacturer is affected by various local economic factors, resulting in the emergence of distinct clustering characteristics.