單人家戶的空間差異再探

Abstract

近十幾年來台灣單人家戶的比例快速上升，截至2010年已有20%以上的家戶屬於單人家戶，成為除了核心家庭以外台灣最主要的家庭結構。而台灣單人家戶的增加不僅僅是反應諸多社會條件變遷的結果，本身亦是無法被忽略的社會變遷趨勢，並已然對社會各層面產生程度不等的影響。本研究的目的即為結合人口普查資料與空間資料，針對台灣單人家戶進行全面性的探索，主要使用2010年人口普查資料，以台灣本島349個鄉鎮市區為空間分析單位，以探索性空間分析中的空間自相關分析為主，釐清台灣本島單人家戶的空間分布模式，爾後進一步使用人口普查資料以及其他資料中的社會經濟指標，對前述空間分析結果進行關聯性分析，分類、解明單人家戶空間聚集區的空間特色，以及各空間聚集區當中的單人家戶樣貌。\n研究結果部分，空間自相關分析顯示台灣單人家戶的空間分佈確實存在群聚的現象，關聯性分析結果則顯示單人家戶比例與既有鄉鎮市區分類方式、鄉鎮市區人口結構與社會經濟指標等均存在相當程度的關聯性，結合上述分析結果，本研究提出單人家戶空間聚集區可被分為「老年單人家戶熱區」、「非老年單人家戶都會熱區」、「非老年單人家戶原鄉熱區」等三大類型，其中以「非老年單人家戶都會熱區」在各方面最為優勢，「非老年單人家戶原鄉熱區」次之，「老年單人家戶熱區」則面臨人口結構、社會經濟等全方面的劣勢。三大單人家戶熱區中人口結構、社會經濟條件等特性均存在諸多差異，顯示單人家戶人口在空間上的明顯異質性。有鑑於不同類型的單人家戶可能面臨相異的困難與產生相對應的社會支持需求，藉由如此分類、辨認單人家戶空間聚集區的模式與其各項條件，應可明確不同地區單人家戶之社會福利需要，作為政策評估與執行之參考。

For the past few decades, the proportion of one-person households in Taiwan has increased rapidly, with over 20% of households being one-person households as of 2010, it has become the second-most dominant family type in Taiwan after the nuclear family. The growth of one-person households in Taiwan is an important trend of social change which has affected various dimensions of society to varying degrees. The aim of this research is to conduct a comprehensive exploration of one-person households in Taiwan by combining census and spatial data, primarily from the 2010 Population and Housing Census of 349 townships, cities and districts on the main island of Taiwan. The research used spatial autocorrelation analysis to clarify the pattern of spatial distribution of one-person households in Taiwan. The results of the spatial analysis were then further analyzed using socioeconomic indicators from the census and other data to classify and explain the spatial characteristics of both the spatial agglomerations of one-person households and one-person households in each spatial agglomeration.\nThe findings of this research show that the spatial distribution of one-person households in Taiwan is indeed clustered, while the association analysis indicates that the proportion of one-person households is associated with the demographic structure and socioeconomic status of townships, cities and districts. Summarizing the analysis results, this research suggests that the spatial clusters of one-person households can be divided into three major types: “elderly one-person households hotspots”, “non-elderly one-person households urban hotspots” and ”non-elderly one-person households indigenous hotspots”. Among them, “non-elderly one-person households urban hotspots” are the most advantageous in terms of both demographic structure and socioeconomic status. "Non-elderly one-person households indigenous hotspots" come the second, "elderly one-family hotspots" are the least advantageous one. The result also points out that there can be numerous differences in demographic and socioeconomic characteristics among the three one-person households hotspots, indicating the obvious spatial heterogeneity of the one-person households population. As different types of one-person households may face different difficulties and would be in need of corresponding social support, this classification and identification of the spatial clustering patterns and spatial characteristics should clarify the social welfare needs of one-person households in different clusters and serve as a reference for policy evaluation and implementation.