住宅次市場界定及住宅價格空間分析： 以新升格之台南市為例Housing Submarket Delineation and Spatial Analysis of Housing Prices: A Case Study of the Newly - Upgraded Tainan Municipality

住宅價格具有空間相依性且受到不同的空間次市場影響頗深。本文以縣市合併升格後之台南市為例，利用集群分析法以及空間自我相關分析法界定住宅價格的空間次市場，並與行政區次市場做比較，探討此三種類型次市場對住宅價格估計的影響。實證結果發現，三種方法所界定的空間次市場範圍雖不盡相同，但仍可勾勒出住宅價格的空間分布，高價位住宅仍聚集在原台南市中心區，隨著距市中心區越遠，價格逐漸降低。再者，本文發現台南市住宅價格存在明顯空間相依性，在價格模型中，以空間分析法所界定之空間次市場對於住宅價格有較顯著的影響且可明顯提升住宅價格估計精確度。

關鍵詞：住宅次市場、住宅價格、集群分析、空間自我相關分析

Housing prices are spatially dependent and are significantly influenced by spatial submarkets. Based on Tainan city, a new special municipality resulting from the merging of the city and the county, this study has used different methods regarding political boundaries, cluster analysis and spatial autocorrelation analysis to classify housing price submarkets and to analyze the prediction accuracy of housing prices. The results show a clear pattern of housing price allocation where higher prices are concentrated in inner city areas and prices decrease with an increase in the distance to the city center. Moreover, it is found that there exists significant spatial dependence among housing prices in the Tainan municipality. Housing submarkets delineated by spatial autocorrelation analysis methods have more significant impacts on housing prices and can clearly improve housing price prediction accuracy.

Key words: housing submarkets, housing prices, cluster analysis, spatial autocorrelation
analysis