住宅特性對空屋形成機率影響之研究

Abstract

過去空屋研究文獻已說明空屋形成機率與住宅屬性具有相關性(Jud and Frew,1990; 彭建文,1994; 林韋宏,2003)，但過去研究通常是以住宅建物本身屬性對空屋形成機率的影響進行分析，並未進一步分析住宅所在的社區、區域特性是否會對空屋形成機率造成影響。本研究重新界定住宅投資性與異質性特性。首先，我們從市場上主要投資客偏好商品角度，選擇影響住宅投資性的變數，包含屋齡、面積、是否座落於整體開發地區、是否座落於捷運周邊地區等四個變數。同時，從住宅存量統計、市場行銷角度選擇反應住宅是否為主流或非典性的異質性變數，包含總樓層、所在樓層、社區總戶數、是否為主力坪數。在估計方法上，我們採用Logit模型，探討臺北市12個行政區、新北市6個行政區中住宅屬性對空屋形成機率的影響。實證發現 3年內新成屋、15坪以下的住宅等投資性住宅的特性，比較其他屋齡、其他面積坪數的住宅之空屋形成機率顯著的相對偏高。臺北市主流4-5層公寓、超過150戶的大規模社區等市場主流商品特性，比較其他建物型態、規模社區較小住宅的空屋形成機率顯著的相對偏低。

The linkage between the housing characteristics and vacant probability has been proved (Jud and Frew, 1990; Lin, Peng, and Lin, 2003), but those researches only used the limited characteristics based on the housing unit without considered the characteristics based on the community or location. This article redefines and classifies characteristics as heterogeneity and investable type with the data in Taipei city and new Taipei city. Based on the investors’ preference, the article chooses 4 characteristics, such as ages, floor area, whether located in new planed area, and whether located in TOD zone, and based on the distribution of housing stock, and marketing, the article chooses 4 characteristics, such as total floor height, located floor height, total houses in the community, whether as main type. The results of a binary logit model show that a close relationship between housing vacancy and housing characteristics, both in heterogeneity and investable type, such as ages, floor area, whether located in new planed area, total floor height, and total units in the community.