›› 2017, Vol. 29 ›› Issue (11): 41-49.

• 经济与金融管理 • 上一篇    下一篇

城市居民住房消费选择中的“新房偏好”效应研究——基于分位数模型的反事实分解

黄静   

  1. 上海师范大学商学院, 上海 200234
  • 收稿日期:2016-08-01 出版日期:2017-11-28 发布日期:2017-11-25
  • 作者简介:黄静,上海师范大学商学院副教授,博士。
  • 基金资助:

    国家自然科学基金项目(71373162;71403059);上海师范大学第七期“城市经济学”和第九期“产业经济学”重点学科建设。

Study on the Effect of “First-hand Housing Preference” in the Housing Choice of Urban Residents

Huang Jing   

  1. School of Business, Shanghai Normal University, Shanghai 200234
  • Received:2016-08-01 Online:2017-11-28 Published:2017-11-25

摘要:

我国城市居民在住房消费选择时是否偏好新建商品房?"新房偏好"效应有多大?文章采用上海市大型微观交易数据,分别建立新建商品房和二手房的Hedonic特征价格模型,利用分位数回归进行参数估计,然后应用反事实分解方法,剥离新旧住房价格差异中的"特征差异",首创性地对高、中、低端不同价位层次住房的"新房偏好"效应进行研究。结果表明,我国城市居民住房购置时存在明显的"新房偏好"消费特征,"新房偏好"是新旧住房价格差异的主要来源。随着房价从低至高,"新房偏好"程度呈U形,在低端住房(40%分位数以下)和极高端住房(90%分位数以上)中"新房偏好"程度更高。

关键词: 住房价格, 特征价格模型, 分位数回归, 反事实分解

Abstract:

In China, urban residents prefer to buy new housing than second-hand housing. This paper studies the first-hand housing preference effect, using a large micro-transaction data based on the quantile regression method and the counterfactual decomposition analysis. The results show that there is an obvious first-hand housing preference effect. And the first-hand housing preference effect is the main source of the difference between the first-hand and the second-hand housing prices. With housing prices ranging from low to high, the first-hand housing preference effect degree is U-shaped. In the low-end housing (below 40% quintiles) and very high-end housing (above 90% quintiles), the first-hand housing preference effect is higher.

Key words: housing price, Hedonic price model, quantile regression, counterfactual decomposition