城市教育配套对住宅价格的影响研究
发布时间：2018-08-20 14:52
【摘要】：随着我国城市居民生活水平日益提高,人们的生活需求已经超越衣食住行,追求教育等更高层次的精神需求。教育资源,尤其是公立教育机构属于地方公共品,往往供给有限,且受到地域限制。为让子女得到优质教育资源,父母往往通过支付各类费用以购买受教育的机会,其中通过购买住房享受便利或优质的教育资源是最有常见且有效的途径之一。随着房地产市场研究的日渐深入,教育对住宅价格的影响机制也被逐步展现出来。虽然研究者在研究思路上有类似之处,但得到的结果却呈现出很大的差异,研究的广度和深度参差不齐。为进一步准确、细致、全面地呈现教育对住宅市场的影响方式,本文重点从教育变量细化和消除邻里效应两方面入手,构建特征价格模型和空间计量模型,分析不同类型教育配套对住宅价格的影响机制。本文的主要结论有： (1)影响学校资本化研究结果的因素较多,本文通过梳理既有文献的研究进展和研究思路,总结发现选取教育变量和消除邻里效应是本研究的两大重点。当前众多国内外文献中的教育变量可以大致分为投入变量、产出变量、其他变量,本文对上述三类中常用变量进行了详细阐述和分析。邻里效应是影响学校资本化结果的主要障碍之一。本文将消除或降低邻里效应的主要方法总结为三类：空间计量模型、边界固定效应法、工具变量,并分别分析了三种方法的发展、使用及优劣势。 (2)根据不同受教育机会的获取方式和量化多样化的特点,笔者选取了11个教育变量,组合构建了12个模型。比较11个教育变量的实证结果发现幼儿园数目、小学质量、初中质量、邻近高中和邻近大学等5个变量组合得到最优结果。实证结果证明教育配套对住宅价格的正效应明显,且消费者在购房时愿意为得到更好或更多的教育资源支付额外费用。 (3)计算得到标准小区的教育特征的边际价格,并通过模型中标准化回归系数确定了各类教育配套的影响程度及其排序。例如,学区内小学提高一个等级,小区均价提高3.666%,小区均价上升770.97元／平方米。以标准化后的回归系数为标准进行比较,得到5个教育变量对住房价格影响程度,其中初中质量和小学质量排名靠比较前,幼儿园数目影响力居中下水平,邻近高中和邻近大学排名靠后。 (4)使用GeoDa软件证实了杭州住宅市场的空间相关性,然后根据空间计量模型和杭州市实际情况,构建空间滞后模型和空间误差模型。结果显示：空间模型拟合效果好于基本特征价格模型,尤以空间误差模型更优。模型中几乎所有教育变量显著,邻里效应或空间相关系引起的学校资本化率虚高在一定程度被去除,回归系数稍减小。最后笔者定量测算了在消除空间相关性后,教育配套对在住宅市场的资本化程度。
[Abstract]:With the increasing improvement of living standards of urban residents in China, people's living needs have exceeded the needs of food, clothing, housing and transportation, and the pursuit of higher spiritual needs such as education. Educational resources, especially public educational institutions, which are local public goods, are often limited in supply and subject to geographical constraints. In order to give their children access to high-quality educational resources, parents often pay all kinds of fees to buy opportunities for education, and one of the most common and effective ways is to enjoy convenient or high-quality educational resources through the purchase of housing. With the deepening of real estate market research, the impact of education on housing prices has been gradually revealed. Although there are similarities in the research ideas, the results are very different, and the breadth and depth of the research are not uniform. In order to further accurately, meticulously and comprehensively present the influence of education on the housing market, this paper focuses on two aspects: education variable refinement and neighborhood effect elimination, and builds a feature price model and a spatial measurement model. This paper analyzes the influence mechanism of different types of education on housing price. The main conclusions of this paper are as follows: (1) there are many factors influencing the research results of school capitalization. It is found that the selection of educational variables and the elimination of neighborhood effect are the two main points of this study. At present, the educational variables in many domestic and foreign literature can be roughly divided into input variables, output variables and other variables. This paper describes and analyzes the three commonly used variables in detail. Neighborhood effect is one of the main obstacles to the result of school capitalization. In this paper, the main methods to eliminate or reduce the neighborhood effect are summarized into three categories: spatial metrology model, boundary fixed effect method, tool variable, and the development of the three methods are analyzed respectively. (2) according to the characteristics of obtaining different educational opportunities and quantitative diversification, the author selects 11 educational variables and constructs 12 models. The empirical results of 11 educational variables showed that the best results were obtained from the combination of 5 variables, such as the number of kindergartens, the quality of primary school, the quality of junior high school, the adjacent high school and the adjacent university. The empirical results show that the positive effect of educational matching on housing price is obvious, and consumers are willing to pay extra cost to get better or more educational resources. (3) calculate the marginal price of the educational characteristics of the standard residential area. Through the standardized regression coefficient in the model, the influence degree and ranking of all kinds of educational matching are determined. For example, the primary school in the school district raised one grade, the average price of the district increased by 3.666 yuan, the average price of the district increased by 770.97 yuan per square meter. With the standardized regression coefficient as the standard, the influence of five educational variables on the housing price was obtained. Among them, the junior middle school quality and primary school quality ranked first, and the number of kindergartens was at the middle and lower levels. (4) the spatial correlation of Hangzhou residential market is verified by using GeoDa software, and then the spatial lag model and spatial error model are constructed according to the spatial metrology model and the actual situation of Hangzhou. The results show that the fitting effect of the spatial model is better than that of the basic feature price model, especially the spatial error model. Almost all the educational variables in the model are significant, the virtual high of school capitalization rate caused by neighborhood effect or spatial relationship is removed to a certain extent, and the regression coefficient is slightly reduced. Finally, the author quantificationally calculates the degree of capitalization in housing market after eliminating spatial correlation.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F293.3;F224
本文编号：2194037
[Abstract]:With the increasing improvement of living standards of urban residents in China, people's living needs have exceeded the needs of food, clothing, housing and transportation, and the pursuit of higher spiritual needs such as education. Educational resources, especially public educational institutions, which are local public goods, are often limited in supply and subject to geographical constraints. In order to give their children access to high-quality educational resources, parents often pay all kinds of fees to buy opportunities for education, and one of the most common and effective ways is to enjoy convenient or high-quality educational resources through the purchase of housing. With the deepening of real estate market research, the impact of education on housing prices has been gradually revealed. Although there are similarities in the research ideas, the results are very different, and the breadth and depth of the research are not uniform. In order to further accurately, meticulously and comprehensively present the influence of education on the housing market, this paper focuses on two aspects: education variable refinement and neighborhood effect elimination, and builds a feature price model and a spatial measurement model. This paper analyzes the influence mechanism of different types of education on housing price. The main conclusions of this paper are as follows: (1) there are many factors influencing the research results of school capitalization. It is found that the selection of educational variables and the elimination of neighborhood effect are the two main points of this study. At present, the educational variables in many domestic and foreign literature can be roughly divided into input variables, output variables and other variables. This paper describes and analyzes the three commonly used variables in detail. Neighborhood effect is one of the main obstacles to the result of school capitalization. In this paper, the main methods to eliminate or reduce the neighborhood effect are summarized into three categories: spatial metrology model, boundary fixed effect method, tool variable, and the development of the three methods are analyzed respectively. (2) according to the characteristics of obtaining different educational opportunities and quantitative diversification, the author selects 11 educational variables and constructs 12 models. The empirical results of 11 educational variables showed that the best results were obtained from the combination of 5 variables, such as the number of kindergartens, the quality of primary school, the quality of junior high school, the adjacent high school and the adjacent university. The empirical results show that the positive effect of educational matching on housing price is obvious, and consumers are willing to pay extra cost to get better or more educational resources. (3) calculate the marginal price of the educational characteristics of the standard residential area. Through the standardized regression coefficient in the model, the influence degree and ranking of all kinds of educational matching are determined. For example, the primary school in the school district raised one grade, the average price of the district increased by 3.666 yuan, the average price of the district increased by 770.97 yuan per square meter. With the standardized regression coefficient as the standard, the influence of five educational variables on the housing price was obtained. Among them, the junior middle school quality and primary school quality ranked first, and the number of kindergartens was at the middle and lower levels. (4) the spatial correlation of Hangzhou residential market is verified by using GeoDa software, and then the spatial lag model and spatial error model are constructed according to the spatial metrology model and the actual situation of Hangzhou. The results show that the fitting effect of the spatial model is better than that of the basic feature price model, especially the spatial error model. Almost all the educational variables in the model are significant, the virtual high of school capitalization rate caused by neighborhood effect or spatial relationship is removed to a certain extent, and the regression coefficient is slightly reduced. Finally, the author quantificationally calculates the degree of capitalization in housing market after eliminating spatial correlation.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F293.3;F224
【参考文献】
相关期刊论文 前1条
1 温海珍,贾生华;住宅的特征与特征的价格——基于特征价格模型的分析[J];浙江大学学报(工学版);2004年10期
，本文编号：2194037
本文链接：https://www.wllwen.com/jingjilunwen/zbyz/2194037.html
