基于Box-Cox转换的Hedonic模型在房产税税基评估中的应用研究

发布时间：2018-04-01 03:40

本文选题：房产税税基　切入点：特征价格模型　出处：《福州大学》2013年硕士论文

【摘要】：在现阶段的中国,房地产的价格具有多重垄断性的特点,其供给多是满足高收入人群的居住和投机需求,而非普通民众的基本住宅保障需求。所以,这需要我国政府实施适当的规制手段来抑制房价的过快增长及提供具有社会保障性的住房来满足人民最基本的生存需要。与前阶段连续出台的干扰市场的行政干预相比,我们认为物业税的开征将更符合市场经济秩序建立的要求。但物业税改革方案为何迟迟没能落地呢?其关键就在于中国缺乏一个切实可行且公允公正的房产税税基评估体系。在国外,特征价格模型被用于房产税税基评估领域由来已久；而在我国,却鲜有资产评估机构将其运用在日常实践中。究其原因,笔者认为是我国缺乏一个统一的房地产业数据库、相关统计资料更是有限。所以,本研究打算从我国国情出发,在不影响评估效果的基础上,减少数据信息的收集量,为我国房产税税基评估体系的建立提供现实的指导意义。本文是在基于主体变量特征价格模型的研究基础上,通过引入Box-Cox转换对原模型加以优化,构建出基于Box-Cox转换和主体变量的特征价格模型。本研究首先运用了基于主体变量的Hedonic模型来计算特征变量的价格；其次采用Box-Cox转换来优化的原模型。期间,将两个模型进行对比分析,发现经过Box-Cox转换的特征价格模型在显著性上优于前者。本文突破了以往研究人员需要投入大量精力去采集小区层次(区位和邻里环境特征)信息的常态,简化了数据的收集范围；同时克服了直接运用特征因素建立模型所面临的样本数据少、解释变量难以涵盖所有隐含特征的困难；在总结了前人对建筑特征研究的基础上,选择了52个变量作为本文住宅特征的指标体系。最后,以福州市晋安区五四北新店板块的二手房作为研究对象,取得有效样本356个,运用改进的特征价格模型对二手房数据进行了实证研究。通过回归分析,确立了各小区的基准均价；并在此基础上,明确了楼层、总层、景观、间距、装修、卧室、客厅、厨房、卫生间等变量对特定房屋的价格调整大小。本研究主要成果如下：(1)构建了引入小区变量的住宅特征指标体系；(2)优化了引入主体变量的多元线性回归模型,构建了福州市住宅特征价格模型；(3)本文将福州住宅市场作为单一市场,分析得出住宅各特征的价格调整系数。
[Abstract]:In China at the present stage, the price of real estate is characterized by many monopolies. The supply of real estate is mostly to meet the housing and speculative needs of high-income people, rather than the basic housing security needs of ordinary people. This requires our government to implement appropriate regulatory measures to curb the excessive growth of house prices and to provide housing with social security to meet the most basic survival needs of the people. We think that the introduction of property tax will be more in line with the requirements of the establishment of market economic order. The key lies in the lack of a feasible and fair property tax base assessment system in China. In foreign countries, the characteristic price model has been used in the field of property tax base assessment for a long time. However, few asset appraisal organizations apply it in daily practice. The author thinks that there is a lack of a unified database of real estate industry in our country, and the relevant statistical data are even limited. Therefore, this study intends to proceed from the national conditions of our country. On the basis of not affecting the evaluation effect, the collection of data information is reduced, which provides practical guidance for the establishment of the property tax base evaluation system in China. This paper is based on the study of the characteristic price model based on the main variables. By introducing Box-Cox transformation to optimize the original model, the feature price model based on Box-Cox transformation and principal variable is constructed. Firstly, the Hedonic model based on principal variable is used to calculate the price of feature variable. Secondly, the original model is optimized by Box-Cox transformation. During this period, the two models are compared and analyzed. It is found that the feature price model transformed by Box-Cox is more significant than the former. This paper breaks through the normal situation in which researchers need to devote a lot of energy to collecting the information of community level (location and neighborhood environment). It simplifies the scope of data collection, and overcomes the difficulty of using feature factors directly to build models with less sample data and less explanatory variables to cover all implicit features. On the basis of summarizing the previous researches on architectural features, In this paper, 52 variables are selected as the index system of the residential characteristics. Finally, the second-hand housing in the North Xindian Plate of the May fourth quarter of Fuzhou City is taken as the research object, and 356 effective samples are obtained. Using the improved characteristic price model to carry on the demonstration research to the second-hand house data. Through the regression analysis, has established the standard average price of each district, and on this basis, has defined the floor, the total floor, the landscape, the spacing, the decoration, the bedroom, the living room, The main results of this study are as follows: 1) A residential characteristic index system with the introduction of community variables is constructed, and the multivariate linear regression model with the introduction of main variables is optimized. This paper takes Fuzhou housing market as a single market, and analyzes the price adjustment coefficient of housing characteristics.
【学位授予单位】：福州大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F299.23;F812.42

【相似文献】