Detecting the Spatially Non-Stationary Relationships between Housing Price and Its Determinants in China: Guide for Housing Market Sustainability

Abstract

Abstract:Given the rapidly developing processes in the housing market of China, the significant regional difference in housing prices has become a serious issue that requires a further understanding of the underlying mechanisms. Most of the extant regression models are standard global modeling techniques that do not take spatial non-stationarity into consideration, thereby making them unable to reflect the spatial nature of the data and introducing significant bias into the prediction results. In this study, the geographically weighted regression model (GWR) was applied to examine the local association between housing price and its potential determinants, which were selected in view of the housing supply and demand in 338 cities across mainland China. Non-stationary relationships were obtained, and such observation could be summarized as follows: (1) the associations between land price and housing price are all significant and positive yet having different magnitudes; (2) the relationship between supplied amount of residential land and housing price is not statistically significant for 272 of the 338 cities, thereby indicating that the adjustment of supplied land has a slight effect on housing price for most cities; and (3) the significance, direction, and magnitude of the relationships between the other three factors (i.e., urbanization rate, average wage of urban employees, proportion of renters) and housing price vary across the 338 cities. Based on these findings, this paper discusses some key issues relating to the spatial variations, combined with local economic conditions and suggests housing regulation policies that could facilitate the sustainable development of the Chinese housing market.

Spatio-Temporal Changes and Dependencies of Land Prices: A Case Study of the City of Olomouc

“…It partially corresponds to the results of Yang [12], where the influence of immigrant population, GDP, and investment in residential buildings were observed. However, the influence of "net income per capita" is not significant for land price changes in Olomouc as compared to results that were obtained by Mou [10] in his analysis, which also includes the influence of average wages of employees on land price in selected cities.…”

“…It is important to emphasize that the results presented in this study are based on a detailed data set for one city: Olomouc. However, the studies mentioned above examine a selected group of cities, and their results show that the significance, direction, and magnitude of the relationships between selected factors can vary across cities [10]. For example, the impact of macroeconomic indicator GDP varied from relatively strong positive to strong negative impacts on residential land price in different cities in China, as observed by Yang [12].…”

“…This is one of the main reasons why land price sustainability issues have been previously addressed by other authors (e.g., [2,3]). The research centres on issues linked to land price volatility [4,5], inflation [3], structure [6], changes [7][8][9], and factors influencing land price [1,[10][11][12]. An estimation of the effect of environmental factors on the real estate prices caused by environmental change has been analysed in the research of Del Giudice and Massimo [13,14].…”

Types of Crime, Poverty, Population Density and Presence of Police in the Metropolitan District of Quito

“…Table 2 shows the results of the OLS regressions. The Koenker (BP) statistics for all models had a p-value larger than 0.05, suggesting homoscedasticity, which means that the model is spatially consistent, while the Jarque-Bera statistics obtained p-values smaller than 0.05, suggesting that some models may be biased [40]. However, OLS models are robust linear models that produce results that are reliable even when the residuals have no normal distribution.…”

Quantitative Estimation and Spatiotemporal Characteristic Analysis of Price Deviation in China's Housing Market

“…The geographically weighted regression (GWR) is a common method for capturing spatial heterogeneity [36], and it outperforms the spatial expansion method in terms of explanatory power and predictive accuracy [37]. It has been applied to the research of China's real estate market [38], but the GWR model are commonly used for cross-sectional data. Although some studies developed expanded version of the cross-sectional GWR analysis to panel data, the method assumes bandwidths and weights generated from the kernel function will remain temporally invariant [39][40][41].…”