Spatial Identification of Housing Vacancy in China

Abstract

The housing vacancy rate (HVR) is an important index in assessing the healthiness of residential real estate market. In China, it is hardly to take advantage of the basic data of real estate information due to the opaque of those data. In this paper, the HVR is estimated to two scales. At the grid level, urban area ratio was calculated by nighttime images after eliminating outliers of nighttime images and night light intensity of non-residential pixels in mixed pixels by a proposed modified optimal threshold method, and built-up areas in each pixel were extracted from the land-cover data. Then, the HVR is calculated by comparing the light intensity of specific grid with the light intensity of full occupancy rate regions. At the administrative scale, the GCI ('ghost city' index) is constructed by calculating the ratio of the total light radiation intensity of a city to the total construction land area of the city. The overall spatial differentiation pattern of the vacant houses in the national prefecture level administrative regions is analyzed. The following conclusions were drawn: vacant housing is rare in certain eastern coastal cities and regions in China with relatively fast economic development. Cities based on exhausted resources, some mountainous cities, and cities with relatively backward economic development more typically showed high levels of housing vacancy. The GCI of prefecture-level administrative units gradually declined from north to south, whereas the east-west distribution showed a parabolic shape. As city level decreased, the GCI registered a gradual upward trend. China's urban housing vacancy can be divided into five categories: industry or resources driven, government planned, epitaxy expansionary, environmental constraint and speculative activate by combining the spatial distribution of housing vacancy with the factors of natural environment, social economic development level, and population density into consideration.

Identifying housing vacancy using data on registered addresses and domestic consumption

“…At last, new families of approaches are the exploitation of open-source social media data of essential facilities (Williams, 2019) and the use of satellite nighttime lighting data coupled with land use information (Wang et al, 2019;Tan et al, 2020;Pan et al, 2020). These approaches are relevant in the face of mass vacancy, as in China, where the vacancy phenomenon reaches rates of an order of magnitude of 20% in the major cities (Tan et al, 2020) and entire neighbourhoods under-occupied.…”

“…The satellite data analysis is relatively accurate compared to pre-existing statistics (Wang et al, 2019), but it has important limitations. While it can identify areas of vacancy and can be implemented regularly, it cannot determine addresses or duration (Pan et al, 2020). The approach is suitable for major cities but not for smaller cities, as the intensity of light radiation is too low (Wang et al, 2019).…”

“…The approach is suitable for major cities but not for smaller cities, as the intensity of light radiation is too low (Wang et al, 2019). The low resolution of the satellite data, about 100 m, coupled with spurious light sources such as urban park lighting and road traffic induce inaccuracies that prevent analysis of the vacancy phenomenon at a more precise scale (Tan et al, 2020;Pan et al, 2020). The method of open-source social media data of essential facilities assumes that residents need essential facilities such as shops and schools.…”

Spatial Pattern of Housing Sales Vacancy in Guangzhou’s Urban District, China

“…The subject of home vacancy has gotten a lot of attention in recent years. Scholars have analyzed the housing vacancy pattern in different cities using nighttime light data [1] , geographic national conditions monitoring data [2] , power consumption data [3] , municipal water consumption data [4] , and other data, and have observed different vacancy pattern characteristics, but most cities show the characteristic that the central urban area is constantly extending to the outer suburbs [2,5] . However, there is currently no research on the sales vacancy pattern in China or elsewhere.…”

Influencing Factors of Housing Sales Vacancy in the Guangzhou Metropolitan Area, China

“…Location factors, including distance from the CBD, living facilities, traffic conditions, educational and medical resources, also have an impact on vacancy rates, as do social factors, such as population density, per capita GDP [3] , and housing purchase policies. Other factors, such as property developers' speculation [4] , property hoarding, and price rigging lead to a high housing vacancy rate. However, there is still a lack of research on the influencing factors of "sales vacancies.…”