-
Northeast China including Liaoning, Jilin and Heilongjiang provinces, is an important geographical region in China (Fig. 1). Administratively, the region is further divided into 30 prefecture-level cities, 4 sub-provincial cities and 2 Prefectures. It covers an area of 787 300 km2 and is home to 108.7 million people in 2019. Northeast China is often referred to as the Rust Belt of China due to the recession of its once-powerful industrial sector (Wang et al., 2014). Over the last two decades, urbanization of the region has slowed down significantly, with the urbanization rate higher than the national average by 19.2 percentage in 1999 and only 2.6 percent higher in 2019. Moreover, just like many Rust Belts in the world, population shrinkage is emerging and intensifying in cities of Northeast China, affecting 52 county-level cities in the 1990s and 94 county-level cities in the 2000s (Li and Mykhnenko, 2018). Nevertheless, the housing market of the region does not seem to have been severely affected as a whole. Since the ending of the allocation of welfare housing in 1998, the average price of commodity house in the region has increased from 1641.7 yuan/m2 to 7861.7 yuan/m2 in 2019.
-
The study investigates the housing price responses under different directions of population change in Northeast China from 1999 to 2018. The 35 cities at the prefecture-level and above are used as analysis units, as it is the smallest unit for which long time-series data can be obtained. To be consistent with previous studies (Glaeser and Gyourko, 2005), the research period is split into decade-long intervals to reflect population change and housing price responses. We calculate the house price and the population size for each unit over the sample period as follows:
$$ {HP}_{i,t}={SH}_{i,t}/{FH}_{i,t} $$ (1) $$ {Pop}_{i,t}={TPop}_{i,t}-{RPop}_{i,t} $$ (2) where HPi, t is the housing price of city i in year t, SHi,t is the sales of commercial house of city i in year t, the data are deflated by using Consumer Price Index (CPI), FHi,t is the floor space of sold commercial house of city i in year t. Popi,t is the population of city i in year t, TPopi,t is the total population of city i in year t, RPopi,t is the rural population of city i in year t.
We use the modeling approach advocated by York (2012) and McGee and York (2018) to analyze whether the effects of population loss and population gain on housing prices are symmetrical or not. Empirical evidence of previous studies reminds us that the relevant factors should be included in the model as well, so we also add the following additional factors affecting the supply and demand of commercial housing. From the perspective of housing supply, the following factors were selected: land price (Landcost), investment in real estate (Investment), completed area of commercial housing (Completed) and new added area of construction land (Construction). From the perspective of housing demand, GDP per capita (GDPper), disposable income per capita (Income) and the number of urban households (Household) were selected as additional exploratory factors. In addition, we add time dummy variables for each year to account for general period effects.
The logic of the modeling approach is to control for other factors that are known to influence housing prices volatility and assess the effect of population loss and population gain separately in one model. To accomplish this, we created separate independent variables for population loss (Poploss) and population gain (Popgain), where the increase variable is coded as 0 if there are decreases and the decrease variable is coded as 0 if there are increases (York, 2012; McGee and York, 2018). Specifically, the model we estimate is as follows:
$$ \begin{split} &H{R_{i,t}} = {{\beta} _0} + {{\beta} _1}Popgain{R_{i,t}} + {{\beta} _2}Poploss{R_{i,t}} +\\ &\;{{\beta} _3}Landcost{R_{i,t}} +{{\beta} _4}Investment{R_{i,t}} + {{\beta} _5}Completed{R_{i,t}} + \\ &\;{{\beta} _6}Construction{R_{i,t}} +{{\beta} _7}GDPper{R_{i,t}} + {{\beta} _8}Household{R_{i,t}} + \\ &\;{{\beta} _9}Income{R_{i,t}} +{{\lambda} _t} + {u_i} + {{\varepsilon} _{i,t}} \end{split} $$ (3) where HRi, t is the housing price change rate of city i in year t,
$ {\;\beta }_{0} $ is the constant,$ {\;\beta }_{1}-{\beta }_{9} $ are the coefficients of the independent variables,$ {R}_{i,t} $ is the change rate of independent variable of city i in year t, λt is the period effect in the year t, ui is the unknown fixed effect for each city and the εi,t is the error term.Other variables included in the model are the annual change rate of the variables under consideration (Table 1), to make it an elasticity model. The raw data of the study was derived from the national, provincial and municipal statistical yearbooks, and the statistical bulletin of sample cities (Table 2). Due to the lack of data in Daxinganling Region, the study conducted an empirical analysis of 35 sample cities with complete data.
Table 1. Variables and their abbreviations
Abbreviations Variables PoplossRi,t Population loss rate of city i in year t PopgainRi,t Population gain rate of city i in year t *LandcostRi,t Landcost change rate of city i in year t *InvestmentRi,t Change rate of real estate investment of city i in year t CompletedRi,t Change rate of completed area of commercial housing of city i in year t ConstructionRi,t Change rate of new added construction land area of city i in year t *GDPperRi,t Change rate of GDP per capita of city i in year t HouseholdRi,t Change rate of urban households of city i in year t *IncomeRi,t Change rate of disposable income per capita of city i in year t Notes: Variables with ‘*’ are calculated at comparable price. PoplossRi,t is equal to population change rate when it is negative, otherwise is 0; PopgainRi,t is equal to population change rate when it is positive, otherwise is 0 Table 2. Data source and description
-
During the study period, the demographic trend of 35 cities in Northeast China has undergone significant changes. It can be seen that the distribution curve of population change shifted leftward over time (Fig. 2), which suggests that a large number of cities that exhibited population growth from 1999 to 2008 experienced population decline from 2009 to 2018. During the period from 1999 to 2008, 28 cities experienced population growth with an average population growth of around 12 percent, and 13 cities of them, mainly distributed along Harbin-Dalian railway (Fig. 3), underwent rapid population growth (average annual growth above 1%). There were seven cities that experienced population shrinkage, most of them are resource-based cities (Fig. 3). In general, population growth was still the mainstream of cities in Northeast China during this period. Population shrinkage is moderate with an annual population loss of less than 1% and only concentrated in a few cities. As a result, the total population of these 35 cities increased by a 4.1million from 1999 to 2008.
Figure 2. Kernel density of population and housing price change of sample cities in Northeast China from 1999 to 2018
Figure 3. Changes in population and housing price of sample cities in Northeast China from 1999 to 2018
Between 2009 and 2018, the number of cities with shrinking populations rose sharply, 31 out of 35 cities experienced population loss. Meanwhile, the severity of population shrinkage also rose significantly, nearly half of these population shrinking cities experienced annual population loss of over 1%, especially for cities located in Heilongjiang Province and Jilin Province (Fig. 3). Only four cities (Shenyang, Dalian, Changchun and Daqing), either large/capital cities or mining cities with unexhausted resources, showed a positive population growth (Fig. 3). Nevertheless, compared to the previous decade, the population growth rate of these cities has decreased significantly. Obviously, population shrinkage had become widespread and intensified among cities in Northeast China from 2009 to 2018. On the contrary, population growth during this period had been rare for cities in the region. In aggregate, the total population of these 35 cities was reduced by 2.4 million from 2009 to 2018.
Different from the demographic change mentioned above, the housing prices of most cities in Northeast China showed an upward trend from 1999 to 2018. After adjusting for inflation, the average house prices of sample cities increased by 150% during this period. Comparatively, housing price of sample cities experienced a higher average growth rate (56%) as population growth during the period from 1999 to 2008 (Fig. 3), especially for cities located in Liaoning Province, the south of Heilongjiang Province and the central of Jilin Province (Fig. 3). While the housing price growth of most cities has not been reversed during the following decade, the average growth rate of housing price slowed down to 43% with the widespread population shrinkage of cities in the region (Fig. 2), the number of cities with house price growth rate less than 50% increased significantly (Fig. 3). Meanwhile, the housing price gap between population growing cities and population shrinking cities was expanding from 468 yuan/m2 in 2009 to 2893 yuan/m2 in 2018.
According to the changes in population and housing price, three categories of coupled relationships between them can be identified (Fig. 4), namely positive growth of population and housing price, negative growth of population and housing price, population shrinkage and housing price rising. From 1999 to 2008, the most common category, including 28 cities, was positive growth of population and housing price. For cities with shrinking population, nearly all of them, except for Baicheng where housing price decreased by 77 yuan/m2, experienced rises of housing prices. During the period between 2009 and 2018, 30 out of 35 cities followed the same pattern of population shrinkage and housing price rising, only Jixi experienced negative growth of population and housing price (decreased by 230 yuan/m2). Overall, positive growth of housing price was prevalent among cities in Northeast China regardless of whether they experienced population growth or shrinkage, and there was no significant linear relationship between changes in population and housing price as manifested in the Fig. 4.
-
The results of the regression analyses between a housing price change and its determinants are reported in Table 3. It can be seen that the effects of population gain and loss on housing price volatility are different in direction and magnitude. The estimated coefficient for PoplossR has been higher than that of PopgainR over the investigated period, implying that a larger decline in housing price associated with population loss than a housing price increase with the same size of population gain. For example, during the period 1999–2008, one percent increase in population, on average, increases the house price by 0.197% while one percent loss of population, on average, decreases the house price by 0.598%. Nevertheless, it should be noted that the asymmetry effect is not as significant as found in previous studies (Glaeser and Gyourko, 2005; Feng et al., 2018; Hashimoto et al., 2020). Such a result indicates that changes in population size did not have a statistically significant impact on housing prices volatility of sample cities from 1999 to 2018, which is similar to Droes and van de Minne (2016) where they found that population change is not a constant determinant of house price dynamics. This confirms the overall increase in housing prices across Northeast China since the commercialization of urban housing, regardless of population growth or shrinkage.
Table 3. Regression results of house price change and its determinants of sample cities in Northeast China from 1999 to 2018
Independent variables 1999–2008 2009–2018 PopgainR 0.197 0.205 PoplossR –0.598 –0.445 LandcostR 0.498*** 0.068*** InvestmentR 0.026** –0.006 CompletedR –0.015*** 0.010** ConstructionR –0.040*** –0.023* GDPperR –0.003 0.387*** HouseholdR 0.087** –0.094 IncomeR 0.005 –0.008 Constant 0.030 0.057*** Observations 315 315 Fixed effects Y Y Time effects Y Y R2(within) 0.670 0.504 Note: ‘***’,‘**’ and ‘*’ indicate significant at the 0.01, 0.05 and 0.1 levels -
According to the regression results of our model (Table 3), most factors influencing housing volatility of sample cities came from the supply side, but their impacts were weakening during the investigated period, which can account for the slowdown of housing price growth rate of sample cities. It can be seen in Table 4 that the supply-side factors influencing housing price were much more variable in the first decade than in the second. Among these factors, the rise of land price and real estate investment played a significant role in promoting housing price in the first decade. Comparatively, the increase of land price, which is an important part of housing price and local fiscal revenue, had a stronger driving effect on the rise of house price. However, in the second decade, with the introduction of the government’s measures to deal with the problems of land finance and high inventory of housing, the growth rate of these two indicators decreased significantly (Table 4). As a result, the effect of real estate investment on house price change was insignificant, and the impact of land price growth on the rise of house price decreased significantly.
Table 4. Growth rate of the supply-side factors influencing housing price of the sample cities in Northeast China from 1999 to 2018 / %
Factor 1999–2008 2009–2018 Land cost 850 47 Real estate investment 874 37 Completed floor space of commercial houses 333 –32 New added construction land area 125 –33 Land supply can directly affect housing supply and then influence housing prices (Li, 2010). Over the two periods, the growth of newly added area of construction land had a significant negative effect on house price change, but the effect decreased slightly due to the reducing supply of newly added construction land during the 2009–2018 period (Table 4). This shows that the elasticity of land supply in the sample cities remained high during the study period and the reduction of land supply did not have a positive impact on house prices as found in other studies (Wen et al., 2017). The change in completed area of commercial housing was similar to that of new added area of construction land, but the impact of this indicator on housing price volatility changed in the two time periods (Table 4). Over the 1999–2008 period, the rapid supply of a large number of housing actually played a negative role in housing price change. During the period 2009–2018, while the completed area of commercial housing decreased significantly, the oversupply of housing has not changed, as the unsold area of commercial housing in 35 sample cities increased from 32.1 million m2 to 62.4 million m2. Against this backdrop, the decline in completed areas of commercial housing did not drive a rapid rise in house price. Instead, the growth rate of house price slowed down with the reduction in the completed areas of commercial housing during this period.
Among the demand-side factors, the changes of GDP per capita had a periodical positive impact on housing price volatility during the investigated period. In the first decade, especially after the implementation of the revitalization strategy of Northeast China, the GDP per capita of sample cities grew significantly (195% on average) under the policy and financial supports from the central government. However, the economic development of Northeast China during the period was mainly driven by external forces, and the development path of over-reliance on heavy industries was still relatively obvious (Li and Zhang, 2012; Wang et al., 2014). A large number of revitalization policies and funds were put into restructuring state-owned enterprises and industrial chain extension of traditional heavy industries, while improvement in urban infrastructure and urban environment progressed slowly, except for the large-scale renovation of shantytowns (Zhang, 2008). Therefore, the development environment of most cities has not been significantly improved with economic growth, which led to the insignificant role of economic growth in driving housing prices in the period. Between 2009 and 2018, more attention was paid to the transformation of economic growth pattern and the cultivation of endogenous development impetus, such as eliminating the backward production capacity of traditional heavy industries and vigorously advancing urban renewal. Although the economic growth rate of sample cities generally slowed down during this period, the quality of economic development has been significantly improved (Zhang, 2013), which had a positive impact on house prices.
As the basic unit of housing consumption, changes in households number have an important impact on housing demand (Lauf et al., 2012). During the investigated period, the household number of sample cities grew significantly as households size became smaller. In particular, the process of household miniaturization was significantly accelerated by the rapid advancement of shantytown renovation during the first decade, as households with multiple generations in shanty towns were largely replaced by small households (Li et al., 2015). The number of urban households increased by 29% (4.661 million households) from 1999 to 2008, which had a significant positive impact on rising house prices. Between 2009 and 2018, with the gradual completion of the shantytown renovation, the growth rate of household numbers in sample cities dropped to 8% on average. Therefore, changes in household numbers did not have a significant impact on house price volatility during this period. In addition, the effect of the increase in disposable income per capita was also insignificant as it did not enhance the housing purchasing power of residents. As shown in Fig. 5, the housing-price-to-income ratio of sample cities still kept rising during the investigated period.
-
Previous explanations about the asymmetric impact are mainly based on the kinked supply proposed by Glaeser and Gyourko (2005), one implicit assumption of these explanations is that housing demand will change accordingly with population change. This is suitable for the housing market in that housing demand is largely determined by changes in population size. For example, the floor space per capita of US increased from approximately 37.2 m2 to 74.3 m2 over the 1891–2010 period, while population increased five times during the same period, the change of floor space per capita is relatively small compared with population change (Moura et al., 2015). Under this background, population changes usually have a significant impact on housing demand, and then asymmetrically affect housing price volatility as housing supply adjusts elastically when population growth and inelastically when population shrinkage.
However, this is not completely consistent with the actual housing situation in Northeast China. After the ending of the allocation of welfare housing in 1998, housing demand, which has been suppressed for a long time with a greatly releasing. The floor space per capita of 35 sample cities in Northeast China was only 9.6 m2 in 1999 and the figure grew to 30 m2 in 2018. Meanwhile, the population of these cities increased by only 4% over the period. Comparatively, the improvement of residential conditions played a greater role in promoting housing demand of Northeast China over the period. Even for cities with shrinking populations, housing demand was kept rising due to the significant improvement of floor space per capita and the moderate population decline (Table 5). Therefore, the negative effect of population loss on housing price was largely offset by the positive effect of residential conditions improvement.
Table 5. Growth rate of population and housing supply/demand of sample cities in Northeast China from 1999 to 2018 / %
Factor Population shrinking cities Factor Population growing cities 1999–2008 2009–2018 1999–2008 2009–2018 Population –4 –9 Population 12 7 Floor space per capita 131 21 Completed floor space of commercial houses 367 –52 On the other hand, the rising housing demand associated with population growth was accompanied by a quicker and larger housing supply. Over the 1999–2008 period, the completed floor space of commercial houses of population growing cities increased by 367%, while the population of these cities only rose by 12% (Table 5). During the following decade, the growth rate of the completed floor space of commercial houses slowed down significantly to destock the oversupply of housing of the former decade. Overall, the high housing supply elasticity, especially over the 1999–2008 period, lead to the weak impact of population growth on housing price appreciation.
-
Although the impacts of population change on housing prices were insignificant during the investigated period, the issue still requires ongoing attention, especially for cities with intensifying population shrinkage. In fact, the commodity housing price of Jixi, a typical population-shrinking city, has declined during the period 2009–2018. Moreover, other population-shrinking cities also face the risk of falling commodity housing prices. Firstly, considering that the improvement of residential conditions in population shrinking cities has slowed down dramatically in the past decade (Table 3), the accelerated population shrinkage would in the end lead to a smaller demand for commercial housing, which might have a negative impact on the price of commercial housing. Secondly, along with the population shrinkage, the number of urban households in some cities, like Hegang, Jixi, Benxi, Liaoyang and Chaoyang, begun to decrease during the recent five years which would further reduce the demand for commercial housing. Thirdly, some population shrinking cities, such as Fuxin, Dandong and Siping, are not effective in destocking commercial housing, the unsold area of commercial housing kept growing, which would lead to difficulties in sustaining the growth of commercial housing prices. Overall, given the weakening impacts of factors contributing to house price appreciation, the negative impact of population shrinkage on house prices may become increasingly significant, declining commercial housing prices could be a medium- or a long-term trend for some population shrinking cities. It is necessary to further explore detailed analysis on the risk of housing price decline in population-shrinking cities, which would provide useful implications on real estate regulation of Northeast China, we leave this for future research.
Responses of Housing Price under Different Directions of Population Change: Evidence from China’s Rust Belt
-
Abstract: Population growth has been widely regarded as an important driver of surging housing prices of urban China, while it is unclear as yet whether population shrinkage has an impact on housing prices that is symmetrical with that of population growth. This study, taking 35 sample cites in Northeast China, the typical rust belt with intensifying population shrinkage, as examples, provides an empirical assessment of the roles of population growth and shrinkage in changing housing prices by analyzing panel data, as well as a variety of other factors in related to housing price, during the period of 1999–2018. Findings indicate that although gap in housing prices was widening between population growing cities and population shrinking cities, the past two decades witnessed an obvious rise in housing prices of those sample cities to varying degree. Changes in population size did not have a statistically significant impact on housing prices volatility of sample cities, because population reduction did not lead to a decline in housing demand correspondingly and an increasing housing demand aroused by population growth was usually followed by a quicker and larger housing supply. The rising housing prices in sample cities was mainly driven by factors like changes in land cost, investment in real estate, GDP per capita and household number. However, this does not mean that the impact of population shrinkage on housing prices could be ignored. As population shrinkage intensifies, avoiding the rapid decline of house prices should be the focus of real estate regulation in some population shrinking cities of Northeast China. Our findings contribute a new form of asymmetric responses of housing price to population growth and shrinkage, and offer policy implications for real estate regulation of population shrinking cities in China’s rust belt.
-
Key words:
- asymmetrical impacts /
- population change /
- housing price /
- rust belt /
- Northeast China
-
Table 1. Variables and their abbreviations
Abbreviations Variables PoplossRi,t Population loss rate of city i in year t PopgainRi,t Population gain rate of city i in year t *LandcostRi,t Landcost change rate of city i in year t *InvestmentRi,t Change rate of real estate investment of city i in year t CompletedRi,t Change rate of completed area of commercial housing of city i in year t ConstructionRi,t Change rate of new added construction land area of city i in year t *GDPperRi,t Change rate of GDP per capita of city i in year t HouseholdRi,t Change rate of urban households of city i in year t *IncomeRi,t Change rate of disposable income per capita of city i in year t Notes: Variables with ‘*’ are calculated at comparable price. PoplossRi,t is equal to population change rate when it is negative, otherwise is 0; PopgainRi,t is equal to population change rate when it is positive, otherwise is 0 Table 2. Data source and description
Table 3. Regression results of house price change and its determinants of sample cities in Northeast China from 1999 to 2018
Independent variables 1999–2008 2009–2018 PopgainR 0.197 0.205 PoplossR –0.598 –0.445 LandcostR 0.498*** 0.068*** InvestmentR 0.026** –0.006 CompletedR –0.015*** 0.010** ConstructionR –0.040*** –0.023* GDPperR –0.003 0.387*** HouseholdR 0.087** –0.094 IncomeR 0.005 –0.008 Constant 0.030 0.057*** Observations 315 315 Fixed effects Y Y Time effects Y Y R2(within) 0.670 0.504 Note: ‘***’,‘**’ and ‘*’ indicate significant at the 0.01, 0.05 and 0.1 levels Table 4. Growth rate of the supply-side factors influencing housing price of the sample cities in Northeast China from 1999 to 2018 / %
Factor 1999–2008 2009–2018 Land cost 850 47 Real estate investment 874 37 Completed floor space of commercial houses 333 –32 New added construction land area 125 –33 Table 5. Growth rate of population and housing supply/demand of sample cities in Northeast China from 1999 to 2018 / %
Factor Population shrinking cities Factor Population growing cities 1999–2008 2009–2018 1999–2008 2009–2018 Population –4 –9 Population 12 7 Floor space per capita 131 21 Completed floor space of commercial houses 367 –52 -
[1] Beauregard R A, 2009. Urban population loss in historical perspective: United States, 1820–2000. Environment and Planning A, 41(3): 514–528. doi: 10.1068/a40139a [2] Bian T, Gete P, 2015. What drives housing dynamics in China? A sign restrictions VAR approach. Journal of Macroeconomics, 46: 96–112. doi: 10.2139/ssrn.2378678 [3] Bureau of Statistics of Heilongjiang, 1999–2019. Heilongjiang Statistical Yearbook 1999–2019. Beijing, China: China Statistics Press. (in Chinese) [4] Bureau of Statistics of Jilin, 1999–2019. Jilin Statistical Yearbook 1999–2019. Beijing, China: China Statistics Press. (in Chinese) [5] Bureau of Statistics of Liaoning, 1999–2019. Liaoning Statistical Yearbook 1999–2019. Beijing, China: China Statistics Press. (in Chinese) [6] China Society of Urban Development, 1999–2019. The Yearbook of China’s Cities 1999–2019, China: Chinese City Yearbook society. (in Chinese) [7] Ding Y, 2019. Housing prices and population dynamics in urban China. Pacific Economic Review, 24(1): 27–45. doi: 10.1111/1468-0106.12271 [8] Droes M, van de Minne A, 2016. Do the determinants of house prices change over time? Evidence from 200 years of transactions data. In: Proceedings of the 23rd Annual European Real Estate Society Conference. Regensburg, Germany. doi: 10.15396/eres2016_227 [9] Editorial Department of China Land and Resources Yearbook, 2000–2013. China Land & Resources Almanac 1999–2012, China: China Statistics Press. (in Chinese) [10] Ministry of Land and Resources of China, 2005–2018. China Land and Resources Statistical Yearbook 2005–2018, China: China Statistics Press. (in Chinese) [11] Feng Y L, Kim T, Lee D C, 2018. Housing price and population changes: growing vs shrinking cities. Accounting and Finance Research, 7(4): 59–65. doi: 10.5430/afr.v7n4p59 [12] Galster G, 2019. Why shrinking cities are not mirror images of growing cities: a research agenda of six testable propositions. Urban Affairs Review, 55(1): 355–372. doi: 10.1177/1078087417720543 [13] Glaeser E L, Gyourko J, 2005. Urban decline and durable housing. Journal of Political Economy, 113(2): 345–375. doi: 10.1086/427465 [14] Hashimoto Y, Hong G H, Zhang X, 2020. Demographics and the Housing Market: Japan’s Disappearing Cities. IMF Working Paper No. 20/200, 2020.09/2021.06.21. [15] Lauf S, Haase D, Seppelt R et al., 2012. Simulating demography and housing demand in an urban region under scenarios of growth and shrinkage. Environment and Planning B: Planning and Design, 39(2): 229–246. doi: 10.1068/b36046t [16] Leading Group Office of National Economic Census of Heilongjiang Provincial People’s Government, 2006, 2010, 2016, 2021. Heilongjiang Economic Census Yearbook 2004/2008/2013/2018, China: China Statistics Press. (in Chinese) [17] Li He, Zhang Pingyu, 2012. Main characteristics and driving factors of industry structure evolution in Northeast China since 1990. Arid Land Geography, 35(5): 829–837. (in Chinese) [18] Li H, Chen Z, Li C, 2015. Post-Project evaluation of China’s urban slum resettlement program: the case of Fushun city. Geography Research Forum, 35: 58–72. [19] Li H, Mykhnenko V, 2018. Urban shrinkage with Chinese characteristics. The Geographical Journal, 184(4): 398–412. doi: 10.1111/geoj.12266 [20] Li He, Lo K, Zhang Pingyu, 2020. Population shrinkage in resource-dependent cities in China: processes, patterns and drivers. Chinese Geographical Science, 30(1): 1–15. doi: 10.1007/s11769-019-1077-4 [21] Li J, Xu Y, 2016. Evaluating restrictive measures containing housing prices in China: a data envelopment analysis approach. Urban Studies, 53(12): 2654–2669. doi: 10.1177/0042098015594594 [22] Li Z G, 2010. Housing conditions and housing determinants of new migrants in Chinese cities. Chinese Sociology & Anthropology, 43(2): 70–89. doi: 10.2753/CSA0009-4625430204 [23] Lieberson S, 1985. Making It Count: The Improvement of Social Research and Theory. Berkeley: University of California Press. [24] Lin Y C, Ma Z L, Zhao K et al., 2018. The impact of population migration on urban housing prices: evidence from China’s major cities. Sustainability, 10(9): 3169. doi: 10.3390/su10093169 [25] Liu X J, Wang M S, Qiang W et al., 2020. Urban form, shrinking cities, and residential carbon emissions: Evidence from Chinese city-regions. Applied Energy, 261: 114409. doi: 10.1016/j.apenergy.2019.114409 [26] Long Y, Wu K, 2016. Shrinking cities in a rapidly urbanizing China. Environment and Planning A: Economy and Space, 48(2): 220–222. doi: 10.1177/0308518X15621631 [27] Long Y, Gao S Q, 2019. Shrinking Cities in China: The Other Facet of Urbanization. Singapore: Springer. [28] Ma Z P, Li C G, Zhang J, 2020. Understanding urban shrinkage from a regional perspective: case study of Northeast China. Journal of Urban Planning and Development, 146(4): 05020025. doi: 10.1061/(ASCE)UP.1943-5444.0000621 [29] Ma Zuopeng, Li Chenggu, Zhang Pingyu, 2021. Characteristics, mechanism and response of urban shrinkage in the three provinces of Northeast China. Acta Geographica Sinica, 76(4): 767–780. (in Chinese) [30] McDonald J F, 2020. How asymmetric are urban housing markets? Some worst cases. The Journal of Economic Asymmetries, 21: e00149. doi: 10.1016/j.jeca.2019.e00149 [31] McGee J A, York R, 2018. Asymmetric relationship of urbanization and CO2 emissions in less developed countries. PLoS One, 13(12): e0208388. doi: 10.1371/journal.pone.0208388 [32] Ministry of Housing and Urban-Rural Development, PRC, 2006–2019. China Urban Construction Statistical Yearbook 2007–2020, China: China Statistics Press. (in Chinese) [33] Moura M C P, Smith S J, Belzer D B, 2015. 120 years of U.S. residential housing stock and floor space. PLoS One, 10(8): e0134135. doi: 10.1371/journal.pone.0134135 [34] National Bureau of Statistics of China, 2000–2020. China City Statistical Yearbook 1999–2019, China: China Statistics Press. (in Chinese) [35] National Bureau of Statistics of China, 1999–2020. China Real Estate Statistical Yearbook 1999–2020, China: China Statistics Press. (in Chinese) [36] National Bureau of Statistics of China, 2000–2015. China Statistical Yearbook for Region Economy 2000–2014, China: China Statistics Press. (in Chinese) [37] Richardson H W, Nam C W, 2014. Shrinking cities: A global perspective. London, UK: Routledge. [38] Sun Wenkai, 2020. Changes in the number of households and the housing demand of Chinese residents. Social Science Journal, (6): 162–168. (in Chinese) [39] Turok I, Mykhnenko V, 2007. The trajectories of European cities, 1960–2005. Cities, 24(3): 165–182. doi: 10.1016/j.cities.2007.01.007 [40] Wang J L, Xu Q, 2017. The influence of floating population on real estate prices: an empirical study of Beijing. In: Proceedings of the 2nd International Conference on Education, E-learning and Management Technology. Pennsylvania, USA: DEStech Publications, Inc., 112–118. [41] Wang Lin, Chen Weilin, 2018. A PVAR study on the relationship between housing price and population mobility. Modern Urban Research, (6): 9–15. (in Chinese) [42] Wang M, Cheng Z M, Zhang P Y et al., 2014. Old Industrial Cities Seeking New Road of Industrialization: Models of Revitalizing Northeast China. Singapore: World Scientific Publishing. [43] Wang X R, Hui E C M, Sun J X, 2017a. Population migration, urbanization and housing prices: evidence from the cities in China. Habitat International, 66: 49–56. doi: 10.1016/j.habitatint.2017.05.010 [44] Wang Y, Wang S J, Li G D et al., 2017b. Identifying the determinants of housing prices in China using spatial regression and the geographical detector technique. Applied Geography, 79: 26–36. doi: 10.1016/j.apgeog.2016.12.003 [45] Wen Le, Peng Daiyan, Qin Yidong, 2017. Land supply, housing prices and peri-urbanization in China. China Population Resources and Environment, 27(4): 23–31. (in Chinese) [46] Wolfgang M, Lisa D, 2008. Shrinking and growing metropolitan areas asymmetric real estate price reactions? The case of German single-family houses. Regional Science and Urban Economics, 38(1): 63–69. doi: 10.1016/j.regsciurbeco.2007.08.009 [47] Wu Kang, Long Ying, Yang Yu, 2015. Urban shrinkage in the Beijing-Tianjin-Hebei region and Yangtze River Delta: pattern, trajectory and factors. Modern Urban Research, (9): 26–35. (in Chinese) [48] Yan Guanghua, Chen Xi, Zhang Yun, 2021. Shrinking cities distribution pattern and influencing factors in Northeast China based on random forest model. Scientia Geographica Sinica, 41(5): 880–889. (in Chinese) [49] Yang Z S, Dunford M, 2018. City shrinkage in China: scalar processes of urban and Hukou population losses. Regional Studies, 52(8): 1111–1121. doi: 10.1080/00343404.2017.1335865 [50] York R, 2012. Asymmetric effects of economic growth and decline on CO2 emissions. Nature Climate Change, 2(11): 762–764. doi: 10.1038/nclimate1699 [51] You H L, Yang J, Xue B et al., 2021. Spatial evolution of population change in Northeast China during 1992–2018. Science of the Total Environment, 776: 146023. doi: 10.1016/j.scitotenv.2021.146023 [52] Zhang Pingyu, 2008. Northeast Regional Development Report 2008. Beijing: Science Press. (in Chinese) [53] Zhang Pingyu, 2013. Urbanization progress, problem and policy in Northeast China since 2003. Bulletin of Chinese Academy of Sciences, 28(1): 39–45, 38. (in Chinese)