What are the Effects of Demographic Structures on Housing Consumption?: Evidence from 31 Provinces in China

Zhang, Yu; Jin, Haiyan; Xiao, Yue; Gao, Yumin

doi:https://doi.org/10.1155/2020/6974276

Mathematical Problems in Engineering

On this page

Abstract Introduction Literature Review Data Conclusion Data Availability Conflicts of Interest Acknowledgments References Copyright Related Articles

Research Article | Open Access

Volume 2020 | Article ID 6974276 | https://doi.org/10.1155/2020/6974276

What are the Effects of Demographic Structures on Housing Consumption?: Evidence from 31 Provinces in China

Yu Zhang,¹Haiyan Jin,¹Yue Xiao,²and Yumin Gao¹

Academic Editor: Vincenzo Bianco

Received07 Nov 2019

Accepted05 Feb 2020

Published16 Mar 2020

Abstract

Housing is closely related to population. The world’s demographics have been changing greatly, like ageing, urbanization, and shrinking household. This paper takes China as the research object to study the effects of demographics on housing consumption. Demographics were considered from three aspects: natural structure, regional structure, and social structure. Based on Life Cycle Theory and Permanent Income Hypothesis, the models on housing consumption were constructed. Results show that demographic structures are important factors affecting housing consumption: (i) child-age dependency ratio (CDR), education level, and family size negatively affect housing consumption, while urbanization rate and old-age dependency ratio (ODR) have positive effects; (ii) CDR positively affects housing consumption through deposit and the indirect effect of ODR on housing consumption through deposit is negative; (iii) the influence of demographics on housing consumption is heterogeneous in different regions. It is relatively close between the central and the west to some extent but quite different in the east.

1. Introduction

Demographic changes lead to very heterogeneous demand responses for different housing attributes [1, 2]. According to the World Demographic Trends (https://www.un.org/zh/development/population/conclusions.shtml) reported by the UN Secretary-General, the global demographic trends by 2050 are as follows: ageing due to prolonged life, continued urbanization, declining fertility, and shrinking household size. Table 1 shows that all of these trends are strongly related to housing market. Demographic analysis can provide a good basis for housing demand assessment over the short term [3]. However, existing literature studies just focused on one or two aspects of demographic structures when investigating the relationship between demographic structures and housing (e.g., [4–6]). Thus, it is necessary to study housing consumption in combination with comprehensive demographics to make a better forecast of housing consumption from the perspective of population.

In 2017, China ranked first in the world with a population of 1.368 billion, accounting for 18.41% of the global population. The large population base reduces the impact of accidental factors and makes the demographic trends more obvious, and these trends are consistent with the global population trends (which will be introduced in detail in Section 2). In addition, the reform of the housing system and housing marketization in 1998 is an important milestone in the Chinese housing market [4]. Before that, housing was allocated to people by the Chinese government. But after 1998, housing became a common commodity in China. The only way for residents to own a house is to make a purchase in the housing market. This means that residential housing consumption is more sensitive to demographic changes since 1998 because citizens are the decision makers of housing consumption. During housing consumption, the characteristics and preferences will affect their behaviour. In summary, China is a good object to study the relationship between demographics and housing consumption and can provide reference for most countries, especially for those where housing is marketized.

Taking Chinese housing system reform in 1998 as a starting point, this paper investigates the effects of demographic structures on housing consumption using the panel data of 31 provinces from 1999 to 2017 in China. The regional differences of these effects among eastern, central, and western China are also revealed. In particular, natural structure, spatial structure, and social structure of population are included in the model on housing consumption and the partial effect of age structure on housing consumption is studied in detail.

The remainder of the paper is organized as follows. Section 2 sorts out the trends of demographics and housing consumption in different regions of China and nationwide. Section 3 reviews existing studies on the effects of demographic structures on housing consumption from three aspects (e.g., natural structure, regional structure, and social structure). How to choose suitable indicators of demographic structures is also fully discussed in this part. Based on life cycle theory and permanent income hypothesis, Section 4 constructs a panel model containing the indicators selected in Section 3 and presents provincial data from 1999 to 2017 in China. In Section 5, we analyse the panel model using fixed effect regression, interpret the direct and partial effect of demographic variables, and test the robustness of the model using the 4 trillion investment plan in 2008 in China. Section 6 concludes with implications.

2. Demographic Trends and Housing Consumption in China

2.1. Nationwide Demographic Trends and Housing Consumption

China’s housing consumption increased rapidly since the abolishment of welfare housing policy in 1998 (housing was allocated to the household by Chinese government before 1998). According to the National Bureau of Statistics of China, housing sales area in 2018 was 14 times that of 1998. The growth rates in different periods are different. From 1998 to 2010, the sales of commodity houses increased by 19.67% annually, while the annual growth rate fell to 5.92% from 2011 to 2018. This may be related to the decline in the share of working-age people from 2011 (shown in Figure 1(a)). From the perspective of human capital, working-age households, who have more housing desire and better affordability, tend to have greater housing consumption [2]. As people aged 45–59 retire gradually, people aged 0–14 step into working age (15–59). According to the Chinese population aged 0–59 in 2017 (Figure 1(b)) and the low birth rate in China, working population is forecasted to decline continually in the next 15 years and housing consumption will be affected directly.

(a)

(b)

China has entered the ageing society since 1999, and its ageing speed tends to be faster. Due to the age dependency of private demand for different forms of housing, population ageing can unfold heterogeneous impacts across different housing markets [19]. Besides, family support for the elderly is the most common in China. The increased pressure on offspring to care for the elderly will curb the housing consumption of the younger generation. According to the WHO, China will become the most ageing country in the world with 35% ageing rate in 2050 (http://www.un.org/esa/socdev/aging/Vienna_intlpanofaction.html). China’s ageing process is significantly faster than other countries. If other factors are not considered, the negative impact of ageing on housing consumption will stick out [34]. In addition, the spatial distribution of ageing is uneven in China. In Figure 2, Heihe-Tengchong Line is a significant demarcation line of ageing. Younger populations are mainly found in the northwest and the elderly are mainly in the southeast [35]. Depending on the degree of ageing, regional housing consumption will vary differently.

The social characteristics of population are also changing. The family size in China fell from 3.58 to 3.03 from 1999 to 2017, and the proportion of population with college degree or above education rose from 2.87% to 12.89% during this period. Also, some other structural changes of population have emerged in China, such as large-scale movements and regional unbalance distribution. Natural and socioeconomic conditions lead to extremely uneven distribution of population [36]. The Chinese Sixth National Census shows that the floating population was as high as 221 million (compared with 121 million in the Fifth National Census), which was almost 20% of the total population. Actually, the inflowing cities were mainly concentrated in eastern provinces and municipalities, like Guangdong, Zhejiang, Shanghai, and Beijing. The unbalanced distribution and different densities in population make regional housing consumption heterogeneous. Consequently, we will divide China into three parts to study the regional difference in the relationship between housing consumption and demographic structures.

2.2. Regional Demographic Trends and Housing Consumption

In the 2016 National Real Estate Development Investment and Sales Situation in China, 31 provinces in China were divided into three parts: East, Central, and West (the East includes 8 provinces, such as Hebei, Liaoning, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong, and Hainan and 3 municipalities, such as Beijing, Tianjin, Shanghai; the Central includes 8 provinces, such as Shanxi, Jilin, Heilongjiang, Anhui, Jiangxi, Henan, Hubei, and Hunan; and the West includes 11 provinces, such as Inner Mongolia, Guangxi, Sichuan, Guizhou, Yunnan, Shaanxi, Gansu, Ningxia, Tibet, Qinghai, Xinjiang, and 1 municipality which is Chongqing). Differences in geographical location have caused great inequalities in the macroeconomic environment. Development level of the East is the highest, the Central is the second, and the West is the last (the per capita GDP in 2017 was 81625, 47766, and 44717 yuan, respectively). Next, we choose Beijing, Hubei, and Sichuan as the representatives of the East, Central, and West, respectively (Figure 3) and roughly compare the demographic trends and housing consumption of these three regions.

To some extent, Beijing, Hubei, and Sichuan have different regional characteristics (Figure 4). Specifically, the demographic trends and housing consumption from 1999 to 2017 in Sichuan (West) and Hubei (Central) were in line with those in China. For example, the trends of housing consumption, urbanization rate, family size, old-age dependency rate, and child-age dependency rate in Hubei (Central) and Sichuan (West) were similar to those across the country. The circumstance may exist because the central and western parts become the majority of China, as Figure 3 shows. While in Beijing, except for the education level, the other demographic variables showed converse trends. Beijing has become a high urbanized city since 1999, and the urbanization rate was almost unchanged. Its strong economic conditions attracted labors, decreased the old-age dependency ratio, and increased the total population. However, the increasing population did not necessarily increase housing sales area, which may be because the income of inflowing labors was not enough to afford the expensive houses in Beijing.

(a)

(b)

(c)

(d)

There are significant differences in demographic factors and housing consumption among the eastern, central, and western regions. In the following, we will conduct a more in-depth analysis by using the macrodata to identify the relationships and regional differences between demographics and housing consumption.

3. Literature Review

Demographic structures can be classified into three categories: natural structure, regional structure, and social structure. Studies on how demographic factors affect housing consumption are sorted out accordingly and shown in Table 1. In particular, among demographic factors, population size, age structure, spatial structure, education structure, and family size have a more significant effect on housing consumption, and these factors are used in this paper to study how demographic structures affect housing consumption.

3.1. Natural Structure and Housing Consumption

Due to the global trend of declining birth rates and increasing life expectancies, the impacts of population size and age structure on the housing market have received a great deal of attention in the existing literature.

3.1.1. Population Size

Changes in population size will alter housing consumption. The general dependence of the housing market on the total growth of population is widely recognized, because the demand for housing must come from a population to be housed [7]. Although population growth does not necessarily translate into effective housing demand, an increase in population often stimulates housing consumption. Kuchay et al. [8] pointed out that rapid population growth is the major contributory factor for the housing expansion of Srinagar. Choi et al. [5] also found that population size affected housing market to a certain extent. Therefore, population size is adopted as an indicator.

3.1.2. Age Structure

Age structure is most closely related to housing consumption in the demographic structures. As early as 1954, Modigliani began to study the relationship between age and consumption and proposed Life Cycle Theory. Campbell [9] and Easterlin [10] focused on the relationship between age structure and housing consumption in the early time period. Carliner [11] believed that income elasticity of housing demand was discrepant during different ages, which resulted in different housing consumptions. Mankiw and Weil [12] first introduced the Life Cycle Theory into the housing market in 1989 and used US Census data in 1970 to estimate housing demand of different ages. While Green and Hendershott [14] computed the partial effect of age on housing demand and suggested that keeping all else constant, the demand for housing tended to be flat or rising slightly with age. More scholars such as McFadden [13]; Holly and Jones [16]; Ermisch [15]; and Myers and Ryu [17] supported Mankiw and Weil’s opinions and found that age structure had a significant effect on housing demand, despite different estimation methods being adopted.

The main trend of age structure in China is ageing and declining birthrate (China’s total fertility rate was declining since 1991 and was 1.635 in 2018). Both ageing [18–20, 37] and birth rate [5, 20, 37] have a close link with housing market. We choose child-age dependency rate (CDR) and old-age dependency rate (ODR) as indicators to reflect the change of age structure. Considering the partial effect mentioned by Green and Hendershott [14], the interactive items between CDR or ODR and deposit are also used to examine whether age structure has a partial effect on housing consumption through deposit.

3.2. Region Structure and Housing Consumption

Spatial migration may have an influence on the local housing market, by changing housing demand and housing consumption. In developed countries, such as Korea [22], America [23], Canada [24], Spain [25], and Iceland [28], immigrants affect housing market significantly and differently. In the United States, Saiz [23] used the IV (instrumental variable) method and concluded that a 1% increase in migration will result in a 1% increase in rents and housing prices. In Spain, immigration led to an average 2% annual increase in housing prices, for a 1.2%–1.5% increase in housing units between 2000 and 2010. In Iceland, 1% net immigration yielded a rise in housing prices as high as 4%–6%. There are differences in migration between China and developed countries. Migration in China mainly occurs between urban and rural areas or between regions. Many Chinese scholars, such as Long and Wu [21]; Chen et al. [4]; Lu et al. [26]; and Wang and Zhang [27], confirmed that there is a long-term and stable positive relationship between urbanization level and housing market. Urbanization process will generate a large number of rigid, investment, and speculative housing demand. Therefore, the urbanization rate is included as the indicator to respect the change of Chinese population spatial structure.

3.3. Social Structure and Housing Consumption

Family size, education level, and other sociodemographic factors play an increasingly important role in making housing purchase decision. Scholars usually study social structure of population from two aspects, that is, education structure based on the individual feature and family size based on the family feature.

3.3.1. Education Structure

High education levels would increase a household’s housing consumption [2]. Green and Hendershottb [14] used census data of 1980 and found that in 1980, the willingness to pay for a constant-quality house was about 50% lower for households in 70-year-olds than for those in 50-year-olds, owing to the surge in education after World War II. Zhou [30] pointed out that families with household heads of high educational level prefer to own private housing, and educational expenditure uncertainty negatively affects housing choice. Based on the microhousehold data in China, Chen and Zhang [31] analyzed dynamic evolution of urban housing prices from the perspective of human capital. They found that every 1% increase in population of higher education rises urban housing prices by 4.6%–7.9%. Obviously, education is significantly related to housing consumption. In this paper, the population share of people with college degree or above is used as the indicator to measure the change in educational structure.

3.3.2. Family Size

The consumption in housing is family-based. Horioka [32] proposed that the number of households, rather than the number of people, is the driving force of housing prices, and the expansion of household size would increase the demand for nonhousing assets, thereby weakening housing consumption. Hugo [33] used statistical methods to analyze the population and family changes in Australia, and results showed that the increase of households would increase total housing demand. Lauf et al. [6] also found that even if the total population showed a downward trend, the miniaturizing households had contributed to the rising demand for housing. The future housing demand in China will be affected by changes in household size. Chen and Chen [34] indicated that family miniaturization will boost urban housing consumption in China in the next 20 to 30 years. Thus, family size is used as another indicator.

Above all, population size, child-age dependency rate (CDR), old-age dependency rate (ODR), urbanization rate, the population share of people with college degree or above, and family size are adopted as indicators in this paper. Besides, the interactive items between CDR or ODR and deposit are also added to examine whether age structure has a partial effect on housing consumption through deposit.

4. Theoretical Framework and Data

4.1. Model Assumptions

Keynes proposed the Absolute Income Hypothesis in The General Theory of Employment, Interest and Money, arguing that people’s consumption is determined by their current disposable income. On this basis, Friedman’s Permanent Income Hypothesis believes that the main determinant of consumption is the persistent income of consumers rather than the current income. Modigliani’s Life Cycle Theory, mainly analyzing the impact on consumption from income and deposit, holds that people plan their living expenses over a longer period to achieve the optimal allocation of consumption throughout the life cycle.

According to Life Cycle Theory and Permanent Income Hypothesis, the model of rational consumer pursuing maximization of intertemporal utility can be expressed aswhere , , and denote the consumption, income, and deposit in the t-term, is the time preference rate, is the return on assets, and represents the life cycle stage, T = {0, 1, 2}, where “0” means childhood, “1” means labor period, and “2” means old age.

The utility function is a quadratic curve. When , the optimal consumption path is

It explains that the consumption level in the t-term depends on the current deposit and the discounted value of the current and expected future income. Further assuming that income follows first-order autoregression (), the simplified long-term consumption function can be obtained:where and are the long-term influence coefficients of income and deposit on consumption.

Housing consumption is an important aspect of consumption. Based on the long-term consumption function, this paper considers the impact of housing supply on housing consumption and establishes the metrology models as follows (We tried to add housing price as a controlling variable, but the results showed that housing prices in all regions did not have a significant impact on housing consumption. In the traditional relationship between supply and demand, price is an important factor affecting demand. But house can not only be used for living but also a good investment. The rise in housing prices has stimulated investment behavior in the real estate industry and maintained the stability of housing consumption):where i denotes the province, t represents the year, HSA, HCA, INCO, DEPO, POPU, CDR, ODR, URB, FS, and EDU represent housing consumption, housing supply, income, deposit, total population, child-age dependency rate, old-age dependency rate, urbanization rate, family size, and education level, respectively, is the unobservable interprovincial effect, and is the error term. And Model (4) does not consider the impact of deposit.

Besides, there may be a partial effect of population age structure on housing consumption through deposit. The squeezing effect of CDR and ODR on consumption will vary among households with different deposits. In households with higher deposit, increased support pressures will result in a reduction in consumption. However, due to its considerable economic foundation, the degree of squeezing effect will be smaller. For families with lower deposit, the sensitivity of consumption to support pressure will be greatly increased. Furthermore, children and deposit can be seen as two types of pension tools that have the same economic function and replace each other: (i) when a family has a multitude of children, parents will reduce deposit that were originally for old-age security and increase consumption; (ii) when the number of children is smaller, parents increase their deposit to support their old age and the consumption decreases.

Above all, this paper introduces the interaction items of CDR and deposit and ODR and deposit into the model. The extended model is

4.2. Selection of Variables and Data

This paper selects the panel data of 31 provinces and municipalities in China from 1999 to 2017 as the research sample. All data come from the China Statistical Yearbook, and the People’s Bank of China. The specific descriptions of each variable are shown in Tables 2 and 3.

With 1999 as the base year, the income and deposit are adjusted by the Consume Price Index, and variables are processed by natural logarithm except the structural variables. Meanwhile, to ensure the robustness of the results, the panel data are checked for panel smoothness. The unit root test is used for the panel stationarity test commonly, mainly including LLC, IPS, Fisher-ADF, HT, and so on. In this paper, LLC, IPS, Fisher-ADF, and HT are all used for the unit root test. The results in Table 4 show that most variables are stationary at 99% levels in more than two tests, indicating the panel data in this paper have good stability.

5. Results of the Fixed Effect for Panel Data

5.1. Analysis of Regression Results

To reveal the effects of demographic structures on housing consumption and the regional differences in effects, we conduct an overall analysis of China at first and then divide 31 provinces and municipalities into East, Central, and West. Based on the panel data, fixed-effect models are established and estimated by Stata15.1.

In this paper, the F-test and Hausman test are carried out in the national and regional models. The F-test results show consistent rejection of the null hypothesis, which means that there are significant differences in the cross sections of the panel data. The results of Hausman test are shown in Table 5. Except for Model 2 in the Central, the other models consistently reject the null hypothesis of the random-effect model. Therefore, the static fixed-effect model is adopted for panel regression (although Model 2 in Central did not pass the Hausman Test, a fixed-effect regression was still performed on it for comparison). And robust cluster regression is used to avoid heteroscedasticity and cross-sectional dependence in the model.

In Model 1, we do not consider the impact of deposit and add income and housing completion area as controlled variables to examine the impact of demographic factors on housing consumption. Then, the deposit is added to test the impact of deposit on housing consumption and to make comparison with Model 3. Finally, in Model 3, we add the interaction items between age structure and deposit to test whether the partial effect of age structure on housing consumption exists.

The estimated results of Model 1, Model 2, and Model 3 are shown in Table 5. It displays that R² of all models reaches 85% and above, indicating the models explain the dependent variable very well, and the conclusions of this paper are highly reliable. Based on the estimated results, the empirical analysis is as follows:(1)Income has a positive effect on housing consumption in the Nationwide, East, and Central. In the West, only Model 1 shows a significant positive impact, reflecting that the influence of income on changes in housing consumption is less, and the impact of demographic factors on housing consumption is more significant. Comparatively, the impact of income in the Central is greater than that in the East, indicating that the increment of housing consumption caused by equal increase of income is larger in the Central than in the East. This can be explained from the perspective of housing affordability. In 2017, the housing price-to-income ratio in the East was 8.8, compared with 6.9 in Central (the data are calculated by the author based on data from the Statistical Yearbook of China 2018). The East has higher housing prices and less income elasticity of housing consumption, and then, the same increase of income causes less increment of housing consumption.(2)In Model 2 of the Nationwide and West, there is a positive impact of deposit on housing consumption. After adding the interaction items between deposit and age structure, the effect of interaction terms on housing consumption is significant, but the impact of deposit turns to be not significant. This shows that the impact of deposit on housing consumption is mainly realized through age structure.(3)Among all models, the completed area of commercial housing has a significant positive impact on housing consumption. It means that with the housing supply increase in China, more housing needs are met, which increases housing consumption. The increase range is higher in the West than in the Central, and the increase range in the Central is higher than in the East.(4)The total population has a negative impact on housing consumption Nationwide and in the Central, which is quite unexpected but reasonable. The dependent variable in the paper is essentially the annual housing consumption. After the rapid development of China’s real estate in the past two decades, the per capita housing area of urban residents in 2017 increased to 36.9 m². As the housing needs of residents are gradually met, the stock of housing demand is decreasing year by year. Furthermore, China’s population growth slows down, the number of new populations decrease every year, and then the new demand for housing decreases. The final result is that although total population increases, the housing consumption declines. This reflects that the total population is not the main factor determining housing consumption in China and improvement demands are the future of housing market.(5)Urbanization has a significant positive impact on housing consumption in the Central and West, as well as Nationwide. In the process of urbanization, a large number of rural people have entered the towns, resulting in the continuous expansion of the urban population and the increasing living demand and investment demand for urban housing. However, this effect is not significant in the East, in line with Chen et al. [4]. The stunning level of economic development in the East has attracted a large amount of floating population, which is included in the resident urban population (the urbanization rate in this paper is obtained from the China Statistical Yearbook, which is the proportion of resident urban population to the total population, not the urban registered population). While in the East, restrictions on the housing purchase of floating population are greater, curbing the housing consumption of floating population. In addition, many eastern provinces already have high urbanization levels in an earlier period. It would be difficult for ordinary rural-urban migrants to obtain local urban household registration status [4]. Some developed provinces, such as Beijing and Shanghai, set a high threshold for local urban household registration [38], further reducing the role of urbanization in promoting housing consumption.(6)ODR has a positive direct impact on housing consumption in the West. According to the Life Cycle Theory, the increase in ODR would reduce the housing consumption. The possible reason for the opposite situation is that China is still in the early stage of ageing, and the pressure to support the elderly is not serious enough to compress the housing consumption of younger generation. Furthermore, helping children purchase houses with “altruism” increases the elderly population’s housing consumption [39]. But the further ageing will curb housing consumption, and the turning point is when the average pension burden of the working-age population equals to the house purchase support of the elderly population. Perhaps, the East and Central are at this point, ODR has no significant impact on housing consumption. In the West, ODR has a negative impact on housing consumption through deposits. The higher the deposit of the elderly, the less they support purchasing their children’s houses, and the promotion of ODR to housing consumption will reduce. If the deposit (after adjusting by Consume Price Index) reaches 1187.38 (=e^^{31.5801/3.366}/10) billion yuan, ODR will affect housing consumption negatively. At present, only Sichuan Province in the West has exceeded this threshold, and the ODR of the other 11 provinces in West is positively affecting housing consumption. Nationally, ODR has a direct positive impact on housing consumption and has no significant indirect impact.(7)CDR has a significant negative impact on housing consumption in the East and West and Nationwide. When CDR increases, the reduction caused by aggravated pressure to raise children is more than the increasing demand from newborns. With the decrease of CDR in China, the housing consumption will continue to increase. After adding the interaction item between CDR and deposit, results show that in the East and Nationwide, the direct impact of CDR on housing consumption is more negative and CDR affects housing consumption through deposit positively. Only after deposit breaking through the bottleneck can the housing consumption of newborns be released. By then, CDR will have a positive impact on housing consumption. In the East, if the deposit (after adjusting by Consume Price Index) reaches 3242.73 (=e^^{23.5904/2.2712}/10) billion yuan, CDR will affect housing consumption positively. At present, only Guangdong Province in the eastern region has exceeded this threshold, but the deposits in Jiangsu, Zhejiang, and Shandong are approaching. Perhaps, because the deposits in the West are not high, the indirect effect of CDR through deposit is not significant.(8)In the Central, family size has a significant negative impact on housing consumption, and the national trend is basically in line with the Central. With the shrinking of family size and the increasing number of households in China, the demand for housing will continue to increase.(9)All models show that the education level has a significant negative impact on housing consumption. This is inconsistent with Eichholtz and Lindenthal [2]. According to Chen and Zhang [31], talent gathering would bring housing prices up, which may curb housing consumption. However, housing prices have no significant impact on housing consumption in the result of this research. There are other reasons for the phenomenon that improvement in education level curbs housing consumption, like education spending squeezing housing consumption [30]. With the popularization of the concept of “Knowledge Change Fate” in China, more people choose to receive a higher education; what followed is an increase in social competitive pressures. The increasing pressure will lead to a surge in education investment in the family and even squeeze housing consumption. The crowding-out effects is greatest in the West, second in the Central, and smallest in the East. One interpretation is that the income in Central and West regions is lower, and the education consumption accounts for a larger proportion, so the crowding-out effect is greater.

5.2. Robustness Test

Next, the paper examines the robustness of the effects of demographic changes on housing consumption. In response to the global financial crisis, China launched a 4 trillion investment plan and a series of policies to stimulate domestic demand in November 2008, mainly by building a large amount of infrastructure, including railways, highways, and others. The policy drove China’s economic growth and caused a great shock to real estate market, resulting in great demand for real estate and rapid growth in housing prices. This paper uses this historical policy to test the robustness and effectiveness of the model and introduces the year 2009 as a dummy variable into Model 3, expanding Model 3 to Model 4. The estimated results are shown in Table 6. It shows that there is no large deviation between Model 3 and Model 4. The investment plan stimulated housing consumption at that time and had a significant positive impact on housing consumption, which was consistent with the fact. The degree of influence was deeper in the Central and West than in the East.

6. Conclusion

Like other countries, demographic variables affect housing consumption from several aspects in China, such as population size, age structure, spatial structure, education level, and family size. At present, China’s annual housing consumption is decreasing. With the rapid development of the housing supply, China’s housing demand is gradually being met, and the impact of population size on demand of new houses is weakened. What really affects the housing consumption is the changing housing demand brought about by the other demographic transition. Among them, child-age dependency ratio (CDR), education level, and family size negatively affect housing consumption, while urbanization rate and old-age dependency ratio (ODR) have positive effects.

Comparing the results of different regions, there do exist significant differences among the eastern, central, and western China (Table 7). The level of development in the East is much higher than that in the Central and West, and the differences between the latter two are not obvious to some extent. The stunning level of economic development in the East has attracted a large number of floating populations, which are included in the resident urban population. However, the stricter restrictions on purchases of them has curbed the housing consumption of the floating population and reduced the role of urbanization rate in promoting housing consumption. Therefore, the positive impact of urbanization on housing consumption is significant in the Central and West, not in the East. Moreover, the negative impact of educational level on housing consumption is greater in the Central and West, followed by the East. This is mainly because education expenditures crowd out housing consumption. And in China, the difference in education expenditure is small, so the East with higher income has less impact.

Besides, inspired by Green and Hendershottb [14], this paper introduces the interaction items between age structure and deposit to analyze the partial effects of age structure. Results show that in the East, although the overall impact of CDR on housing consumption is negative at present, it can positively affect housing consumption through deposit. As CDR increases, increased support costs will crowd out housing consumption, but the household’s demand for housing space increases at the same time. When deposit reaches a high level, the housing demand of newborns will be released. In the East, if the deposit reaches 3242.73 billion yuan, CDR will affect housing consumption positively. Likewise, the interaction item between ODR and deposit has a significant negative impact on housing consumption in the West. Older people’s support for their children’s house purchases can increase housing consumption [39], but their preference for pension savings may weaken this help. When the deposit in the West exceeds 1187.38 billion yuan, the overall impact of ODR on housing consumption will become negative.

Using the 4 trillion investment plan in 2008 to conduct a robustness test, the results not only prove the robustness of the model but also confirm the feasibility of adjusting housing consumption through policies. At present, the growth rate of housing consumption in China is slowing down. For Chinese policymakers, it is a good choice to adjust the real estate market according to the trends of demographic structures, so as to realize housing consumption and maintain the stable development of real estate market. According to the findings of this article, we can draw some interesting policy implications: (i) nationally, further liberalization of free education or an improvement in the quality of public education may reduce the education expenditure of households, thereby releasing housing consumption; (ii) adjusting household deposit by changing bank interest rates can alleviate the impact of changes in the age structure on the housing market; (iii) the harsh household registration system in the past severely restricted the housing demand of floating population, and liberalizing the household registration system gradually may help translate housing demand into housing consumption, especially in eastern China.

Data Availability

The data used in this paper were mainly derived from the China Statistical Yearbook and the People’s Bank of China, which can be found at http://www.stats.gov.cn/english/Statisticaldata/AnnualData/ and http://www.pbc.gov.cn/en/3688006/index.html, respectively.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

This research was supported by the Fundamental Research Funds for the Central Universities (no. 2019CDSKXYJSG0041) and the National Natural Science Foundation of China (no. 71740027).

References

E. Belsky, “Demographics, markets, and the future of housing demand,” Journal of Housing Research, vol. 18, no. 2, pp. 99–119, 2009.
View at: Google Scholar
P. Eichholtz and T. Lindenthal, “Demographics, human capital, and the demand for housing,” Journal of Housing Economics, vol. 26, pp. 19–32, 2014.
View at: Publisher Site | Google Scholar
P. Chris and F. Joe, “Demographic trends and changing housing systems in Northern Ireland,” Housing Studies, vol. 33, no. 8, pp. 1264–1285, 2018.
View at: Publisher Site | Google Scholar
J. Chen, F. Guo, and Y. Wu, “One decade of urban housing reform in China: urban housing price dynamics and the role of migration and urbanization, 1995–2005,” Habitat International, vol. 35, no. 1, pp. 1–8, 2011.
View at: Publisher Site | Google Scholar
C. Choi, H. Jung, and L. Su, “Population structure and housing prices: evidence from Chinese provincial panel data,” Emerging Markets Finance and Trade, vol. 55, no. 1, pp. 29–38, 2019.
View at: Publisher Site | Google Scholar
S. Lauf, D. Haase, R. Seppelt, and N. Schwarz, “Simulating demography and housing demand in an urban region under scenarios of growth and shrinkage,” Environment and Planning B: Planning and Design, vol. 39, no. 2, pp. 229–246, 2012.
View at: Publisher Site | Google Scholar
W. S. Thompson, “Population growth and housing demand,” The ANNALS of the American Academy of Political and Social Science, vol. 190, no. 1, pp. 131–137, 1937.
View at: Publisher Site | Google Scholar
N. A. Kuchay, M. S. Bhat, and N. Shafi, “Population growth, urban expansion and housing scenario in Srinagar City, JK, India,” Journal of Geography and Regional Planning, vol. 9, no. 1, pp. 1–11, 2016.
View at: Publisher Site | Google Scholar
B. O. Campbell, “Long swings in residential construction: the postwar experience,” The American Economic Review, vol. 53, no. 2, pp. 508–518, 1963.
View at: Google Scholar
R. A. Easterlin, “Long swings in United-States demographic and economic-growth-some findings on the historical pattern,” Demography, vol. 2, no. 1-2, pp. 490–507, 1965.
View at: Publisher Site | Google Scholar
G. Carliner, “Income elasticity of housing demand,” The Review of Economics and Statistics, vol. 55, no. 4, pp. 528–532, 1973.
View at: Publisher Site | Google Scholar
N. G. Mankiw and D. N. Weil, “The baby boom, the baby bust, and the housing market,” Regional Science and Urban Economics, vol. 19, no. 2, pp. 235–258, 1989.
View at: Publisher Site | Google Scholar
D. L. McFadden, “Demographics, the housing market, and the welfare of the elderly,” in Studies in the Economics of Ageing, pp. 225–288, University of Chicago Press, Chicago, IL, USA, 1994.
View at: Google Scholar
R. Green and P. H. Hendershott, “Age, housing demand, and real house prices,” Regional Science and Urban Economics, vol. 26, no. 5, pp. 465–480, 1996.
View at: Publisher Site | Google Scholar
J. Ermisch, “The demand for housing in britain and population ageing: microeconometric evidence,” Economica, vol. 63, no. 251, pp. 383–404, 1996.
View at: Publisher Site | Google Scholar
S. Holly and N. Jones, “House prices since the 1940s: cointegration, demography and asymmetries,” Economic Modelling, vol. 14, no. 4, pp. 549–565, 1997.
View at: Publisher Site | Google Scholar
D. Myers and S. Ryu, “Aging baby boomers and the generational housing bubble: foresight and mitigation of an epic transition,” Journal of the American Planning Association, vol. 74, no. 1, pp. 17–33, 2008.
View at: Publisher Site | Google Scholar
E. Levin, A. Montagnoli, and R. E. Wright, “Demographic change and the housing market: evidence from a comparison of Scotland and England,” Urban Studies, vol. 46, no. 1, pp. 27–43, 2009.
View at: Publisher Site | Google Scholar
N. Hiller and O. W. Lerbs, “Aging and urban house prices,” Regional Science and Urban Economics, vol. 60, pp. 276–291, 2016.
View at: Publisher Site | Google Scholar
B. Malmberg, “Low fertility and the housing market: evidence from Swedish regional data,” European Journal of Population/Revue européenne de Démographie, vol. 26, no. 2, pp. 229–244, 2010.
View at: Publisher Site | Google Scholar
F. Long and G. Wu, “The impact of urban population on real estate investment,” China Civil Engineering Journal, vol. 36, no. 9, pp. 65–70, 2003, in Chinese.
View at: Google Scholar
Y. G. Lee and E. Huh, “Consumption and saving behavior of older and younger baby boomers in Korea,” Early Childhood Education Journal, vol. 25, no. 4, pp. 507–526, 2004.
View at: Publisher Site | Google Scholar
A. Saiz, “Immigration and housing rents in American cities,” Journal of Urban Economics, vol. 61, no. 2, pp. 345–371, 2007.
View at: Publisher Site | Google Scholar
A. H. Akbari and Y. Aydede, “Effects of immigration on house prices in Canada,” Applied Economics, vol. 44, no. 13, pp. 1645–1658, 2012.
View at: Publisher Site | Google Scholar
L. Gonzalez and F. Ortega, “Immigration and housing booms: evidence from Spain,” Journal of Regional Science, vol. 53, no. 1, pp. 37–59, 2013.
View at: Publisher Site | Google Scholar
M. Lu, B. Chen, and H. Ou, “Rationality or Bubble? an empirical study on urbanization, migration and housing price,” The Journal of World Economy, vol. 36, no. 1, pp. 30–54, 2014, in Chinese.
View at: Google Scholar
Z. Wang and Q. Zhang, “Fundamental factors in the housing markets of China,” Journal of Housing Economics, vol. 25, pp. 53–61, 2014.
View at: Publisher Site | Google Scholar
L. Elíasson, “Icelandic boom and bust: immigration and the housing market,” Housing Studies, vol. 32, no. 1, pp. 35–59, 2017.
View at: Publisher Site | Google Scholar
Y. Lin, Z. Ma, K. Zhao, W. Hu, and J. Wei, “The impact of population migration on urban housing prices: evidence from China‘s major cities,” Sustainability, vol. 10, no. 9, 3169 pages, 2018.
View at: Publisher Site | Google Scholar
J. Zhou, “Uncertainty and housing tenure choice by household types: evidence from China,” China Economic Review, vol. 22, no. 3, pp. 408–427, 2011.
View at: Publisher Site | Google Scholar
B. Chen and C. Zhang, “Human capital and housing prices in Chinese cities,” Social Sciences in China, vol. 5, pp. 43–64, 2016, in Chinese.
View at: Google Scholar
C. Y. Horioka, “Tenure choice and housing demand in Japan,” Journal of Urban Economics, vol. 24, no. 3, pp. 289–309, 1988.
View at: Publisher Site | Google Scholar
G. Hugo, “Implications of demographic change for future housing demand in Australia,” Australian Planner, vol. 42, no. 2, pp. 33–41, 2005.
View at: Publisher Site | Google Scholar
Y. Chen and X. Chen, “The influence of population ageing on China’s urban housing demand,” Economic Theory and Business Management, vol. 5, pp. 45–58, 2013, in Chinese.
View at: Google Scholar
L. Wang, R. Wu, H. Liu et al., “Spatial patterns and regional differences of population ageing in China based on the county scale,” Progress in Geography, vol. 35, no. 8, pp. 921–931, 2016, in Chinese.
View at: Google Scholar
M. Li, B. He, R. Guo, Y. Li, Y. Chen, and Y. Fan, “Study on population distribution pattern at the county level of China,” Sustainability, vol. 10, no. 10, p. 3598, 2018.
View at: Publisher Site | Google Scholar
H. Jin, “The effect of demographic structure on housing demand in Chongqing,” in Proceedings of the ICCREM 2018, Charleston, SC, USA, August 2018.
View at: Publisher Site | Google Scholar
F.-L. Wang, “Reformed migration control and new targeted people: China’s hukou system in the 2000s,” The China Quarterly, vol. 177, pp. 115–132, 2004.
View at: Publisher Site | Google Scholar
D. N. Weil, “The saving of the elderly in micro and macro data,” The Quarterly Journal of Economics, vol. 109, no. 1, pp. 55–81, 1994.
View at: Publisher Site | Google Scholar

Copyright

Copyright © 2020 Yu Zhang et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

PDF Download Citation

Download other formats

Order printed copies

Views

8719

Downloads

1459

Citations