The effect of a real estate market boom on China’s real economy : a cross-city perspective

(1)

i

The effect of a real estate market boom on

China’s real economy- a cross-city

perspective

Master thesis MSc Business Economics- Real Estate Finance & Asset Management Finance University of Amsterdam

Hao Sun (11843144)

Supervisor: dhr. dr. M.I. (Martijn) Dröes

Abstract

This study investigates a real estate market booms’ impact on China’s real economy between 2006-2016, based on 35 Chinese main cities’ data from NBSC 1_{. To get more general results, we}

design two models to test both the direct effect and indirect effect. Our results indicate that house prices have no significant direct effect on GDP in the short run, while in a longer period (three years lagged) increasing house prices do have a negative effect on GDP growth. From the results of the direct effect model, we find that real estate investments and government expenditures are key determinants of Chinese GDP growth. The indirect model’s results show that house prices do have a negative effect on house transactions, while house transactions can stimulate real estate investment. From our findings, we can further conclude that comparing with Western markets, the Chinese real estate market is more relevant and important to the real economy, but also more distorted due to strong government intervention.

Keywords: real estate boom, house price, house transaction, real estate investment, GDP growth, government expenditure, China, main Chinese cities

(2)

ii

Statement of Originality

This document is written by Student Hao Sun, who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents

Acknowledgments

I would like to take a moment to thank everyone that helped me during the process of writing my thesis. I would like to thank my University supervisor Martijn I. Dröes, who helped me greatly through the whole process. Secondly, I would like to thank my supervisor at Rabobank Björn Giesbergen, who gave me a lot of support and steered me in the right direction during my research. His support has made me a stronger person, and I will forever be grateful. I would also like to thank my other colleagues at Rabobank; they were always willing to help me and offering valuable advice. I feel extremely lucky to have the opportunities to write my thesis at Rabobank and enjoyed every second of it. Finally, I must express my very profound gratitude to my parents for providing me with unfailing support and continuous encouragement through my years of study. This accomplishment would not have been possible without them. Thank you.

(3)

iii

1. Introduction

In 1979, after more than 20 years’ seclusion apart from the Western world, China started the great reform and opening-up policy to build an open and dynamic market. Benefitting from more foreign investment and the development and deepening of the domestic capital market, China has been an impressive performance. China is now the second largest economic entity regarding GDP (Gross Domestic Product) and even the first economy in PPP (purchasing power parity) terms in the world. During the past forty years, the Chinese government sustains pragmatism to keep the economy growing fast and left a bunch of contradictions and side effects such as a high total debt-to-GDP ratio (around 260% of GDP by end-2017), an environmental pollution problem, insufficient domestic demand, and the economic structural transition problem (from labor-intensive to technology-intensive).

Many researchers (Glaeser et al., 2017), found that in China’s case, construction activity has led to more economic progress. Especially for the residential real estate market, it has become a common view that stability of this marketAs such, the real estate market plays an essential role in China’s economic development and will continue to be one of the key determinants of China’s future economy. There are many discussions about the overestimation of China’s housing prices because of the world-top level house price in China’s big cities and the incredible high affordability index (price-to-income level) in China’s real estate market. In most Chinese cities, the local governments need to make economic plans based on the central government requirements and past behaviors. The local governments treat the plans as political tasks and will try their best to reach the goals. As an essential component of the economy, the real estate investment is often used by the local governments to achieve their GDP plans (Wu & Zhang, 2007). However, according to Xu Zhong2_{, the head of People’s Bank of China’s research department, the}

real estate is no longer suitable to be used as a macro regulatory tool. He emphasized that under China’s political and economic system, using real estate as a macro regulatory tool usually accompanies faster house price growth which will lead to a huge negative effect on the economy. He believes that there are several risks related to the excessive growth of

(5)

2 house prices, including: bringing some cities’ house prices far beyond the reasonable level, a net crowing-out effect on consumption and investment, moderating TFP (Total Factor Productivity), intensifying the income gap, an over-leveraged economy, and thus threatens the stability and healthiness of China’s economy. It is therefore important to know the relation behind a real estate boom and the real economy to find out whether an increase in housing prices is driving the economic growth or not.

In this research, we measure the real estate boom by house price: if the growth rate of a city’s house price is greater than 20% or two standard deviations of the city-specific distribution of house price growth in a given year, the period will be defined as a real estate boom period. Gross domestic product (GDP) will be used as the measure the economy and the inflation will be controled by the time fixed effect. Our research

question is what is the impact of a real estate boom on the real economic development.

As China is a giant country with more than 600 cities, a simple investigation on the national level will have relatively limited observations, and some key influential factors might be hidden in the country level. Considering this, we believe the main Tier-1, Tier-2, and Tier-3 cities3_{will be most representative to reflect the relationship between housing}

prices and GDP. The reason why we only look at the main cities is that the main cities are the most representative examples of China’s economic model and their economic data are relatively complete comparing with small cities. For the data source, we collect data from the national statistical office(NBSC). The advantage of NBSC’s city-level data is that the data is more accurate and have a similar statistical standard through the sample period, though the data is only available on an annual basis and the number of real estate market and economic indicators is limited.

One big question in this research is the causality between house prices and GDP. The common view is that real economic developments will drive the house price growth, but no light is shined on the impact of the house price on the real economy. To get valuable results, we need to keep an eye on the interrelationship between the house price and GDP and test both the direct effect and indirect effect. In the direct effect model, we test the

3 _{Based on Yicai Global, the Chinese city tier system,}_{Tier-1, Tier-2 and Tier-3 cities mean the bigger,}

medium and smaller cities, ranking by five dimensions: (1) concentration of commercial resources, (2) the extent to which a city serves as a commercial hub, (3) vitality of urban residents, (4) diversity of lifestyle, (5) future dynamism

(6)

3 direct relationship between house prices and GDP growth with a one period lag. While in the indirect model, we study the one period lagged indirect effect of house prices on GDP through separate house transaction and real estate investment channels.

The results of our direct effect model show that there is no significant lagged effect between house prices and GDP, while the real estate investment and government expenditure are most essential determinants of GDP growth. A one percent increase in real estate investment increases GDP by 0.062 percent, while a one percent increase in the government expenditure will increase GDP by 0.116 percent. There are strong momentum effects in GDP and house price. GDP growth in one year has an effect of 0.155 percent on next year’s GDP, while for house price the autoregressive effect is -0.15 percent.

Since no significant relationship is found between house prices and GDP growth. We design an indirect model to further test our hypotheses. The results of the indirect effect model show that house price, house transaction, real estate investment, and GDP have more significant relationships compared to the direct model in some cases. A one percent increase in house price will decrease the residential transaction by 0.24 percent while a one percent increase of residential transaction will stimulate the real estate investment to increase by 0.265 percent. A one percent increase in real estate investment will increase the GDP by 0.066 percent, which is similar to the result of the direct model. No inverse relationships are found in the indirect model. Based on the findings of the direct effect and indirect effect model, we further test the long run effect of house price on GDP and find there is a significant negative relationship, given that a one percent increase in house price will lead to 0.6 percent decrease in three-year-later GDP. Our results (see Table 1) indicate that the real estate market has a strong influence on China’s real economy. We find that more investments in the real estate market can stimulate the GDP growth in the short run, but that increasing house prices are harmful to the economy in the long run. Our research suggests that the housing booms in China aggravate the pressure of a future economic downturn.

The structure of this thesis is as follows. In the next chapter, we will briefly introduce the characteristics of the Chinese real estate market to give readers an overview of this market. Then we will present literature reviews which are relevant to our studies in

(7)

4 chapter 3. After that, we list our hypotheses with argumentation and present the methodology of our research in chapters 4 and 5. Data analysis and results will be displayed in the chapters thereafter. In the last part, we will give a conclusion and discuss the limitations of our research as well as suggestions for further study.

Table 1: Overview of Hypotheses results Hypothesis 1 The real estate investment growth

has a positive effect on GDP growth in the short run (less than one year).

Proved Significant at 99% level

Hypothesis 2 The real estate price growth has a negative impact on GDP growth in the long term (more than one year).

Proved Significant at 99% level

Hypothesis 3 The real estate boom has a negative relationship with the private consumption.

Not proved

No significant relationship

Hypothesis 4 The real estate boom will attract more investment in the real estate sector.

Not proved

Hypothesis 5 The real estate boom will increase the local government expenditure.

Not proved

Hypothesis 6 The real estate boom has a negative impact on the total export & import.

Not proved

2. Real estate market with Chinese characteristics

Since the global financial crisis in 2008, the economic growth of China is slowing, and the government is trying hard to avoid a hard landing of the economy by promoting industrial upgrading, stimulating domestic demand, and more significant monetary and fiscal stimulus. The most growing concern in optimizing the economic structure is the unprecedented housing market boom since 2008 and the potential bubbles in the Chinese real estate market. As the growth of the broader economy is slowing down in the past

(8)

5 years (comparing with more than 10% GDP growth around the 2000s), the real estate market is playing a more important and critical role regarding economic growth. From Figure 1 we can see that the real estate investment contribution to GDP has grown rapidly in China, from only 4% in 1997 to a peak point of 15% in 2015. After 2015, this number slightly fell due to stricter control of the central government, but the level is still relatively high. Comparing with more mature real estate markets like the United States (US) and United Kingdom (UK), the Chinese real estate market only has a short history of no more than 30 years Before 1978, property trading was not allowed in China. Only after 1988, the Chinese real estate market is officially commercialized, and people have not experienced a real recession in either economy or real estate.

Housing prices in most Chinese cities have increased by more than 10% per year since 2003 without any real downturn, while the stock market has experienced several large fluctuations and performed relatively bad in returns compared with real estate investment in recent years (see Figure 2). Giving the fact that there are few reliable alternative investment opportunities for Chinese investors (Zou, 2016) due to capital control and financial fraud, houses have become the most important asset and the first choice investment for Chinese households. It seems that homeowners will never lose money in the Chinese residential real estate market since the price never really falls.

From 2006-2016, the real estate developers built more than 20 billion square meters of new residential houses and the residential area per capita has reached 40.8 m2_{in 2016,}

which is no more than 18 m2_{in 2006}4_{. The statistical data shows that the supply of}

residential real estate should be sufficient to fill the increasing demand and keep the price at a reasonable growth level. However, housing prices continue to rise, and fewer younger households can afford large houses. Though most of the newly-built residential communities have incredible high vacancy rates, the developers would rather leave the houses empty than lower the prices. The developers will face pressure and protest from both their competitors and original house owners if they want to lower the price (李磊, 2009). Compared to the European housing markets where house prices and turnover are

4 _{Data resource: National Bureau of Statistics of China. Residential area per capita = Total residential area /}

(9)

6 highly correlated (Dröes and Francke 2016), the supply and demand mechanism seems to be an invalid adjustment for the house price in Chinese housing market.

Figure 1- Real Estate Investment/GDP and residential real estate investment/GDP (%)

Source: CEIC and National Bureau of Statistics

Figure 2: Chinese House Prices monthly growth vs. Stock Market index monthly return (%) Source: NBSC -10 -5 0 5 10 15 20 25 30 Jan -14 A p r-14 Ju l-14 O ct -14 Jan -15 A p r-15 Ju l-15 O ct -15 Jan -16 A p r-16 Ju l-16 O ct -16 Jan -17 A p r-17 Ju l-17 O ct -17 Jan -18 A p r-18 Sto ck exch an ge m o n th ly r etu rn (% ) H o u se p rice gro w th r at e (% )

Tier-1 cities house price growth Tier-2 cities house price growth Shanghai stock Exchange monthly return

(10)

7 Comparing with Western governments, the Chinese government has more frequent and stronger interventions over the real estate market. The local governments control and regulate the housing market through three main kinds of policies: purchasing limitations (e.g., only residents with local hukou can buy properties, each household can only buy one property), mortgage rate and minimum down payment (e.g., households with more than one property have to pay higher down payments). Currently, the central government is planning to introduce a property tax which is seen as crucial to cool house prices and tame the potential housing bubble.

3. Literature review

Most of the literature about the real estate market effect on the real economy focus on the components of GDP like consumption and investment based on datasets from the US, UK or Europe d. Studies about the Chinese real estate market mainly concern the possibility of housing bubble burst and the potential impact on China and the world economy. Due to the unique political system and market mechanism of China that both the economy and market are highly planned and controlled by the government, the result for Chinese market may be very different with the research covering Western real estate markets. Based on our models and conclusion, we will discuss the economic meaning behind the differences and try to present an original view of the Chinese market.

3.1 House price relationship with GDP and GDP components

Based on the UK real estate market, Attanasio et al. (2009) found that increasing housing prices and private consumption growth are tightly synchronized, and the house price has a stronger relationship with consumption for younger households than older ones. For the homeowners, the growing house price increases the value of their perceived wealth and collateral, while for non-owners, an increase in house price will lead to a reduction in their expected future wealth.

(11)

8 In Chen, Han, and Zhao’s (2009) studied of the Chinese real estate price volatility effect on the real economy, they found that a fast-increasing house price hurts the development of the real economy: a large volatility of housing prices will lead to weak domestic demand, unbalance investment portfolio and hurt the manufacturing industry. Similarly, Liu and Chen (2011) used macroeconomic data to test the real estate market volatility effect on three primary economic drivers: total consumption, total investment, and net export. They found that the increasing housing price has a negative impact on consumption and a positive effect on net export, while the impact on investment is not significant. Another research by Tang, Xu, and Ba (2010) also found a significant negative relationship between the house price and total sales of consumer goods. Based on a report of the European Central Bank (ECB, 2017) about China’s economic growth the Chinese real estate boom, solid household income growth, high saving rates, and limited alternative investment options have made real estate the most attractive asset and investment for households in comparison to bank deposits and the stock market.

Li and Chen (2014) did further research on the effect of housing prices on Chinese household consumption. They studied the wealth effect of housing assets and productive fixed assets based on household surveys. In this paper, they assumed that the housing assets are nonproductive and not used for investment purpose. They found that the household consumption is not affected by the rising house price, while the productive fixed assets have significant wealth effect. Their findings imply that the increasing house price does not make households wealthier. However, Li and Chen’s conclusions may have a limitation as more and more housing assets are used as investment goods rather than consumption goods, especially in main Chinese cities, while they only consider the housing assets for self-usage.

Phang (2004) studied the relationship between house price and aggregate consumption in Singapore case based on the life-cycle-permanent income framework of consumption5_.

The relevant studies of OECD countries show that changes in housing price are positively related to the aggregate consumption, while in Phang’s study, she found no evidence that increasing house price will positively affect aggregate consumption in Singapore.

5 _{Developed by Modigliani and Brumberg (1954) and Friedman (1957), a standard way to think about}

(12)

9 Moreover, she found the declines in expected house price growth have a significantly strong negative impact on the aggregate consumption.

Chen and Qiu (2011) built a Bewley model including lifecycle, housing price and housing demand to analyze the impact of rising house prices on the household saving rate and wealth inequality. They found that the growing house price will increase the investment demand of wealthy households for houses, which in turn will further increase the house price. In a certain period of the lifecycle, the young households’ saving rate significantly increase because they have to save more money to afford the down payment for buying houses. The saving rate of young household in that certain period is getting higher as the housing prices keep increasing. They also pointed out that the rising house price leaves fewer households being able to buy their properties and most of the mid-low income urban residents are suffering welfare loss. While in Zhao, Liang, and Li’s research (2013) based on CHIPs6_{data, they found some contrary results compared with}

Chen and Qiu’s results. Zhao et al. (2013) found that there is a negative relationship between the house price and the household saving rate. Their results show that no evidence is found to support the Hypothesis that home renters will save more to buy a house. For homeowners with more than one property, the rising housing prices will negatively affect the saving rate.

Huang, Wu, and Du (2008) did an empirical study based on the national level data testing the relationship between real estate market and macro-economy in China. They found that the relationship between house price and macro-economy is unbalanced in short run but stable in the long run. There is no significant relationship between interest rate and house price in the Chinese market. The historical house price and GDP are the main factors that influence GDP and house price, which mean there exists an interaction relationship between GDP and house prices. However, the real estate market is a regional market, the relationship between house price and macroeconomy in regional level is not covered in their paper.

3.2 Relationships within real estate indicators

(13)

10 Dröes and Francke (2016) used a panel vector autoregressive model to test what are the determinants of correlation between house prices and turnover rate in European markets. They found a strong feedback mechanism between prices and turnover. They also found that the momentum effect is an important factor that correlates prices and turnover. Case, Pollakowski, and Wachter (1997) used hedonic price model to study the impact of transaction frequencies on house price. They found that properties with higher transaction frequency also have higher house price. Wit, Englund, and Francke (2012) studied the correlation between house price and transaction volume in Dutch housing market. They found that the shocks of market fundamentals such as the mortgage rate have an immediate and significant effect on the transaction volume, but a gradual impact on the house prices.

He, Zhan, and Wan (2009) studied the herding behavior in Chinese commercial housing market. They found that there is serious herding behavior in Chinese real estate market and the herding behavior is one of the critical determinants that accelerate the change of house price and sales. Sun, Shen, and Zhao (2007) found a positive relationship between house price, house transaction, and investors’ belief. They believed that the positive feedback trading is a key factor of the continuing rising house prices. However, their research focuses on the Chinese real estate market before 2007, when the government intervention was not strong.

3.3 Chinese housing market

Real estate has become the most critical engine of China’s fast-growing GDP in the past decade. The Chinese real estate market is susceptible to both overbuilding and price imbalance problem (Chivakul et al., 2015). On the demand side, though the experts keep warning investors for the risk of overvaluation, housing is still the most attractive instrument given the fact of historical robust capital gains, capital account restrictions, low real interest rates, and most important, a lack of alternative investment opportunities (Ding et al., 2017). While on the supply side, as the local government financing highly relies on land sales, the local government’s tight control over land supply aggravates the distortion. Ding et al. (2017) believed the house price already exceed the fundamental level and the bubble is expanding to smaller cities. They warned that if the prices rise further beyond the fundamental levels, the possibility of a sharp and costly correction will

(14)

11 increase, which will undermine financial stability, weaken economic growth, and leave the government less room to correct the crisis.

Glaeser et al. (2017) compared the Chinese real estate market with the US real estate market, and in statistical level, they thought the Chinese real estate market is a classic housing bubble in many respects: the sharply increasing house prices, enormous new supply, and pervasively high vacancy rate. They believed the key factor to avoid the burst of the bubble is controlling the house price by muting supply. However, it would be a hard decision for policymakers as selling construction land is an essential income for local governments and reduce constructions will cause a series of social problems like a higher unemployment rate (张雷, 2012). The government’s decisions on the housing market will be an essential determinant of whether the housing crash is inevitable.

4. Hypotheses

In this chapter, we will list the hypotheses for our research and briefly explain our reasons and logic based on previous studies and Chinese characteristics. The hypotheses are as follows.

Hypothesis 1: The real estate investment growth has a positive effect on GDP growth in the short run (less than one year).

Assuming this, more investment in real estate market can stimulate the economic growth by stimulating the related industries and total consumption. The boom of real estate market can drive the development of both upstream and downstream industries like steel, construction materials, machinery, chemical industry, ceramics, textile, home appliances. In the short run, the growth of real estate price should have a positive effect on GDP growth.

Hypothesis 2: The real estate price growth has a negative impact on GDP growth in the long term (more than one year).

(15)

12 Based on the supply and demand model, when real estate price increases, the demand for construction materials will exceed the supply. The price of real estate related materials will increase due to the imbalance between demand and supply and increase the construction cost (X. Chen, H. Han, and Y. Zhao, 2009), which will discourage real estate investment. In the long run, the growing house price may have an inverse impact on GDP growth.

Hypothesis 3: The real estate boom has a negative relationship with the private consumption.

In China society, the decision to buy a house is always related to the whole families. For young people who are going to buy their first property, their parents, sometimes even grandparents will raise money to help them buy houses. The phenomenon is especially common for rural-urban migrants in China. Being a homeowner is virtually the most necessary thing from a social point of view and tightly related to the reputation and “face” of the whole family7_{. As a result, the higher housing price put pressure on the whole}

society and hurts people’s purchase power (Liu and Chen, 2011).

Hypothesis 4: The real estate boom will attract more investment in the real estate sector.

More Chinese investors are willing to invest in the real estate market seeing the continuous boom in the housing market, while the downturn in business investment and sharp volatility in stock markets make investors lose their patience and confidence in the other industries. However, C. Liu and C. Chen’s study in 2011 did not find significant evidence for this. Considering the high developing sector before 2010 and lagged effect of real estate market, we may have new findings using the most recent data.

Hypothesis 5: The real estate boom will increase the local government expenditure.

According to a report in The Wall Street Journal (Fong, 2017), due to the overbuilding problem, the unusual high vacancy rate is threatening the stabilization of the real estate

7 _{THE HOUSEHOLD AS SECURITY: STRATEGIES OF RURAL-URBAN MIGRANTS IN CHINA}

(16)

13 market. To stabilize the housing market, local governments have become the biggest buyers in real estate market. In Fong’s report, she points out that in 2016, more than 18% of all the residential space sold in China was purchased by local governments or with government subsidies. We believe the increasing house price will push up the local government expenditures.

Hypothesis 6: The real estate boom has a negative impact on the total export & import.

In first-tier cities and some second-tier cities, the housing price-to-income ratio exceeds 10, which means owning a property will cost more than 30 years’ income. The increasing house price does people not only have to work harder but also have stronger incentive to seek higher-paid opportunities. More employees skip from job to job and employers must increase salaries to keep their workers. The labor cost of the Chinese market is rising fast during the past decade. A direct consequence of higher labor cost is the increasing price of Chinese products, which will lower China’s price competitiveness and therefore harm exports.

5. Methodology

In this chapter, we will show the methodology for our research. To give more general results for this research, we design two models: one directly studies the relationship between real estate indicators and GDP in first difference with a panel AR(1) model. The other one studies the relationships between the real estate indicators and checks if there is longer lagged effect from real estate market indicators to GDP growth. Both models use the municipality economic data collected from the national bureau statistic of China.

5.1 Direct effect model

To deal with the potential causality problem, we will use the following reduced form of bivariate panel vector autoregressive, PVAR (1) model with house price and economic indicators (consumption, investment, and net export) as dependent variables:

(17)

14 Equation (1): [ 𝛥𝑙𝑛(𝑌𝑖𝑡) 𝛥𝑙𝑛(𝐻𝑜𝑢𝑠𝑒 𝑝𝑟𝑖𝑐𝑒𝑖𝑡)] =[ 𝛾1 𝛼1 𝛾2 𝛼2] [ 𝛥𝑙𝑛(𝑌𝑖𝑡−1) 𝛥𝑙𝑛(𝐻𝑜𝑢𝑠𝑒 𝑝𝑟𝑖𝑐𝑒𝑖𝑡−1)]+ [ λ1 λ2] Δln(X it-1) + [ β1,𝑖 β2,𝑖] + [ τ1,𝑡 τ2,𝑡 ] + [ ε1,it ε2,it ]

Where 𝑌𝑖𝑡 is GDP, i is city, and t is time. The 𝛾1and 𝛼2 are coefficients that capture the

autoregressive components in house price and real economic indicator GDP. The coefficients 𝛾2 and 𝛼1 show the lagged feedback between house price and GDP. The β𝑖

and τ𝑡 are the fixed effects of city and time. Due to the unique planned partial market

economic system of China, the macroeconomic indicators, especially GDP growth, are highly affected by the government plans. The Chinese real estate market is characterized by strong government intervention. They control and regulate the housing market through three main policies: purchasing limitations (e.g., each household can only buy one property in some cities), mortgage rate and minimum down payments. Always, both the announcement and implementation of government policies affect the transactions and prices. By including fixed effects (year dummies and city dummies), we can control the average differences across years and cities on any observable and unobservable predictors.

Xit-1 are other real estate indicators and macroeconomic variables including real estate

investment, transaction volume, population, total household savings, government expenditures, total import & export, and total social consumption in retail. To avoid the potential endogenous problem, we also use one period lagged Xit in Equation (1).

We will estimate several expanding versions of the model: First, based on model (1), we will replace Y (GDP) with private consumption, private savings, government expenditures, and total import & export respectively to test the relationships between house price and GDP components. The use of these independent variables will be slightly different based on the dependent variable Y. For example, the disposable income is an important factor for private consumption and private savings. Second, we estimate a more parsimonious model that only includes Tier-1 cities for a robustness check. Third, we estimate the model for the period after 2012, the year when Xi Jinping became China’s president, for robustness check as well.

(18)

15

5.2 Indirect effect model

Furthermore, we also want to know the relationship between the real estate indicators and test whether there is a longer lagged effect between house price growth and GDP growth. The fundamental logic behind this is as follows (C. Liu and C. Chen, 2011):

1) House price growth will hold-up the transaction growth through lower affordability and more strict policies from the government, especially in case of excessive housing price increases.

2) Higher or lower growth in housing transactions will stimulate the investment in real estate market.

3) Real estate investment growth will stimulate the GDP growth.

The models based on the level-relationships will be:

Equation (2): ∆ln(House transactioni,t)= α1∆ln(House Pricei,t-1) + α2 ∆ln(House

Transactioni,t-1)+ α3∆ln(Household Savingsi,t-1)+ α4∆ln(New Residential Stocki,t-1)+

α5∆ln(Incomei,t-1)+ α6t + α7i + εi,t

Equation (3): ∆ln(Real Estate Investmenti,t)= ß1∆ln(Real Estate Investmenti,t-1) +

ß2∆ln(House Transaction i,t-1) + ß3∆ln(New Real Estate Stock i,t-1)+ ß4 ∆ln(Household

Savingsi,t-1) + ß5t + ß6i +ε’i,t

Equation (4): ∆lnGDPi,t = γ1∆ln(GDPi,t-1) + γ2∆ln(Real Estate Investmenti,t-1) + γ3

∆ln(Household Savingsi,t-1)+ γ4∆ln(Government Expenditurei,t-1) + γ5 ∆ln(Populationi,t-1)+

γ6 ∆ln(Import & Exporti,t-1)+ γ7∆ln(Consumption Retaili,t-1)+γ8t + γ9i + ε’’i,t

We will do several expanding versions for the indirect effect model. First, we check whether there is any inverse causality within real estate indicators. Second, we add more

House

price

House

transaction

Real estate

investment

GDP

(19)

16 lagged period for variable investment in real estate and house price to check the longer-term effect of real estate market on the real economy.

6. Data

As mentioned in the introduction part, due to the limitation of the incomplete statistical system, our research will mainly focus on the Tier-1, Tier-2, and Tier-3 cities. Also, these main cities are more representative and close-related to the real Chinese economy. For the data source, we have two choices: collecting data from each cites’ statistical office or getting data from the national statistical office. The advantage of collecting data from each cites’ statistical office is that they have the detailed quarterly data. However, the correctness of these data may be a serious problem. While for collecting city-level data from national statistical office, though the data is more accurate and have the same standard through the chosen period, the data is only available in annual level and the number of indicators is limited. After collecting data from both sources, we find that although the quarterly data from city-level statistic offices are ideal for our model because of more detailed indicators, the quality of this data is a big problem8_{. The}

statistical standards are not the same across cities. For example, the currency used for economic data is not consistent through the sample period, and the statistical methods for some macroeconomic indicators like consumption and investment changed in 2014. The reliability of the data is another concern for the data from city statistical offices. For example, year-on-year growth rates shown in the quarterly reports, are not always equal to the growth rates calculated using real data. For instance, in 2010, the import & export data of Shenzhen suddenly jumped by more than ten times compared with 2009, but the quarterly report only shows an increase of 10 percent. Considering the reasons above, we decide that data from city-level statistical office is not appropriate for our research. For the reasons mentioned above, we will use city-specific data from the national statistical office for our research.

6.1 General data description

8 _{For quarterly data from city-level statistic office, we choose 10 cities, including Beijing, Shanghai,}

(20)

17 The variables include real estate investment, import & export, housing prices, housing transaction volumes, GDP, population, private consumption, disposable income, and government expenditures. The city-specific difference may include geography, custom, history, and weather. The cause of yearly fixed effect can be both internal trend (like policy events) and external economic environment.

Table 2 – House price, GDP, and other macroeconomic variables (level)

Variable Mean Std.Dev. Min Max

GDP 5434 4934 281.6 28179

Investment in Real estate 882.6 803.4 29.38 4177

Commercial real estate transaction 1120 891.6 106.4 6257

Residential real estate transaction 979.2 778.3 98.44 5105

Commercial real estate price (per m2) 7371 4831 2022 45146

Residential real estate price (per m2) 7146 4936 1940 45498

Total household savings 4108 4377 88.65 28012

Total retail consumption 2221 1987 108.7 11005

Population 714.2 546.7 144.7 3392

Average income 47679 19136 16911 122749

Total Import&Export 54600 100831 443.9 537359

Number of higher education students 40.37 23.20 3.598 105.7

Local government revenue 635.0 867.5 14.16 6406

Local government expenditure 795.1 1022 32.25 6919

New Residential stock 710.5 567.6 45.57 3386

Total new real estate stock 956.0 786.8 60.95 4630

Sample period 2006~2016 (11 years)

Number of cities 35

Number of observations 385

Notes: The sample includes 35 main Chinese cities, including 17 tier-1 cities, 15 tier-2 cities, and three tier-3 cities.

There are 35 cities9 _{included in the dataset, and most of the 35 cities are the capital city of}

their provinces. Consistent data about investment and separated data for import and export are missing due to statistic data changing. Considering the different statistical standards and recording currency in each cities’ database, we only use the available data from the national office which record the period from 2006 to 2016. The newest data for 2017 is not available for all 35 cities in the national statistical database due to the data modification problem. Moreover, if more than three years of housing market data are missing, such as Lasa, we exclude the city from the sample. The strongly balanced

9 _{Beijing, Changchun, Chengdu, Chongqing, Dalian, Fuzhou, Guangzhou, Guiyang, Haikou, Hangzhou,}

Harbin, Hefei, Hohhot, Jinan, Kunming, Lanzhou, Nanchang, Nanjing, Ningbo, Qingdao, Shanghai,

Shenyang, Shenzhen, Shijiazhuang, Taiyuan, Tianjin, Urumchi, Wuhan, Xi’an, Xiamen, Xining, Yinchuan and Zhengzhou.

(21)

18 dataset of 35 cities for 11 years provides us with sufficient observations to analyze the relationship between real economy and house price. Table 2 reports the descriptive statistics of real estate variables and macroeconomic variables in level, while Table 3 shows the descriptive statistics in (log) differences. Note that we also include the data about commercial real estate for the robustness check and compare which real estate market indicator has stronger prediction power. More important, we must be aware that Chinese market is a unique emerging market with a state-directed development model.

Table 3 – House price, GDP, and other macroeconomic variables (Differences)

Variable Mean Std.Dev. Min Max

Δ log GDP 0.129 0.0612 -0.287 0.315

Δ log Investment in Real estate 0.166 0.152 -0.634 0.671

Δ log Residential real estate transaction 0.0709 0.268 -0.900 0.983

Δ log Residential real estate price 0.100 0.106 -0.195 0.463

Δ log Total household savings 0.127 0.187 -2.087 2.533

Δ log Total retail consumption 0.146 0.0441 0 0.301

Δ log Total Import&Export 0.107 0.285 -0.866 1.967

Δ log Commercial real estate transaction 0.0782 0.245 -0.652 0.686

Δ log Commercial real estate price 0.0972 0.109 -0.451 0.449

Δ log New Residential stock 0.0356 0.318 -0.905 1.612

Δ log Total new real estate stock 0.0532 0.299 -0.920 1.618

Sample period 2006~2016 (11 years)

Number of cities 35

Number of observations 350

Notes: The sample includes 35 main Chinese cities, including seventeen tier-1 cities, fifteen tier-2 cities, and three tier-3 cities.

The economy is not only affected by the market but also driven by the government plans and policies. The potential revision of the economic data might lead to a biased result. To enhance the critical power of our analysis, we will also discuss whether our results make sense from an economic point of view, based on China’s situation.

6.2 GDP growth and House price growth

Table 4 shows several important results about the GDP and house price growth in Tier-1, Tier-2, and Tier-3 cities. The average GDP growth rate in Tier-1 cities (12.60%) is smaller than in Tier-2 and Tier-3 cities (13.14%), while the house price is rising faster in Tier-1 cities (10.19%). The number of average house boom episodes during the sample period is also higher in Tier-1 one cities, with on average 2.35 house booms in 11 years.

(22)

19 In Shenzhen, the number of boom periods is as high as 5, which leads to a much faster house price growing than its GDP growth. There are six cities (Beijing, Shanghai, Shenzhen, Nanjing, Shijiazhuang, and Xiamen) which have higher house price growth rate than GDP growth rate in our sample.

Figure 3-The average annual growth of GDP and house price in Tier-1 and Tier-2 cities

Source: NBSC

From Figure 3 we can see that the average GDP growth rates in Tier-1 and Tier-2 cities have a similar trend and Tier-2 cities have higher GDP growth than Tier-1 cities after 2008. The house price growth rate is more fluctuate in Tier-1 cities. In 2008, when the global financial crisis broke up, the house price growth rate in Tier-1 cities has a more significant crash than in Tier-2 cities. Similarly, when the central government implements the strictest policy to control the residential real estate price, the Tier-1 cities’ house price growth rate suffers more lose than Tier-2 cities. Tier-2 cites’ house prices are less sensitive to shocks than Tier-1 cities. We also find that the GDP growth has a similar lagged trend with house price growth. The increase or decrease of house price growth will lead to an increase or decrease of one-year-later GDP growth.

0 0.05 0.1 0.15 0.2 0.25 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

(23)

20 Table 4 – GDP growth, House price growth, and number of house booms in Tier-1 and Tier-2 cities

Tier-1 cities Tier-2 and Tier-3 cities

City house booms GDP growth House price growth City house booms GDP growth House price growth Beijing 4 0.1151 0.1351 Changchun 2 0.1223 0.0916 Changsha 3 0.1659 0.0930 Fuzhou 4 0.1315 0.1023 Chengdu 2 0.1487 0.0746 Guiyang 3 0.1656 0.0925 Chongqing 3 0.1513 0.0908 Haikou 3 0.1279 0.1199 Dalian 1 0.0963 0.0762 Harbin 1 0.1069 0.0929 Guangzho u 1 0.1169 0.0977 Hefei 2 0.1765 0.1176 Hangzhou 3 0.1190 0.0999 Jinan 1 0.1096 0.0929 Nanjing 3 0.1331 0.1432 Kunming 1 0.1270 0.0919 Ningbo 2 0.1106 0.0833 Lanzhou 1 0.1266 0.0896 Qingdao 1 0.1139 0.0810 Nanchang 2 0.1303 0.0926 Shanghai 3 0.0980 0.1303 Nanning 2 0.1448 0.0935 Shenyang 1 0.0773 0.0764 Shijiazhuan g 2 0.1073 0.1300 Shenzhen 5 0.1210 0.1637 Taiyuan 2 0.1070 0.0845 Tianjin 2 0.1388 0.1018 Urumchi 2 0.1324 0.1059 Wuhan 2 0.1526 0.1022 Xiamen 4 0.1176 0.1342 Xi'an 1 0.1446 0.0731 Hohhot 1 0.1260 0.0871 Zhengzhou 3 0.1383 0.1101 Xining 2 0.1489 0.0948 Yinchuan 1 0.1574 0.0711 Average 2.3529 0.1260 0.1019 Average 2 0.1314 0.0992 Sd. 1.1695 0.0235 0.0267 Sd. 0.9701 0.0203 0.0164

Note: the house boom episodes are defined as deviations from a regional specific standard. We classify an episode as a boom if the growth rate of house price is greater than 20% or two standard deviations of the city-specific distribution of house price growth in a given year.

6.3 Other real estate indicators

We have four real estate market indicators in our dataset, including real estate investment, real estate transaction volume, average real estate price per m2_{and new increased real}

estate stock. We will mainly focus on the real estate price, but also control the other three indicators to increase the precision of our model. In the research for the Western real estate market, strong relationships are found within these real estate indicators (Tsatsaronis & Zhu, 2004). We will also test whether the similar relationships can be found in the Chinese real estate market and try to exclude the interaction effect in our model.

(24)

21

6.4 Other macroeconomic indicators

There are several other macroeconomic factors that we use in the model as exogenous control variables. In general, economic concept, GDP is composed of consumption, investment, net export, and government expenditure. However, due to the limitation of Chinese data, we cannot find the same indicators for these four components. In our dataset, we have total private consumption in retail, total household savings, total import and export, and government expenditures. We use these four variables as control variables because they are closely related to the four components of GDP.

6.5 Stationarity

Table 5-Stationary: Panel Unit Root Tests

Level Difference Variable Inverse chi-squared p-value Inverse chi-squared p-value log GDP 6.739 1.000 158.911 0.000

log Real estate investment 77.511 0.252 125.659 0.000

log Residential transaction 133.626 0.000 436.874 0.000

log Residential price 25.257 1.000 145.595 0.000

log Total household savings 84.008 0.121 354.220 0.000

log Consumption in retail 7.869 1.000 122.077 0.000

log Export&Import 82.281 0.150 209.444 0.000

log New residential stock 106.292 0.003 277.785 0.000

log Government expenditure 29.139 1.000 167.290 0.000

Sample period 2006-2016 (11 years)

Number of cities 35

Null Hypothesis: All panels unit root. Alter: At least one panel stationary

Notes: Fisher type of test is based on the Dickey-Fuller test. All the tests on the level and difference are based on the logarithmic values to test cross-sectional dependence and include one lag, to test serial correlation, and a trend.

Table 5 suggests that most log variables are non-stationary except for residential transaction and new residential stock. While in first difference, all variables are stationary when including a trend. Based on the stationary test, we decided to use all variables in the first difference in the regression analysis.

(25)

22

7. Results

This section describes the empirical results found in our research. In section 7.1, we will explain the results of the direct model (Equation 1). Then, sections 7.2 shows the results for the expanded direct model, relationships between house price and GDP components (government expenditure, total household savings, total consumption in retail, and total import & export). After this, section 7.3 elaborates on the results for the indirect model (Equation 2, 3, and 4), the relationship between real estate indicators and longer lagged effect between house price and GDP growth.

7.1 The direct model

Table 6 reports the regression results on direct effect model based on Equation (1). Column 1 contains the results when the full sample is included in the regression equations. The results in Column 1 show that the growth of house price and house transaction are not key determinants of GDP growth while the real estate investment has a significantly positive effect on GDP growth. A one percent increase in real estate investment will lead to a rise in GDP by 0.0624 percent. This can be explained by the fact that higher investment in the real estate market can drive the development of both upstream and downstream industries which will stimulate GDP growth. For house transactions and price growth, their effects on related industries are smaller and smoother in the short run. As a result, they have an insignificant impact on GDP growth. Government expenditure is another key determinant of GDP growth, for which a one percent increase in government expenditure will lead to a rise of GDP by 0.116 percent. For determinants of house price, the results show that the growth of house price is not affected by the lagged GDP or transaction growth. This is very different from the research in Western real estate markets. However, since the Chinese housing market is not as transparent as the Western markets and the capricious government policies distorted the demand-supply mechanism, the house transaction volume fails to explore the real estate market information whether the demand and supply are balanced. We can also see that previous growth in house prices growth has an inverse impact on the growth of current house prices. This might be the case because a higher house price will lead to lower affordability and stricter control of the government which will restrain house price growing.

(26)

23 Population growth is one of the key motivations for the growth of house prices. A one percent increase in population will lead to an increase in house prices by 0.306 percent. We also find that there are strong momentum effects in GDP and house prices. GDP growth in one year has an effect of 0.155 percent on next year’s GDP, which means it will take about seven years (1/0.155) for shocks in GDP to be fully absorbed. While for house price the autoregressive effect is -0.15 percent.

In Column 2, 3, and 4, we separate the data into only Tier-1 cities, only Tier2&3 cities, and only period after 2011 as a robustness check. The results show that in Tier-1 cities, the investment in real estate market has a larger effect on GDP growth than in Tier-2&3 cities. In Tier-1 city, a one percent increase in real estate investment will increase GDP by 0.132 percent, while in Tier-2&3 cities, the effect is only 0.024. Similarly, government expenditure and momentum effect of GDP are also stronger in Tier-1 cities. The momentum effect of house price is larger in Tier-2&3 cities, house price in one year has an effect of -21.9 percent on next year’s house price growth in Tier-2&3 cities. For the period after 2011(Column 4), when Xi Jinping became the president, the government expenditure’s role in stimulating the GDP growth is getting more important compared with the whole sample period.

It should be noted that the results of Equation (1) show that both population growth and total household savings have a negative effect on the GDP growth. The large population and high savings rate advantages used to benefit China’s economic development in the early stage. In the current phase, these two advantages are no longer the impetus for GDP growth in main cities, while the government expenditure and investment in real estate have become the primary drivers. The excessive population growth in big cities leads to a problem of labor oversupply, especially when growing real estate prices increase the producing cost and squeeze out the low revenue industries, which are usually labor intensive, from big cities. The imbalance of the labor market will cause relatively lower labor salaries growth and further decrease the consumption power. Inordinate population growth will engender resource shortage and productivity-inefficiency which are also harmful to the economic growth (Wang et al., 2009). The negative relationship between total household saving growth and GDP growth is also strange compared with the European experience. Since the Chinese saving banks are purely controlled by the central

(27)

24 T able 6 – T h e r el at ion shi p b et w ee n h ouse p ri ce an d GDP i n 3 5 m ai n C hin es e ci ti es (B ot h ti m e fi xed e ff ec t and ci ty fi xed ef fec t ar e inc lude d )

(28)

25 government, local governments and state-owned enterprises are more easily accessible for bank financing while for non-state-owned companies it is harder to get loans from saving banks (Xu et al., 2013). Due to the serious corruption and nontransparent problem of state-owned enterprises and local governments, the use of loans is inefficient and even wasted (Lin et al., 2016). The inverse effect of household savings on GDP can be partially explained by the ineffectual use. To conclude, Hypothesis 1 is proved.

7.2 The expanded direct model

Table 7 – House price effect on GDP components (both time fixed effect and city fixed effect are included)

Based on Equation (1), we expand the model by replacing dependent variable GDP with its components, namely government expenditure, total household savings, total private

(29)

26 retail consumption, and total import & export. The results are shown in Table 7. In Hypothesis 5, we assume that a real estate boom will increase the government expenditure. However, the results show that house transaction and real estate investment will increase the government expenditure, but the house price itself does not. For government expenditure, increasing the real estate investment by one percent will raise government expenditure by 0.088 percent. Real estate transaction also has a positive relationship with government expenditure, one percent more house transactions lead to a rise in government expenditure by 0.04 percent. For the other three GDP components, total household savings, total private retail consumption, and total import & export, we could not find any significant relationships with real estate indicators. As a result, Hypothesis 3, 5, and 6 are not proved in our expanded model.

7.3 The indirect model

Table 8 reports the regression results for Equation (2), (3), and (4). Column 1 shows that the increasing house price has a significant negative effect on house transaction growth. With a 1% higher residential price, the residential transaction will decrease by 0.525%. The effect of government expenditures on house transaction is positive; this can be partially explained by the government purchase and subsidy behavior on the residential real estate market. Column 2 shows the relationship between house transaction and real estate investment. An increase in house transaction of one percentage point increases real estate investment by 0.104 percent. We find no significant evidence to prove Hypothesis 4 that real estate boom attracts more investment in real estate market. This might be a result of the market bias due to strong government intervention. Column 3 is a parsimonious model for Equation (1) that is only based on real estate investment and macroeconomic indicators. The regression estimates are very similar. This validates that in the short run, the house price and house transaction are not the key determinates for GDP growth. In Column 4, we add more lagged period for variable investment in real estate and house price to check the longer-term effect of a real estate boom on GDP. We find that the impact of real estate investment is positive and diminishing over time. The three-year lagged house price growth has a negative effect on GDP growth – a one percent increase in house price decreases GDP growth by 0.06%three-year later. The increasing real estate price rises the production cost and labor cost and will squeeze out low-revenue industries from the main cities. The industry transfer from big cities to small

(30)

27 Table 8 – The relationship between real estate indicators and long-run effect of house price on GDP (both time fixed effect and city fixed effect are included)

cities usually take several years. Thus the negative effect of higher real estate price over GDP growth will not expose immediately. The linear fit of the model with three-period lagged house price and real estate investment is relatively good: the independent

(31)

28 variables can explain 71.5 percent of the variation in GDP. In Column 5 we do a robust check with only the period after 2011 when Xi become the president. It turns out that our primary results do not change. Hypothesis 2 that the real estate price growth has a negative impact on GDP growth in the long term (more than one year) is proved based on the results of Column 4 and 5.

To sum up, Hypothesis 1 and 2 are proved due to significant and robust results, while Hypothesis 3, 4, 5, and 6 are not proved since the results are not significant.

8. Conclusion

The real estate development in China has become an essential determinant of the Chinese economy in the past decade. Many papers have been focusing on studying whether there is a real estate bubble in the Chinese market, and what the potential impact could be on China’s economy in case a real estate bubble would burst. In our research, we regard the Chinese real estate market oppositely by looking at the effect of a real estate boom on China’s economy.

In the introduction part, we refer to Xu Zhong’s speech about the overuse of the real estate market as a macro regulatory tool. While his concern is more general and intuitive, we use an empirical method in our research to analyze the relationship between a real estate boom and China’s real economy. Our results suggest that real estate investments and government expenditures are the key determinants that drive GDP growth, while in Western markets, in general, the investment in real estate is not considered as a key factor for GDP growth. We also find that the house price growth does not significantly affect GDP growth in the short run (one year lagged), but in the longer term (three-year lagged), the growth of house price has a significant negative influence on GDP growth. Furthermore, compared with Western real estate markets, because of strong government intervention in China, the residential transaction volume in the Chinese real estate market could not expose the relationship between supply and demand and fails to explain the changes in house prices. We also find that the real estate boom does not have significant relationships with GDP components (government expenditure, total private consumption,

(32)

29 total household savings, and total import & export). This finding is different with major European studies but in line with Pang’s (2004) study based on Singaporean experience.

There are also several limitations to our research and some suggestions for further research. First, data reliability and availability are the biggest limitations of our research. The falsification of Chinese economic data is inevitable due to political necessity (Chow, 2006). Second, given that most detailed city-level macroeconomic data do not have consistent statistical standards throughout the sample period, we have to exclude those data or use other associated variables to replace the unqualified ones. For example, net export is a more proper indicator for GDP, but we only have data for the total import & export. Another interesting finding is that population growth and total household saving growth have an adverse impact on GDP growth. In previous research (Wang & Yao, 2003), large population and high saving rate are primary advantages of China. As such, further research should study such relationships between population, savings, and GDP.

References

Attanasio, O. P., Blow, L., Hamilton, R., & Leicester, A. (2009). Booms and Busts: Consumption, house prices, and expectations. Economica, 76(301), 20-50

Case, B., Pollakowski, H. O., & Wachter, S. M. (1997). Frequency of transaction and house price modeling. The Journal of Real Estate Finance and Economics, 14(1-2), 173-187.

Cerutti, E., Dagher, J., & Dell'Ariccia, G. (2017). Housing finance and real-estate booms: a cross-country perspective. Journal of Housing Economics, 38, 1-13.

Chivakul, M., Lam, M. R. W., Liu, X., Maliszewski, W., & Schipke, M. A. (2015). Understanding residential real estate in China (No. 15-84). International Monetary Fund.

Chow, G. (2006). Are Chinese official statistics reliable?. CESifo Economic Studies, 52(2), 396-414.

(33)

30 Ding, D., Huang, X., Jin, T., & Lam, W. R. (2017). The Residential Real Estate Market in China: Assessment and Policy Implications. Annals of Economics & Finance, 18(2).

Dröes, M. I., & Francke, M. K. (2016). What causes the Positive Price-Turnover Correlation in European Housing Markets? The Journal of Real Estate Finance and Economics, 1-29

ECB Economic Bulletin, Issue 7 / 2017 – Article China’s economic growth and rebalancing and the implications for the global and euro area economies

Glaeser, E., Huang, W., Ma, Y., & Shleifer, A. (2017). A real estate boom with Chinese Characteristics. Journal of Economic Perspectives, 31(1), 93-116.

HUANG, Z. H., WU, C. F., & DU, X. J. (2008). Empirical Study on Interaction between House Price, Interest Rate and Macro-economy in China [J]. China Land Science, 7, 38-44.

Lin, C., Morck, R., Yeung, B., & Zhao, X. (2016). Anti-corruption reforms and shareholder valuations: Event study evidence from China (No. w22001). National Bureau of Economic Research.

Phang, S. Y. (2004). House prices and aggregate consumption: do they move together? Evidence from Singapore. Journal of Housing Economics, 13(2), 101-119.

Tao Li, & Binkai Chen (2014). Real Assets, Wealth Effect and Household Consumption: Analysis Based on China Household Survey Data. Economic Research Journal, 3, 006.

Tsatsaronis, K., & Zhu, H. (2004). What drives housing price dynamics: cross-country evidence.

Wu, F., & Zhang, J. (2007). Planning the competitive city-region: The emergence of strategic development plan in China. Urban Affairs Review, 42(5), 714-740.

(34)

31 Wu, Y., & Li, Y. (2018). Impact of government intervention in the housing market: evidence from the housing purchase restriction policy in China. Applied Economics, 50(6), 691-705.

Yanbin, C., & Zhesheng, Q. (2011). How Does Housing Price Affect Household Saving Rate and Wealth Inequality?[J]. Economic Research Journal, 10, 25-38.

Yan Wang, & Yudong Yao (2003). Sources of China's economic growth 1952–1999: incorporating human capital accumulation. China Economic Review, 14(1), 32-52.

Xiaolua Wang, Ganga Fan, & Pengb Liu (2009). Transformation of Growth Pattern and Growth Sustainability in China [J]. Economic Research Journal, 1, 4-16.

Xu, N., Xu, X., & Yuan, Q. (2013). Political connections, financing friction, and corporate investment: Evidence from Chinese listed family firms. European Financial Management, 19(4), 675-702.

Zhao, X. L., Liang, W. Q., & Li, S. (2013). Can Growing Housing Prices Explain High Household Saving Rates in Urban China? An Empirical Analysis Using CHIPs Data. China Economic Quarterly, 13(1), 81-102.

W.Zou, (2016). Analysis of Financial Fraud of Listed Company in China: A Case Study of Dandong Xintai Electric Co., Ltd. American Journal of Industrial and Business Management, 6(09), 931. 陈秀梅, 韩和林, & 赵元兵. (2009). 我国房地产价格波动对经济的影响分析——兼论我国房地产宏观调控. 价格理论与实践, (8), 54-55. 贺京同, 战昱宁, & 万志华. (2009). 房地产市场中的羊群行为及其对商品房交易量的影响. 浙江大学学报 (人文社会科学版), 39(2), 172. 刘纯彬, & 陈冲. (2011). 我国房地产价格上涨的宏观结构效应分析. 中央财经大学学报, (10), 48-53.

(35)

32 唐志军, 徐会军, & 巴曙松. (2010). 中国房地产市场波动对宏观经济波动的影响研究. 统计研究, (2), 15-22. 李磊. (2009). 地产商降价惹了谁?. 北京房地产, (2), 40-41. 孙小琰, 沈悦, & 赵建军. (2007). 投资者行为, 正反馈交易与房地产价格异常波动. 預測, 26(5), 59-63. 张雷. (2012). 转机看过来! 房地产市场困境与转机. 房地产导刊, 4, 015.

Chinese cities buy off housing glut with borrowed money, Oct 13, 2017, by Dominique Fong, The Wall Street Journal, https://www.wsj.com/articles/chinese-cities-buy-off-housing-glut-with-borrowed-money-1507887001

THE HOUSEHOLD AS SECURITY: STRATEGIES OF RURAL-URBAN MIGRANTS

IN CHINA

The effect of a real estate market boom on China’s real economy : a cross-city perspective