Housing and household wealth inequality: Evidence from the People's Republic of China

Li, Sheng; Li, Jie; Ouyang, Alice Y. Working Paper Housing and household wealth inequality: Evidence from the People's Republic of China ADBI Working Paper, No. 671 Provided in Cooperation with: Asian Development Bank Institute (ADBI), Tokyo Suggested Citation: Li, Sheng; Li, Jie; Ouyang, Alice Y. (2017) : Housing and household wealth inequality: Evidence from the People's Republic of China, ADBI Working Paper, No. 671, Asian Development Bank Institute (ADBI), Tokyo This Version is available at: https://hdl.handle.net/10419/163171 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/3.0/igo/ ADBI Working Paper Series HOUSING AND HOUSEHOLD WEALTH INEQUALITY: EVIDENCE FROM THE PEOPLE’S REPUBLIC OF CHINA Sheng Li, Jie Li, and Alice Y. Ouyang No.671 February 2017 Asian Development Bank Institute The Working Paper series is a continuation of the formerly named Discussion Paper series; the numbering of the papers continued without interruption or change. ADBI’s working papers reflect initial ideas on a topic and are posted online for discussion. ADBI encourages readers to post their comments on the main page for each working paper (given in the citation below). Some working papers may develop into other forms of publication. ADB recognizes “China” as the People’s Republic of China. Unless otherwise stated, boxes, figures and tables without explicit sources were prepared by the authors. Suggested citation: Li, S., J. Li, and A. Y. Ouyang. 2017. Housing and Household Wealth Inequality: Evidence from the People’s Republic of China. ADBI Working Paper 671. Tokyo: Asian Development Bank Institute. Available: https://www.adb.org/publications/housing-and-household-wealthinequality-evidence-prc Please contact the authors for information about this paper. Email: [email protected], [email protected], [email protected] Financial support from the China National Social Science Fund (Project No. 16BJY167), Key Project sponsorship by the People’s Republic of China Ministry of Education (Project No. 14JZD016), and financial support by the Fundamental Research Funds for the Central Universities are greatly acknowledged. All errors remain ours. Sheng Li and Jie Li are associate professors at the Central University of Finance and Economics, Beijing. Alice Y. Ouyang is a professor at the Central University of Finance and Economics, Beijing. The views expressed in this paper are the views of the author and do not necessarily reflect the views or policies of ADBI, ADB, its Board of Directors, or the governments they represent. ADBI does not guarantee the accuracy of the data included in this paper and accepts no responsibility for any consequences of their use. Terminology used may not necessarily be consistent with ADB official terms. Working papers are subject to formal revision and correction before they are finalized and considered published. Asian Development Bank Institute Kasumigaseki Building, 8th Floor 3-2-5 Kasumigaseki, Chiyoda-ku Tokyo 100-6008, Japan Tel: +81-3-3593-5500 Fax: +81-3-3593-5571 URL: www.adbi.org E-mail: [email protected] © 2017 Asian Development Bank Institute ADBI Working Paper 671 Li, Li, and Ouyang Abstract We examine the issue of the widening wealth inequality in the People’s Republic of China (PRC) from the perspective of housing. Using China Household Finance Survey (CHFS) data from 2011, we find that the PRC’s wealth inequality including housing is much larger than income inequality. Housing value appreciation, in particular, contributes to wealth inequality by allowing households to enjoy equity market premium through investing more in equity markets and taking a higher position in risky assets. JEL Classification: E44, O11, O15 ADBI Working Paper 671 Li, Li, and Ouyang Contents 1. INTRODUCTION ................................................................................................... 1 2. EMPIRICAL MODEL AND DATA ............................................................................ 2 2.1 Empirical Model .......................................................................................... 2 2.2 Data and Methodology ................................................................................ 4 3. EMPIRICAL RESULTS........................................................................................... 7 4. HOUSEHOLDS’ WEALTH DECOMPOSITION ...................................................... 10 4.1 How Do Households Get Wealthier? Wealth Decomposition ....................... 12 5. CONCLUSIONS................................................................................................... 14 REFERENCES ............................................................................................................... 15 ADBI Working Paper 671 Li, Li, and Ouyang 1. INTRODUCTION The PRC’s economic reform since 1978 has brought not only rapid economic growth, but also enlarged income inequality. The inequality problem started to get more serious in the mid-1980 when the government focused their reform efforts in the urban sector.1 Based on the estimate in Wang and Sebastian (2011), there were 336 million Chinese people living on under $2 per day in 2008, while there were 960,000 millionaires in 2010, each with more than $1.6 million in personal wealth. The PRC has become one of the most unequal countries in the world, in a short period of less than 3 decades. Among the driving forces behind inequality, housing stands out as a significant one. Even in the early phase of housing reform transitioning from state allocation to market-determined supply, housing subsidies had become a major contributor to the PRC's urban inequality, according to the study of Khan and Riskin (1998) using data from the China Household Income Project (CHIP).2 Homeownership can be helpful in household wealth accumulation. Using longitudinal data from the Panel Study of Income Dynamics (PSID) between 1984 and 2001, Di et al. (2007) find that those who owned homes and owned for longer periods of time had significantly higher household net wealth by 2001. The findings are suggestive of a positive influence of ownership over long periods on net wealth. In addition, each year of ownership is associated with approximately 2% of increase in household income and doubling the length of ownership increases household income by about 11% (Di 2007). However, the impact of homeownership varies by income status, with each additional year of homeownership being associated with $15K more in wealth holdings for high-income households and roughly $6 to $10K more in wealth holdings for lowand middle-income households (Turner and Luea 2009). Along this line of literature, we are interested in why homeownership may function differently in wealth accumulation, possibly widening income inequality. More specifically, we empirically test whether housing price appreciation can help household's financial market investment, particularly investment in equity markets. If this is the case, equity market premiums, coming along with housing price appreciation, leads to more inequality. There are theories linking the distribution of wealth and financial market investment. In the model of Aghion and Bolton (1997), individuals are assumed to be able to engage in specific productive projects. Only entrepreneurs with sufficiently high levels of personal wealth will be able to finance their “project.” With a simple indivisibility, the initial wealth distribution will determine how many individuals will be able to undertake such projects. In models with capital market imperfections, credit constraints will prevent the poor from undertaking profitable indivisible investments. Pastor and Veronesi (2016) examine the channels through which financial markets and business ownership affect inequality. In their model, investment risk and differences in financial market participation are the principal drivers of income inequality. 1 See Wang et al. (2015) and Knight (2013) for comprehensive review on the issue of the PRC’s income inequality. 2 See Wang (2011) for the discussion on how the PRC’s removal of price distortions, originating from state misallocation, allowed households to increase their consumption of housing and led to an increase in equilibrium housing prices. 1 ADBI Working Paper 671 Li, Li, and Ouyang Investmentor asset-induced inequality tends to self-reinforce over time. There is a fundamental constraint on poverty reduction: the poor lack of access to the assets necessary for increased productivity and income. The World Bank, targeting poverty reduction, should put more emphasis on the distribution of assets, both physical and human capital (Birdsall and Londono 1997). In the meanwhile, this justifies the importance of asset inequality in the discussion of income inequality.3 Meanwhile, unlike stocks and bonds, owner-occupied housing provides significant consumption benefits (Henderson and Ioannides 1983). Acquisition of such housing is thus driven by both consumption and investment motives. Some experts argue that this dual role leads to an overinvestment in housing (Brueckner 1997). While the asset substitution argument explains why homeowners would lower their equity proportion in net worth, there is also diversification effect of owning two risky assets: home equity and stocks. Compared with a homeowner, a renter's risk exposure depends only on his holding of risky stocks. With a low return correlation between risky stocks and home equity, a homeowner reduces his stockholding in net worth but holds a riskier liquid financial portfolio (Yao and Zhang 2005). We organize the paper as follows. The next section explains empirical model and data while section 3 contains empirical results. Section 4 uses decomposition method while the final section concludes. 2. EMPIRICAL MODEL AND DATA 2.1 Empirical Model To examine whether the rise of housing values may deteriorate the PRC’s wealth inequality through extending homeowners’ investment choices and further increasing their total wealth from the investment on other financial assets, we first construct a benchmark model to test if change of housing values may influence households’ total wealth as below: 𝑊𝑒𝑎𝑙𝑡ℎ𝑖=𝛼0+𝛼𝐻𝐻𝑜𝑢𝑠𝑖𝑛𝑔𝑖+𝛼𝑐𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑖+𝜖𝑖 (1) where 𝑊𝑒𝑎𝑙𝑡ℎ𝑖 is household’s net wealth (in logarithm) in 2011. Based on different coverage, two kinds of net wealth are used in the paper, i.e., net wealth including all housing values (up to three housing units) (𝑊𝑒𝑎𝑙𝑡ℎ_𝑎𝑙𝑙𝑖) and non-housing net wealth (𝑊𝑒𝑎𝑙𝑡ℎ_𝑛𝑜ℎ𝑜𝑢𝑠𝑒𝑖). 𝐻𝑜𝑢𝑠𝑖𝑛𝑔𝑖 is a proxy of changes of housing values, measured by either the change of total housing values (in logarithm) (𝐻𝑜𝑢𝑠𝑒𝑣𝑎𝑙𝑢𝑒𝑖) that captures the real appreciation/depreciation of housing units by deducting housing units’ initial acquisition cost from the current value reported by the respondent in 2011, or the rate of returns that capture the average annual yield of holding housing units (𝐻𝑜𝑢𝑠𝑒𝑟𝑒𝑡𝑢𝑟𝑛𝑖).4 3 Deininger and Olinto (1999): Asset inequality—but not income inequality—has a relatively great negative impact on growth and also reduces the effectiveness of educational interventions. This means that policy makers should be more concerned about households' access to assets, and to the opportunities associated with them, than about the distribution of income. 4 For some reasons, some respondents report the current housing value as zero. The minimum values of annual yield are thus calculated as –1, i.e. –100%. 𝐻𝑜𝑢𝑠𝑒𝑟𝑒𝑡𝑢𝑟𝑛𝑖 is calculated as the simple average of annual yields of households’ total (up to 3) housing units. 2 ADBI Working Paper 671 Li, Li, and Ouyang Table 1: Gini Coefficients Household Head Labor Income Household Net Wealth Household Non-Housing Wealth Total Sample 0.65 (8,381) 0.74 (8,205) 0.84 (7,749) Region: East 0.67 (3,952) 0.71 (3,924) 0.82 (3,807) Central 0.53 (2,498) 0.59 (2,444) 0.77 (2,280) West 0.55 (1,931) 0.64 (1,837) 0.81 (1,662) Development level: Urban 0.64 (5,171) 0.70 (5,099) 0.82 (4,967) Rural 0.61 (3,210) 0.70 (3,106) 0.80 (2,782) Note: The values in brackets are the number of observations with non-negative values that are used to draw the Lorenz curves in Figures 1–3 and the calculation of corresponding Gini coefficients. Housing tenure choice is considered as a key determinant of household wealth accumulation (Di et al. 2007). In this study, we control the housing tenure choice by including a dummy variable of 𝑅𝑒𝑛𝑡𝑒𝑟𝑖. One would expect that a renter accumulates less wealth than a homeowner. 𝑀𝑢𝑙𝑡𝑖ℎ𝑜𝑢𝑠𝑖𝑛𝑔𝑖 is the dummy to further distinguish households with several housing units from households with one or no housing units. A homeowner with multiple housing units has the potential to accumulate more wealth than a homeowner with only one housing unit or renter. Considering the different types of housing units may influence the wealth accumulation, we define a dummy variable, 𝐶𝑜𝑚𝑚𝑜𝑑𝑖𝑡𝑦𝑖, to separate commodity housing from the remaining types such as affordable housing, inheritance or gifts, purchased at below market prices, financed housing, self-built, demolition/relocation, and others. Both 𝐻𝑜𝑢𝑠𝑖𝑛𝑔𝑐𝑜𝑠𝑡𝑖 and 𝐿𝑜𝑎𝑛𝑖 are used to proxy households’ budget constraint. While 𝐻𝑜𝑢𝑠𝑖𝑛𝑔𝑐𝑜𝑠𝑡𝑖 is the initial cost that households pay for the housing, 𝐿𝑜𝑎𝑛𝑖 is the unpaid loans for a household to acquire housing units, including mortgage and other loans. Both variables are in logarithm. Moreover, the features of different households and households’ head may also influence households’ wealth accumulation and investment decision, and thus should be controlled in this study. These control variables include household head’s age, gender, education level, marital status, migrant status, investment attitude, household size, income, and geographic location. According to life-cycle consumption theory (Modigliani 1966), the younger households like to borrow than save to smooth life-cycle consumption relative to the elder ones. Hence, we expect to observe a generalized inverted U pattern, indicating that household wealth peaks at middle age. To test this effect, 𝐴𝑔𝑒𝑖 and 𝐴𝑔𝑒𝑖 2 are both added into the empirical model. In the model of estimating households’ wealth, household head’s permanent income is considered as an important impact factor (Choudhury 2002). Under the assumption that the head’s permanent income is generally not accessible, his or her education achievement (𝐸𝑑𝑢𝑐𝑎𝑡𝑖𝑜𝑛𝑖) is taken as a proxy. 𝐺𝑒𝑛𝑑𝑒𝑟𝑖 is the dummy to identify household head’s gender. 𝐷𝑖𝑣𝑜𝑟𝑐𝑒𝑖 is the dummy variable to control for household head’s marital status. One would expect a negative impact of divorce on household wealth since it is possible to split the wealth between divorced couple. 𝐼𝑛𝑐𝑜𝑚𝑒𝑖 is household’s income (in logarithm), used to control for differences in household income in determining household investment. We also expect that migrant households (𝑀𝑖𝑔𝑟𝑎𝑛𝑡𝑖) display different patterns from native households. Self-reported attitude to risk (𝑅𝑖𝑠𝑘𝑎𝑡𝑡𝑖𝑡𝑢𝑑𝑒𝑖) is considered as an important household demographic feature (Campbell 2006). The higher the value is, the more conservative risk attitude the household has. Household size (𝐹𝑎𝑚𝑖𝑙𝑦𝑠𝑖𝑧𝑒𝑖) is also controlled because the number of dependents may affect a 3 ADBI Working Paper 671 Li, Li, and Ouyang household’s capacity to save and its motivation to save (Di et al. 2007). Rural and Province are all geographic control variables. Households living in rural areas may exhibit quite different wealth accumulation patterns than those in urban areas. 2.2 Data and Methodology In this study, we use the data collected in the first round of the China Household Finance Survey (CHFS) in 2011. 5 The dataset provides micro-level financial information of more than 8,000 PRC households in 25 provinces. CHFS data allow us to have a comprehensive understanding of each respondent household’s assets and liabilities, including the information associated with housing and financial assets. In the survey, a respondent may report detailed information of the housing unit he or she rents or as many as three housing units owned. In addition to the housing information, a household was also inquired to report the relevant information about the investment in other financial assets, such as checking, savings, stocks, bonds, etc. Table 2: Household Portfolios of China Household Finance Survey Data Number of Households Holding the Asset Percentage of Total Sample (%) Mean Asset Share (%) Owner-occupied housing 7,570 89.71 93.91 Other housings 1183 14.02 51.99 Housing mortgage 846 10.03 29.20 Checking account 4,364 51.72 9.87 Saving account 1,304 15.45 20.54 Stocks 622 7.37 11.72 Bonds 57 0.68 5.99 Funds 351 4.16 5.41 Derivatives 1 0.01 1.97 Bank financial products 76 0.90 9.78 Non-RMB assets 101 1.20 2.97 Gold 54 0.64 5.63 Other liabilities 2,929 34.71 44.23 Net wealth 8,438 100 100 Note: The data of net wealth reported only for households with non-negative net worth and households have ownership of the asset. Households’ net wealth (𝑊𝑒𝑎𝑙𝑡ℎ_𝑎𝑙𝑙𝑖) is calculated by adding the value of private business, value of at most three housing units,6 value of all automobiles owned, value of 12 categories of durable goods, value of luxury goods, account balance of checking, saving, stock, bond, fund, future, warrant, other derivatives, financial product, non-RMB assets, gold, cash, lending, and eliminating bank and/or other loans for private 5 The CHFS data set is provided by the Survey and Research Center of China Household Finance, Southwestern University of Finance and Economics, Chengdu, PRC. For more detail about the dataset, please see Gan et al. (2013). Updated data from the second round were not available when we conducted the research. 6 For a renter, the value of housing units is set as zero. 4 ADBI Working Paper 671 Li, Li, and Ouyang impact that a 1% change in the net asset of housing will decrease the inequality level among households who own housing by 2.76%. When we turn to the non-housing wealth, the difference between the impacts of small changes in the risky assets upon inequalities in these two groups is large. A 1% increase in risky assets lowers inequality for the group owning housing by only 0.82% but decreases inequality for the not-house-owning group by 6.66%. Table 9: The Distribution of Wealth and Its Decomposition by Factor in Urban PRC House-Owning Samples Mean Share %>0 Gini % Change Wealth_all 549,932 1.000 97.86 0.701 – Housevalue 444,810 0.839 98.71 0.689 –2.76 Wealth_nohouse 105,252 0.161 91.66 0.927 2.76 Income 52,619 0.437 96.39 0.630 –3.88 household business asset 22,972 0.026 11.50 1.432 2.44 automobile 15,282 0.011 14.66 0.939 0.17 durable and luxury goods 14,338 0.033 98.27 0.743 –0.38 risk free assets 15,940 0.011 96.31 0.965 0.59 risk assets 34,044 0.087 12.39 0.818 –0.82 Loan –35,978 0.063 33.14 0.931 –3.62 Loan for housing –22,026 0.049 21.55 0.951 –1.00 Loan for household business –10,770 –0.009 9.96 0.986 –1.81 Loan for automobiles –467 0.007 2.47 0.994 –0.05 Loan for durable and luxury goods –16 0.000 0.04 1.000 0.00 Loan for risk assets –89 0.000 0.16 0.999 0.00 Loan for education –141 –0.006 1.32 0.994 –0.05 Other loan –2,574 0.022 4.76 0.988 –0.71 Not-House-Owning Samples Mean Share %>0 Gini % Change Wealth_all 73,977 1.000 91.35 0.946 – Housevalue – – – – – Wealth_nohouse 73,977 1.000 91.35 0.946 0 Income 52,209 6.511 92.26 0.708 –47.64 household business asset 15,989 0.070 11.35 1.253 5.46 automobile 8,522 0.037 8.90 0.982 –0.45 durable and luxury goods 7,230 0.356 94.45 0.704 –4.49 risk free assets 12,728 0.035 96.26 0.970 –0.49 risk assets 29,674 0.455 10.58 0.858 –6.66 Loan –11,728 5.205 14.84 0.975 –27.07 Loan for housing – – – – – Loan for household business –8,241 5.211 5.55 0.993 –18.26 Loan for automobiles –528 0.014 1.55 0.996 –0.88 Loan for durable and luxury goods 0 0.000 0.00 . . Loan for risk assets 0 0.000 0.00 . . Loan for education –122 –0.030 1.29 0.993 –0.27 Other loan –2,837 0.010 6.97 0.977 –7.66 Note: "Share" accounts for the proportion of households with nonzero net wealth. 11 ADBI Working Paper 671 Li, Li, and Ouyang 4.1 How Do Households Get Wealthier? Wealth Decomposition To explore housing owning difference in household wealth, we apply a regressionbased decomposition method of Blinder–Oaxaca-type decomposition (Blinder 1973; Oaxaca 1973), which allows the decomposition of housing owning difference in the amount of household wealth into a part that is caused by differences in observable characteristics and a part that is explained by differences in estimated coefficients. Consider the following linear regression model, which is estimated separately for the groups 𝑔=ℎ,𝑛: 𝑊𝑖𝑔 =𝑋𝑖𝑔𝛽𝑔+𝜀𝑖𝑔 (2) where 𝑊𝑖𝑔 denotes wealth of household 𝑖 in group g, 𝑋𝑖𝑔 is a vector of observable characteristics, 𝛽𝑔 represents a vector of parameters to be estimated and 𝜀𝑖𝑔 is a standard error term. For these models, classical Blinder–Oaxaca decomposition proposes the decomposition as: 𝑊ℎ− 𝑊 𝑛=�𝐸𝛽ℎ(𝑊𝑖ℎ|𝑋𝑖ℎ)− 𝐸𝛽ℎ(𝑊𝑖𝑛|𝑋𝑖𝑛)�+�𝐸𝛽ℎ(𝑊𝑖𝑛|𝑋𝑖𝑛)− 𝐸𝛽𝑛(𝑊𝑖𝑛|𝑋𝑖𝑛)� (3) where 𝑊 𝑔=𝑁𝑔 −1 ∑𝑊𝑖𝑔 𝑁𝑔 𝑖=1 and 𝑋𝑔=𝑁𝑔 −1 ∑𝑋𝑖𝑔 𝑁𝑔 𝑖=1 . 𝐸𝛽𝑔�𝐶𝑖𝑔�𝑋𝑖𝑔� refer to the conditional expectation of 𝐶𝑖𝑔 evaluated at the parameter vector 𝛽𝑔. The first term on the right-hand side of equation (3) is the component of the difference in household wealth between housing owner and non-housing owner households that is due to differences in observable characteristics. The second term represents the difference that is due to differences in coefficient estimates. The first step of our econometric methodology consists of estimating equation (1) using the benchmark model. In the previous wealth decomposition studies (Bodenhorn and Ruebeck 2007; Johns 1990; Juster et al. 2005), household heads’ age, age square, marital status, immigrant status, employment status, education level, household size, geographic location, household income, and inheritances are considered as factors impacting households’ wealth. In this study, we also include all these factors except for inheritances, which is not available in the survey. Estimated results are reported in Table 10. The second step is the wealth decomposition. Results of the decomposition analysis are reported in Table 11. Although most factors that have impact on households’ wealth used in the traditional wealth decomposition studies have been included, the model still does not do a good job in predicting the wealth difference between households owning housing and those owning non-housing. For the household net wealth in column 1, around 98% of the households’ wealth difference is due to differences in coefficients (unexplained part) and only 2% by different observable characteristics. The results do not change much when we add in a few other factors such as 𝐻𝑜𝑢𝑠𝑒𝑣𝑎𝑙𝑢𝑒𝑖, 𝐻𝑜𝑢𝑠𝑖𝑛𝑔𝑐𝑜𝑠𝑡𝑖, 𝐿𝑜𝑎𝑛𝑖, 𝑀𝑢𝑙𝑡𝑖ℎ𝑜𝑢𝑠𝑖𝑛𝑔𝑖, 𝐶𝑜𝑚𝑚𝑜𝑑𝑖𝑡𝑦𝑖. When we turn to household non-housing wealth in column 2, the increase of 0.024 indicates that the difference in all traditional factors listed account for about 1/4 non-housing wealth gap. The unexplained part accounts for around 3/4 of the non-housing wealth gap. These results confirm all these traditional factors play a less important role in determining households’ wealth, but they do play a role in determining the non-housing wealth gap. It suggests housing tenure choice is essential to the gap of households’ wealth in all. 12 ADBI Working Paper 671 Li, Li, and Ouyang Table 10: Determinants of Household Wealth in Urban PRC OLS House-Owning Households Not-House-Owning Households Wealth_all Wealth_nohouse Wealth_all Wealth_nohouse Gender –0.0774*** –0.0134 –0.0473 –0.0314 [0.0245] [0.0169] [0.0704] [0.0495] Age 0.00888 –0.00331 0.0244** 0.0153* [0.00546] [0.00380] [0.0120] [0.00820] Age2 –0.00007 0.00003 –0.000187* –0.00011 [4.88e-05] [3.20e-05] [9.99e-05] [6.80e-05] Education 0.145*** 0.0803*** 0.0699*** 0.0485*** [0.00846] [0.00614] [0.0255] [0.0173] Divorce 0.00492 0.004 –0.0946 –0.0716 [0.0778] [0.0485] [0.151] [0.0925] Income 0.229*** 0.151*** 0.179*** 0.103*** [0.0122] [0.00867] [0.0313] [0.0215] Migrant –0.335*** –0.00429 0.113 0.0801 [0.0618] [0.0539] [0.107] [0.0739] Riskattitude –0.0586*** –0.0541*** –0.0953*** –0.0809*** [0.00908] [0.00650] [0.0304] [0.0210] Familysize 0.0257*** –0.0107* 0.0298 0.0289* [0.00785] [0.00552] [0.0273] [0.0164] Rural –0.694*** –0.119*** –0.0321 –0.0247 [0.0278] [0.0162] [0.0818] [0.0539] Constant 10.75*** 10.44*** 8.649*** 10.34*** [0.197] [0.136] [0.489] [0.351] Province Y Y Y Y Observations 7,128 7,122 666 676 R-squared 0.542 0.305 0.302 0.251 Note: Standard errors are reported in brackets. *, **, and *** indicated the significance level of 10%, 5%, and 1%, respectively. Table 11: Owning Housing Wealth Gap: Oaxaca–Blinder Decomposition Results Wealth_all Wealth_nohouse Coefficient In % of Δ Coefficient In % of Δ Δ 1.550*** 0.093*** (0.040) (0.0255) Explained 0.031 2 0.024 25.81 (0.044) (0.029) Unexplained 1.519*** 98 0.069*** 74.19 (0.047) (0.435) 13 ADBI Working Paper 671 Li, Li, and Ouyang 5. CONCLUSIONS In this paper, we depict household income and wealth inequality status in the PRC. According to our calculations based on CHFS data of 2011, the Gini coefficients have reached a high level of 0.65 for household income, and a higher level of 0.74 for household net wealth. Income inequality is not the only contributor to the gap of households’ wealth. Further exploration of the households’ asset portfolio and homeownership reveals that housing assets account for the largest share of total household wealth for homeowners, and the national homeownership rate is about 90%. These stylized facts shed light on the possibility that housing value appreciation in the past decades may have contributed to enlarged household wealth gap. We focus on analyzing whether and how housing value appreciation deteriorates the PRC’s household wealth inequality. The empirical results confirm that housing value appreciation does contribute to household wealth inequality. The stratified analysis further indicates that housing value appreciation has a greater positive impact on non-housing wealth for wealthier households. The housing, after housing value appreciation, can be refinanced to pay for equity market investment, allowing households to enjoy more equity market premium. This may enlarge wealth inequality. 14 ADBI Working Paper 671 Li, Li, and Ouyang REFERENCES Aghion, P., and P. Bolton. 1997. A Theory of Trickle-Down Growth and Development. Review of Economic Studies 62(2): 151–172. Birdsall, N., and J. Londono. 1997. Asset Inequality Matters: An Assessment of the World Bank's Approach to Poverty Reduction. AEA Papers and Proceedings 87(2): 32–37. Blinder, A. 1973. Wage Discrimination: Reduced Form and Structural Estimates. Journal of Human Resources 8: 436–455. Bodenhorn, H., and C. Ruebeck. 2007. Colourism and African–American Wealth: Evidence from the Nineteenth-Century South. Journal of Population Economics 20: 599–620. Brueckner, J. 1997. Consumption and Investment Motives and the Portfolio Choices of Homeowners. Journal of Real Estate Finance and Economics 15(2): 159–180. Campbell, J. 2006. Household Finance. The Journal of Finance 61(4): 1553–1604. Case, K., J. Quigley, and R. Shiller. 2005. Comparing Wealth Effects: The Stock Market versus the Housing Market. Advances in Macroeconomics 5(1): 1534–6013. Chudhury, S. 2002. Racial and Ethnic Differences in Wealth and Asset Choices. Social Security Bulletin 64(4): 1–15. Deininger, K., and P. Olinto. 1999. Asset Distribution, Inequality, and Growth. Policy Research Working Papers 93(4): 1091–1113. Deng, Y., D. Zheng, and C. Ling. 2005. An Early Assessment of Residential Mortgage Performance in China. Journal of Real Estate Finance and Economics 31(2): 117–136. Di, Z. 2007. Do Homeowners Have Higher Future Household Income? Housing Studies 22(4): 459–472. Di, Z., E. Belsky, and X. Liu. 2007. Do Homeowners Achieve More Household Wealth in the Long Run? Journal of Housing Economics 16: 274–290. Gan L., Z. Yin, N. Jia, S. Xu, S. Ma, and L. Zheng. 2013. Data You Need to Know about China: Research Report of China Household Finance Survey 2012. Springer. Henderson, J., and Y. Ioannides. 1983. A Model of Housing Tenure Choice. American Economic Review 73: 98–113. Jones, D. 1990. Current Wealth Constraints on the Housing Demand of Young Owners. The Review of Economics and Statistics 72(3): 424–432. Juster, F., J. Lupton, J. Smith, and F. Stafford. 2005. The Decline in Household Saving and the Wealth Effect. The Review of Economics and Statistics 87(4): 20–27. Kacperczyk, M., J. Nosal, and L. Stevens. 2014. Investor Sophistication and Capital Income Inequality. NBER Working Paper No. 20246. Khan, A., and C. Riskin. 1998. Income and Inequality in China: Composition, Distribution and Growth of Household Income, 1988 to 1995. The China Quarterly 154: 221–253. 15 ADBI Working Paper 671 Li, Li, and Ouyang King, M., and J. Leape. 1998. Wealth and Portfolio Composition: Theory and Evidence. Journal of Public Economics 69: 155–193. Knight, J. 2013. Inequality in China: An Overview. The World Bank Research Observer 29: 1–19. Logan, J., Y. Fang, and Z. Zhang. 2010. The Winners in China's Urban Housing Reform. Housing Studies 25(1): 101–117. Modigliani, F. 1966. The Life Cycle Hypothesis of Saving, the Demand for Wealth and the Supply of Capital. Social Research 33(2). Oaxaca, R. Male–Female Wage Differentials in Urban Labor Markets. 1973. International Economic Review 14: 673–709. Ortalo-Magne, F., and S. Rady. 2006. Housing Market Dynamics: On the Contribution of Income Shocks and Credit Constraints. The Review of Economic Studies 73(2): 459–485. Pastor, L., and P. Veronesi. 2016. Income Inequality and Asset Prices under Redistributive Taxation. NBER Working Paper No. 21668. Turner, T., and H. Luea. 2009. Homeownership, Wealth Accumulation and Income Status. Journal of Housing Economics 18: 104–114. Wang, C., G. Wan, and D. Yang. 2015. Income Inequality in the People's Republic of China: Trends, Determinants, and Proposed Remedies. Journal of Economic Surveys 28(4): 686–708. Wang, S. 2011. State Misallocation and Housing Prices: Theory and Evidence from China. American Economic Review 101: 2081–2107. Wan, G., and I. Sebastian. 2011. Poverty in Asia and the Pacific: An Update. ADB Economics Working Paper No. 267. Xie, Y., and X. Zhou. 2014. Income Inequality in Today’s China. Proceedings of the National Academy of Sciences 111(19): pp. 6928–6933. Yao, R., and H. Zhang. 2005. Optimal Consumption and Portfolio Choices with Risky Housing and Borrowing Constraints. The Review of Financial Studies 18(1): 197–239. Yitzhaki, S. 1994. Economic Distance and Overlapping Distributions. Journal of Econometrics 61(1): 147–159. 16