The causes of income inequality in urban China: A household assets perspective

Abstract

Despite income inequality in China being a heated topic in scholarship and public discussion, little research has explored the trend in income inequality since the mid-2010s or focused on the implications of household assets on income inequality. Using pooled cross-sectional data from the Chinese General Social Survey, we estimate income inequality at the household level from 2012 to 2021 and further examine the effects of two types of household assets (housing assets (HA) and financial assets (FA)) on income inequality in urban China, as well as the moderating effects of socioeconomic status (SES) and regional marketization. The results indicate that national income inequality was sustained at a high level between 2012 and 2021, while income inequality in urban households was higher than that in rural households. Both HA and FA were positively related to income inequality, and the latter assets had a stronger effect on income inequality. The association between the two types of household assets and income inequality strengthened along with an increase in family SES and regional marketization levels. The study highlights that FA, relative to HA, are a more important contributor to income inequality among urban households, thus extending the literature on inequality and social stratification in contemporary China by revealing the causes underlying rising income inequality from the household assets perspective.

Keywords

Income inequality household assets China socioeconomic status marketization

Introduction

China has changed from a poor, underdeveloped country to the world's leading emerging economy since the introduction of its reform policies in the late 1970s. While decades of rapid economic development in China have led to a sustained increase in household income, income inequality rose dramatically during the economic transition. According to scholars’ estimations based on nationally representative surveys, the Gini coefficient—a common measure of income distribution at the national level—was just above 0.3 in the early 1980s in China, close to the most egalitarian Nordic countries, but it had reached a level above 0.5 by 2013 (Xie and Zhou, 2014), which is above the international “warning line” of 0.4. The rapid increase in income inequality during the economic transition has put China among the most unequal countries in Asia and, indeed, the world (Zhou and Song, 2016), with various economic, political, and social problems arising as a result (Knight, 2014).

Income inequality in China, in sharp contrast to the egalitarianism before the economic reform, has drawn extensive attention from social science scholars (Li and Sicular, 2014). Although the fact that income inequality in China reached very high levels by the early 2010s has been supported by substantial evidence (Li and Sicular, 2014; Piketty et al., 2019; Ravallion and Chen, 2007; Xie and Zhou, 2014), few studies have examined the trend in income inequality in China from the mid-2010s to the present. Indeed, in addition to the deepening of the reform of the system of income distribution (e.g. tax reform, the raising of the minimum wage, protection of rural land rights, etc.), the central government has launched a national campaign of poverty alleviation since 2017, and thus eliminated absolute poverty according to the 3% threshold set by the World Bank. However, income inequality has not been significantly alleviated by these policies because relative poverty has been consistently rising (Wan et al., 2021). Therefore, although the official data indicate a slight fall in national income inequality during the past five years, some researchers have questioned the decline and suggested that it is either not robust or not significant (Luo et al., 2020). Thus, whether income inequality has declined in recent years remains up for debate.

Moreover, a large body of literature suggests that China's income inequality is mainly dominated by institutional factors, such as the urban–rural divide and regional variation in economic well-being (Li and Sicular, 2014; Luo et al., 2020; Xie and Zhou, 2014). However, as housing, the primary component of household assets, was identified as a new mechanism for wealth generation in China since the 2000s (Walder and He, 2014), there has been a growing interest in the role of housing assets (HA) in income inequality (Wu, 2019). Nevertheless, little research focuses on the role of financial assets (FA) (e.g. stocks, bonds, funds, etc.) on income inequality. In China, in addition to the investment in housing, households tend to participate in financial markets and invest in FA. According to the China Household Finance Survey conducted in 2011 and 2019, FA have increased to the second largest component of total household assets (Qiao and Cai, 2023; Yuan et al., 2022). Since FA generally have a higher return on investment than other assets (Langley, 2021), they have a more influential role in the accumulation of households’ income and wealth than HA in China (Piketty et al., 2019). In terms of the dominant position of HA and the growing importance of FA in the household asset structure, it is of great importance to investigate whether the two types of household assets contribute to income inequality among Chinese households.

It is noted that the opportunity to hold certain household assets, including HA and FA, rests on family and regional characteristics (e.g. financial literacy, political background, and the coverage of financial institutions) (Ge et al., 2021; Liao et al., 2017; Honohan, 2008). Concerning family characteristics, socioeconomic status (SES) plays a prominent role in holding assets, including HA and FA, due to the economic cost of purchasing assets (Liu et al., 2023; Wang et al., 2020). In addition, asset value has a close relationship with the scale of the asset market in certain areas, which depends on the development of a market-oriented economy (Chen, 2015; Xin and Yang, 2023). This means that household assets may be unevenly distributed among households because of differences in family SES and regional marketization levels, thus widening the disparities in household income.

Our study analyzes pooled cross-sectional data from the Chinese General Social Survey (CGSS) during the period from 2015 to 2021, a nationally representative sample of Chinese households, to address three research questions: (a) How has income inequality in China changed in recent years, especially since 2015? (b) Is income inequality related to household assets? If so, what is the difference in the impacts of HA and FA on income inequality? (c) Do the relationships between the two types of household assets and income inequality vary by family SES and regional marketization? The rapid transformation in household asset structures combined with remarkable economic and social changes in contemporary China provides a unique opportunity to explore these questions and thus to examine the trend in Chinese income inequality and its causes.

Literature review

Income inequality in China during the economic transition

The initiation of market-oriented economic reform at the end of the 1970s triggered a sustained increase in household income growth in China. The average real annual growth of income per capita in rural and urban China reached 7.6% and 7.4%, respectively, from 1978 to 2015 (National Bureau of Statistics, 2016). At the same time, income inequality has also risen dramatically, which has driven scholars to make efforts to measure income inequality in China. The official data provided by the National Bureau of Statistics (NBS) and survey data collected by several Chinese university survey organizations are the two main sources of calculating China's income inequality in previous studies. Using the NBS microdata, Ravallion and Chen (2007) found that despite a decline in inequality in the early 1980s (1981–1985) and the mid-1990s (1994–1997), the national Gini coefficient increased from about 0.31 in 1981 to 0.45 in 2001. This finding was supported by empirical studies using the China Household and Income Project (CHIP) that used a subsample of the NBS surveys and a more comprehensive definition of income. For instance, results based on the CHIP showed that the Gini coefficient for household income as a whole increased from 0.381 in 1988 to 0.462 in 1995 and then to 0.471 in 2002 (Knight, 2014). In addition to the Gini coefficient, income inequality can be measured by the gap between the bottom and top of the income distribution (Piketty and Saez, 2003). According to the CHIP surveys, the ratio of the share of income of households in the highest 10% to the share of income of households in the lowest 10% surged from 7.3 in 1988 to 19 in 2002, suggesting that a widening income gap between the poorest and the richest over this period (Zhou and Song, 2016).

The NBS statistics showed that income inequality in China has continued to rise since 2002. The national Gini coefficient based on the NBS data increased from 0.479 in 2003 to 0.491 in 2008 (Zhou and Song, 2016). Estimates based on the CHIP survey conducted in 2008 indicated that the Gini coefficient for household income reached a level of about 0.5 (0.490) in 2007 (Li and Sicular, 2014). However, according to official estimates based on the NBS, the Gini coefficient for household income declined slightly from 0.491 in 2008 to 0.474 in 2012 (Wang et al., 2014). This positive change in income distribution is treated because of the implementation of a range of measures to reduce disparities between socioeconomic groups under the leadership of Hu Jintao and Wen Jiabao, including the elimination of agricultural taxes and fees, poverty alleviation in rural areas, and social welfare policies (Li and Sicular, 2014). However, empirical studies based on university-based surveys have questioned the decline and suggested that the standard estimates of inequality in China were underestimated. According to estimations from multiple data sources, including the China Family Panel Studies, Chinese General Social Survey, Chinese Household Finance Survey, and China Labor Force Dynamics Survey, the Gini coefficients for household income were in the range of 0.530–0.611 during the period from 2010 to 2012 (see Xie and Zhou, 2014). This finding indicates that income inequality in China had not only reached a very high level but had also surpassed most other countries at similar stages of economic development.

The use of official data or household surveys to measure income inequality can suffer from underestimation because it is difficult to collect information on top-income individuals, which poses the greatest challenge to the estimation of the exact income inequality (Davies et al., 2011). The underestimation of top incomes is increasingly important in China because the group with the highest incomes has been increasing due to rapid economic growth since the mid-1990s. Scholars concerned about the understatement of income inequality in China during the past decade have corrected for top incomes in income distribution research. Using the 2002 and 2013 waves of the CHIP surveys, Li and Wan (2015) found after including the samples of top incomes that the Gini coefficient for household income increased by 20% in 2002 and 28% in 2013. Piketty et al. (2019) estimated a revised measure of income inequality by combining tax data on top-income individuals with household surveys and national accounts. According to their estimations, the share of national income earned by the top 10% of the population in China increased from 27% in 1978 to 41% in 2015, while the share earned by the bottom 50% dropped from 27% to 15%.

The findings above indicate that income inequality in China ranks among the highest in the world and was approaching United States levels in the early 2010s (Xie and Zhou, 2014). Yet there is an optimistic expectation that income inequality in China will stabilize and even slightly fall due to new policies (e.g. the reform of the household registration system, targeted poverty alleviation, anti-corruption campaign, etc.) intended to moderate income inequality (Li and Sicular, 2014; Luo et al. 2020; Han et al., 2022). However, existing research has not documented the trend in income inequality in China since the mid-2010s due to data limitations. Contrary to the positive outlook for income inequality, there is the possibility of an increase in income inequality because of emerging risk factors, such as surging housing prices, widening disparities in educational attainment, and uneven access to digital technologies (Arshed et al., 2019; Fuller et al., 2020; Tchamyou et al., 2019). What is more alarming is that the COVID-19 pandemic, may have exacerbated income inequality. Indeed, Chinese households were hit harder by the COVID-19 outbreak, with unprivileged families more likely to experience income losses (Qian and Fan, 2020). Another problem is that the existing literature on income inequality in China is concentrated at the group level by calculating the Gini coefficient (Li and Sicular, 2014), urban–rural income inequality (Sicular et al., 2007), or the income group in various deciles (Golley and Meng, 2012), which cannot show the income inequality at the household level. As such, our study estimates the recent trend in income inequality in China using a nationally representative sample covering the pre-COVID-19 pandemic period and post-COVID-19 pandemic period.

Household assets and income inequality in China

Sociologists have long been interested in the role of the wage income gap in overall income inequality because participating in the labor market is the main path to earning a living for most ordinary people (McCall and Percheski, 2010; Khan and Riskin, 2005). Yet, along with the development of capital markets and increasingly diversified assets, capital income, a reward of investing in a specific tangible asset (e.g. rent, interest, and dividends), is becoming an important source of household income and its impact on household wealth accumulation is greater than wage income (Piketty and Saez, 2014). Studies conducted in developed and developing countries have generally documented that the contribution of income sources to inequality is unbalanced, and capital income has a larger marginal impact on income inequality than other types of income, such as wages and welfare (Fräßdorf et al., 2011; Lynch, 2003; Ranaldi and Milanović, 2022). This impact of capital income is especially evident in China (Li et al., 2020; Luo et al., 2020).

Given that household assets are the basis of capital income, many scholars suggest that inequality caused by capital income should be investigated with a focus on assets (Ning and Jiang, 2018). In addition to the absolute scale of assets, asset types matter for understanding the causes of income inequality because the return on investment varies for asset types. Generally, household assets can be broadly divided into two categories: FA and non-FA. Housing, as the main form of non-FA, can produce appreciable economic benefits, such as rental income, and enhance access to home equity credit, thus increasing households’ total income (Sierminska and Takhtamanova, 2012). Moreover, housing disparities have been established as a contributor to rising income inequality (Christophers, 2021). For instance, a cross-national study using harmonized data on advanced economies showed that the rewards of investing in housing are the central determinant of gaps in household income (Pfeffer and Waitkus, 2021). In addition, FA (e.g. bank interest, bond interest, and dividends) also provide investment returns and have a much higher return on investment than that of HA on average (Killewald et al., 2017). In effect, the growing importance of FA to household income has been well-recognized and widely debated (Killewald et al., 2017). Nevertheless, the impacts of FA on income inequality have received little attention.

The significance of household assets to income accumulation and inequality may be more evident in transitional economies, especially in China. Before the economic reform that began in the late 1970s, China had a collectivist ideology and planned economy, in which housing was administratively distributed on egalitarian terms, the financial market was prohibited, and thus no families obtained income by investing in private assets (Xie et al., 2009). However, following the introduction of a market-oriented reform policy, alongside the rapid privatization of public enterprises, housing commercialization was enacted in the 1990s that aimed to turn housing from welfare into a commodity, and it was thoroughly implemented from 1998 to 2003 (Sato, 2006). The housing commercialization reform led to a rapid rise in the home ownership rate from about 55% in the early 1990s to above 90% in 2015 (Chen et al., 2019). Moreover, Chinese households’ housing wealth dramatically increased with the booming housing market and soaring property prices since 2003. According to a report issued by the People's Bank of China (2020), the share of HA in the total assets of urban households in China was 59.1%, which was much higher than that of other major economies in the world, that is, 24.3%, 48.9%, 20.8%, and 36.7% for the United States, United Kingdom, Germany, and Japan, respectively, in 2018. Meanwhile, housing reform triggered an uneven distribution of home ownership and high-quality housing, which gave rise to a widening gap in housing wealth between housing owners and non-owners (Song and Xie, 2014; Walder and He, 2014; Wu, 2013). Therefore, HA have not only become a new mechanism for household income generation, but also a major dimension of social inequality in contemporary China.

In China, the establishment of the stock market in the early 1990s, such as the founding of the Shanghai Stock Exchange and the Shenzhen Stock Exchange, marked the emergence of financial markets. With the continuous development of the capital markets, financial products are becoming more diverse and complex, and a growing number of Chinese households possess FA by actively participating in financial markets. Studies conducted in China show that FA have significantly enhanced the accumulation of household income, and their impact is much greater than that of HA (Ning and Jiang, 2018). However, as investors are typically required to possess certain essential preconditions for entry into financial markets, including financial literacy, economic resources, and risk preference (Zou and Deng, 2019), the opportunity to participate in financial markets is not even across all households. Thus, only a small number of households can participate in financial markets, a trend which is exacerbated in China due to the excessive concentration of HA in households’ total assets. According to the statistics provided by the Survey and Research Center for China Household Finance (2019), the proportion of FA to total assets in China in 2018 was 11.8%, in contrast to 42.6% for the United States in 2016 (calculated by the Survey of Consumer Finance). Fueled by the effect of FA on income accumulation and the disparities in access to the financial market, there may be a tremendous gap in household income by investing in FA in China.

To sum up, HA and FA are sources of household income and contribute to household income growth. Meanwhile, these two types of household assets have been treated as important dimensions of social stratification in contemporary China. However, rigorous empirical analyses of their effects on income inequality in China have been scant. Moreover, although FA have a greater impact on household income accumulation than HA, which is a more important contributor to income inequality it is not clear, as there is little empirical evidence so far.

Finally, our study focuses on urban residents in China, for several reasons: Housing in rural China is not transacted on the real estate market since rural residents only have land use rights (Deng, 2013); by contrast, urban residents can invest in HA and obtain income due to the private ownership of their houses; and in addition, urban residents, relative to rural residents, have more access to financial products due to the more developed financial markets in urban areas. Given these factors, there may be a more evident relationship between household assets and income inequality in urban areas. Our study seeks to estimate the causes of income inequality among Chinese urban households from the household assets perspective.

Family SES and regional marketization as moderators

The SES represents personal capability, economic resources, and social capital (Conger and Donnellan, 2007), which affect one's willingness and ability to invest in certain assets. Previous studies in China have indicated that people with a higher SES, such as members of the Communist Party of China (CPC), professionals, and residents holding urban household registration (hukou), possess more housing and housing that is newer and of higher quality (Song and Xie 2014; Walder and He 2014; Wu 2013). In addition, individuals with a higher SES have better health, more wealth, and higher levels of educational and occupational status (Campbell, 2006; Rosen and Wu, 2004; Shum and Faig, 2006), which are correlated with financial market participation. Empirical studies in China have demonstrated the association between investment in FA and socioeconomic characteristics including investment experience, financial literacy, political background, and subjective SES (Ge et al., 2021; Liu et al., 2023; Yin et al., 2015; Zou and Deng, 2019). Another study conducted in China indicated that low-income families have little opportunity to participate in the capital market to acquire FA and returns on investment (Ning et al., 2016).

Moreover, there is variation in the channels to obtain income for families with different levels of SES. Compared with families with lower SES who rely on wage income, for families with higher SES capital income is the main source of income (Piketty and Saez, 2014). Studies in the United States showed that the incomes of the richest 10% of households are mainly from capital income, whereas poor households are unable to acquire such income for a long period, which is the primary cause of the widening gap between the rich and the poor since the 1970s (Piketty and Saez, 2003). Similarly, although capital income experienced slow growth from 1988 to 2009 in China, the distribution of capital income is highly unequal compared to wage income, and capital income is substantial for high-income families (Chi, 2012). One study in China showed that capital income from investing in housing contributes more to income inequality because of the soaring real estate prices in recent years (Ning et al., 2016). Given these findings, it is reasonable to expect that SES could moderate the relationship between household assets and income inequality. Specifically, the ability to hold household assets may differ for families with higher and lower SES, and higher SES can provide an incentive for investing in assets to obtain capital income, thus leading to greater income inequality.

The associations between household assets and income inequality may also depend on marketization levels. China's market-oriented reform in the late 1970s has brought profound social and economic transformations. One of the principal aspects of the market transition in China has been the shifting of the allocation of economic resources from the government to the market (Nee, 1989). However, there is a gradation in the marketization process across provinces in mainland China since economic reform has diffused from the coast to the interior provinces. Hence, despite the tremendous economic growth, the opportunity to enjoy benefits from marketization is unevenly distributed across regions, which has become a driver of regional inequality during the economic reform (see Wu, 2019). Thus, regional marketization levels are also a factor underlying income inequality in China. For instance, using province-level panel data spanning from 1978 to 2004, Hao and Wei (2010) found that marketization levels played a significant role in the income gap between the inland and coastal provinces. Moreover, income inequality is stimulated by regional marketization levels. He and Wu (2017) found that occupational segregation has a positive effect on wage income inequality between men and women in China, and the effect would increase with the prefectural level of marketization.

With the process of marketization reform, the housing market and financial market gradually emerged, which provides households with the opportunity to invest in commercial apartments and financial products. However, in regions with higher levels of marketization there are more policies and institutions that can promote economic resource allocation by the market mechanism, thus creating more competitive and developed housing and financial markets (Fan et al., 2019). Hence, people who live in regions with higher levels of marketization are more likely to hold HA because there are enough housing resources and more mortgage products that facilitate the obtaining of lower mortgage rates (Wang et al., 2020). For instance, Chen (2015) found that households in regions with a high degree of property rights protection have more HA and property income. Furthermore, a more developed financial market in highly marketized regions not only provides abundant financial products, but also allows people to leverage financing and actively invest in FA (Xin and Yang, 2023). Therefore, differences in regional marketization levels are likely to contribute to gaps in household assets across Chinese households. Although households in higher-marketization-level regions have more opportunities to hold assets, there may be a widening income gap among households in these regions since income inequality can be driven by gaps in household assets. In summary, regional marketization levels may strengthen the effect of household assets on income inequality.

Given the discussion and findings above, it is reasonable to believe that family SES and marketization can moderate the impacts of household assets on income inequality. Thus, another purpose of this study, is to ascertain whether family SES and marketization function in the relationship between household assets and income inequality in urban China.

Methods

Data

The data used in our analysis came from the CGSS, which is a cross-sectional survey conducted by the National Survey Research Center (NSRC) at Renmin University of China. The CGSS is designed to be the Chinese counterpart of the General Social Survey in the United States (Bian and Li, 2012). Using a multistage probability-proportional-to-size sampling method, each wave of the CGSS covers more than 10,000 sample households and family members from 28 provincial units in China. The baseline CGSS survey was in 2003, followed by ten follow-up waves in 2005, 2006, 2008, 2010–2013, 2015, 2017, 2018, and 2021.

Because data on HA and FA were only collected since the 2012 wave, we analyze the data from six waves of CGSS (2012, 2013, 2015, 2017, 2018, and 2021). The original data from these six waves were collected from 67,888 households. We first restricted our sample to adults aged 18 years and above that resided in urban areas at the time of the survey, which resulted in the exclusion of over 37% of the initial sample (25,428 observations). During data cleaning, 5309 observations were dropped because of missing data on household income variables, and 2307 were dropped because of missing data on other variables. We employed the multiple imputation method to impute data and obtained a seven-year pooled cross-sectional dataset of 41,783 observations. Additional analysis of excluding missing data on household income revealed similar results to those reported below (results not shown but available upon request).

Measures

Dependent variable: Income inequality

Different from the Gini coefficient, which is mainly used in the measurement of income inequality at the group level, income inequality at the household level can be measured by the relative deprivation in income, which highlights that income inequality comes from the comparison of households with high-income reference groups (Kakwani, 1984). This approach to measuring income inequality has been used in a growing number of empirical studies on Chinese households (He et al., 2021; Li and Xiao 2021; Qin et al., 2021). Commonly employed income deprivation indexes are the Yitzhaki index (Yitzhaki, 1979) and Kakwani index (Kakwani, 1984). The mean of the Yitzhaki index for households in the reference group is the mean income multiplied by the Gini coefficient of the reference group (Hastings, 2019), and the mean of the Kakwani index for households in the reference group is equal to the Gini coefficient of the reference group (Adjaye-Gbewonyo and Kawachi, 2012). The specific formula for the Yitzhaki index is shown in the following equation:

\begin{matrix} Y i t z h a k i_{i} = \frac{1}{n} \sum_{j = i + 1}^{n} (x_{j} - x_{i}) = γ_{x_{i}}^{+} [(μ_{x_{i}}^{+} - x_{i})], \end{matrix}

(1)

where

Y i t z h a k i_{k}

is the Yitzhaki index for the household i;

x_{i}

represents the income of household i;

x_{j}

is the incomes of all households j whose incomes are higher than household i’ s; n is the total number of households in the sample;

μ_{x_{i}}^{+}

is the average income of the sample whose income was higher than

x_{i}

in the reference group; and

γ_{x_{i}}^{+}

is the percentage of samples whose income was higher than

x_{i}

in the reference group.

The Kakwani index was developed from the Yitzhaki index, which is the ratio of the Yitzhaki index divided by the mean income of the reference group. The specific calculation formula for the Kakwani index is as follows:

\begin{matrix} K a k w a n i_{i} = \frac{1}{n μ_{X}} \sum_{j = i + 1}^{n} (x_{j} - x_{i}) = γ_{x_{i}}^{+} [\frac{(μ_{x_{i}}^{+} - x_{i})}{μ_{X}}], \end{matrix}

(2)

where

K a k w a n i_{i}

is the Kakwani index for the household i; X denotes a reference group; i, n,

x_{i}

, and

x_{j}

are the same as in the equation (1); and

μ_{X}

represents the average income of all samples in the reference group.

The value of the Yitzhaki index is greater than zero, and the range of the Kakwani index is from zero to one. The larger the two indexes of household i, the more unequal is household i. Note that the Kakwani index not only overcomes the Yitzhaki index's sensitivity to changes in the income scale. In addition, it can make up for the indecomposability and non-addition shortcomings of the Gini coefficient (He et al., 2021). Thus, we use the Kakwani index as the main measure of income inequality.

We use households’ total income, based on the CGSS question “What was your family's total income last year?” as the main measure of household income, and then estimated the income inequality at the household level by calculating the Kakwani index of the total income (KITI). Given that household assets are the main source of capital income, we employed the Kakwani index of household capital income (KICI) as an alternative measure for a robustness check.

Independent variables: HA and FA

HA is a continuous variable measured by the number of houses possessed by the surveyed households, in line with previous studies (Asadullah et al., 2018). FA is determined based on the CGSS question “Does your family currently invest in stocks, funds, bonds, futures, warrants, or currency?” (the answers are either “Yes” or “No”). Consistent with the literature (Cheng et al., 2023), from this, we generate a dichotomous variable that equals 1 if a household owns at least one type of financial asset and 0 otherwise.

Moderators: Family SES and regional marketization

Distinct from singular dimensions of SES, such as education, occupation, and income, the international socioeconomic index (ISEI) uses a comprehensive scale of those indicators for the measurement of SES (Ganzeboom et al., 1992). The ISEI has good reliability and validity and has been widely employed in social science research globally (Hooijsma and Juvonen 2021; Pronina et al. 2021). In line with prior studies (Liu et al., 2020), we used the highest ISEI score among a given family's members to measure that family's SES. Marketization was measured by the marketization index at the provincial level, which is calculated by Xiaolu Wang and his associates at China's National Economic Research Institute (see Wang et al., 2018). The higher the score, the higher the marketization level.

Covariates

Consistent with previous works on income inequality that consider demographic and socioeconomic factors (Li et al., 2020), we controlled for: (a) gender, which is a binary variable that equals 0 and 1 if the respondents are female and male, respectively; (b) age, which is coded as a continuous variable in years; (c) marital status, which is coded as a dichotomous variable: 0 indicating not married and 1 indicating married; (d) self-rated health, which is measured by a 5-point Likert scale coded from 1 (unhealthy) to 5 (very healthy); (e) educational level, which is measured by years of formal education; (f) CPC membership, which is a dichotomous variable: 0 = not CPC member, 1 = CPC member; (g) working status, which is measured by whether the respondent had more than one hour of paid employment in the last week, with 0 indicating “not working” and 1 indicating “working;” (h) hukou status, which was coded as a dichotomous variable: 0 indicating rural hukou and 1 indicating urban; and (i) family size, which is measured by the number of co-residing family members. In addition, regional characteristics may affect the income gap in a certain region. Since the public data of CGSS only provide the province-level location of the surveyed households, we controlled for province-level variables drawn from the China Statistical Yearbook, including gross domestic product per capita (GDPPC) and the ratio of social security expenditure to total fiscal expenditure (RSSE). Finally, we included province and survey year fixed effects to control for unobservable regional characteristics and time-varying traits. The standard errors reported are all clustered at the provincial level. The statistical results of the variables above are shown in Table 1.

Table 1.

Descriptive statistics of variables using the urban sample.

Variables	M/percentage	SD	Range	Observations
Dependent variables
KITI	0.56	0.06	0.48–0.65	41,783
KICI	0.78	0.04	0.73–0.84	41,783
Independent variables
HA	1.13	0.67	1–20	41,783
FA
Yes	14.09%			5886
No	85.91%			35,897
Mediating variables
Family socioeconomic status	40.00	11.58	16–90	41,783
MI	7.54	1.71	2.33–9.97	41,783
Control variables
Gender
Male	47.55%			19,867
Female	52.45%			21,916
Age	48.96	17.25	18–118	41,783
Marital status
Married	75.38%			31,494
Unmarried	24.62%			10,289
Self-rated health	3.69	1.01	1–5	41,783
Education level	10.28	4.47	0–19	41,783
CPC membership
CPC member	14.35%			5996
Non-CPC member	85.65%			35,787
Working status
Working	52.19%			21,807
Not working	47.81%			19,976
Hukou status
Urban	67.19%			28,072
Rural	32.81%			13,711
Family size	2.88	1.44	1–50	41,783
GDPPC (10,000 CNY)	6.40	3.31	1.60–16.42	41,783
RSSE	0.13	0.04	0.08–0.28	41,783

Notes: M = mean; SD = standard deviation; KITI = Kakwani index of total income; KICI = Kakwani index of household capital income; HA = housing assets; FA = financial assets; MI = marketization index; CPC = Communist Party of China; GDPPC = gross domestic product per capita; RSSE = ratio of social security expenditure to total fiscal expenditure.

Our statistical analysis consists of three parts. First, using the total sample from CGSS, we estimate the change in income inequality levels between 2015 and 2021 by calculating the Kakwani index of total household income and capital income in urban and rural areas. Second, we pooled all waves of the CGSS data and restricted them to the urban sample, and then examined the basic association between HA and income inequality, controlling for all covariates in urban China. Hierarchical linear models (HLM) were established to account for both the individual, household, and provincial levels of heterogeneity due to the data clustering effects. The HLM models enable us to estimate patterns of variation within and across provinces simultaneously by allowing intercepts and, eventually, slopes to vary (Raudenbush and Bryk, 2002). In the HLM models, we kept the individual and household variables used as the determinants of household income at the first level and GDPPC and RSSE as contextual variables at the second level. Furthermore, we observed the effects of the two types of assets in different years by using single-year data, which helps to identify the causes of income inequality in urban China. Finally, we added the interaction with two moderators and two types of assets to predict income inequality in the basic model, thus testing the moderating effects of family SES and regional marketization in urban China. We used Stata version 16 for data cleaning and model analysis.

Results

The trend in income inequality in China

We estimate the trend in national income inequality as measured by the Kakwani index. As is shown in Figure 1, in general, both KITI and KICI had been increasing during the survey period, with the mean score rising from 0.484 in 2012 to 0.646 in 2021 and 0.726 in 2012 to 0.843 in 2021, respectively, suggesting that there was a significant rise in income inequality between 2012 and 2021. It should be noted that, in addition to an increase in KITI and KICI between 2012 and 2018, there was a continuous upward trend in KITI and KICI during the period from 2018 to 2021, thus indicating that income inequality increased after the COVID-19 pandemic. We also used a conventional measure of inequality, that is, the Gini coefficient, to show the changes in income inequality (see Figure 2). During the initial period from 2012 to 2021, we observed a significant increase in the Gini coefficients of total income and capital income, which displayed a similar trend in the income inequality as measured by the Kakwani index.

Figure 1.

Income inequality in China measured by the Kakwani index, 2012–2021.

Figure 2.

Income inequality in China measured by the Gini coefficient, 2012-2021.

Table 2 presents income inequality by urban–rural division. We found that the KITI and KICI of urban households were higher than those of rural households in each survey year, indicating that relative to households in rural China, there was a greater income gap in households in urban China. Moreover, the gap in the income inequality between urban households and rural households expanded over time.

Table 2.

The trend in household income inequality in China by urban–rural division.

Survey year	KITI			KICI
Survey year	Urban	Rural	Urban–rural	Urban	Rural	Urban–rural
2012	0.474	0.445	0.029	0.733	0.702	0.031
2013	0.487	0.452	0.035	0.743	0.703	0.040
2015	0.595	0.524	0.071	0.813	0.777	0.036
2017	0.638	0.535	0.103	0.850	0.780	0.070
2018	0.646	0.551	0.095	0.855	0.783	0.072
2021	0.745	0.561	0.184	0.901	0.787	0.114

Note: KITI = Kakwani index of total income; KICI = Kakwani index of capital income.

Household assets and income inequality in urban China: A multi-level analysis

The results using the HLM models are presented in Panel A of Table 3. In addition, to show the extent of the effect of two types of household assets (i.e. HA and FA) on income inequality, the standardized regression coefficients of dependent variables are also reported, in Panel B of Table 3. The advantage of the standardized regression coefficients is that they can eliminate the influence of different dimensions of variables on the estimated coefficients, which allows us to directly compare the estimated coefficients for different dependent variables in the same model (Wang et al., 2023). The absolute value of the standardized regression coefficients could reflect the extent of the influence of dependent variables on the independent variable. The larger the absolute value, the greater the effect of the independent variable on the independent variable.

Table 3.

Hierarchical linear models on the association between household assets and household income inequality in urban China.

Panel A
	DV: KITI
	2012–2021	2012	2013	2015	2017	2018	2021
	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6	Model 7
HA^a	0.038***	0.029***	0.016***	0.045***	0.033***	0.032***	0.128***
	(0.003)	(0.006)	(0.003)	(0.006)	(0.005)	(0.005)	(0.016)
FA^a	0.070***	0.030***	0.026***	0.092***	0.058***	0.055***	0.273***
	(0.006)	(0.003)	(0.006)	(0.013)	(0.010)	(0.010)	(0.037)
Control variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Observations	41,783	6948	7020	6470	8043	8728	4574

Panel B
	DV: KITI
	2012–2021	2012	2013	2015	2017	2018	2021
	Model 8	Model 9	Model 10	Model 11	Model 12	Model 13	Model 14
HA^b	0.061***	0.106***	0.053***	0.091***	0.079***	0.078***	0.120***
	(0.004)	(0.010)	(0.010)	(0.013)	(0.011)	(0.011)	(0.015)
FA^b	0.161***	0.118***	0.143***	0.272***	0.185***	0.187***	0.310***
	(0.013)	(0.029)	(0.034)	(0.038)	(0.032)	(0.033)	(0.042)
Control variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Observations	41,783	6948	7020	6470	8043	8728	4574

Notes: KITI = Kakwani index of total income; HA = housing assets; FA = financial assets; maximum likelihood estimations are reported; standard errors are shown in parentheses; the coefficient estimates of the control variables and constants are not shown in Table 3; +p < 0.1, *p < 0.05, **p < 0.01, ***p < 0.001 (two-tailed tests).

Refers to non-standardized regression coefficients.

Refers to standardized regression coefficients.

The results of Panel A of Model 1 in Table 3 suggested that both HA and FA were positively associated with KITI (b = 0.038, p < 0.001, b = 0.070, p < 0.001, respectively) in urban China during the survey period from 2012 to 2021. Specifically, after controlling for the effects of all covariates, every one-unit increase in HA was associated with a 0.038-point increase in KITI, and there was a 0.07-point increase in KITI between households holding FA and households not holding FA. We further estimated the coefficients for HA and FA on KITI in each wave of the CGSS, thus examining the association between household assets and income inequality over time. The results in Panel A of Models 2–7 in Table 3 show that the coefficients of both HA and FA on KITI were significantly positive, suggesting that a stable relationship between HA and FA and income inequality in urban areas during the survey period.

Furthermore, Panel B in Table 3 shows the standardized regression coefficients of the dependent variables. The results of Model 8 show that the standardized regression coefficient for HA is almost equal to that of FA (0.106 vs. 0.118), suggesting that HA and FA had similar effects on income inequality in the early 2010s. However, there are larger gaps in the standardized regression coefficients for HA and FA in Models 9–14, indicating that FA became an increasingly important factor underlying income inequality among urban households when compared to HA. Additionally, comparing the standardized regression coefficients across models, it can clearly be seen that the effect of FA on income inequality significantly increased from 2012 to 2021, but there was only a slight increase in the effect of HA on income inequality during the same period, as evidenced by the change in the magnitude of the coefficients of FA and HA in Model 9. Taken together, these results strongly indicate that FA play a more important role than HA in income inequality in contemporary urban China.

The moderating effects of family SES and regional marketization

Table 4 shows the moderating effects of family SES and regional marketization on the link between household assets and income inequality among urban households using HLM models. The significant interaction effect of family SES in Models 1–2 (b = 0.004, p < 0.001, b = 0.006, p < 0.001, respectively) indicates that both HA and FA had stronger effects on income inequality among households with higher SES than among households with lower SES in urban China. In addition, Models 3–4 also reveal a significant interaction effect of regional marketization (b = 0.025, p < 0.001, b = 0.050, p < 0.001, respectively), suggesting that the effect of household assets on income inequality was more severe in regions with higher levels of marketization than in regions with lower levels of marketization in urban China.

Table 4.

Hierarchical linear models on the moderating effects of family socioeconomic status (SES) and regional marketization in urban China.

Variables	DV: KITI
Variables	Model 1	Model 2	Model 3	Model 4
HA	0.122***		0.155***
	(0.010)		(0.013)
FA		0.194***		0.335***
		(0.020)		(0.029)
Family SES	0.005***	0.001***
	(0.000)	(0.000)
MI			0.018***	0.004***
			(0.005)	(0.001)
Interactions
HA * Family SES	0.004***
	(0.000)
FA * Family SES		0.006***
		(0.000)
HA * MI			0.025***
			(0.002)
FA * MI				0.050***
				(0.003)
Control variables	Yes	Yes	Yes	Yes
Observations	41,783	41,783	41,783	41,783

Notes: KITI = Kakwani index of total income; HA = housing assets; FA = financial assets; MI = marketization index; maximum likelihood estimations are reported; standard errors are shown in parentheses; the coefficient estimates of the control variables and constants are not shown in Table 4; +p < 0.1, *p < 0.05, **p < 0.01, ***p < 0.001 (two-tailed tests).

The significant moderating effects of family SES and marketization level represent the influences of socioeconomic and regional characteristics on the link between household assets and income inequality. On the one hand, the likelihood of holding household assets is positive related to total family economic resources (Liu et al., 2023). Hence, higher SES families gain more economic benefits from their investments in HA and FA compared to lower SES families who have little opportunity to invest in these two types of assets, thus leading to an increasing income gap among families with different levels of SES. On the other hand, the development of a market-oriented economy, which is also positively related to the regional marketization level, is beneficial to the rise of asset prices and expansion of asset markets (Chen, 2015; Xin and Yang, 2023). This means that, in relation to those families who live in areas with lower levels of marketization, families who live in areas with higher levels of marketization are more likely to invest in HA and FA and then gain the resulting economic returns, thus exacerbating income inequality among families at the regional level.

Robustness check

We measured income inequality using KICI to test the robustness of the effects of HA and FA on income inequality. As presented in Panel A in Table 5, the estimated coefficients of HA and FA were significantly positive in all models, indicating that both HA and FA had a significant positive effect on capital income inequality. Panel B in Table 5 shows the standardized regression coefficients of HA and FA. The results indicate that the magnitudes of the standardized coefficients of FA were greater than those of HA in all models, which is consistent with the results of Panel B in Table 3 using KITI as the dependent variable.

Table 5.

Hierarchical linear models on the association between household assets and property income inequality in urban China.

Panel A
	DV: KICI
	2012–2021	2012	2013	2015	2017	2018	2021
	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6	Model 7
HA^a	0.205***	0.004***	0.010***	0.175***	0.234***	0.228***	0.443***
	(0.011)	(0.001)	(0.003)	(0.028)	(0.026)	(0.024)	(0.052)
FA^a	0.252***	0.005***	0.012***	0.314***	0.275***	0.364***	0.474***
	(0.023)	(0.001)	(0.003)	(0.059)	(0.055)	(0.051)	(0.123)
Control variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Observations	41,783	6948	7020	6470	8043	8728	4574

Panel B
	DV: KICI
	2012–2021	2012	2013	2015	2017	2018	2021
	Model 8	Model 9	Model 10	Model 11	Model 12	Model 13	Model 14
HA^b	0.087***	0.019***	0.022***	0.077***	0.101***	0.102***	0.126***
	(0.005)	(0.004)	(0.005)	(0.012)	(0.011)	(0.011)	(0.015)
FA^b	0.154***	0.020***	0.046***	0.205***	0.162***	0.233***	0.164***
	(0.014)	(0.005)	(0.011)	(0.038)	(0.032)	(0.033)	(0.043)
Control variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Observations	41,783	6948	7020	6470	8043	8728	4574

Notes: KITI = Kakwani index of total income; KICI = Kakwani index of capital income; HA = housing assets; FA = financial assets; maximum likelihood estimations are reported; standard errors are shown in parentheses; the coefficient estimates of the control variables and constants are not shown in Table 5; +p < 0.1, *p < 0.05, **p < 0.01, ***p < 0.001 (two-tailed tests).

Refers to non-standardized regression coefficients.

Refers to standardized regression coefficients.

The results in Table 6 were estimated by multiple linear regression models based on the ordinary least squares method. We found that there was a significant association of HA and FA with KITI in all models of Panel A in Table 6. Moreover, FA had a stronger effect on KITI than HA when comparing the magnitudes of the standardized regression coefficients of the two independent variables (see Panel B in Table 6). These results indicate that both HA and FA contribute to income inequality in urban China, while FA, in relation to HA, play a more significant role in income inequality.

Table 6.

Ordinary least squares regressions on the association between household assets and household income inequality in urban China.

Panel A
	DV: KITI
	2012–2021	2012	2013	2015	2017	2018	2021
	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6	Model 7
HA^a	0.039***	0.029***	0.016***	0.046***	0.034***	0.033***	0.127***
	(0.006)	(0.003)	(0.004)	(0.012)	(0.009)	(0.009)	(0.033)
FA^a	0.069***	0.030***	0.026***	0.092***	0.058***	0.055***	0.266***
	(0.008)	(0.005)	(0.006)	(0.018)	(0.015)	(0.013)	(0.050)
Control variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Observations	41,783	6948	7020	6470	8043	8728	4574

Panel B
	DV: KITI
	2012–2021	2012	2013	2015	2017	2018	2021
	Model 8	Model 9	Model 10	Model 11	Model 12	Model 13	Model 14
HA^b	0.059***	0.110***	0.052***	0.090***	0.076***	0.078***	0.120***
	(0.009)	(0.011)	(0.014)	(0.024)	(0.021)	(0.022)	(0.031)
FA^b	0.158***	0.190***	0.131***	0.279***	0.186***	0.190***	0.323***
	(0.019)	(0.027)	(0.030)	(0.055)	(0.050)	(0.045)	(0.058)
Control variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Observations	41,783	6948	7020	6470	8043	8728	4574

Notes: KITI = Kakwani index of total income; HA = housing assets; FA = financial assets; maximum likelihood estimations are reported; standard errors are shown in parentheses; the coefficient estimates of the control variables and constants are not shown in Table 6; +p < 0.1, *p < 0.05, **p < 0.01, ***p < 0.001 (two-tailed tests).

Refers to non-standardized regression coefficients.

Refers to standardized regression coefficients.

Discussion and conclusion

The economic transition initiated in China in the late 1970s resulted in a rapid increase in income among Chinese households, alongside which came an equally rapid increase in income inequality. Although numerous studies have provided detailed information on income inequality in China up to the early 2010s, there is little research on the trend in income inequality from the mid-2010s to the present. In addition, the effects of institutional factors on income inequality have received extensive attention from researchers of various backgrounds. However, empirical research has rarely examined the effects of household assets on income inequality. Hence, identifying the causes of income inequality from the household assets perspective is crucial because household assets are the main source of capital income, which has been recognized as an important factor underlying income gaps among Chinese households.

Using pooled cross-sectional data from the six waves of the CGSS, this study estimates the levels of income inequality from 2012 to 2021, and further examines the effects of two types of household assets (i.e. HA and FA) on income inequality in urban China, as well as the moderating effects of family SES and regional marketization. We find that income inequality, measured by the Kakwani index of total income and capital income, continuously increase in urban and rural areas between 2012 and 2021, while income inequality for urban households was higher than that for rural households. The trend for income inequality since 2010 has been investigated by recent studies using other data sources. For instance, according to Kanbur et al.'s (2021) estimations using the data from the China Family Panel Studies, the Gini coefficients of household income per capita were 0.522 in 2010, 0.501 in 2012, 0.500 in 2014, 0.508 in 2016, and 0.521 in 2018. Our results, estimated using CGSS data, indicate that income inequality increased from 2012 to 2015, which contradicts the estimations of Kanbur et al. (2021) during the period from 2012 to 2014. We believe that differences in the data sources and measures between the two studies may account for the different estimations of income inequality. However, Kanbur et al. (2021) found an increase in income inequality during the period from 2014 to 2018, which is consistent with our results. Therefore, it can be asserted based on different data sources that income inequality has been sustained at a high level since the mid-2010s.

Regression results show that both HA and FA are related to income inequality, and the latter assets have a stronger effect on income inequality. This suggests that FA, relative to HA, are a more important factor underlying income inequality in urban China. Moderating effect analysis indicates that the association between the two types of household assets and income inequality strengthens along with increases in family SES and regional marketization levels. Our study adopted comprehensive robustness testing methods to cross-validate the research conclusions, including replacing the measure of income inequality and changing the estimation model. The research conclusions were consistent, indicating the stability and reliability of the research conclusions.

Our study has several limitations. First, the data were cross-sectional and thus could not be used to identify the causal relationship between household assets and income inequality. Future research should use a longitudinal design to make stronger causal inferences and identify how changes in household assets affect the trajectory of income inequality. Second, we used the number of HA possessed by households and whether households possess HA and FA as the measure of household assets. However, the GGSS provides no data pertaining to the monetary value of these HA and FA and whether the surveyed households obtain returns by investing those assets. Such information should be collected in the future because it is conducive to establishing the relationship between household assets and income inequality. Third, since the CGSS public data contain only province-level residential information, the control variables at the province level may be too rough to control other factors that the respondents face. Therefore, it is very necessary to merge more detailed information at the county level or prefecture level with the micro survey data to control the effect of regional characteristics in the future. Finally, because of the lack of detailed information on household assets in the CGSS data we do not explain and verify why FA, relative to HA, are the stronger contributor to income inequality. We infer that FA, compared to HA, have a higher return on investment, thus playing a more influential role in household income growth. Hence, there may be a significant income gap between households with FA and households without FA. This should be tested in future studies using more detailed data.

Despite these limitations, our study contributes to the existing literature on the change in and determinants of income inequality in China. First, this study shows the latest circumstances of income inequality in China, complementing the literature on measuring income inequality in China. Second, to our knowledge, this is the first study to investigate the effects of HA and FA on income inequality, thus extending the literature on inequality and social stratification in contemporary China by revealing the causes underlying rising income inequality in China from the household assets perspective. Third, our study highlights that inequality in family SES and regional marketization levels widen disparities in household income, which furthers our understanding of the logic of income inequality evolution and enlightens policymakers developing strategies aiming at alleviating income inequality.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research was supported by the Major Program of the National Social Science Fund of China (grant number: 22&ZD191), the Youth Program of the National Social Science Fund of China (grant number: 23CSH043), and the Philosophy and Social Science Research Program of Heilongjiang Province (grant number: 22SHB168).

Contributorship

Jiankun Liu designed the research. Xiaobin He performed the analysis. Jiankun Liu, Xiaobin He, and Yinxi Dong wrote and revised the manuscript.

ORCID iD

Jiankun Liu

References

Adjaye-Gbewonyo

Kawachi

(2012) Use of the Yitzhaki Index as a test of relative deprivation for health outcomes: A review of recent literature. Social Science & Medicine 75(1): 129–137.

Arshed

Anwar

Hassan

, et al. (2019) Education stock and its implication for income inequality: The case of Asian economies. Review of Development Economics 23(2): 1050–1066.

Asadullah

Xiao

Yeoh

(2018) Subjective well-being in China, 2005–2010: The role of relative income, gender, and location. China Economic Review 48(3): 83–101.

Bian

(2012) The Chinese General Social Survey (2003–2008) sample designs and data evaluation. Chinese Sociological Review 45(1): 70–97.

Campbell

(2006) Household finance. The Journal of Finance 61(4): 1553–1604.

Chen

(2015) Protection of property rights and property income: Evidence from micro households. Modern Economic Science 37(4): 29–38 (in Chinese).

Chen

Miao

(2019) Housing and subjective class identification in urban China. Chinese Sociological Review 51(3): 221–250.

Cheng

Wei

Wang

, et al. (2023) Does financial market participation eradicate household energy poverty? Energy Economics 122: 106687.

Chi

(2012) Capital income and income inequality: Evidence from urban China. Journal of Comparative Economics 40(2): 228–239.

10.

Christophers

(2021) A tale of two inequalities: Housing-wealth inequality and tenure inequality. Environment and Planning A: Economy and Space 53(3): 573–594.

11.

Conger

Donnellan

(2007) An interactionist perspective on the socioeconomic context of human development. Annual Review of Psychology 58(1): 175–199.

12.

Davies

Sandström

Shorrocks

, et al. (2011) The level and distribution of global household wealth. The Economic Journal 121(551): 223–254.

13.

Deng

(2013) Land development right and collective ownership in China. Post-Communist Economies 25(2): 190–205.

14.

Fan

Wang

(2019) Institutional reform and economic growth of China: 40-year progress toward marketization. Acta Oeconomica 69(S1): 7–20.

15.

Fräßdorf

Grabka

Schwarze

(2011) The impact of household capital income on income inequality—a factor decomposition analysis for the UK, Germany and the USA. The Journal of Economic Inequality 9(1): 35–56.

16.

Fuller

Johnston

Regan

(2020) Housing prices and wealth inequality in Western Europe. West European Politics 43(2): 297–320.

17.

Ganzeboom

Graaf

Treiman

, et al. (1992) A standard international socio-economic index of occupational status. Social Science Research 21(1): 1–56.

18.

Chen

Zou

, et al. (2021) Political background and household financial asset allocation in China. Emerging Markets Finance and Trade 57(4): 1232–1246.

19.

Golley

Meng

(2012) Income inequality and carbon dioxide emissions: The case of Chinese urban households. Energy Economics 34(6): 1864–1872.

20.

Han

(2022) Anti-corruption and poverty alleviation: Evidence from China. Journal of Economic Behavior and Organization 203(1): 150–172.

21.

Hao

Wei

(2010) Fundamental causes of inland–coastal income inequality in post-reform China. Annals of Regional Science 45(1): 181–206.

22.

Hastings

(2019) Who feels it? Income inequality, relative deprivation, and financial satisfaction in US states, 1973-2012. Research in Social Stratification and Mobility 60: 1–15.

23.

(2017) Marketization, occupational segregation, and gender earnings inequality in urban China. Social Science Research 65(1): 96–111.

24.

Zhou

, et al. (2021) An empirical analysis of the impact of income inequality and social capital on physical and mental health-take China’s micro-database analysis as an example. International Journal for Equity in Health 20(1): 1–14.

25.

Honohan

(2008) Cross-country variation in household access to financial services. Journal of Banking and Finance 32(11): 2493–2500.

26.

Hooijsma

Juvonen

(2021) Two sides of social integration: Effects of exposure and friendships on second-and third-generation immigrant as well as majority youth’s intergroup attitudes. International Journal of Intercultural Relations 80(1): 51–63.

27.

Kakwani

(1984) The relative deprivation curve and its applications. Journal of Business and Economic Statistics 2(4): 384–394.

28.

Kanbur

Wang

Zhang

(2021) The great Chinese inequality turnaround. Journal of Comparative Economics 49(2): 467–482.

29.

Khan

Riskin

(2005) China’s household income and its distribution, 1995 and 2002. The China Quarterly 182: 356–384.

30.

Killewald

Pfeffer

Schachner

(2017) Wealth inequality and accumulation. Annual Review of Sociology 43(1): 379–404.

31.

Knight

(2014) Inequality in China: An overview. The World Bank Research Observer 29(1): 1–19.

32.

Langley

(2021) Assets and assetization in financialized capitalism. Review of International Political Economy 28(2): 382–393.

33.

Xiao

(2021) The impact of income inequality on subjective environmental pollution: Individual evidence from China. International Journal of Environmental Research and Public Health 18(15): 8090.

34.

Wang

, et al. (2020) The role of public pensions in income inequality among elderly households in China, 1988–2013. China Economic Review 61: 101422.

35.

Sicular

(2014) The distribution of household income in China: Inequality, poverty and policies. The China Quarterly 217(1): 1–41.

36.

Wan

(2015) Evolution of wealth inequality in China. China Economic Journal 8(3): 264–287.

37.

Liao

Xiao

Zhang

, et al. (2017) Financial literacy and risky asset holdings: Evidence from China. Accounting and Finance 57(5): 1383–1415.

38.

Liu

Peng

Luo

(2020) The relation between family socioeconomic status and academic achievement in China: A meta-analysis. Educational Psychology Review 32(1): 49–76.

39.

Liu

Zhang

(2023) The class logic of household investment in risky financial assets. Sociological Review of China 11(5): 171–192 (in Chinese).

40.

Luo

Sicular

(2020) The long-term evolution of national income inequality and rural poverty in China. China Economic Review 62: 101465.

41.

Lynch

(2003) Estimates of income and income inequality in the United States and in each of the fifty states: 1988–1999. Journal of Regional Science 43(3): 571–588.

42.

McCall

Percheski

(2010) Income inequality: New trends and research directions. Annual Review of Sociology 36(1): 329–347.

43.

National Bureau of Statistics (2016) China Statistical Yearbook 2016. Beijing: China Statistics Press.

44.

Nee

(1989) A theory of market transition: From redistribution to markets in state socialism. American Sociological Review 54(5): 663–681.

45.

Ning

Jiang

(2018) Review of research on property income in China. Economic and Political Studies 6(2): 221–235.

46.

Ning

Luo

(2016) Analysis on the causes of property income inequality of the Chinese residents in the transition period: From the perspective of internal structure and household features. Economic Research Journal 4: 116–128 (in Chinese).

47.

Pfeffer

Waitkus

(2021) The wealth inequality of nations. American Sociological Review 86(4): 567–602.

48.

Piketty

Saez

(2003) Income inequality in the United States, 1913–1998. The Quarterly Journal of Economics 118(1): 1–41.

49.

Piketty

Saez

(2014) Inequality in the long run. Science 344(6186): 838–843.

50.

Piketty

Yang

Zucman

(2019) Capital accumulation, private property, and rising inequality in China, 1978–2015. American Economic Review 109(7): 2469–2496.

51.

Pronina

Hübscher

Holler

, et al. (2021) Interactional training interventions boost children’s expressive pragmatic abilities: Evidence from a novel multidimensional testing approach. Cognitive Development 57: 101003.

52.

Qian

Fan

(2020) Who loses income during the COVID-19 outbreak? Evidence from China. Research in Social Stratification and Mobility 68: 100522.

53.

Qiao

Cai

(2023) Financial assets and happiness: Evidence from the China Household Finance Survey. Applied Economics Letters 30(4): 466–471.

54.

Qin

, et al. (2021) Income, relative deprivation and the self-rated health of older people in urban and rural China. Frontiers in Public Health 9: 658649.

55.

Ranaldi

Milanović

(2022) Capitalist systems and income inequality. Journal of Comparative Economics 50(1): 20–32.

56.

Raudenbush

Bryk

(2002) Hierarchical Linear Models: Applications and Data Analysis Methods. Thousand Oaks, CA: Sage Publications.

57.

Ravallion

Chen

(2007) China’s (uneven) progress against poverty. Journal of Development Economics 82(1): 1–42.

58.

Rosen

(2004) Portfolio choice and health status. Journal of Financial Economics 72(3): 457–484.

59.

Sato

(2006) Housing inequality and housing poverty in urban China in the late 1990s. China Economic Review 17(1): 37–50.

60.

Shum

Faig

(2006) What explains household stock holdings? Journal of Banking and Finance 30(9): 2579–2597.

61.

Sicular

Ximing

Gustafsson

, et al. (2007) The urban–rural income gap and inequality in China. Review of Income and Wealth 53(1): 93–126.

62.

Sierminska

Takhtamanova

(2012) Financial and housing wealth and consumption spending: Cross-country and age group comparisons. Housing Studies 27(5): 685–719.

63.

Song

Xie

(2014) Market transition theory revisited: Changing regimes of housing inequality in China, 1988–2002. Sociological Science 1: 277–291.

64.

Survey and Research Center for China Household Finance (2019) China urban household wealth health report 2018. Available at: https://chfs.swufe.edu.cn/info/1321/2581.htm (accessed 14 September 2023).

65.

Tchamyou

Asongu

Odhiambo

(2019) The role of ICT in modulating the effect of education and lifelong learning on income inequality and economic growth in Africa. African Development Review 31(3): 261–274.

66.

The People’s Bank of China (2020) Survey of household assets and liabilities of urban residents in China. Shanghai Business 5: 5 (in Chinese).

67.

Walder

(2014) Public housing into private assets: Wealth creation in urban China. Social Science Research 46(1): 85–99.

68.

Wan

Liu

(2021) China’s poverty reduction miracle and relative poverty: Focusing on the roles of growth and inequality. China Economic Review 68: 101643.

69.

Wang

Wan

Yang

(2014) Income inequality in the People’s Republic of China: Trends, determinants, and proposed remedies. Journal of Economic Surveys 28(4): 686–708.

70.

Wang

(2023) Can increasing economic complexity improve China’s green development efficiency? Energy Economics 117: 106443.

71.

Wang

Fan

(2018) Marketization Index of China’s Provinces: NERI Report 2018. Beijing: Social Sciences Academic Press.

72.

Wang

Huang

, et al. (2020) Housing wealth inequality in China: An urban–rural comparison. Cities 96: 102428.

73.

(2013) Redrawing the boundaries: Work units and social stratification in urban China. Chinese Sociological Review 45(4): 6–28.

74.

(2019) Inequality and social stratification in postsocialist China. Annual Review of Sociology 45(1): 363–382.

75.

Xie

Lai

(2009) Danwei and social inequality in contemporary urban China. Research in the Sociology of Work 19: 283–306.

76.

Xie

Zhou

(2014) Income inequality in today’s China. Proceedings of the National Academy of Sciences of the United States of America 111(19): 6928–6933.

77.

Xin

Yang

(2023) Provincial marketization and individual financial literacy in China. Journal of Neuroscience, Psychology and Economics 16(2): 60–79.

78.

Yin

Gan

(2015) Financial availability, financial market participation and household portfolio choice. Economic Research Journal 3: 87–99.

79.

Yitzhaki

(1979) Relative deprivation and the Gini coefficient. The Quarterly Journal of Economics 93(2): 321–324.

80.

Yuan

Puah

Yau

JTH

(2022) How does population aging impact household financial asset investment? Sustainability 14(22): 15021.

81.

Zhou

Song

(2016) Income inequality in China: Causes and policy responses. China Economic Journal 9(2): 186–208.

82.

Zou

Deng

(2019) Financial literacy, housing value and household financial market participation: Evidence from urban China. China Economic Review 55(1): 52–66.