Introduction and Background

When China implemented its open-door policy and began its economic reform in 1978, it had very few ties to the global marketplace. Thirty years later, it is a rising global economic power. In 2008, the world evidenced the success of China’s economic reform as it hosted the Olympics. Accompanying high economic growth, the living standards of the Chinese people have improved greatly. However, being a vast nation that just started to focus its attention on economic growth only 30 years ago, it still faces challenges.

China’s exploration of its economic reform experienced a lot of trial-and-errors, for example, the reform of its social security system. The system was established in the 1950s, when the nation decided to begin large-scale industrial development. At that time, the system covered the employees of the government and the state enterprises and employers were responsible for employees’ retirement, housing, and medical expenses.

After the Cultural Revolution, in 1978, China implemented its open-door policy and started to carry out a series of economic reforms, including the reform of its social security system, with the goal of establishing a socialized and independent security system. A social pooling fund was created to facilitate market competition and alleviate employers’ burden of providing comprehensive benefits to its employees.

In the early 1990s, in an attempt to solve the emerging problems associated with the aging population, the government revamped such system into a combination of the social pooling fund and individual retirement accounts. In late 1990s, a unified enterprise employee pension program was established. Under this system, the government, the employees, and the employer all contribute to the employee’s pension account.

During social security reform, medical benefits, pension, and housing benefits were separated. The original social security system became the new pension system. The medical benefits were commercialized and insurance companies started to play a role in the medical field. The housing acquisition reserves account (the housing fund) was established. The reserves account is a unique fund that belongs to individual employees, contributed to equally by both the employee and the employer, and managed by the local fund management committee. The members of the committee include the municipal government, banks, the workers’ union, and employee and employer representatives. The fund is available to all employees and is granted with no obligations. Individual balance in the account can be used as the housing down payment and buyers with such a fund receive a discount in the mortgage interest rate when borrowing from such a reserve. A series of financial reforms also took place, during the same period of time, to maintain economic stability and to establish a market for the mobilization and efficient allocation of financial resources. Chinese people began to face more financial choices than ever.

Chamon and Prasad (2008) analyzed a series of data from the Urban Household Survey conducted by the National Bureau of Statistics. It was found that in 2005, the average Chinese urban household had three members, received an income of 31,450 Chinese Yuan ($4,602), consumed $3,572, and saved 22.4% of their disposable income. The average household spent 7% of their disposable incomes on durable goods and only 3.4% household had at least one automobile. The proportion of households that owned or partially owned their houses jumped from 17% in 1990 to 86% in 2005 largely as a result of the housing reform. Among those who owned or partially owned a house, 58% purchased it during the housing reform.

Traditionally, Chinese households spend a significant proportion of their income on food but a relatively low percentage on housing and health care. Using the National Bureau of Statistics of China in 2002, Baldacci et al. (2010) found that Chinese urban households allocated 36.4% of their total expenses to food, 16.9% to housing, and 6.6% to health care. These percentages are consumption allocations after the reform in health care. Literature on Chinese consumer finance, although very limited, concluded that Chinese households save a high proportion of their income. Cui (2010) explored the risk of household saving rate and attributed the precautionary savings mainly to inadequate social safety net, rising health care, and education costs for households. The author argued that it was hard for Chinese households to borrow against future income due to the underdeveloped financial system, population aging, and rising income disparity. All these factors contributed to the high household saving rate.

Respect for tradition, being conservative, thrifty and prudent are some of the Chinese core cultural values (Fan 2000). It is likely that it will take some time before Chinese people adapt to and accept the new financial concepts including investment and credit. Consumer education has become an important task for the nation. In 2008, sponsored by the Citi Foundation, the China Center for Financial Research (CCFR) in Tsinghua University of China launched an annual, national “Survey of Chinese Consumer Finance and Investor Education” (SCCFIE). Data will be collected every year for at least 3 years. This survey collected qualitative and quantitative data including the distribution of assets and debts, income and expenses, consumption and investment patterns, demand for financial products and services, and consumer finance education. The purpose of this study is to provide an overview of family financial well-being in China, in particular, an initial analysis of Chinese urban households’ balance sheet and their income and expense statement.

Methods

Data Collection

Questions in the 2008 SCCFIE survey were similar to the Survey of Consumer Finances (SCF) in the US since the latter has been improved over time and widely used by researchers. The Acorn Marketing & Research Consultants income and expenditure survey in Singapore was also used to frame the survey questions. United States and Singapore are more advanced in consumer finance than China and consumers in those countries are more informed than Chinese consumers. Since the cultures and history of the three countries are also different, questions in the SCCFIE survey were selected and modified based on Chinese culture and current economic situation. For example, most variables in the data were collected as categorical variables. Chinese are prudent (Fan 2000) and they are usually very careful in giving their detailed information to someone they do not know. In order to increase the survey response rate, multiple choice questions were used instead of open-ended questions to collect household financial information. Another example is the inclusion of the housing acquisition reserves account as part of total household assets. In the survey, the balance of such an account was included in the respondent’s total assets. Data were collected by a major data collection firm in Beijing.

The survey used the probability proportional to size (PPS) without replacement method to select sample households from selected cities. The population was stratified by an administrative division: municipalities directly under the Central Government, sub-provincial cities, and prefecture-level cities. The mayor of a municipality is equal in status to a governor of a province. Two out of a total of four such cities were selected in this study. The mayor of a sub-provincial city has the same status as the vice-governor of a province; and five out of a total of 15 such cities were selected. The administrative rank of a prefecture-level city is below a province but above a county. There were a total of 268 such cities in China and eight were selected for this study. As a result, 15 cities, located in the East, Central, West, and Northeast part of China, were selected to reach a geographically representative sample of Chinese urban households. Districts in these cities were randomly chosen based on their population distributions. The households were then randomly sampled from these chosen districts. The study was designed to examine the financial well-being of Chinese urban households in the three administrative divisions. Due to the sampling structure, results of this study cannot be generalized to rural households.

In this study, a “household” is defined as a single individual or a core couple and all other persons living in the household that are financially interdependent. The individual who was most knowledgeable about the household’s finance and could make household financial decisions was selected to be interviewed. These individuals were referred to as the financial respondent in this study. The total number of respondents was 2,095.

Errors may occur at various stages during data collection and the data entry process due to interviewers’ misunderstanding of the protocol, the questionnaire, or a respondent’s answers. Efforts were made to eliminate these types of errors by providing a mandatory project-specific training before implementing the survey. Respondents could misunderstand the survey questions as well. To deal with this issue, interviewers received intensive training before the interview and interview guidelines were strictly enforced during the face-to-face interviews. Follow-up phone calls to interviewed households were made after the interview to ensure that information was provided by the interviewee. Interviewers were required to explain every item which appeared to be unclear to the respondents.

The Weight Variable and Treatment of Missing Values

Since the proportion of cities selected from each of the three levels varied and the number of sample households selected from each city also differed, a weight variable was constructed and employed in the descriptive analyses so that the results from the analyses of the sample households in this study would be representative of the total population of Chinese urban households.

The weight variable is the reciprocal of the sample selection probability and is defined as

$$ {\text{weight}}_{i} ={\frac{{{\text{\#\;of\;households\;in}}\;k}}{{{\text{\#\;of\;households\;in\;all\;sampled\;cities\;in}}\;k}}} \times {\frac{{{\text{\#\;of\;households\;in\;city}}\;j}}{{{\text{\# of sampled households in city}}\;j}}} $$

where i is the ith sampled household, j is the jth city where the ith sampled household lives, and k is the kth administrative division where the jth sampled city belongs.

The total number of households living in each city at or above the prefecture-level was taken from the 2008 China Urban Life and Price Yearbook.

Missing values were also an issue. When dealing with missing values, original questionnaires were checked first to identify whether the value was missed by the data entry staff. If the value was indeed missing on the original questionnaire, such item was coded as missing. Treatment of missing values includes deletion and imputation. There are quite a few methods to impute missing data, substitution with a measure of central tendency (mean, median, mode, etc.), regression imputation, and multiple imputation, to name a few. Since only 15 households had missing values in the variables used in the analyses, they were excluded from the analyses in this study resulting in the total sample size of 2,080. Ideally, if households with missing values are excluded from the study, the characteristics of those who answered all questions and those who did not should be compared to examine the existence of possible systematic bias between these two types of households. However, due to the small proportion (0.7%) of cases deleted, the comparison was not performed since results should not be substantially biased.

Results

The weight variable was assigned to each respondent so that the total households interviewed were representative of the total Chinese population living in all cities at the three administrative levels. Therefore, results at the household level were representative of such population.

Demographic Characteristics: Respondents and Households

Of all respondents, as shown in Table 1, 69.6% were 40 years old or younger. The majority of the respondents (72.0%) were married, 26.2% were never married, and the remaining 1.9% were divorced or widowed. Over half (55.2%) of the total respondents obtained no more than a high school education; 42.2% received a bachelor’s or an associate degree; and 2.6% had a graduate degree. Information on the highest education level of household members was also collected. Results show that the financial respondent was not necessarily the most educated in the household. Among the respondents, 5.0% worked for the government; 14.1% had a managerial position in a business while 32.8% worked as non-managerial employees; and 16.5% were self-employed; teacher, doctor, and lawyer accounted for a share of 7.4%.

Table 1 Financial Assets by selected household characteristics
Full size table

Over half (51.5%) of households had one dependent child and 40.9% of the households had no dependent children. Only 7.6% had two or more dependent children. The majority (56.2%) of households had no financially dependent elders, 38.3% supported one or two elders, and the remaining 5.5% had more than two financially dependent elders.

The majority (52.3%) of the households lived in the eastern part of China; 22.7% resided in the west. Of all the households, 14.6% were from municipalities directly under the Central Government; 16.4% were from sub-provincial cities directly under the provincial government; and the rest (69.0%) were from prefecture-level cities.

Household Financial Characteristics

The distribution of financial assets, nonfinancial assets, total assets, liabilities, net worth, income, and expenses are shown in Tables 1, 2, 3, 4, and 5. All values were originally in Chinese Yuan (RMB) but subsequently converted to US dollars (USD) using the exchange rate on December 31, 2008 (1 USD = 6.8346 RMB). These values provide an overall perspective of the financial status of Chinese urban households.

Table 2 Nonfinancial assets by selected household characteristics
Full size table
Table 3 Liabilities and net worth by selected household characteristics
Full size table
Table 4 Income by selected household characteristics
Full size table
Table 5 Expenses by selected household characteristics
Full size table

Almost all (98.9%) households reported ownership of some financial assets, the average amount of which was $23,000 (Table 1). All households reported possession of some nonfinancial assets. The mean of nonfinancial assets and total assets were $58,300 and $81,100, respectively (Table 2). All medians were substantially lower than the mean, indicating a right-skewed distribution of all three assets. The majority (88.9%) of the households had no liabilities (Table 3). Among those who had some debt, their average balance was $15,400. Due to the low percentage of households that had any liabilities, the distribution of net worth was similar to that of total assets.

As shown in Table 4, more than half (67.5%) of households had an annual after-tax income below $10,000. About one-third (30.4%) received between $10,000 and $40,000 annual income after tax. Only 1.6% received between $40,000 and $100,000 and 0.5% of households’ after-tax income exceeded $100,000. Total expenses averaged at $5,100 for all households (Table 5).

Assets

In 2008, all urban households had some kind of assets, and the mean value of total assets was $81,100 (Table 2). The distributions of total assets were different across demographic groups. The average amount of total assets first increased with age, peaked at the 35–40 age group, and then decreased for older respondents. Married households and those with a male financial respondent had higher levels of total assets, as compared with unmarried households and those with a female financial respondent, respectively. The mean value of total assets was substantially higher for households whose respondent had a higher education; except that doctoral degree holders owned fewer assets than high school graduates.

Average total assets varied greatly by respondents’ occupation, with the business managerial employees holding the highest level ($122,900) and those who were not currently working owning the lowest level ($61,900) of total assets. Among all urban households, those with one dependent child and those with two dependent elders had the highest levels of total assets across corresponding demographic groups. Total asset holding was higher for households with a higher after-tax income and for those who had a higher net worth. The only exception is that households with a zero or negative net worth held an average of $12,600 in total assets, about twice as much as that held by those whose net worth was between $0 and $10,000.

The inequality of asset holding across geographic locations was evidenced by the substantial difference in the average amount of total assets held by households living in East China and by those from the other areas. Households living in the East had a mean value of $105,400, which was more than twice as the amount owned by those in the West and the Northeast. Households living in municipalities that are directly under the Central Government held a substantially higher average amount of total assets than did those living in sub-provincial cities and prefecture-level cities.

Total assets included financial assets and nonfinancial assets. The proportion of households who owned any financial asset was 98.9%, and the average value of financial assets was $23,000 (Table 1). The percentage of households who held any nonfinancial asset was 100%, and the average nonfinancial assets amounted to be $58,300 (Table 2). Results in Tables 1 and Table 2 show that across demographic groups, the distributions of financial assets and nonfinancial assets were quite similar to that of total assets.

Financial Assets

Liquid Assets. Savings, CDs and cash value of life insurance were categorized as liquid assets. Table 1 shows that the majority of households had savings accounts and CDs (96 and 73%, respectively).

For many demographic groups, the proportion of households with savings remained at or near 100% with minor differences across groups. The patterns of the mean value of savings across demographic groups are a mixture of a reverse U-shape and a positive-sloped line. The average savings first increased and then decreased after the peak, as age, educational attainment, the number of dependent children, and the number of dependent elders increased. By marital status, single households held the highest level of savings ($7,500), closely followed by married households ($7,300), and those divorced or widowed had $5,000 in savings on average.

Income and wealth had a positive relationship with the mean value of savings. The distribution of savings also differed by geographic location and administrative division. Households living in East China held an average of $8,500 in savings, followed by those in the West ($7,100), the Central ($5,200), and the Northeast ($2,700). Households from prefecture-level cities held the highest mean balance of savings ($8,500), and those from sub-provincial cities had the lowest level of savings ($3,600).

In addition, Table 1 shows the ownership and average value of CDs and cash value of life insurance. Overall, 73.0% of all households had CDs, and the average balance was $10,000. There was no obvious trend among various demographic groups with a few exceptions. The average balance of CDs was substantially lower for those widowed or divorced ($4,000) than for those who were never married ($10,300) and married households ($9,900). Holdings of CDs increased with net worth and income except that those with zero or negative net worth and those whose after-tax income was less than $2,000 had a slightly higher holding than the next category. It is not surprising to find that the mean balance of CDs and life insurance differed by geographic location and administrative division. The proportion of households holding life insurance was 50.6%, and the mean cash value was $3,400. The patterns of differences in the mean cash value of life insurance across demographic groups were not clearly indicated in Table 1, but those with a master’s degree, in the middle age group, with at least four dependent elders, with a high after-tax income, with a high level of net worth had a substantially higher average value than their counterparts.

Investment Assets. Investment assets included stocks, mutual funds and government bonds. Investment asset ownership was relatively low (Table 1). The overall mean value of stocks, mutual funds and government bonds was $7,700, $4,400 and $4,300, respectively. Stock ownership increased with age, peaked at the 35–40 group, and then decreased. Both stock ownership and balance were highest for households whose financial respondent held a master’s degree. Only 6.7% of those without high school diploma had stocks, but the mean balance was $2,500. A smaller proportion of doctoral degree holders owned stocks than bachelor degree holders, and the average value of their stocks was only $1,500. As expected, both stock ownership and the mean balance rose substantially with net worth and after-tax income. There was considerable inequality of stock ownership and the mean value across administrative divisions. More than half of households living in municipalities directly under Central Government owned stocks, 18.4% of those residing in sub-provincial cities held stocks, and 24.7% of those from prefecture-level cities had some stocks.

The ownership and balance of mutual funds shared a similar distribution as that of stocks. For government bond ownership, only 7.5% of all households owned such asset, and the mean balance was $4,300. A higher percentage of older households than younger households held some government bonds, except that those in the 60+ age group had the lowest ownership. However, the average value was highest for those in the 35–40 age group. Doctoral degree holders did not have any government bonds, but master’s degree holders had the highest level of government bond ownership (19.4%). As compared with their counterparts, a higher proportion of households with a higher net worth and those with more after-tax income owned government bonds and they also had a higher value of such bonds. Government bond ownership was substantially higher for those living in the East and in municipalities. On the other hand, there were no obvious differences in the mean value across geographic locations and administrative divisions.

Housing Fund, Pension and Other Financial Assets. The overall housing fund ownership was 37.7%, which seems unusually low since contribution to this account for each employee is mandatory for both employers and employees. The mean value of the housing fund was $4,300 for those who had such a fund. The proportion of households holding housing funds and pensions increased with age until the 35–40 age group, and then decreased thereafter. Similar to other categories of financial assets, a higher proportion of those married, with a master’s degree, with a higher after-tax income, with a higher level of net worth, living in East China, and living in municipalities owned housing funds and pensions and had a relatively higher mean value in these accounts than their counterparts. The ownership of other financial assets was lowest among all financial asset categories (6.3%).

Nonfinancial Assets

Automobiles. Table 2 shows that of all households, 15.8% owned at least one automobile, with a mean value of $21,500. Auto ownership was higher for households whose respondent was younger, except for the youngest age group whose ownership rate was 10.2%. A higher percentage of households with a male financial respondent owned an automobile than their counterparts that had a female respondent. More than half of master’s degree holders owned at least one automobile. However, the automobile ownership rate was only 21.7% for those with a bachelor/associate degree, 11.5% for high school graduates, and 4.4% for those who did not complete high school.

As compared with other households, a higher percentage of households whose financial respondents were a government employee (25.9%) or held a managerial position in a business (33.2%) reported ownership of at least one automobile. In the dependent elder category, the percent ownership of automobiles was highest for households that provided financial support to at least four elders. Those with at least three dependent children were less likely than their counterparts to own an automobile. Automobile ownership rose with net worth except that 19.2% of the lowest net worth group ($≤0) owned at least one automobile.

Ownership of automobile was also higher for households with a higher income except for the lowest income group, whose ownership was slightly higher than the next lowest income group. Households living in the eastern part of China reported the highest auto ownership. Auto ownership was the highest for households from prefecture-level cities (18.8%), followed by those living in municipalities (12.2%), and those from sub-provincial cities (6.3%).

Major Durable Goods. Slightly more than half of all households (51.3%) owned major durable goods, and the mean value of major durable goods was $4,800 for all households (Table 2). The ownership of major durable goods was higher for households whose respondent was older. Of all households, those with a respondent of less than 25 years old or one who was divorced or widowed were the least likely to report a value for durable goods. Both the ownership and value of major durable goods was higher for households with a higher income or net worth, except for the lowest net worth group. Households living in East China or a municipality under Central Government were more likely to have major durable goods and to have a higher mean value of such goods.

Residential Property. An overwhelming majority (85.5%) of all households owned residential properties, with a mean value of $57,600. Residential property ownership was higher for households whose respondent was older, but the mean value showed an M-shape in this category. The mean value and ownership of residential properties did not differ with the gender of respondent. Married households had the highest percent ownership of residential properties and the mean value of their residential properties ($58,800) was also the highest among all. Respondents with a doctoral degree were the most likely to report a residential property ownership. Those with a master’s degree were the second most likely to own residential properties; but the mean value of such assets held by these households was drastically higher ($83,900) than all other households. Business managerial employees have the highest residential property ownership (92.4%) with a mean value of $76,000. Ownership of residential properties did not differ significantly among households with different number of dependent children or elders.

Households with higher net worth also had a higher percentage ownership of residential properties and the mean value of their residential properties had a positive relationship with net worth except for the zero and negative net worth group. The mean value of residential properties was also generally higher for households with a higher after-tax income. Households with the highest after-tax income also had the highest reported value of residential properties ($112,000).

As compared with other households, a higher proportion of those living in Northeast China and those living in municipalities reported ownership of residential properties. However, the mean value of residential properties owned by households in the East was significantly higher than all other households. The mean value of residential properties owned by households from municipalities was $125,900, which was drastically higher than the other two groups.

Business Assets. Business asset ownership was extremely low (12.2%) with a mean value of $13,300 for all households. Households with a younger respondent (<35 years) were more likely to own business assets. However, the mean value of such assets topped in the oldest group (60 years or older) at $43,900. Business asset ownership for households with a male respondent was almost twice as high as their counterparts with a female respondent. Mean value of business assets peaked for households whose financial respondent completed high school; however, the highest ownership appeared in the master’s degree group (17.5%). As expected, a higher proportion of self-employed respondents (29.4%) owned business assets than others and the mean value of their business assets ($18,800) was the highest among all.

Relatively higher levels of business asset ownership were noted among those with higher net worth, although a spike in business asset ownership was noted for those with zero or negative net worth. The mean value of business assets rose substantially with net worth. Both the mean value and ownership rate were higher for households with a higher after-tax income except that households in the lowest income group had a higher mean value than households with after-tax income between $2,000 and $10,000.

Households living in Central China were more likely to report ownership of business assets than households in other geographic locations. However, households in West China owned more business assets (mean value of $23,800) than other households. Both ownership (15.7%) and mean value of business assets ($13,400) were the highest for households living in prefecture-level cities.

Other Nonfinancial Assets. Other nonfinancial assets included gold, silver, jewelry, antiques, collectibles, arts and others. Slightly more than half (51.3%) of the total households reported an ownership of other nonfinancial assets with an overall mean value of $2,800. Master’s degree holders and business managerial employees reported the highest percent ownership of such assets and they owned more of such assets.

Both ownership and mean value of other nonfinancial assets were higher for households with a higher after-tax income except for the second highest income group. Mean value of other nonfinancial assets was higher for households with higher net worth except that the mean value for the $10,000–$20,000 net worth group was significantly higher than the next two groups with a higher net worth. Households living in the East (68.1%) and those living in municipalities (73.7%) reported the highest ownership of other nonfinancial assets than other households in their category.

Liabilities

Table 3 shows the percent ownership and the average value of each type of liabilities by household characteristics. Generally, Chinese urban households tend to bear low levels of liabilities (11.1%). For those holding liabilities, the average amount of liability owed was $15,400. The distributions of the ownership and mean value of liabilities varied across different demographic groups. Since the number of households in each category is quite small, these results should not be generalized to the general population.

The proportion of household holding liabilities was 15.9% for the 25–34 age group, which was the highest among all age groups, followed by the <25 years old group. The lowest liability ownership was 1.6% for the 60+ age group. Households with an older respondent had a higher debt balance than those whose respondent was younger. The liability ownership was the highest for married households (12.1%) among the marital status category. However, those who were never married had the highest liability balance ($17,300). Households whose financial respondent was more educated had a higher liability balance on average. Those with a bachelor’s degree held an average of $18,500 in liabilities, about twice as much as the balance held by those who did not complete high school ($9,200).

The level of liabilities differed among occupation groups. The mean value of liabilities was highest for the business manager group ($18,900). One noticeable result is that only 6.7% of the non-managerial group had liabilities, but their average debt balance was $14,500, the fourth highest among all groups. Liability ownership was higher for households with more dependent elders. The proportion of households holding liability rose substantially with income except for the highest income group who were debt free. Households living in Central China had the highest liability ownership (16.6%) and those living in the East had the highest balance. Liability ownership was the highest for households living in prefecture-level cities (12.9%) but those living in a municipality held the highest balance ($28,300).

Mortgages

As shown in Table 3, the mortgage ownership for urban households was low (8.7%), although the overall residential property ownership was 85.5%. For those who had a mortgage balance, the average balance was $17,100. Most households with a respondent of 60 years or older owned residential properties; however, only 1.6% of them had a mortgage balance. Only 0.8% of the households had a respondent with a doctoral degree but 44.4% of them had a mortgage balance. The teacher/doctor/lawyer group was the most likely (22.6%) to have mortgage debt as compared with other occupation groups, followed by business managerial employees (14.2%). Mortgage ownership was higher for households with a higher income except for the highest income group who did not have a mortgage balance.

Auto Loan, Short-Term Consumption Loan and Other Debt

As compared with the 15.8% ownership of automobile, the 0.7% ownership rate of auto loan also seemed to be low. The ownership of short-term consumption loans was 1%. The ownership of other debts was relatively higher but still at a very low 2.2% level. The proportions of households with auto loans across various demographic groups were all <6%. The percentages of households with short-term consumption loans and other debts across different groups were also relatively low except for those in the lowest net worth group (28.2%).

Net Worth

The distributions of net worth differed across demographic groups (Table 3). A value-weighted percentage (VWP) was composed to depict the share of value-weighted net worth in each demographic group. The VWP for group i is defined as:

$$ {\text{VWP}}_{i} = {\frac{{\sum\nolimits_{j = 1}^{L} {V_{ij}*W_{ij} } }}{{\sum\nolimits_{i = 1}^{K} {\sum\nolimits_{j = 1}^{L} {V_{ij} *W_{ij} } } }}},\quad i = 1,2, \ldots ,K;\;\;j = 1,2, \ldots ,L$$

where i is the ith group in each category, j is the jth observation in each group, L is the total number of observations in that group, K is the total number of groups in each category, V ij is the value of the jth observation in the ith group, and W ij the corresponding weight of V ij .

This index was widely used in this section to compare with the percentage of households. The average value of net worth was $79,400 for all households. The 35–40 age group comprised 19.8% of the total urban households; however, they possessed 25.1% of the net worth of all households. On average, households with a female financial respondent had a lower level of net worth than those with a male respondent; and married households had the most and the divorced or widowed group owned the least amount of net worth.

The mean value of net worth was higher for households whose respondent received a higher education; except that doctoral degree holders owned a lower level of net worth than those with a bachelor’s degree. The low net worth was due to their low total assets ($48,500, about 3.5 times lower than the amount owned by master’s degree holders) and relatively high liabilities (44.4% debt ownership and an average of $13,600 debt balance). The VWP of net worth for college or associate degree holders exceeded the percentage of households they represented. In contrast, the VWP of net worth for households whose financial respondent did not complete high school were lower than the percentage of households represented by them. The level of net worth differed noticeably among occupation groups. The mean value of net worth peaked in the business manager group, who accounted for 14.1% of all households but owned 21.3% of total net worth.

Those with one dependent child had the highest mean value ($83,500) and the highest VWP (54.1%) of net worth. Households with two financially dependent elders tended to have the highest average amount of net worth with a corresponding VWP of 24.9%, while comprising 21.5% of all households. For households with less than $10,000 after-tax income, the VWPs associated with their net worth were lower than their share of the total households. On the contrary, the VWPs of the higher income groups (at least $20,000 per year) were at more than twice higher than their percentages of total households.

Accounting for 52.3% of all households, those living in East China held disproportionally greater shares of total net worth (VWP = 68.3%). The mean value of their net worth ($103,700) exceeded two times such value for households living in the West ($47,800) and the Northeast ($45,300). Households residing in municipalities comprised 14.6% of all households; however, they owned 27.8% of all net worth.

After-Tax Income

Wages and Salaries

As shown in Table 4, the majority of urban households (86.2%) reported wages and salaries as one source of income. The overall mean wages and salaries was $6,000 for all households who reported such an income. Mean wages and salaries did not vary across age groups except that the youngest age group (younger than 25) tended to earn less than the rest of the households. However, average wages and salaries were higher for households whose respondent obtained a higher education, but with a sizable drop from master’s degree holders to doctoral degree holders. Households whose respondent was self-employed earned the least ($5,000) and business employees with a managerial position earned the most ($8,000). Households living in East China and those living in a municipality earned considerably more than the rest of the households ($7,100 and $9,700, respectively).

Social Security and Housing Funds

Slightly less than half (47.6%) of the total households reported income in the form of social security and housing funds; and the mean value of such an income for these households was $1,300. Only 30.9% of households whose respondent was 60 years or older received such income. Education had a positive relationship with the amount of such income received with only one exception in the doctoral degree holders group; however, the ownership was higher for households with a more educated respondent. Business managerial employees were the most likely to receive (67.7%) social security and housing funds among all occupation groups and had the highest amount of social security and housing funds ($1,800). Both ownership and mean value of social security and housing funds were higher for households with higher net worth or a higher after-tax income. Households in the East and those living in a municipality had the highest ownership and mean value of this type of income.

Business Income

One-third (33.2%) of the total households reported business income in year 2008. These households had an average business income of $6,500. The percentage reporting a business income was lower for households with an older respondent. Mean household business income went up with respondent education except for a drop in the master’s degree holders group. As expected, households with a self-employed respondent received the highest amount of business income ($9,700) and showed the highest ownership (68.2%). Households living in Northeast China were the least likely to receive business income. Administrative division did not make a difference in the average amount of business income.

Investment Income

Over one-fourth (27.4%) of the households received investment income. For these households, their average income from investments was $2,300. All age groups received approximately the same amount of investment income except that the 41–50 age group received substantially less ($1,700). Households whose financial respondent received a doctoral degree were the most likely to report an investment income (64.2% ownership rate); however, their average investment income ($700) was substantially lower than that of those whose respondent received a master’s degree ($5,100). Business managerial employees group were the most likely to receive investment income with an average of $3,000. Deviation in the amount of investment income within the dependent child(ren) and dependent elder(s) groups was quite small. Household living in the East and those living in municipalities were the most likely to report the receipt of investment income and they received the highest amount of such income.

Expenses

Food appeared to be the single most important expense (Table 5). All households reported such an expense. Chinese urban households spent an average of $1,800 on food in 2008. Expenses on food were higher for households with a more educated respondent, more dependent children, or a higher after-tax income. Households living in the east and those residing in municipalities spend the most on food ($2,000 and $2,500, respectively), which is related to the higher cost of living in East China. Age did not seem to affect household expenses, which is consistent with the results of Ying and Yao (2006). Using data collected in Wuhan, China, the authors concluded that there are no substantial differences in consumption between older and younger Chinese households.

Table 5 showed that 59.1% of all households reported expenses on children’s education. The percentage of total households that reported at least one dependent child was also 59.1% (Table 1), it is highly likely that all households with at least one dependent child allocated a proportion of their income to the education of such child(ren). The amount spent on children’s education was higher for households whose respondent obtained a higher level of education; except that those doctoral degree holders spent slightly less than their counterparts who received a bachelor or a master’s degree. Expenses on clothing, communications, and children’s education were higher for households with higher net worth or a higher level of after-tax income. The only exception was that the highest income group (more than $100,000) spent slightly less on children’s education than those that received between $40,000 and $100,000 after-tax income. After-tax income had a positive relationship with all kinds of expenses (Table 5).

Discussion

The one-child policy has been implemented since 1980s; however, there were still 7.6% of the total sample households that had two or more dependent children. Although the one-child policy applies to most urban residents, there have been exceptions to this policy. To name a few: (1) if one of the spouses is an ethnic minority, they can have two children; (2) if both spouses are the only child to their own parents, they can have two children; (3) if the first child is born to be disabled, the couple can have a second child. Moreover, when a divorced or widowed parent marries to another single parent, their new family will have more than one child. The majority of households had no financially dependent elders, which suggests that the traditional extended families are declining in number and nuclear families are becoming the most common family arrangement.

The 35–40 age group had the highest percentage of stock ownership, but the 41–50 age group had the highest mean value of stocks. According to Lai (2008), in the U.S., the older age group (retired or approaching retirement) is the main player in the stock markets. However, results in this study showed that stock ownership and the mean balance peaked at the middle age group, which could further imply that the age-equity profile differs between Chinese and Americans. The 51.3% of overall pension account ownership was indicative of the reform of the pension fund system that originated from the privatization of state-owned enterprises. As the reform reaches more areas, especially the Northeastern region and the cities at the prefecture level, the ownership of stock is expected to increase.

Ownership of residential properties was the highest (90.6%) for households in Northeast China. However, the mean value ($75,800) of residential properties owned by households in the East was significantly higher than all other households. The differences in the ownership and value of residential properties may be due to the regional difference in the cost of living, especially real estate prices. The northeast part of China was traditionally an underdeveloped industrial region and home of many large heavy industry employers, where cost of living has been low and the possibility of multiple household members working for the same employer has been high. With the help of the housing fund, it is likely that many households in this region were able to purchase a home. On the contrary, the east part of China developed faster during the economic reform and housing price in some large cities in this area increased rapidly. In Beijing and Shanghai, the average housing price exceeded $220 per square foot (China Economic Net, December 7, 2009) and a 1,500 square feet condo would cost about $330,000, which is about 30 years of a typical household’s annual after-tax income. Therefore, it is not surprising to see a lower ownership and a higher average value of own residential properties of households living in this region.

Less than one-tenth of Chinese urban households had a mortgage balance, although the overall residential property ownership was 85.5%. The gap between these two proportions implied that Chinese urban households may not like to owe money. For those who had a mortgage balance, the average balance was only $17,100. It is also possible that, due to the special employee benefits provided by many employers at the beginning of the economic reform, employees were given an opportunity to purchase housing from their employer with a very low price and, therefore, did not need to take a mortgage loan.

Urban households in China hardly borrowed. Only about one-tenth (11.1%) of the total households had any type of liability. The low holdings of each type of debt reflects the underdevelopment of the Chinese credit market. It also implies that since credit is relative new to Chinese households, they are not comfortable having debt yet. Being conservative and taking a long range view are two core Chinese cultural values, which can explain Chinese households’ preference to financing long term investments such as a home purchase rather than to using debt to fund short term consumption needs such as buying a car. However, Baek and Hong (2004) found that the level of consumer debt increased significantly during the last decade in China. Since credit cards, mortgage, and installment loans are becoming more and more popular in China, it is reasonable to expect that the percentage of Chinese households holding liabilities will increase gradually following a similar pattern as that in the U.S.

The annual after-tax income levels may seem very low (an overall average of $10,200) after converting Chinese Yuan to U.S. dollars. However, the average annual salary was $3,561 for urban employees in 2007 (Xinhua News Agency April 2, 2008). Therefore, household wages and salaries should be compared with such an average in examination of households’ relative economic status in terms of income.

The overall average of annual expenses was $5,100 and the overall average of annual after-tax income was twice as much. This indicates a 50% saving rate, which is about twice as high as what was found in Wang et al. (2004) and Qin and Ren (2008). The high saving rate is consistent with the Chinese culture of being thrift, conservative, and taking a long range view. The high saving rate is also indicative of suppressed household consumption.

The percent of households reporting dependent children and the percent reporting expenses on children’s education was the same, indicating that it is highly probable that all households with dependent children are financial supporting them to obtain education. This is consistent with the Chinese culture that children’s education is one of the most important saving motives. As the number of dependent elders increased, total expenses first increased and then decreased after the peak ($6,300). Care for elders is a tradition in China and it is more likely for adult children to be responsible for providing this care in China as compared with other nations. After-tax income was higher for households with more dependent elders. However, when the number of dependent elders exceeded two, household after-tax income went down. Therefore, it is not surprising to see the increasing level of total expenses when households have dependent elders and the decreasing expense level when the household has more than two dependent elders.

Conclusion and Implications

The Survey of Chinese Consumer Finance and Investor Education was informative and provided an overall perspective of the Chinese urban households’ assets, liabilities, income, and expenses. Results were indicative of the current household economic well-being in China. In summary, China has experienced great improvements in the consumer finance field. Results in this study indicated that compared with nonfinancial assets, financial assets accounted for a smaller proportion of the total household assets. Residential properties were the most important asset for the majority of households in urban China. Results from this study also showed that a very small percentage of households had any kind of debt, which indicates that the credit market in China is currently underdeveloped and Chinese households are generally reluctant to finance current consumption against their future income. Wages and salaries were the main source of income, which accounted for 58.8% of the total after-tax income. Households spent only about half of their after-tax income and saved or invested the other half. This result is consistent with previous research, which found that the overall consumption rate in China was only 48.6% in 2008 (Xu et al. 2010).

Although China has seen great economic improvements since 1978, a lot still need to be done in the consumer finances field. China will experience a faster development in this field in the near future; and its urban households will have more choices in the financial products and services. A good understanding of consumer finance is needed for Chinese households to make rational financial decisions and improve their well-being.

As compared with the results of 2007 SCF of the U.S. (Bucks et al. 2009), there are considerable differences in household financial situation between these two countries. For American households, the 55–64 age group had the highest percentage of stock ownership (Bucks et al. 2009), whereas results in this study showed that stock ownership peaked at the 35–40 age group in China. The most significant different between Chinese and American households is the percentage of debt ownership. About 80% of American households held some kind of debt. However, only 11.1% of Chinese households had debt. In addition, less than one-tenth of Chinese urban households had a mortgage while 85.5% households owned residential properties. In contrast, 70.9% of U.S. homeowners had a mortgage balance (Bucks et al. 2009). It is worth noticing that average net worth of U.S. households increased with age except for a considerable drop between the 65–74 and the 75 or older age group. However, in China, the pattern of net worth across age groups followed a U-shape with the peak at the 35–40 age group. Moreover, in the U.S., the 45–54 age group had the highest annual income; whereas in China, the 35–44 age group had the highest annual income.

Further research should explore the possible reasons for these differences. This study also sheds light on the financial behavior of Americans with Chinese origin, and can help public policy makers develop strategies to improve economic well-being of Chinese immigrants, especially the older population, which is the most economically vulnerable group (Sharpe 2008).

There are some limitations in this study. China is currently in its economic reforms, changes of household financial well-being would be indicative of the effectiveness of such reforms. For example, the changes of debt holdings can reflect the effectiveness of the development and promotion of the credit market. Also, the aging of population can affect the overall economic growth (Abdel-Ghany 2008). However, data used in the analyses is cross-sectional, which limits the ability to conduct analysis to track changes of individual household financial status, although the mean changes in the population should be reasonably observed with multiple years of data.

Another limitation results from the design of the survey. Most variables in the data were collected as categorical variables to increase the response rate, including all asset categories, which results in the difficulty in estimating the real values. Residents in rural areas were not included in the data collection although most Chinese citizens live in those areas. Information about ethnic background was not collected in the survey. There are 55 minorities in China with different cultures and life styles. Results from the total population may not adequately represent certain minorities.

Although there are some limitations in this study, the data are currently the most detailed national survey regarding household finance in China. This study is the first to provide an overview of the household financial situation in transitional China. Future research should continue to investigate consumer finances. Longitudinal data should be collected to observe changes of Chinese household finances overtime and examine the effectiveness of economic policies and reforms on consumer finances. Rural Chinese households should be studied and minorities should be represented in future data collection efforts to provide a more comprehensive understanding of Chinese consumer finance. Future research should also be done to examine characteristics of households that affect their knowledge and attitudes of consumer finances and factors that contribution to household financial decision-making.

The low percent ownership of debt and the low size of debt owed indicate the possibility that Chinese urban households are not utilizing debt to level their life-cycle consumption. To Chinese households, credit and debt are relatively new concepts (Chen 2004). Education is needed to help consumers understand the benefit and risks of using credit. The low ownership of investment assets (28.0% for stocks, 7.5% for bonds, and 16.3% for any kind of mutual funds) calls for consumer education on financial products and their risks to ensure a good understanding of the development of the capital market and help them better utilize financial products and services to improve their economic well-being and reach their financial goals.

Chinese government should play an active and leading role in supporting financial education to the general public. Economic incentives can be given to employers who provide financial education to their employees. Employers could provide free or affordable financial education to employees. Financial practitioners should be trained to help consumers better understand financial products including investments and credit so that consumers can make informed and rational financial decisions.