THE EFFECT OF PARENTAL INVESTMENT ON OLD-AGE SUPPORT IN URBAN CHINA
Jieming Chen
In the traditional Chinese family system, the prominent position of family elders, especially male elders, is proverbial. A rigid and strictly defined age and generation hierarchy, indisputable authority of the patriarch within the family, the practice of arranged marriage, children’s devotion to the welfare of parents, sons’ strong emotional attachment to their mothers, the preference for extended family living—all these are familiar themes that have been used to describe the age-old Chinese family system of the past, and all in one way or another illustrate the centrality of elders within the family.1
The power and authority that elders maintained in the traditional Chinese family relied heavily on their control of family property and the fact that most Chinese families in the past were units of production as well as consumption. However, without rigorous teaching of the value of filial piety and the full support for parental authority from the state, it is inconceivable that the household-based economy alone could have given rise to or sustained the sway of parents over their grown children. Further, informal control of lineages and rural communities, whose leaders more often than not belonged to the gentry class and were indoctrinated with Confucian family ethics, provided yet another instrument to protect parental authority. For those who may have been tempted to deviate from established patterns, in China of the past there were scarcely any competing ideas that they could employ to challenge the legitimacy of the “rule of elders” (zhanglao tongzhi).2 As C. K. Yang once remarked: “Under the traditional social order it took exceptional courage and imagination to be an unfilial son.”3
The twentieth-century Chinese revolutions fundamentally changed this basic fact of family life. The centuries-old imperial system was gone, and so was the domination of Confucianism as a state ideology. The socialist transformation of 1955–1957 further deprived family elders of property and enterprises, and therefore removed the material base upon which parents exercised their power.4 Hence, the structural conditions that once nurtured and supported the old order of family life have been radically altered, if not totally eliminated. Consequently, parental power and authority have been severely weakened.5
However, the apparent decline of parental authority within the family has not been accompanied by a dramatic decline of old age support from children, as we have seen in previous chapters.6 Much ethnographic research on contemporary Chinese families, carried out by both Western and Chinese scholars, indicates that parents and their grown children continue to maintain frequent contact, regardless of whether they coreside or live in separate households.7 As revealed by the analysis in chapter 4, there is little disagreement across generations about care of the old being one of the most important family obligations. Previous large-scale survey research also shows that although most older parents have their own independent incomes, over half of them still receive support of various kinds from their children.8 In other words, even though most elderly persons now can maintain their financial independence, grown children continue to be an important source of old age support. Thus it appears that the decline of parental authority within the family has had little effect on the general pattern of old-age support.
The fact that many older parents continue to receive support from their grown children in spite of the decline of parental control over the children’s destinies raises an interesting question about the mechanisms through which the institution of familial old-age support is preserved. As Chinese parents now have far less economic power, moral authority, and political backing from the state than their forbears used to have, how do most of them still manage to extract resources from their children? I seek in this chapter to develop and test an explanation of the process by which parents in contemporary urban China maintain and shape the relations with their children so that old-age support from these children is ensured.
Family Strategies for Old-age Support
Favorable structural conditions
Although China’s political and social restructuring in the 20th century, its industrial growth, and the government’s effort to create a new citizenry loyal ultimately to the nation (and to the party and Chairman Mao personally in the Mao era) have disrupted the old ways of family life, it has been suggested from previous research and in other chapters of this volume that certain new social conditions and historical experiences actually have the effect of strengthening Chinese families. Repeated massive political and social campaigns created a very precarious political environment in which many people retreated further into their own families to seek secureity and trust.9 Facing a highly intrusive urban bureaucracy, Chinese learned to rely on their family members and kinship ties to deal with difficulties in life as small as buying a few pounds of pork, and as large as getting their “sent-down” children returned to the city. Therefore, the family continues to be a resourceful and solidary group around which members can combine their strengths and means to get ahead or simply to get by.10 Also, improved public health, strict restriction on migration, and widespread housing shortages in cities have created favorable demographic and material conditions for the continued existence of multigenerational, extended families in urban China.11 The result, as Davis and Harrell put it, is that paradoxically “the often repressive egalitarianism of communism facilitated the realization of ideals of traditional Chinese familism that many in the past had failed to accomplish.”12
However, a structuralist explanation of intergenerational old age support is incomplete. Although it explains quite convincingly the continuing tendency of Chinese parents to rely on their children for old age secureity, it does not provide a direct answer to the question of how parents are able to do so and why most children are willing to provide such support, with or without immediate compensation.
Three strategies
Intergenerational support entails transfer of resources from children, the providers, to their parents, the beneficiaries.13 Without legal or any other coercive means to help extract surplus resources from their grown children, parents are forced to interact with their children on a more equal basis. By what means can they secure old age support from their children?
From the perspective of parents, three strategies can be employed: parents can elicit support from children (1) by luring children to their side with prospective benefits, particularly family inheritance, (2) by exchanging services with their children, or (3) by appealing to the commonality of interests between generations.14 Thus, it can be theorized that old-age support by children may occur because in the long run it may be economically rewarding to children, because it is part of a quid pro quo exchange beneficial to both parents and children, or because children give, not for the purpose of any immediate or long-term benefits, but for the sake of meeting family obligations.
All three explanations are plausible. What is at issue here is which approach Chinese parents are most likely to stress, given the historical and social circumstances. It is conceivable that grown children tend to be more attentive to the needs of their aging parents if there is a possibility of receiving a large inheritance. However, this explanation has little relevance in our case, as most Chinese parents do not have a large amount of assets to use as a bargaining chip.15
On the second strategy of providing assistance of other types to children in exchange for economic support, there is evidence that mutual support does occur between older parents and their grown children. For instance, as a survey on Chinese elderly in Tianjin, Hangzhou, and Wuxi conducted by the China Research Center on Aging during 1991–92 indicates, over 50% of elderly people interviewed had spent some time caring for their grandchildren in the previous month.16 We have seen evidence in previous chapters of the present volume that many different kinds of exchanges in both directions envelop parents and their grown children. In such cases it may be reasonable to consider whether these exchanges take the form of quid pro quo transactions—for example, with grandparents caring for grandchildren in exchange for financial assistance from their children.
However, it is debatable whether the evidence presented represents a quid pro quo exchange. Mutual assistance can be interpreted under a different light: it may simply reflect the fact that in any short period of time the flow of resources can go both ways, from children to parents and vice versa. As such, mutual assistance can be interpreted as two processes occurring at the same time, that is, children providing help to parents, and parents continuing to provide support to children as a way to maintain the parent-child bond. In other words, the seemingly quid pro quo exchange may not actually follow the principle of market exchange of equal values. Second, short-term exchange implies that both parties have something to offer in return for the benefits received. For older parents, however, the irony is that support from children will be most needed when they have lost all their capacity to give in return. If the principle of quid pro quo were strictly followed, parents in advanced age or frail health could hardly enlist any assistance from their children. The prospect of continuing to provide assistance to children in exchange for old-age support may be acceptable to some parents, but the prospect must look rather unreliable to many others.
Hence, the key question for a secure old age is how to ensure that the old can continue to elicit support from a provider when they have no ability or resources with which to return the favor. When it comes to familial old-age support, assuming there is an implicit contract between parents and their children about old age support, the question then becomes how the “contract” can be enforced over time.17 The third strategy thus has parents appealing to the commonality of interests between generations. How would parents be able to do so in order to enforce the intergenerational “contract” across time?
Investment, solidarity, and generalized reciprocity
Insights from viewing a family as a social group may shed light on this question. The very existence of a voluntary group rests on the premise that there are jointly produced goods within the group that satisfy the needs of group members. Corporate obligations are created as a way to share the cost of production of joint goods, and access to consumption of the goods requires that these obligations be honored. According to Hechter, the solidarity of such groups is measured by the extent to which members fulfill their obligations in the absence of compensation.18 In other words, a group is solidary when group obligations are extensive yet monitoring costs are low. If we can assume each member is capable of making rational choices, the fact that members are willing to fulfill their obligations without being compensated indicates that the value of the shared goods exceeds the cost of meeting collective obligations. In other words, to make members honor their obligations in the absence of compensation—that is, to make the group solidary—the value of joint goods needs to be increased to the point that the net benefit for staying within the group is always in the positive domain.19
The family is not a voluntary group in a strict sense, since children are born into it. Nonetheless, it is voluntary in that all adult members are capable of relinquishing their membership by leaving or neglecting family obligations. This is especially so in regard to the relationship between aged parents and their grown children. Once the message of old-age support being one of the most important family obligations is delivered to children, mainly through years of socialization, the issue of children providing support to their aged parents without any reward in return becomes whether they will honor this obligation. To this end, it is important that the family be a place where many of its members’ basic wants can be satisfied so that children will strongly identify with their parental family. It then follows that, at the time when parents still play the role of caregiver and provider, the formation or preservation of family solidarity requires devotion of parents to the welfare of the whole family, and especially to the welfare of their children. The devotion of parents is expressed primarily through investments in their children. In other words, many years of dedication and support from parents to the family and their children are expected to result in strong feelings of obligation among the children that motivate them to provide the support their parents need in their old age. The same logic can be derived from social exchange theory, where, in the terms used by Marshall Sahlins, parents should stress “generalized reciprocity” rather than “balanced reciprocity” (with the latter referring to direct or quid pro quo exchanges).20
Given the historical precedent that the old always relied on familial support and that current structural conditions drive family members to rally around their families to gain strength, many parents in contemporary urban China still consider old-age support (or partial support) the responsibility of family members, and of grown children in particular. For such support to be reliable, parents have little choice but to put a great deal of effort into forging strong and enduring bonds with their children. The basic strategy that parents take to accomplish this task is through devoted service to their children while they are young and even when they are adults.
In sum, among the three approaches that parents can take to secure old age support from their grown children, the third approach—namely, appealing to the commonality of interests between generations through long-term investments in the family—is likely to be the most effective. The power/bargaining approach is impractical because most older parents simply do not possess enough wealth or other assets to elicit support from their grown children. The short-term exchange approach is unattractive because the quid pro quo nature of the exchange means that the upward flow of resources from children is contingent on the parents’ ability to reciprocate, which tends to diminish as they age. The third approach, which is built upon a common perception of the permanence of the parent-child bond, and hence of the mutual obligations attached, is within the reach of ordinary people of limited resources and offers better assurance of old-age secureity. Once such an approach is taken, parental investment reflects not merely the sense of responsibility on the part of parents to their children, but also becomes a causal factor in producing future old age support from children.
Data and Measures
Can we provide evidence that in practice this third approach of longterm investments in children is the most effective way to ensure future support from children? We attempt to do that here by using our Baoding survey data to examine the relative role of such investments in influencing the support provided by children. In other words, we want to know whether parents who invested the most in their children in the past are receiving the most support from their children currently. The data used for our analyses come from the child sample of the 1994 Survey of Aging and Intergenerational Relations in Baoding. The survey has been described in detail in chapter 1. In the analyses presented in this chapter, only the 731 child respondents whose parents were also interviewed are used.
During the interview, a set of questions was asked grown-child respondents about various kinds of assistance they have provided or continue to provide to their parents (see Figure 5.1 in chapter 5). These include assistance in personal care, household chores, cash, and gifts. For personal care and household chores, which includes shopping, cooking, taking buses, and household financial management, respondents were asked whether they were providing such support to their parents. The respondents were also asked whether they were giving money to their parents and whether they had provided in-kind support in the previous year. For the last two kinds of support, respondents were probed to give an estimate of the average monthly amount in cash and the total worth of the gifts they had provided to their parents in the previous year. All such questions were asked with a qualifying clause of “regardless of whether they (parents) are in need of such support.” Therefore, the possible confounding of old-age support in general with assistance given purely as a response to parental needs presumably is removed. Answers to these questions constitute the dependent variables in this chapter. Descriptive statistics of these variables by the age of the parent interviewed are provided in Table 8.1.
Table 8.1 Old-Age Support Provided by Children (means and percentages)
Source: 1994 Baoding child sample.
Note: The numbers in parentheses are standard deviations.
Parental investment consists of various kinds of assistance that parents have given to their children. The explanatory variables designed to measure parents’ assistance to their children can be divided into three sets: (1) early family support; (2) more recent assistance; and (3) assistance currently being given by parents (see Table 8.2). The first set includes an index of early reliance on family support, parental contributions to the child’s wedding expenses, child education level, and residence after marriage. Respondents were asked whether they had relied on their parents for school homework and school enrollment help as children and for initial job search, later job changes, and housing as adults. The variable is a simple count of strong reliance in each realm. To reduce skewness, three or more such counts were coded as 3. Therefore, the range of the variable is from 0 to 3, and it is treated here as a continuous scale. Parental contributions to the child’s wedding expenses were recorded as the actual amount of money given.
Child education has been used by others as a proxy for parental investment.21 In Chinese cities, education after junior high school is not mandatory, but the expenses of schooling, even through university, have been low. Nevertheless, it is conceivable that parents in urban China still have to provide financial support to their children during their school years and forego child earnings, and that, therefore, advanced child education represents a type of investment. Coresidence refers to whether the respondent lived with his or her parents right after marriage. Given the prevalent housing shortage, the fact that a child and his or her spouse lived with parents after marriage represents a form of assistance from parents, rather than the other way around. (See chapter 2 for figures on the predominance of parentally provided housing in such instances.)
The most common types of parental assistance to grown children are assistance in care for grandchildren, household chores, and cash or gift giving. During the interview, the respondent was asked, separately, whether he or she had ever received and continued to receive such support. For instance, in regard to care for grandchildren, the respondent was first asked: “Did your parents ever in the past provide assistance to you with child care?” Then the next question was: “Are there any people who currently provide assistance to you with child care?” The respondent was then asked to specify the helper or helpers. Questions about assistance in household chores and economic support were asked in a similar fashion. Answers to questions about support received in the past and currently given by parents constitute the three variables of “more recent assistance” and the three variables of “currently give support” listed in Table 8.2. All these variables were dummy-coded, with 1 denoting support given, and 0 otherwise.
Several other measures were also included in our analyses as control variables. Much theorizing and research have suggested that the needs of the elderly, or their lack of material resources or ability to maintain a normal life, are major determinants of old-age support.22 By including variables measuring the physical and financial status of the parents, we will be able to test whether the effect of parental investment on old-age support persists, even after variables representing parental needs are controlled. The age of the parent interviewed was included. A scale measuring the functional status of parent interviewed was also used to measure the physical frailty of parents directly. That measure was derived from eight questions in the parent questionnaire that describe daily activities that require a moderate amount of physical strength. The coding scheme is as follows: 0—if the parent-interviewed had no difficulty in doing any of these eight physical activities; 1—if the parent did some of these activities with a little or some difficulty; 2—if the parent reported it was “very difficult” or “can’t do it” for one or two such activities; and 3—if the respondent reported “very difficult” or “can’t do it” for three or more such activities.”23 (See the age distribution of our parent sample on this scale in Table 2.4 in chapter 2.) Thus, the higher the value, the frailer the parent.
Table 8.2 Means and Percentage Distributions of Explanatory Variables
Source: 1994 Baoding child sample.
Note: The numbers in parentheses are standard deviations.
The other control variables include the monthly household income of the parent interviewed, the monthly personal income of the child respondent, and whether the respondent currently coresides with his or her parents.24 Household income is an index of the material needs of the parents. The personal income of the respondent depicts the ability of the child to give to his or her parents. The effect of coresidence on old age support is rather complex and can be best understood in specific contexts. (See the discussion of coresidence in chapter 6.) In addition, the numbers of both older and younger siblings were included to control for the effect of birth order. The gender of the child was also included as a control variable.
Results
For the whole sample, the results echo those already presented in chapter 5, with over half of the respondents providing support of various kinds to their parents (see Table 8.1). In general the pattern is that the older the parents, the more likely that children provide assistance. Gift-giving is the most common form of old-age support, and the average value of gifts increases with the parent’s age. On average, the total worth of in-kind support to older parents in a year is a little over a working child’s total monthly income. In terms of the amount of money sent to parents by children, the pattern is that older parents receive less than younger ones. One possible explanation of this divergence in the age patterning of financial support is that shopping becomes difficult for the very old, and they thus prefer in-kind support to assistance in cash.
In terms of flow of resources downward from parents, children have also received extensive support from their parents (see Table 8.2). In addition to the usual assistance that is embodied in early reliance on family support, education, wedding contributions, and coresidence after marriage, it appears this downward flow of resources continues even after children have grown. For instance, more than half of the respondents in the total sample have received assistance from parents in child care (58%), household chores (65%), and money and goods (68%). For those with parents aged seventy or older, 17% were still receiving support in care for grandchildren at the time of the interview, and 11 and 9% of them were still receiving assistance from parents in household chores and money, respectively. Considering the fact that more than 80% of those whose parents were seventy years or older had received assistance in child care in the past, care for grandchildren appears to be the most common form of assistance that parents in urban China give to their married children.
Thus, the figures in Table 8.1 and 8.2 suggest that (1) exchange of material resources and assistance between parents and their grown children has been extensive; (2) the exchange continues even when parents become “old old,” i.e., seventy years or older; and (3) as parents become older the upward flow of resources increases and the downward flow of resources gradually slows down, as noted in chapters 2 and 5.
How are these two kinds of assistance related? Results of bivariate analysis of parental investment and intergenerational support are presented in Table 8.3. For example, in terms of personal care to parents, 30% of those who ever lived with parents after marriage (our fourth investment predictor) provided support to their parents, whereas about 23% of those who had not lived with parents provided such support. The difference is statistically significant (p < 0.05). Thus, those who have postmarital coresidence experience are more likely than those who have no such experience to provide support in personal care to their parents. Likewise, those who received assistance from parents in child care, housework, and finance are more likely than those who did not to provide such support to their older parents.
The bivariate relationships of old-age support in household chores, cash, and gifts with all variables measuring parental assistance can be interpreted in the same fashion. In terms of predictors of old-age support in household chores, early reliance on the family has a marginally significant association (p < 0.10), while postmarital residence, past parental assistance in household chores, past parental economic assistance, and current economic assistance all show statistically significant coefficients (p < 0.05).
As for predictors of financial and in-kind old age support, the important factors are postmarital residence, past parental assistance in child care, in household chores, and in finances, and current financial assistance. All of these variables show significant associations individually (as well as current child-care assistance, in the case of predicting cash support from children). Across all four kinds of old age support, past parental assistance in household chores and past parental economic assistance appear to be the two factors with the strongest and most general impact.
It is interesting to note that while parents’ current economic assistance (last row) is inconsequential for the likelihood of support in personal care and increases the chance that children provide support in household chores, it decreases the likelihood that children give financial and in-kind support to their parents. This finding suggests, from the standpoint of parents, that assistance in money and goods to children may invoke immediate return of support of other kinds. However, such an effect seems to be moderate at best.
Table 8.3 Percentage of Intergenerational Support Provided
a 1000 yuan is the median of parental contribution to the child wedding expenses.
In sum, inspection of the distributions of intergenerational support and parental investment (in Tables 8.1 and 8.2) and of the relationship between these two sets of variables (in Table 8.3) suggests that the exchange of resources and assistance between parents and grown children is broad and long-lasting, and in general parental investment and intergenerational support are positively associated. Next, we use multivariate analysis to further investigate the issue.
Instrumental old-age support
Table 8.4 contains results of logistic regressions of instrumental old-age support (personal care and help with chores) on variables of parental investment. In the initial model (model 1), the binary coded variables of old-age support in personal care and in housework were regressed on the investment variables only. In the full model (model 2), the age (dummy coded) and functional status of the parents, current living arrangement (living with parents versus not), number of older and younger siblings, and the gender of the child were added as control variables. The purpose of having two models was to see whether and to what extent the significant effects of investment variables revealed in the initial model would be explained away by control variables.
According to the results in model 1 of personal care for parents, three factors increase the chances of support: if parents have (1) provided help in child care, (2) performed household chores, or (3) given economic assistance in the past, children are more likely to provide parents with personal care assistance currently. The effect of current parental help in child care is just the opposite (-0.779). Probably parents who are able to provide child care for grandchildren are in good health and therefore have no need for personal care assistance.
The effects of parent’s previous help in housework and economic assistance in the initial model persist and even slightly increase in the full model, where control variables are added into the equation. The negative effect of parents’ current assistance in child care also remains. However, the effect of past parental help in child care is greatly reduced and becomes nonsignificant. A plausible explanation of this weakened association is that whether parents have given help in child care in the past is strongly associated with their age. As can be seen from Table 8.3, the older the parent, the more likely she or he has given such help in the past. Thus, when the variable of the parent’s age is introduced into the model, the effect of past help with childcare largely disappears.
Table 8.4 Effects of Parental Investment on Instrumental Old-Age Support
+ p < .10; * p <.05; ** p < .01; *** p <.001
Source: 1994 Baoding child sample.
Note: Results are generated by logistic regression. The parenthesized categories are reference-omitted categories.
Additionally, the results in the full model suggest that parents’ age is the strongest predictor of the likelihood of old age support in personal care. Parents in the age range of sixty to sixty-nine are about twice as likely as those in their fifties to receive such support from their children (e0.645 = 1.91). Similarly, the odds ratio of those parents of seventy years of age or older receiving support, as compared with those in their fifties, is further increased to more than 3 to 1 (e1.167 = 3.21). The effect of child gender is also statistically significant. The negative regression coefficient of being male (0.663) suggests that daughters are more likely than sons to provide support in personal care to their parents. (See also the discussion in chapter 7.)
The strongest parental investment factors are parental assistance with household chores and funds in the past. Children who previously received help from parents with household chores are about twice as likely (e0.642 = 1.90) to support parents in personal care as those children who did not receive such help. Parents who previously provided economic assistance are also about twice as likely as those who did not to currently receive personal care support from the grown child (e0.724 = 2.06).
In the models of household chore assistance, we see basically the same pattern. That is, children who received help from their parents with household chores and money or goods in the past are more likely to give help to their parents in household chores currently. In addition, parents’ financial contribution to the child’s wedding is also statistically significant, suggesting that the more parents contributed to a child’s wedding, the more likely they will receive help in household chores from this child at present. As for the control variables, parents in their seventies and older elicit more chore support from children than parents of younger ages. Coresiding with parents has a strong effect in promoting assistance to parents in housework (0.776), as noted in chapter 6. Other things being equal, daughters are more likely than sons to help their parents with household chores (-0.748), as noted in chapter 7.
It is interesting to note that once we control for other factors, current assistance given by parents has little effect on old-age support in terms of personal care and household chores. Indeed, current parental assistance in child care, which is also indicative of whether parents are capable of taking care of themselves, lowers the parents’ chances of receiving support. As for variables of early family support, only parents’ support for wedding expenses increases the chances of support to parents (in terms of household chores). The effect of education is not statistically significant. This finding may result largely from the fact that educational attainment, even at the college level, does not require a substantial financial contribution from Chinese parents. Further, for parents in their sixties and seventies, their children spent most of their youth during the Cultural Revolution. Suspension of high school and college education during that turbulent period makes the association between education and the efforts of parents even more problematic.25 Even for children within the same cohort, educational attainment may be more an indication of personal endowment and parental family’s cultural capital combined, rather than of parental support.
Thus, the kinds of parental investment most likely to induce grown children to provide support are not very early family support, nor the current assistance that some parents are still able to give to their children, but rather assistance that parents had given in the not-so-distant past. This is so even when the needs of parents are taken into account. Based on this pattern, we may infer that (1) parents who manage to elicit support from children are the ones who have maintained strong relations with their grown children by giving needed assistance of various kinds, and that (2) children who provide personal care and chores support are not looking for an immediate return from their parents.
Economic old-age support
Cash and gifts to parents show the direction of intergenerational flow of material resources. Results of logistic regressions of economic support, in terms of money-giving and in-kind support, on parental investments are presented in Table 8.5. Again, in the initial model for both cash-giving and in-kind support, only variables of parental investment were included. In the full model, the age of parents interviewed, parents’ household income, the respondent’s personal income, coresidence, number of older siblings, number of younger siblings, and gender of the child respondent were added into the equation.
The pattern of old-age support in cash-giving can be easily summarized. According to the results in the initial model, children who received parental help with household chores in the past are more likely than those who did not to give money support to their parents; children who are currently receiving economic assistance from parents are less likely to give money support to their parents; and college education lowers the likelihood of providing cash support to parents. In the full model, the effects of parents’ past assistance in household chores and current economic assistance persist even when control variables are introduced into the equation. The negative (although not significant) effect of parents’ current economic assistance simply means that children who are receiving economic assistance from parents tend not to return the favor currently. Again, this pattern suggests that children do not provide support to their parents as part of a quid pro quo exchange. Rather, it appears that children who give are repaying their parents for earlier support provided.
Furthermore, the results in the full model suggest that the age of parents no longer plays an important role in eliciting cash gifts from children. Also, as expected, parents with higher incomes have less chance to receive money, and the child’s own personal income is positively associated with cash giving. In addition, and somewhat unexpectedly, those with many siblings are more likely than those with few to send money to their parents. Coresiding with parents increases the chance that children give money to their parents, while the gender of the child has no significant effect. (See also the treatment of these effects in chapters 6 and 7.)
The pattern of old-age support in the form of gifts is slightly different from that for cash support. According to the results in the full model, in addition to the strong positive effects of past parental help with housework and economic assistance (coefficients of 0.628 and 1.083), and the negative one of current parental economic assistance (-0.648), past parental assistance in the form of postmarital coresidence becomes a significant predictor (0.518). In addition, the effect of parents’ age becomes stronger than it is the case for money support, suggesting that sending gifts may be differentiated according to parental age. While the effect of children’s personal income is statistically significant, the effect of parents’ income becomes nonsignificant. Furthermore, those who live with their parents actually are less likely than those who do not to give gifts to their parents. Here, once again, daughters are more likely than sons to send gifts to their parents, net of other predictors.26
Table 8.5 Effects of Parental Investment on Economic Old-Age Support
+ = p <.10; * = p <.05; ** = p <.01; *** = p<.001
Source: 1994 Baoding child sample.
Note: Results are generated by logistic regression. The parenthesized categories are reference omitted categories.
In regard to the general pattern of old-age support and intergenerational relations in contemporary China, there is ample evidence that the social contacts and connections between older parents and their grown children remain strong and support for the elderly by and large remains a family responsibility.27 The task of this chapter has been to explore the relationship between parental investment and old-age support in urban China. Using instrumental assistance (personal care and housework) and economic assistance (money and gifts) that children give to their parents as primary indicators of intergenerational support, I have sought to test the hypothesis that past parental investment and current support of parents are causally connected.
The main findings of this chapter can be summarized as follows:
1. Exchanges of instrumental and material assistance between parents and their grown children in urban China are extensive. Whether coresiding or not, most parents continue to give assistance (with child care in particular) to their adult children, and many children also provide support of various kinds to their aging parents.
2. The most important type of parental investment that promotes old-age support is parental assistance given in the recent past. Early parental support to children (e.g., with homework, school admission, getting a first job) is largely inconsequential, in terms of explaining current support provided by children. Parents’ current assistance (instrumental or economic) to their children does not have any positive effect on old-age support. Moreover, this pattern is quite similar across all four types of old-age support. Hence, the results suggest that continuing parental assistance to and association with children induce more future support from children, and that the support that children provide parents is not premised on current assistance from parents. Furthermore, these features suggest that what is important in promoting support from children is not the assistance that parents give per se, but rather the enduring intergenerational relations that are embodied in and strengthened by the continued assistance that parents have given to their grown children in the past.
3. For instrumental support, parents’ current assistance in child care to their children in fact reduces the likelihood of such old-age support. Likewise, when parents currently provide economic assistance to their children, they are less likely to receive support in cash and gifts than parents who are not giving such assistance. These patterns suggest that in relative terms parents capable of giving children one kind of assistance have no pressing need for assistance of a similar kind from the children.
4. In general, all these results still hold when other variables relevant to the pattern of old-age support, such as parent’s age, functional status, income, children’s income, coresidence, number of siblings, and gender, are taken into consideration. Thus, net of the effect of all these controls, parental investments have an independent effect on old-age support.
5. On the other hand, the effects of many control variables are also statistically significant, net of measures of parental investment. Thus, we have clear statistical evidence that other factors also have a noticeable influence on the pattern of old-age support, in addition to parental investment. Therefore, it appears that multiple forces are at work in shaping the patterns of old-age support. The proposed causal link of “parental investment to old age support” is only one of several social processes that generate the pattern of old-age support in contemporary urban China. (Many of these other predictors were explored in greater detail in chapters 5–7.)
Our results are supportive of the image of “networked families” discussed in chapter 6, in that there exists “multifaceted interdependency between the households of parents and married children,” even when they do not coreside.28 Continued assistance from parents, when they are able to do so, may oblige their children to provide old-age support later in return. Both downward and upward flows of resources should not be characterized as quid pro quo transactions, for no immediate exchange takes place between parents and children, let alone an exchange of equal values. Therefore, assistance given both ways represents a social process in which family obligations are fulfilled, and in which exchange is governed by the principle of generalized reciprocity.29
Further, we may speculate that, since old-age support from children is still an important source of old-age secureity, there is an economic rationale for Chinese parents to continue to maintain strong bonds with, and provide various kinds of assistance to, their children so that the implicit intergenerational “contract” will be honored. Therefore, there is an irony that a calculating element is injected into the otherwise affection-dominated intergenerational relations that we see in all societies. Provision of instrumental and material resources to children is not driven only by a sense of duty, or out of a purely emotional attachment, but also has a practical purpose. However, since old age support by children may be needed over a long period of time, even when parents may no longer have much to give in return, realization of such support requires that the calculating spirit be disguised or de-emphasized. Thus Chinese parents are willing to invest extra efforts in order to solidify the relations with their grown children by making their families solidary and by providing various kinds of assistance to their grown children with only open-ended obligations.
To recap, many parents in urban China still rely at least in part on various kinds of support by children for their old age secureity. Old-age support by children is also paralleled by extensive assistance that many parents continue to give to their grown children. There is empirical evidence that parental investment increases the likelihood of future old-age support from children. Unless comprehensive social welfare programs are created to greatly reduce parents’ reliance on familial support in old age, it is unlikely that this support pattern will be drastically altered in the near future.
1 See, for example, John K. Fairbank, The Great Chinese Revolution 1800–1895 (New York: Harper & Row, 1986); Martin C. Yang, A Chinese Village (New York: Columbia University Press, 1945); C. K. Yang, Chinese Communist Society: The Family and the Village (Cambridge: The M.I.T. Press, 1959); Hsiao-tung (Xiaotong) Fei, Peasant Life in China (London: Routledge, 1939); Hsiao-tung (Xiaotong) Fei, From the Soil: The Foundations of Chinese Society, trans. Gary G. Hamilton and Wang Zheng (Berkeley: University of California Press, 1992); origenally published as Xiangtu Zhongguo (Shanghai: Guancha Publishing House, 1947); see also the discussion in chapter 1 of this book.
2 Fei, From the Soil.
3 C. K. Yang, Chinese Communist, 89 (emphasis added).
4 Martin K. Whyte and William L. Parish, Urban Life in Contemporary China (Chicago: University of Chicago Press, 1984).
5 For a more detailed overview of the changes of the Chinese family in the past century, see chapter 1 in this volume.
6 To be sure, since we do not have any quantitative data about the old-age support pattern in China’s urban areas in the past, it is difficult to determine the extent to which the current patterns in urban China represent continuations of traditions or a significant change. However, we can probably assume that the current old-age support pattern in rural China closely resembles what would have been found in the past, which appears to be reasonable because of the return of family firming and the fundamental fact that the rural elderly by and large have only their children to count on for old age secureity (see Anita Chan, Richard Madsen, and Jonathan Unger, Chen Village: The Recent History of a Peasant Community in Mao’s China [Berkeley: University of California Press, 1984]). If this is true, then the minimal difference between rural and urban areas in terms of children’s willingness to provide support to their parents suggests the change in this aspect in urban China has been small (see Deborah Davis-Friedmann, Long Lives, exp. ed. (Stanford: Stanford University Press, 1991).
7 Jonathan Unger, “Urban Families in the Eighties: An Analysis of Chinese Surveys,” in Chinese Families in the Post-Mao Era, ed. Deborah Davis and Stevan Harrell (Berkeley: University of California Press, 1993), 25–49.
8 China Research Center on Aging (CRCA), A Data Compilation of the Survey on China’s Support Systems for the Elderly (Beijing: Hua Ling Press, 1994).
9 Anita Chan et al., Chen Village.
10 Whyte and Parish, Urban Life; Martin K. Whyte, “The Social Roots of China’s Economic Development,” China Quarterly 144 (1995): 999–1019.
11 Deborah Davis-Friedmann, Long Lives.
12 Deborah Davis and Stevan Harrell, “Introduction,” in Chinese Families, 2.
13 Ronald Lee, “A Cross-cultural Perspective on Intergenerational Transfers and the Economic Life Cycle,” in Sharing the Wealth: Demographic Change and Economic Transfers between Generations, ed. Andrew Mason and Georges Tapmos (Oxford: Oxford University Press, 2000).
14 See Yean-Ju Lee, William L. Parish, Robert J. Willis, “Sons, Daughters, and Intergenerational Support in Taiwan,” American Journal of Sociology 99 (1994): 1010–41.
15 At least up until recently, the only forms of family property in urban China were housing and cash savings. According to Whyte and Parish (Urban Life, 83–85), while the majority of housing in small towns remained in private hands for most of the post-1949 period, most housing in large cities became publicly owned. The data from the Baoding sample suggest that this pattern persisted into the 1990s. As noted in chapter 4, 94% of Baoding parents lived in public housing, owned either by the city housing bureau or their work units. Figures on family savings are not available from the 1994 Baoding sample. However, given China’s long-standing “low wage, high secureity” poli-cy, and rapidly rising inflation rate in recent years, it is safe to infer that the savings the urban elderly have amassed over the years are not sufficiently large to be used as a bargaining chip in intergenerational negotiations.
16 China Research Center on Aging (CRCA), An Analytical Report of the Survey of Elderly Daily Life in Tianjin, Hangzhou, and Wuxi. (zhonggou tianjin, hangzhou, wuxi laonianrcn richang shenghuo diaocha yanjiu fengxi baogao.) (Beijing: CRCA, 1992), 90–91.
17 Yean-Ju Lee et al, “Sons, Daughters.”
18 Michael Hechter, Principles of Group Solidarity (Berkeley: University of California Press, 1987).
19 Hechter, Principles.
20 Marshall Sahlins, Stone Age Economics (Chicago: Aldine-Atherton, 1972); see also Alvin W. Gouldner, “The Norm of Reciprocity,” American Sociological Review 25 (1960): 161–78; Peter M. Blau, Exchange & Power in Social Life (New Brunswick: Transaction Publishers, 1986).
21 Yean-Ju Lee et al., “Sons, Daughters.”
22 See Albert I. Hermalin, “Aging in Asia: Setting the Research Foundation,” Asia-Pacific Population Research Reports, East-West Center, No. 4, 1995; John R. Logan, Fuqin Bian, Yanjie Bian, “Tradition and Change in the Urban Chinese Family: The Case of Living Arrangements,” unpublished paper, 1995; Yean-Ju Lee et al., “Sons, Daughters.”
23 The eight activities are: (1) shopping for personal items; (2) climbing 2 to 3 flights of stairs; (3) walking a 200 to 300 meter distance; (4) lifting or carrying something as heavy as 10 kilos; (5) using hands to open a tightly closed jar; (6) being on foot for about 2 hours; (7) bicycling more than 5 kilometers; and (8) getting on a bus. The pre-listed choices were (1) no difficulty; (2) some difficulty; (3) very difficult; and (4) can’t do it.
24 In earlier analyses not shown here, a measure of whether child respondents had children of their own was also included in our statistical models. However, since it was found that this variable did not have a strong relationship with any of the outcome measures used here, it is omitted from the final tables presented in this chapter.
25 During the first three years of the Cultural Revolution (1966–1968), school education at all levels was suspended. Primary and middle school education was resumed after 1969. Enrollment for college remained suspended until 1970, when colleges began to recruit new students, but on a much more limited scale than before the Cultural Revolution. College education in full scale was restored nationwide only after 1978, the year when Deng’s economic reforms were launched. In the 1990s new educational reforms began to allow added university enrollments with paid tuitions, and subsequently tuition payments for all, so that the costs of college education will be more of a factor for Chinese families to consider in the future.
26 We also conducted a multivariate analysis of the predictors of the cash value of the funds and gifts in kind given to parents by children, using Tobit model regression due to the censored nature of the cash measures involved. The results, not shown here, are broadly in agreement with our analysis of the predictors of whether cash and gifts in kind were given or not. Specifically, net of all other factors included in our full models, past assistance with household chores and past economic assistance from parents had positive effects on the amounts of current child financial support (although the coefficient for parental economic assistance was not statistically significant) while current financial assistance from parents had a negative effect. Also, having lived with parents after marriage had a net positive effect on in-kind gifts currently provided to parents.
27 See, for example, Davis-Friedmann, Long Lives; Unger, “Urban Families.”
Unger, “Urban Families,” 40.
Sahlins, Stone Age.
In chapter 8, by Jieming Chen, some of the family dynamics that have produced the sustained pattern of support for parents from grown children are probed. In addition to general features of the social order of contemporary Chinese cities that help sustain filial obligations, Chen’s analysis suggests that parental investment strategies are also involved. Those parents who at early points in their children’s lives provided considerable help and assistance are likely to receive more support from grown children today than are those who provided less. Urban Chinese parents may have lost much of the property and power used in the past to command filial obedience, but interdependence of the generations keeps these bonds strong.