• 5. Support Received By The Elderly In Baoding: The View From Two Generations
• Chapter
• University of Michigan Press
• pp. 121-141
• View Citation
• Additional Information

5

SUPPORT RECEIVED BY THE ELDERLY IN BAODING: THE VIEW FROM TWO GENERATIONS

Albert I. Hermalin and Shiauping R. Shih

Previous chapters have already indicated that there does not appear to be a “crisis” in support for the elderly in Baoding. However, we still do not have a full picture of the amount of assistance and support older residents of that city are receiving, and of how central (or peripheral) are members of their families, and their grown children in particular, in providing such support. Chapters 5–8 present detailed analyses of a variety of dimensions of, and influences on, the provision of support to the Baoding parents we interviewed. The analyses in these chapters rely primarily on a set of questions included in our 1994 Baoding survey about assistance parents were currently receiving in four areas of daily life: physical care (with activities such as bathing, dressing, and going to the bathroom), household assistance (with activities such as shopping, meal preparation, transportation, or managing finances), financial support, and material support (the provision of goods such as food items or clothing). Obviously, these simple questions do not cover the full range of possible areas in which older Baoding residents could be needing or receiving support and assistance. However, these four questions do capture key elements of the three major “currencies” of exchange—space, money, and time.¹ (Living arrangements reflecting the currency of space are treated in chapter 6.) Furthermore, these questions were specifically patterned after the questions used in the Taiwan surveys in 1989 and later years, thus permitting the Taiwan-PRC comparisons that will be presented in chapter 10.² For all of their simplicity, then, we think these questions provide a sound basis for drawing conclusions about the patterns of assistance provided to Baoding parents.

The nature of the Baoding Survey on Aging and Intergenerational Relationships and the characteristics of the parents and children were described in chapters 1–2. Because of the broad aims of this chapter, we note that the first two sets of analyses will be based on the full set of 1002 parents interviewed but that the third section, which focuses on agreement between parents and children, will of necessity be restricted to the 731 cases in which a matching parent-child dyad was successfully interviewed.

Figure 5.1 describes the structure of the questions on support received by parents and provided by children that will form the basis of the analyses to follow. The left side of the figure gives the questions addressed to the parents. For four types of support—physical care, assistance with household activities, financial support, and provision of material goods—respondents were asked to indicate whether they were receiving that type of assistance. If so, they were asked who provided such assistance and they were allowed to name up to four specific sources including children, other family members, and nonfamilial sources such as neighbors or community and governmental sources. They were also asked to identify the primary provider and to evaluate the sufficiency of the support that they received. Those who said they did not receive a given type of support were asked if they needed that type of support.

To ensure that the nature of the question did not lead to an overemphasis on familial as against nonfamilial sources of support, there was an additional question (not shown) that asked if any of these types of support was provided by a governmental agency or work unit. The questions addressed to the children as potential providers paralleled those asked of the parents and will be examined further in the last section of this chapter.

The sociocultural underpinnings of the support arrangements of the elderly in Chinese societies and the transformations that may occur as a result of socioeconomic and demographic changes have been discussed in detail in earlier chapters and are further pursued in chapter 9. In translating these theoretical considerations into the analysis of microdata, the focus is on the characteristics of those receiving and providing support. We would expect, for example, that those elderly who have greater needs, as reflected in advanced age, poor health, poverty, or widowhood, are more likely to be recipients. We would also expect support to be a function of the size of the support network and their ability to provide support, as reflected in numbers and sex-composition of children and their socioeconomic characteristics. In terms of social change, although the absence of longitudinal data will limit any strong inferences, the microdata can provide some indication of whether governmental sources have started to supplant familial sources for the four types of support under consideration. In addition the data can reveal whether there have been changes in traditional providers in terms of sons versus daughters and in the role of daughters-in-law as a result of the growing role of women in the labor force and other social changes.

The next section examines the types of support received by persons age fifty years and over and the sources of that support. Although the characteristics of the sample have been described in detail in chapter 2, it is helpful to keep the main contours in mind. First, as the data in chapter 2 show, the age structure is sharply tapered, with about half the sample between fifty and sixty, somewhat more than a third between sixty and seventy, and one-seventh age seventy or older. The sample is almost equally divided between men and women, with a slight preponderance of men, and with a high proportion of respondents currently married. They are widely distributed in educational attainment, with close to a half of the sample having primary schooling or less, but a quarter with a senior high school (or its equivalent) or college education. Despite the young age of the sample, more than three-fifths are no longer working, reflecting in large part the relatively early retirement age in urban China (as described in chapter 3).

Reflecting the young age structure, 29% of respondents live only with their spouse and another 29% live only with unmarried children (usually with a spouse as well). Most of the remainder live with married children. Almost half consider their health as good or very good, but more than a third say it is only fair, and a sixth report that their health is poor or not too good. Finally, the previous high levels of fertility in China are evidenced by the large number of children among these respondents. Over 70 % have three or more children alive. Due to this high fertility and low geographic mobility, all but a handful have one or more children in Baoding, as noted in previous chapters.

Types of Supports Received and Their Providers

Table 5.1 displays the proportion receiving each type of support, by age and sex. Overall less than 4% of the Baoding elderly are receiving personal care support, but substantial proportions are receiving each of the other types, ranging from about one-quarter for financial support to about one-third for assistance with household activities and receipt of material goods. The proportions receiving support are uniformly ordered by age and sex. For each type of support the proportion receiving assistance increases with age, with the sharpest increment usually occurring after age seventy. For each type of support and within each age category, women are more likely to receive assistance than men. Both patterns suggest that the provision of support is responsive to needs, as discussed further below.

Table 5.2 presents the percentage distribution of main providers for those elderly who receive each type of support, separately for men and women. For the small number of men receiving personal care support, almost two out of three specify their wives as the main provider, while one out of four name a son. For women receiving physical support, the pattern is quite different: almost 50% name a daughter, 16% each name their husband or a daughter-in-law, and only 11% a son. For help with household activities there is a similar asymmetry between the main providers named by men and women. For men, not surprisingly, wives are named by 57%, with sons and daughters accounting for most of the remaining providers. For women, husbands are mentioned by less than 20%, while sons, daughters, and daughters-in-law are each indicated by 20 to 30% of the women receiving this type of support.

For those receiving financial support or assistance with material goods, the pattern of main providers is much more similar for men and women. Spouses are rarely mentioned, most likely because couples view themselves as pooling or sharing resources, rather than transferring to one another. Both men and women list a son as the main provider in about 60% of the cases of financial support, with daughters listed about a third of the time. For material goods, sons and daughters are mentioned with about equal frequency by both men and women. Outside of sons and daughters, few other sources of financial or material goods support are noted.

Respondents were also asked to name all sources of support on these dimensions and Table 5.3 presents the percentages who named each of the sources shown. (Since there are multiple responses, the percentages add to more than 100.) The main effect of Table 5.3 is to show more clearly the role of children and daughters-in-law in the provision of physical assistance and help within the household, and the widespread involvement of both sons and daughters in the provision of financial and material support. The asymmetry in the role of husbands and wives in providing personal care or household assistance to the other is still present, however, as is the tendency for male respondents to name sons as providers rather than daughters or daughters-in-law, in contrast to the response pattern of women.

Tables 5.2 and 5.3 also reveal the absence of certain support providers. Beyond spouses, children, and daughters-in-law, other relations (siblings, sons-in-law, etc.) are rarely mentioned, nor are friends, neighbors, or other unrelated individuals. Formal sources of support such as a work unit or other governmental entity also are rarely mentioned, although the nature of the question permitted such replies. Apparently, although the work unit may be a major economic influence in terms of provision of pensions, housing, and the like, it is not thought of as the source of assistance in the particular spheres under examination here.

The detailed questions on sources of support and specific providers permit analyses of the structure of support from a number of dimensions. For example, it is possible to determine how many types of support older respondents receive, what combinations predominate, and how concentrated or diffuse the providers of multiple support are. Select results of these analyses are presented in chapter 10, where the patterns of support in Baoding are contrasted with those in Taiwan.³

The other questions about support shown in Figure 5.1, which ask whether those not receiving support need assistance, and whether the amount of support received is adequate, permit estimates of the levels of need and the sufficiency of support. In comparison with the proportions receiving support shown in Table 5.1, the proportions reporting that they needed help and were not receiving any ranged between 1% and 4% across the four types of support. Similar ranges were observed for the proportions reporting needing more help than they were receiving. Although these percentages are small, taken together they are not trivial relative to the percentage of persons who receive assistance, and this suggests the possibility of shortfalls in the support network of some Baoding elderly. However, our main conclusion is that the great majority of Baoding parents feel they are receiving as much assistance as they need, and that family members, and grown children in particular, are the primary providers of such assistance.

Table 5.4 examines in a multivariate framework the characteristics affecting the receipt of each type of support, with the exception of personal care, for which the number of cases is insufficient. These analyses test the theoretical expectations of the patterning of support discussed earlier. In general the expected relationships emerge. The likelihood of receiving assistance with household chores, finances, and material goods increases with the age of the respondent and is greater among women and those not currently married (though the marital status effect is not statistically significant). The effects of the other characteristics tend to vary by the type of assistance in ways that generally reflect needs and opportunities. For example, the education of the respondent has no effect on the receipt of household assistance or material goods, but there is a sharp negative gradient on the receipt of financial support, presumably because those with greater education have more resources of their own. Health makes a difference in the receipt of assistance with household chores but not for the other two forms of assistance, again reflecting the likelihood of greater need for help with shopping, transportation, etc. among those with poor health. The differential effects of household composition on the receipt of support display the importance of the situation-specific context. Those coresiding with married children are much more likely to receive help with household activities, reflecting both the ready availability of assistance and the possibility that coresidence was prompted by the need for this type of help. By contrast, those coresiding with an unmarried child or children only are the parents least likely to be receiving assistance with material goods, presumably because at this stage parents are mainly providing support to children rather than vice versa. The importance of the size of the provider network emerges quite clearly in the positive relationship between number of children living nearby and the likelihood of receiving assistance with finances and material goods.

Children as Support Providers

The previous section has noted that children, both sons and daughters, are an important source of all types of support. They predominate as providers of financial and material goods support, and they play a considerable role as well with physical support and household assistance.

In this section we inquire whether certain children, as defined by their sociodemographic characteristics and by their family composition, are more likely to supply assistance than others. Table 5.5 presents the characteristics of all the children of the 1002 parents who are the focus of this analysis. In total, these parents have 3226 children; most are between twenty-three and thirty-five years old, married, have one or more siblings, have senior high or more education, and are working. It is worth noting that only 11% live outside of Baoding, while 27% are coresiding with parents. The remaining three-fifths reside in Baoding but are not coresiding with parents.

Table 5.6 indicates for the children in each category the proportion who have been named as a provider, whether as the main provider or not. The table is restricted to the elderly who receive support and have more than one child. Perhaps the most striking finding revealed in this table is the relatively high proportion of children that were identified as support providers in almost every category. There are some differences that emerge by characteristic and type of support. Daughters are consistently more frequently named than sons; the youngest children (under twenty-two), who are also less apt to be married, are least frequently named as providing assistance. Beyond these differences, however, there are few sharp patterns that emerge from this tabulation, especially when the small numbers involved in certain categories (e.g., children who are illiterate) are kept in mind.

Accordingly, we move to a multiple regression in Table 5.7 and show the log-odds of being named as a support provider by the elderly respondent for household, financial and material assistance. The units of analysis in these models are all children from families that contain two or more children when the parents indicate receiving support from at least one child. We omit personal care support because of the small number of recipients.

Certain characteristics of the child exert significant influence over the likelihood of their providing support. A female child is significantly more likely to provide support of each type than a male child. The likelihood of providing support appears to increase with the age of the child, and the oldest children are significantly more likely to provide material goods and assistance with household chores than the youngest child. The size of the sibship has a significant effect on each type of support, but not always in the same direction. Children with a larger number of siblings are less likely to provide household assistance or financial assistance, presumably because other siblings are providing such support, but they are more likely to provide material goods. There is no difference by education of child in the provision of material goods but, compared to those who are illiterate (a very small group), those with education tend to be more likely to provide household and financial assistance. Division of labor across children is suggested by the pattern of assistance according to the location of children vis-a-vis parents. Compared to children who coreside with parents, those living in Baoding but not coresiding are more likely to provide assistance with needed household activities, and along with children outside of Baoding, are more likely to provide financial assistance. On the other hand those outside of Baoding are less likely to provide material goods. (For further analysis of the effects of child distance on the provision of support to parents, see chapter 6.)

In general, within the context of the high proportions of children in each socioecononiic and demographic category being involved in provision of assistance (as shown in Table 5.6), the multivariate analysis of Table 5.7 suggests that children’s resources (as reflected in age and education) and location come into play, as do issues of division of labor across the sibship. The strong role of daughters in each form of support is noteworthy and may signal an emerging pattern at sharp variance with past practice. Nevertheless, as Table 5.2 has shown, sons may be the major providers of financial assistance.⁴ (See the more detailed discussion of support from sons versus daughters in chapter 7.)

Agreement between Children and Parents on Support Arrangements

A distinguishing feature of the 1994 Baoding survey is that both the elderly and a randomly selected child were interviewed. To this point all the data presented have reflected reports from the elderly. For the 731 pairs of parents and children who were interviewed, we can determine whether the parent named the interviewed child as a support provider and whether that child indicated that he or she provided support of each type to the parent. Hence, we can examine the degree of agreement between parent and child reports of the assistance provided to parents.

Table 5.8 provides the basic distributions for agreement on each type of support. Although there is a fairly high level of agreement, ranging from 50% for household assistance to 77% for physical support, this arises mainly from the agreement by parent and child that the parent did not receive support from that child.

Levels of disagreement are also high, however, and they arise mainly from the situation in which the parent did not indicate receipt of support from the interviewed child, but the child reported providing assistance to the parent. In all these cases, it should be noted, the parent may be receiving assistance from other children or sources, and we are examining here the agreement on transfers between a specific child and parent. Thus a parent may know that the specific child interviewed provided some support, but failed to mention that child because the degree of assistance was slight compared to the amount received from other sources.

As before, we turn to multivariate analysis to examine the characteristics of parents and children that are associated with higher levels of agreement on the reporting of receipt and provision of support. In Table 5.9, the dependent variable is dyad agreement, coded “1” if the parent and child agree and “0” if they disagree.

There are few patterns that carry across all types of support. For physical support and assistance with household activities, agreement is lower when the child is a female. Agreement on physical support is also significantly lower if the child is working, and it decreases with age of parent. If the parent is not receiving any support there is more agreement, as one might expect.

For agreement on the provision and receipt of household assistance, agreement is significantly higher when the child is not coresident and when the parent is in poor health. Non-coresidence is also positively associated with agreement in the case of financial support. The effect of numbers of children is seen for financial and material support. Agreement is significantly lower as the number of children increases, suggesting possible selectivity as to whom the parent names vis-a-vis the interviewed child. Older parents show more disagreement with children on financial and material support, and mothers are more likely to disagree than fathers on material support. Disagreement on material support is also more likely when the eldest child is interviewed, suggesting there is normative pressure to display filial piety.

Overall, Table 5.9 indicates that both characteristics of parents and children come into play in affecting agreement in reporting of support, as well as family characteristics such as the size of the sibship and living arrangements. These patterns suggest that the situation-specific context, the expectations of parents, and differential perceptions of what constitutes support all come into play in determining the level of agreement in reporting.

The levels of disagreement reported here point to the value of collecting data on transfers from both parents and children, since all too often our estimates of the level of intergenerational support come from only one generation. Multiple sources can provide sounder estimates of the magnitude of support and point to data collection strategies that will reduce the level of generational disagreement.

Conclusions

This chapter has examined the degree to which the elderly and near elderly in Baoding are receiving four types of support, the sources of that support, and the characteristics associated with the receipt and provision of support. It also investigated the degree to which parents and children agree in their reports of the exchanges taking place.

Although relatively few are receiving personal care assistance, a substantial portion of those age fifty and over in Baoding are receiving assistance in household activities, finances, or material goods, with the percentages ranging from one-quarter for finances to one-third for material goods (such as food and clothing). This assistance comes overwhelmingly from family members, with both sons and daughters figuring prominently in each type of assistance. Taken together, these figures indicate that the needs of older Baoding residents are not primarily met through their work units or governmental programs, and that family ties and traditions of filial piety remain very strong, as indicated in earlier chapters.

When the characteristics of those receiving assistance are examined in a multivariate context, it appears that receipt is responsive to need as evidenced by the role of gender, age, health, and education; to the situation-specific context as revealed by the importance of household composition; and to the size of the pool of potential providers, as reflected in the number of children in and around Baoding.

The characteristics of the children who do and do not provide assistance were also explored. A striking finding was that a high proportion of children in all socioeconomic and demographic categories were named as providers by parents receiving support.⁵ In the multivariate analysis, daughters were significantly more likely to provide support than sons for each type of support, and younger children were somewhat less likely to assist than older children. The size of the sibship and location of the child also entered into the picture differentially according to the type of support in question.

Finally, this chapter took advantage of the fact that the survey interviewed parents and a randomly selected child and obtained enough detail on the nature of transfers to test the degree of agreement between parents and children in their accounts. Overall agreement ranged from 50% to almost 80% of the dyads across the four types of support, with most of this agreement arising from the circumstance in which both the parent and child agree that no support was provided. At the same time, there was a substantial level of disagreement, stemming mostly from situations in which parents did not name the child in question but the child claimed to be providing support.⁶ Interestingly, dyads involving daughters were more likely to be in this category when support in the form of personal care or household assistance was at issue.

It thus appears that, although daughters are frequently named by parents as providers (see Tables 5.2 and 5.3) and they emerge more often as providers than sons when characteristics of children are examined (Tables 5.6 and 5.7), their contributions may still be underestimated to some extent by their parents, judging by the pattern of disagreement in parent-child reports. This parental “discounting” of the assistance from daughters will be examined further in chapter 7. To better understand the multiple sources of these disagreements will require further research.

Overall, the analyses in this chapter confirm the persistence of strong family traditions in urban China. A high proportion of the elderly, commensurate with needs, are receiving support from family members, and a high proportion of children are involved in the provision of that support. Even the tendency of children to report doing more than parents recall may signal the importance the children attach to being seen as maintaining their obligations of filial piety, the “family altruism” ethic discussed in chapter 4. The explicit values that parents and children express, analyzed in chapters 4 and 11, also point in this direction. These conclusions reinforce once again the preliminary generalizations discussed in chapter 2: The large majority of Baoding elderly feel that their needs for support and assistance are being met, and that their families, and their grown children in particular, are central in providing such care. Intergenerational bonds and assistance remain strong despite the tumultuous changes that characterized China’s twentieth century.

NOTES

¹ See Beth J. Soldo and Martha S. Hill, “Intergenerational Transfers: Economic, Demographic, and Social Perspectives,” Annual Review of Gerontology and Geriatrics 13 (1993): 187–216; A. I. Hermalin, M. B. Ofstedal, and M. C. Chang, “Types of Support for the Aged and Their Providers in Taiwan,” in Aging and Generational Relations Over the Life Course, ed. T. K. Hareven (Berlin: Walter de Gruyter, 1996).

² Taiwan Provincial Institute of Family Planning, Population Studies Center and Institute of Gerontology, University of Michigan, 1989 Survey of Health and Living Status of the Elderly in Taiwan: Questionnaire and Survey Design, Comparative Study of the Elderly in Four Asian Countries, Research Report No. 1 (Ann Arbor: University of Michigan Population Studies Center, 1989).

³ Other analyses of similar data from Taiwan can be found in Hermalin, A. I., M. B. Ofstedal, and M. C. Chang, “Types of Support for the Aged.”

⁴ It should be noted that these analyses do not take into account that children within the same family are not independent and that, ideally, one would like to reflect the “joint” characteristics of the children. A simultaneous-equations approach to handle this problem has been developed by D. A. Wolf and could be incorporated into future analyses (Wolf, “Caregiving within Family Networks,” unpublished manuscript (1988); Wolf and B. J. Soldo, “Household Composition Choices in Older Unmarried Women,” Demography 28 (1988): 383–403). This approach has been used to study living arrangements in Taiwan by M. B. Ofstedal, “Coresidence Choices of Elderly Parents and Adult Children in Taiwan” (Ph.D. diss., University of Michigan, 1995). Chapter 11 addresses the non-independence issue by selecting a random child within each family in analyzing the Taiwan data, as does Chapter 7 in analyzing the Baoding data.

⁵ It should be noted that the amount of support in terms of money or time is not measured here, but simply whether a child is listed by the parent as a support provider.

⁶ A similar pattern was found in Taiwan in a related analysis. See C. L. Roan et al., “Intergenerational Contact and Support in Taiwan: A Comparison of Elderly Parents’ and Children’s Reports,” Elderly in Asia Research Report Series, No. 96–36 (Ann Arbor: Population Studies Center, University of Michigan, 1996).

During the interview, child respondents were asked whether they were currently providing assistance in personal care, household chores, cash giving, and economic support in kind to their parents (for the format of these questions, see 16 Answers to these five questions constitute the first set of the dependent variables. The first two variables measure assistance in services, or “instrumental support” in the jargon of gerontology. The second two variables, assistance in cash and assistance in kind, measure economic old-age support. The variable of “other support” can be understood as a “grab bag” measure, probably capturing any kind of assistance that is less tangible than either instrumental or economic support. For all these variables, the coding is 1 if support is provided, and 0 otherwise.

The chapters in Part III of the volume then analyze the nature of the exchanges between generations, exchanges which appear to be both a source of, and reflection of, continuing strong filial obligations. In chapter 5, by Albert Hermalin and Shiauping Shih, the exchanges between parents and children in regard to physical care, help with household chores, financial assistance, and the provision of food or other goods are analyzed in detail. The general picture presented in this chapter reinforces the account in chapter 2, with most parents (and children) enmeshed in complex webs of intergenerational exchanges. Although there is some indication in these data that children claim they provide more support to parents than parents report receiving from them, at the same time very few parents claim they have needs for assistance that their grown children are not meeting.