• 10. Patterns Of Intergenerational Support In Urban China And Urban Taiwan
• Chapter
• University of Michigan Press
• pp. 255-275
• View Citation
• Additional Information

10

PATTERNS OF INTERGENERATIONAL SUPPORT IN UPBAN CHINA AND URBAN TAIWAN

Albert I. Hermalin, Mary Beth Ofstedal, and Shiauping R. Shih

Introduction

This chapter builds on the analyses of intergenerational support presented in chapter 5 by providing a comparative perspective on patterns of support provided to the elderly in urban China and urban Taiwan. As described in the previous chapter, China and Taiwan share a strong cultural tradition based on the Confucian family system, which emphasizes the importance of multigenerational coresidence, familial support of the elderly, and other aspects of filial piety. During the latter half of the twentieth century, however, the two societies underwent rapid and divergent political, economic, and social changes that are likely to have had a profound impact on family relations. The availability of comparable survey data on levels and sources of support in China and Taiwan provides us with an opportunity to examine how these different institutional arrangements have affected patterns of intergenerational support in the two places.

In both the 1989 Taiwan and 1994 Baoding surveys, detailed information was collected on the characteristics of all those residing in the same household as the elderly respondent. The data collected included characteristics, location, and frequency of contact with close relatives not in the household, and the exchanges that took place between the respondent and others on several dimensions. More specifically, the respondent was asked separately whether he or she received assistance with activities of daily living (bathing, dressing, going to the toilet); assistance with certain instrumental activities (household chores, shopping, meal preparation, transportation, or managing finances); financial assistance; and material assistance in the form of food, clothing, or other goods. (For the structure and wording of the support questions, see Figure 5.1 in chapter 5.) In each case the respondent was asked to identify all persons who provided any support as well as the most important provider. (In Baoding, the respondent could name up to four providers; in Taiwan, there was no limit on the number of providers that could be reported.) Those receiving a given type of support were asked to characterize the adequacy of that support, and those not receiving it were asked if they needed that type of assistance. In this way, the survey provided the basis for assessing overall needs. Although the questions as presented did lead respondents to think of individual providers, they were free to name specific service providers or organizations. In addition, a follow-up question asked specifically if any of the sources of support of the type discussed were provided by “government, religious, or other private services” (Taiwan survey wording). Positive responses were recorded along with other providers under the appropriate form of assistance. Finally, although not the subject of this analysis, it should be noted that a detailed inventory of what the elderly do for family members and others was also obtained and provides the basis for analyzing the extent of reciprocity within families.

As noted in chapter 9, we have made two specific adjustments to the Taiwan and Baoding samples to bring them into closer alignment. First, the data for Baoding are restricted to those age sixty years and over to match the age range for Taiwan; secondly, the Taiwan sample was restricted to those residing in the five major cities of Taiwan (Taipei, Keelung, Taichung, Tainan, Kaohshiung) to create an urban subset. Of course, the two samples may still differ considerably on many other characteristics, and as appropriate these differences will be controlled by cross-tabulation or multivariate analysis. Our interest is in comparing contemporary urban China and Taiwan, while we realize that many factors will contribute to observed differences. Finally, as noted in chapter 9, the urban elderly in Taiwan are subdivided into “Taiwanese” and “Mainlanders” in all of the analyses presented here. Having spent part of their life in each country, and being subject to special constraints, the Mainlanders provide an opportunity to test further the question of the persistence of cultural traits in the face of special types of demographic and socioeconomic change.

The results presented in this chapter focus on a detailed analysis of (1) the degree to which the elderly in Baoding and in urban Taiwan receive support of various types, (2) the sources of support for the elderly, (3) the nature of the support network in terms of its size and breadth, and (4) how the likelihood of receiving certain types of support varies with the characteristics of the elderly. In addressing these topics, we first present a series of cross-tabulations that show, for each country separately, the level of support received for each of the four types by key sociodemographic characteristics and the distribution of support across providers. We then employ logistic regression analysis to examine the factors that are associated with the likelihood of receiving support of a given type in a multivariate context.

As a preliminary stage in the multivariate analysis, we estimated separate models for urban China and urban Taiwan in order to examine how the patterns of association for individual sociodemographic factors differ across the two countries. Because a key interest in this analysis is on cross-country differences in the likelihood of receiving support, we then pooled the two samples and estimated the same set of logistic regression models, adding indicators for country of residence and several interactions that emerged as important variations in the country-specific models. Furthermore, in order to allow for differential probabilities of support for Mainlanders and Taiwanese, as revealed in previous studies,¹ we included two dummy variables to represent the country/ethnicity effects. The first variable represents Mainlanders (coded 1 if Mainlander, 0 otherwise), and the second represents Taiwanese (coded 1 if non-Mainlander residing in Taiwan, 0 otherwise), with Baoding residents serving as the contrast group for each. For purposes of brevity, we present only the pooled regression results in this chapter. Finally, in order to facilitate interpretation of the regression results (especially for models containing interaction terms), we provide a set of predicted probabilities for specific subgroups of respondents in the final table.

Types of Support Received by the Elderly in China and Taiwan

Table 10.1 provides a detailed description of the levels and patterns of support received by the elderly in urban China and urban Taiwan. The percentage receiving each type of support is presented separately for Baoding, Taiwanese, and Mainlander elderly, and is further broken down by key socio-demographic characteristics. The bottom row presents the overall percentages.

Overall only a small percentage of the elderly in each country are receiving physical care, a reflection no doubt of their generally young age and good health as indicated in the previous chapter. The figure for Baoding, 6.7%, exceeds Taiwan somewhat (4.5% for Taiwanese and Mainlanders combined), and this differential holds up in most categories. In both urban Taiwan and Baoding the percentage receiving physical care increases with age and poor health. Beyond these two characteristics, few of the patterns hold across all three of the groups that are compared. Women are more likely to receive physical care than men in Baoding, but this pattern does not hold for either the Taiwanese (for whom the proportions are roughly equal for men and women) or the Mainlanders in Taiwan (for whom the percent is slightly higher among men). The pattern for Mainlanders may be explained in part by the fact that Mainlander men tend to be much older, on average, than Mainlander women. Those living with a married child or with other relatives are generally more likely to receive this form of assistance (except among Mainlanders residing with a married child), reflecting in part the need for coresidence among many who require physical care.

Much sharper differences between the Taiwan and Baoding elderly appear with regard to the percent receiving household assistance. Approximately two-thirds of the Taiwanese elderly and three-fifths of Mainlander elderly report assistance with these tasks compared with less than two-fifths in Baoding. It should be recognized that, as asked, this question does not necessarily match the usual approach to ascertaining needs for the Instrumental Activities of Daily Living (IADL), which often inquires whether assistance is required for shopping, transportation, etc., because of a health problem. In the form asked here, respondents may reply that they are receiving help because it is customary for someone else (e.g., the daughter-in-law) to do the cooking or shopping, and this arrangement may have been in operation for a long time.² In this connection it is worth noting that in Taiwan the percent receiving assistance is highest among males (81% for Taiwanese and 63% for Mainlanders) and among those who are living with children (particularly married children for the Taiwanese) or who have children living nearby. In Baoding, coresidence with married children is also important (although the percent is lower than for urban Taiwan), but the patterns by age, gender, marital status, and health are quite different from Taiwan and suggest a pattern more indicative of need.

The elderly in Baoding and urban Taiwan also differ considerably in the proportion receiving financial assistance, ranging from three-quarters for the Taiwanese and nearly one-half among Mainlanders to one-third for the elderly of Baoding. This broad difference may reflect the relative well-being of the elderly vis-a-vis their children in each location. In Baoding, the provision of pensions, highly subsidized housing, and other support to the elderly from their work units may reduce the need and opportunity for cash transfers. At the same time, more of the children in Taiwan are able to provide both basic funds and extra income to their parents, given the longer period of strong economic growth there, in contrast to the more recent emphasis on a market economy and private enterprise in China. The pattern of differentials in Baoding by gender, marital status, and health does suggest that this form of assistance responds strongly to needs and not simply to customary obligations. Although all three groups are similar in that younger persons, men, and those reporting very good health are less likely to receive financial assistance than their counterparts, the gradients across categories are generally steeper in Baoding than among the other two groups of elderly. The patterns for other characteristics are more variable; for example, coresidence with a married child is associated with greater likelihood of support in Taiwan than in Baoding, while having children nearby seems to be more of a factor in Baoding and among the Mainlanders than among the Taiwanese.

The contrast between urban Taiwan and Baoding on household and financial assistance should not be interpreted as a generalized lower level of family obligation in China, and this point is reinforced by the differential observed for material assistance. Overall, 45% of the elderly in Baoding compared to 41% of Taiwanese and 20% of Mainlanders report receiving this type of assistance. In both Baoding and urban Taiwan the differentials within categories suggest that this assistance tends to be oriented toward those in need.

The levels and patterns of support displayed in Table 10.1 must be interpreted with caution pending multivariate analyses, but they do suggest that a number of factors are at play. These include broad institutional arrangements of the type discussed in chapter 9, in addition to the needs of the elderly, the ability of children and others to meet these needs, and the feasibility of doing so in terms of living arrangements and location of kin.

Sources of Support for the Elderly in China and Taiwan

As described previously, the Confucian family tradition places great emphasis on family obligations and filial piety. To what degree does the family remain the major locus of assistance in contemporary China and Taiwan, given the major changes in political, social, and economic structures characterizing both countries?

As noted previously, in addition to identifying the providers, respondents were also asked to specify the main provider. Although respondents could name nonrelatives and nongovernmental and governmental services, as the data below make clear, they overwhelmingly named close family members as their main sources of support.

Tables 10.2A and 10.2B present the distributions of main providers reported by those receiving each type of support. Table 10.2A focuses on the provision of physical care and household assistance, types of support generally requiring close proximity of the provider (i.e., either coresident or nearby). Among the small number receiving physical care in Baoding, the spouse is the most frequently named main provider (41%), followed by a daughter (24%). Among the urban Taiwanese, daughters-in-law and daughters are named somewhat more frequently than spouses (22% versus 16%). (Only twelve Mainlanders reported receiving physical care, and two-thirds named their spouse as the main provider.)

The differential between the Baoding elderly and the urban Taiwanese in the importance accorded to the daughter-in-law emerges even more clearly in the provision of household assistance. Almost half of the urban Taiwanese name the daughter-in-law as the main provider compared to only 11% of the Baoding elderly. In contrast, almost half the Baoding elderly name a son or daughter, nearly one-quarter each, while only 11%) of the elderly Taiwanese do so. Spouses are named with almost equal frequency by these two groups. These contrasts may reflect actual differences in the patterns of assistance, with Taiwanese much more likely than Baoding elders to have a coresident daughter-in-law who is not in the labor force. On the other hand, they also might reflect customary differences in ways that filial obligations are acknowledged, with sons receiving “credit” in Baoding for services that they provide directly as well as indirectly through their wives. Additional research will be required to sort out these competing explanations.

The Mainlanders in Taiwan have a distinct pattern of household assistance, relying to a very great extent on spouses to the exclusion of other providers, reflecting perhaps the “thinner” family and support network of this group compared with the others. As noted in chapter 9, Mainlanders are less likely to coreside with a married child than their Taiwanese and Baoding counterparts; they also have fewer children overall and fewer children living nearby. In addition, because the Mainlanders tended to marry and have children at relatively late ages, their children are younger on average. Hence, even though a sizeable proportion of Mainlanders coreside with unmarried children (33%, as shown in Table 9.2 in chapter 9), these children may be too young to make any significant contribution in terms of household assistance.

In Table 10.2B, the focus is on the main providers of financial and material assistance, types of support that can be more easily carried out at a distance. Compared with patterns of physical care and household assistance, the patterns here are more similar among the three population groups. Spouses tend to be infrequently named, probably because pooling of resources for couples is expected and not considered “assistance” per se. For financial support, sons are singled out by all three groups as the predominant provider, but it is worth noting that daughters are named more frequently in Baoding than in urban Taiwan. For material assistance, sons and daughters are named with almost equal frequency by each group of elderly. In Baoding, few other main providers are indicated; in Taiwan, sons and daughters account for about two-thirds of main providers, with the remaining third being dispersed across other relatives or combinations of individuals.

The first section in Table 10.3 provides one way of summarizing the detail in Table 10.1, i.e., by measuring the number of supports received by the elderly in each location. Reflecting the considerable degree of assistance provided, 67% of the elderly in Baoding and 87% of those in urban Taiwan are receiving one or more of the four types of support considered here. The distribution of the Mainlanders in Taiwan closely resembles the pattern for Baoding. Close to two-fifths of the elderly in Baoding and of the urban Mainlanders are receiving two or more types of support, while among the urban Taiwanese elderly nearly two-thirds are receiving two or more forms of assistance. Thus, the general magnitude of assistance appears high in both China and Taiwan.

The second section of Table 10.3 speaks to the breadth of the support network for those who are receiving each type of support, and here we see a clear patterning across the three groups compared. In each instance, those in Baoding are more likely to have multiple providers than those in Taiwan, and the Mainlanders are the most likely to rely on a single provider for any given type of support. For the Mainlanders this no doubt reflects the relatively small size of their kin network, given that the Mainlanders are comprised of a disproportionate number of unmarried males and couples with few adult children. The urban Taiwanese and those in Baoding are fairly close in their distributions of total number of providers (for all supports combined); however, this overall equivalence is confounded by the fact that Taiwanese elderly receive a larger number of different types of support than the Baoding elderly. As noted above, the patterns from Table 10.1 suggest that support in Baoding is more closely tied to need, and in such cases there may be more explicit sharing across potential providers.

The last section of Table 10.3 addresses the question of diffusion of support across main providers. Is the same person called on to carry major responsibility when an older person requires multiple types of care, or is there a specialization or selection among providers such that different individuals assume primary responsibility for different types of assistance? The figures shown in this panel represent the percent who have the specified number of providers among persons who are receiving at least two types of support (and are, therefore, at risk of having more than one main provider). Diffusion of support is greatest among the urban Taiwanese, for whom nearly three-quarters have two or more main providers, with at least 16% having three or more main providers. The elderly in Baoding show the lowest level of diffusion, with two-fifths reporting a single provider who carries the bulk of the responsibility for all types of support that are received (ranging from two to four types). Hence, although the elderly in Baoding appear to have broader networks in terms of the overall numbers of persons from whom they receive support (as shown in the middle panel of Table 10.3), the primary responsibility for support appears to be more concentrated in Baoding. Whether this reflects a more well-defined division of labor within families in Taiwan or mere differences in perceptions or valuations on the part of older parents as to who is viewed as the most important provider is an intriguing question, but one that we are unable to answer with the data at hand.

Though our discussion has focused primarily on differences in patterns of support across the three groups of elderly, from a broader perspective these distributions appear to be variations on a single theme, which is all the more remarkable given the distinct histories each has experienced. In particular, although there are some differences in the identification of the main provider for each type of support (especially for physical care and household assistance), in the implied care-giving burden, and in the breadth of support networks, there is considerable consistency across the three groups in terms of the high degree of reliance on family. This concentrated reliance, coupled with the substantial proportions receiving assistance, indicates that the family is still a primary provider of assistance, and that family obligations are still accepted and acted on to a high degree in both urban Taiwan and China.

Factors Associated with Receipt of Household, Financial and Material Assistance

The final set of analyses presented in this chapter follow up on the patterns displayed in Table 10.1 by examining the factors associated with the likelihood of receiving support in a multivariate context. A logistic regression model is employed to estimate the log-odds of receiving a given type of support as a function of a set of independent variables. In keeping with the conceptual model that was described in some detail in chapter 9 and utilized in chapter 5, the independent variables included in the models represent the respondents’ need for support (e.g., age, marital status, health, and socioeconomic status) and their potential for receiving support, in terms of the availability and location of their kin.

Table 10.4 presents the estimated regression coefficients and associated significance levels for three separate regression models, each pertaining to a different type of support. Physical care was excluded from the multivariate analysis due to the very small number of recipients. With the exception of age, which is represented as a continuous variable in the model, all of the independent variables are categorical and are represented by dummy variables (with the omitted group indicated by the dashed lines). The coefficients represent the amount of increase (if positive) or decrease (if negative) in the log-odds of receiving support for persons who fall in the specified category relative to the omitted one. As noted previously, in order to evaluate the extent to which there are differences across countries that are not accounted for by other factors included in the model, the Baoding and Taiwan samples were pooled, and dummy indicators were included in the model to capture differences between Baoding residents and both Taiwanese and Mainlanders residing in urban Taiwan. On the basis of results of preliminary analyses, a number of interactions were tested for each model; those that were significant in the multivariate models were retained, and they are shown toward the bottom of the table. To facilitate interpretation of the interaction effects, we computed a set of predicted probabilities for each type of support, and these are presented in Table 10.5. The results from Tables 10.4 and 10.5 will be discussed in tandem here.

The results shown in Table 10.4 suggest that assistance is more likely to go to those who are older, less educated, and in poorer health (although not every contrast is statistically significant), reflecting most plausibly greater need among such individuals. Those with more children nearby generally receive more assistance, reflecting the importance of the size of the potential provider pool, as discussed at the outset. Household composition, because of its special character, also has different effects across types of support. Those coresiding with a married child are more likely to receive household assistance, as might be expected given the time-intensive nature of this support (thus reinforcing one of the conclusions from chapter 6). Those coresiding with an unmarried child (or children) are less likely to receive material assistance. All of these associations hold for Baoding, Taiwanese, and Mainlander elders, as well as for subgroups defined by gender and marital status. Although the effects of country and ethnicity were found to be conditional on other characteristics (as is described in detail below), the general pattern is one of higher levels of household help and financial assistance in Taiwan and of material assistance in Baoding, confirming the patterns detected earlier in Table 10.1. Among older persons in Taiwan, the Taiwanese have a slightly higher likelihood of receiving support of each type than the Mainlanders, although these differences between the groups were not tested explicitly in the models.

Several of the characteristics considered in these models vary across subgroups in their effects on receipt of support. For example, among the elderly in Baoding, there are essentially no differences in the likelihood of receiving household assistance by gender or marital status (see Table 10.5). Very different pictures emerge for Taiwanese and Mainlander elderly, however. Among the Taiwanese, men are significantly more likely to report receiving household assistance than women, and this is particularly true for married men. For Mainlanders, the effect of gender is only evident among those who are married, and unmarried, Mainlander males have the lowest likelihood of receiving support of all subgroups examined in Table 10.5. An additional finding of interest is that, other things being equal, women in the three groups have equivalent probabilities of receiving assistance with household tasks. Taken together, the results show that married Taiwanese and Mainlander men report the highest levels of household assistance received, followed by unmarried Taiwanese men. In contrast, unmarried Mainlander men report the lowest level of such assistance. Unmarried Mainlander males are comprised of a disproportionate number of never-married persons (44%), and their lack of a family network may pose special vulnerabilities for this group as they grow older and begin to require support of various types.

Moving to financial assistance, we observe cross-country differentials in the effects of work status and living arrangements. As shown in Table 10.5, among elderly Taiwan residents, work status is negatively associated with receipt of financial support (the probability of receiving financial support is nearly double for persons who are retired or not working compared to those who are currently working). In contrast, for Baoding elderly, there is no association between work status and receiving financial support, other things being equal, and this finding again highlights the extensiveness of the pension system in Baoding relative to Taiwan. Looking across country and ethnic groups, we see the likelihood of receiving financial support is equivalent among those who are currently working. However, among those not working (a large majority of these elderly), the likelihood of receiving financial support is much higher in Taiwan than in Baoding, and slightly higher among Taiwanese compared to Mainlanders.

Differentials between Baoding and urban Taiwan were also observed with respect to the association between living arrangements and receipt of financial support. In Baoding, living arrangements show only a modest association with financial support, with those living alone having a slightly higher likelihood of receiving support than those living with married or unmarried children. In Taiwan, the opposite is true—elders living alone or with a spouse only are least likely to receive financial support (although the probability is still quite high, and much higher than for their counterparts in Baoding), and those living with children (especially married children) have the highest likelihood of receiving this type of support. This pattern is somewhat surprising, but it may suggest that coresidence in Baoding is either more responsive to constraints on the availability of housing than to financial needs per se, or that it is driven more by the financial needs of coresident children than of their parents. A final variation of interest relates to the differential effect of marital status for males and females. Specifically, among men, those who are married are more likely to receive financial support than those who are not married; for women, in contrast, the unmarried have a slight advantage. This pattern was observed across all country and ethnic groups.

Taken together, the levels and patterns of support for the elderly revealed in these tables point to several conclusions. First, aside from assistance with physical care, which does not appear to be widely required, a substantial proportion of the elderly population in Baoding and urban Taiwan are receiving household, financial, and/or material assistance. In general, the share receiving assistance is proportionately highest among urban Taiwanese and lowest among those residing in Baoding, with the Mainlanders in urban Taiwan occupying an intermediate position. However, as noted in the above discussion, the relative standing of each group varies with the specific type of assistance, and also across subgroups as defined by gender and marital status, household composition, and work status. Finally, the level of support varies with characteristics of the elderly in ways that suggest it is responsive to needs, as well as to the size of the provider network, and perhaps to the location of kin and their financial capabilities. Nevertheless, substantial numbers of the elderly in every socioeconomic and demographic category were receiving assistance, indicating a broad societal consensus on family obligations at work in both China and Taiwan.

Summary

What can we conclude about the strength of familial obligations in China and Taiwan as manifest in the levels of assistance to the elderly? In the absence of careful over-time data it is difficult to draw strong conclusions, but certain patterns emerge quite clearly. First, in both China and Taiwan, a high percentage of the older respondents are receiving at least one type of support of the four types covered in this chapter, and this support comes overwhelmingly from family members. The relative absence of credit to work units, employers, local leaders, or service organizations in both countries speaks both to the lack of formal assistance, and to a mind-set that views pensions or other formal provisions as outside the realm of “assistance.” To most respondents, discussion of “assistance” conjured up family transfers, despite the opportunity and invitation to name other sources. In addition, the fact that a sizable proportion of the elderly in all socioeconomic categories was receiving some form of assistance reflects the presence of broadly held societal values. Where consistent variations did emerge (e.g., by age and health status), the results suggest that assistance is also oriented to those most in need.

The elderly in Taiwan are generally more likely to receive assistance than those in China, but this is not true for every form of support, which suggests that levels of income of parents and children, living arrangements, and local conditions affect the type of assistance that is provided. Within urban Taiwan, the Taiwanese are more likely to receive assistance than the Mainlanders, reflecting to some degree differentials in need, but also the greater breadth of the Taiwanese provider network. Finally, although not shown here, our previous research on the sufficiency of assistance that is received does not suggest that significant portions of the elderly are being neglected in terms of household, financial, or material support. (See chapter 5 for discussion of sufficiency of support in Baoding.)³

This review does not speak to the importance of the transfers the elderly receive, nor does it indicate the content and magnitude of the support. Other data obtained in the Taiwan survey asked respondents to indicate the major or most important source of income, and 51% reported children as against earnings from work, pensions, savings, etc., indicating that the financial assistance received is far from incidental. In Baoding, in contrast, the data analyzed in the present volume (see particularly chapter 2) suggest that financial assistance from grown children is more often supplemental, rather than primary. Indeed, only a little over 13% of Baoding parents over age sixty reported that their children collectively provided more than half of their income in the year prior to our survey.

It is also important to keep in mind that considerable assistance flows from older parents to their children and grandchildren. Earlier analysis revealed that 34%) of the elderly in Taiwan were giving one of the forms of assistance reviewed here to their children or others.⁴ An analysis of the 1996 Taiwan aging survey (which included the surviving members of the 1989 survey cohort plus a new sample of persons age fifty to sixty-six) showed that 30% of persons fifty years or over who have at least one grandchild were providing childcare for one or more grandchildren, and 58% were providing indirect care by undertaking household duties in households with young grandchildren.⁵ The Baoding survey results similarly indicate a high level of assistance provided from parents to grown children (see particularly chapters 2 and 8 in the present volume).

These findings and our focus-group studies in Taiwan and elsewhere in Asia point to emerging areas of accommodation between elderly and children regarding the exchange of assistance. Older respondents in the focus-group discussions often recognize, for example, that working daughters-in-law cannot take time to make their favorite dishes and carry out other traditional functions, yet they recognize and appreciate the help they do receive, just as the younger participants recognize the assistance the elderly are providing to alleviate their busy schedules. It would appear, therefore, that the types of support to the elderly reviewed in this chapter are not just relics of some earlier arrangements that are slowly vanishing, but rather part of a vibrant family system that is adjusting and accommodating to changing needs and conditions while still playing a vital role for its members. Future studies might well focus not so much on the presence or absence of assistance, but on its content, and on the reciprocities across generations that provide needed benefits to all family members.

Finally, the present analysis provides evidence related to one of the primary research questions of the Baoding survey project: Have China’s multiple revolutions produced distinctive alterations in family patterns, changes quite different from those experienced by Chinese living in Taiwan? At one level the contemporary patterns of familial assistance to aging parents look quite similar in Baoding and urban Taiwan, as we have stressed in the preceding pages. Neither revolutions and socialism (in the PRC) nor capitalist development and global culture (in Taiwan) have substantially eroded the strong filial orientations that stem from the Chinese tradition. With regard to Baoding, in particular, much of the analysis in preceding chapters has stressed the theme of the survival and even reinforcement after 1949 of traditional filial obligations.

However, beneath these broad similarities we see clear differences in the way familial support is provided to aging parents in these two Chinese locales. Furthermore, there is a patterning to these differences, with parents in urban Taiwan (particularly the Taiwanese, more than the Mainlanders) relying more heavily on support from grown children than in Baoding. Taiwanese parents also receive more support from sons and daughters-in-laws in comparison with daughters than is the case in Baoding. Although these differences are subtle rather than stark, they suggest that family relations in urban Taiwan are in some ways more “traditional” than in Baoding. “Having sons to guard against old age” remains more important today in urban Taiwan than in Baoding. The distinctive institutional legacy of Chinese socialism—including work unit-provided pensions, virtually universal employment of women, and the absence until recently of family-owned enterprises and other property—has produced somewhat more evolution away from “traditional” Chinese family patterns than has resulted from the higher level of economic development in Taiwan. So the answer is yes, China’s revolutions have made a difference and have produced distinctive features of family life. Chapter 11 will consider whether similar conclusions can be derived from some of the attitudinal data collected in the surveys in Baoding and in urban Taiwan.

NOTES

¹ 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); A. I. Hermalin, M. B. Ofstedal, and M. C. Chang, “A Comparison of Alternative Approaches to Measuring Needs and Levels of Support Received among the Elderly in Taiwan,” in Readings in Population Research: Policy, Methodology and Perspectives, eds. P. Krishnan et al. (Delhi, India: B. R. Publishing Corp., 1992).

² In the health section of the Taiwan questionnaire, respondents were asked if because of a health or physical problem they had any difficulty shopping for personal items, managing money, using the telephone and several other items related to household activities. An analysis of these items vis-a-vis those asking respondents if they received any household assistance revealed that many respondents who did not report any difficulty in executing these tasks nevertheless reported receiving assistance (Hermalin, Ofstedal and Chang, “A Comparison”). This finding lends support to the hypothesis that to some degree, reports of household assistance are reflective of long-standing arrangements rather than help provided because of physical limitations.

³ See Hermalin et al., “Types of Support,” for Taiwan findings. A possible exception involves the provision of physical care in Taiwan. Given the small number of cases that receive physical care assistance it is difficult to draw firm conclusions, and more research is needed.

⁴ A. I. Hermalin, M. B. Ofstedal, and M. L. Lee, “Characteristics of Children and Intergenerational Transfers.” Elderly in Asia Research Report No. 92–21. Population Studies Center, University of Michigan (Ann Arbor: University of Michigan, 1992).

⁵ A. I. Hermalin, C. Roan, and A. Perez, “The Emerging Role of Grandparents in Asia” (paper presented at the annual meeting of the Population Association of America, Chicago, April 1998).

Chapter 10, by Albert Hermalin, Mary Beth Ofstedal, and Shiauping Shih, then compares the pattern of exchanges between parents and grown children in Baoding and in urban Taiwan, in a manner somewhat parallel to the one presented in chapter 5. In chapter 11, by Jennifer Cornman, Jieming Chen, and Albert Hermalin, several categories of attitudes between parents and children of Baoding and urban Taiwan are compared, in a manner somewhat parallel to that presented in chapter 4. In general, the results presented in these chapters show that while strong filial obligations and extensive exchanges between generations are characteristic of both Chinese settings, at the same time there are a number of ways in which the results in urban Taiwan and in Baoding diverge. To oversimplify somewhat, we found that although Taiwan is clearly a more economically developed locale than the PRC or even Baoding specifically, in certain ways the patterns in the Taiwan data appear more “traditional” than their counterparts in Baoding.