| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
RESEARCH ARTICLE |
Section on Socio-Environmental Studies, National Institute of Mental Health, Bethesda, Maryland
Address correspondence to Carmi Schooler, Ph.D., Section on Socio-Environmental Studies, Intramural Research Program, National Institute of Mental Health, National Institutes of Health, 6101 Executive Boulevard, Suite 360, Room 362, MSC 8408, Bethesda, MD 20892-8408. E-mail: carmi.schooler{at}nih.gov
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
OLD age is a time when many people often need the help of others in dealing with personal and financial difficulties (Wolfson, Handfield-Jones, Glass, McClaran, & Keyserlingk, 1993
). Not all elderly people are willing to ask for such help, even from their children (Blieszner & Mancini, 1987; Cohler, 1983). In fact, asking one's children for help may imply a change from a relatively independent "parent" role to a more dependent "elder" role. In this study we use an unusual body of longitudinal data to examine some of the determinants of older parents' willingness to ask their adult children for help. In particular, we ask what characteristics of parenting practices and experiences with a 13- to 25-year-old child would lead parents to feel free to ask that child 20 years later for help with a financial or personal problem.
Assuming that older parents' choices of where to turn for help for a personal or financial problem are affected by their perceptions of where they are likely to get such help (see Pillemer & Suitor, 2006), we find that two related research literatures become highly relevant. One examines the determinants of relationship quality between aged parents and aging children; the second examines how the quality of such relationships may affect adult children's actual and perceived willingness to help their parents. Reviewing these two literatures, Suitor, Pillemer, Keeton, and Robinson (1995) suggested that there are several general theories of motivation for such caregiving. One set of theories, which includes both exchange theory (George, 1986) and equity theory (Cicirelli, 1991), centers on the norm of reciprocity. According to these theories, adult children provide help for their parents either to repay their parents for the time and effort their parents spent taking care of them in the past or in the expectation of receiving a future inheritance. Direct reciprocity between generations is evident when high levels of shared activity time between parents and children lead to higher levels of support to parents in later life (Silverstein, Conroy, Wang, Giarrusso, & Bengtson, 2002).
Attachment theory (e.g., Cicirelli, 1991) is seen by Suitor and colleagues (1995) as providing yet another possible source of motivation for adult caregiving of aging parents. According to this theory, the attachment relationship formed between parent and child during infancy lasts through adulthood so that "once the attachment bond has formed, the adult child wants to protect the parent whose life is threatened by sickness or injury" (Suitor et al., p. 235). Suitor and colleagues view attachment theory as one in which the sense of attachment derives not only from the infant's bonding with the parent, but also from parental behaviors and family practices that affect parent–child attachment after infancy. Suitor and colleagues argue that the care that adult children provide their parents when motivated by attachment lasts longer and is of higher quality than parental care motivated by the more limited motives of exchange or obligation.
Two types of parenting style have long been known to affect the quality of child–parent attachment during infancy and childhood (see Ainsworth & Bowlby, 1991). Secure attachment is associated with parenting that emphasizes emotional warmth, responsiveness, communication, and parent–child trust. Insecure attachment, in contrast, is associated with less responsive, more intrusive, and controlling parenting (Belsky, Jaffee, Hsieh, & Silva, 2001).
Although the concept of attachment was originally developed in the context of infant–parent relations, it has more recently been considered to underlie parent–child relations when the child is considerably older. Belsky and colleagues (2001) have suggested that the warm and trusting relationships that characterize secure attachment may continue to underlie congenial and satisfying relations between parents and their adult children in later years. In contrast, earlier coercive and controlling behavior on the part of the parent might be expected to lead to strain and conflict in later years.
Research on intergenerational relationships suggests that warm-responsive and intrusive-controlling parenting styles may each differentially affect the quality of relations between parents and their adult children in later years. The results of several retrospective studies have demonstrated that early parent–child relationships that were warm and supporting are associated with closer, more responsive later relationships; in contrast, earlier coercive, controlling parenting is associated with relationship difficulties in later years (Rossi & Rossi, 1990; Whitbeck, Hoyt, & Huck, 1994; Whitbeck, Simons, & Conger, 1991). More recently, Belsky and colleagues (2001) utilized a prospective methodology and found that supportive childrearing environments are associated with better parent–child relationships in later years.
Lang and Schütze (2002) maintain that filial autonomy—"the establishment of a sense of independence from one's parents and an understanding of one's parents' actual wishes"—is "associated with a resilient motivation to support parents even when experiencing much strain in the relationship" (p. 663). It seems highly plausible that such a sense of independence would be an expected outcome of the authoritative parenting style described by Baumrind (1966), in which parental responsiveness is critical.
The Current Study
Nearly all of the research on the effects of childhood parent–child relationships on later adult parent–child relationships has focused on the effects of such relationships on the way that children help and relate to their elderly parents. To our knowledge, no research has focused, as does our present research, on how parent–child relationships affect the ways in which older parents interact with their adult children in later years. In this study we investigate a specific aspect of this general question by examining how the nature of childhood parent–child relationships affects the readiness and willingness of older parents to ask their now-adult children for help with a personal or financial problem.
We examine the influence of two aspects of parenting—responsiveness and restrictive dominance—by using a longitudinal prospective design. Responsiveness refers to a parenting style that is emotionally close, warm, supportive, and responsive to a child's needs. Restrictive dominance refers to a parenting style that emphasizes strict obedience and conformity to parental rules and wishes. These characteristics of parental behavior or style have been intensively investigated and are among the most studied in the developmental literature (e.g., Ainsworth & Bowlby, 1991; Baumrind, 1966). We assessed parental style by means of interviews with a child and his or her parents in 1974. Subsequently, we asked parents in 1994 if there was anybody that they would ask for help and, if so, who this would be.
Both socioeconomic status (SES) and SES-related parental values played an important part in the Kohn–Schooler research program (Kohn & Schooler, 1983), the origin of the present study. In addition, it is theoretically plausible that both SES and SES-related parental values might have independent effects on a parent's willingness to ask for aid with a personal or financial problem. Therefore, we include both SES and high-status parental values as predictors of parental willingness to ask for help.
We examine possible requests for two kinds of help: help with a personal problem and help with a financial problem. We do so for three reasons. First, different variables might predict one's willingness to ask for different kinds of help. For example, one could imagine that characteristics of an interpersonal relationship might matter more when a person considers asking for help with a personal problem than for help with a financial problem. Second, one might expect to find that the predictors of the two kinds of help might differ for fathers and mothers. Finally, asking about two types of help was intended to keep us from the type of inappropriate generalization of findings that might occur if we had asked about only one.
Our general hypothesis is that parents whose relationships with a 13- to 25-year-old child were seen, at the time that they were occurring, as positive by both the parents and the child will more readily see that child as a source of personal and financial support 20 years later. Such parents will consequently feel freer to then ask that child for help with a financial or personal problem than those whose parenting practices and experiences were not seen as positive.
|
METHODS |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
). In 1974, we selected a strict probability subsample of one fourth of those men in the original sample (883) who were then no older than 65 years. Of these 883 men, 820 (93%) were located (i.e., we either knew where they lived or that they had died). Of the 785 men who were still alive, 687 (88%) were available and reinterviewed. We also interviewed the wives of the men then married (n = 555), and the child who had been the focus of the earlier inquiry (n = 352). This was the only time period in which we collected data from children. For the 1994 follow-up, we located 650 of the 687 (95%) households that had taken part in the 1974 survey. Of the 352 households in which both of the parents and the focused-on child had been previously interviewed, 315 (90%) of the parental couples were reinterviewed in 1994. The fathers ranged in age from 51 to 84 years in 1994 (M = 68.46 years, SD = 7.36). Mothers ranged in age from 55 to 87 years of age in 1994 (M = 66.08, SD = 7.22). The median education for fathers and for mothers was attainment of a high school diploma. The children, when interviewed in 1974, ranged between 13 and 25 years of age (M = 18.51, SD = 3.72). Of the 315 children included in the sample analyzed here, 211 were living at home with their parents at the time of the 1974 interview.
As described more fully in Schooler and Mulatu (2001), there is some evidence that female, older, unemployed, less educated, and African American individuals may be underrepresented among those interviewed in 1994. Consequently, one should use caution when generalizing our results. Nevertheless, because we analyzed the data for men and women separately, and, as we shall see, background social status variables do not affect the tendency to ask for help from children, our overall findings would seem to be reasonably reliable.
Data Collection
Data presented here are from responses to a subset of questions from much larger face-to-face interviews conducted in participants' homes. In 1964 and 1974, the National Opinion Research Center (NORC) conducted the interviews; in 1994, the NORC and the Cygnus Corporation (Rockville, MD) conducted the interviews. The NORC trained and supervised all interviewers. Participants in 1994 were paid $50 for their participation.
Measures
Responsiveness
We used items from the 1974 interview as indicators for two first-order factors (i.e., Closeness and Understanding, and Family Communication) and a second-order factor (i.e., Caring Openness). These three factors, in turn, served as indicators for the third-order Responsiveness factor (see Figure 1). Questions concerning Closeness and Understanding involved asking children about the degree to which (a) they were like their parents, (b) they would like to be like their parents, (c) they were close to their parents, (d) their parents were interested in why they did something of which the parents disapproved, (e) they felt their parents understood them, and (f) their parents treated them fairly. Children rated each parent separately. Items concerning Family Communication involved the children's ratings of the degree to which children believed each parent was interested in what the children had to say, the degree to which they were interested in what their parents had to say, and the frequency with which parents disagreed about what the children should do.
|
Restrictive dominance
Items from the 1974 interview that were used to measure parental Restrictive Dominance included ratings of parental strictness, restrictiveness, likelihood of laying down the law, and likelihood of dominating. We derived items from responses to four questions, each of which was answered by both parents (describing themselves and their spouses separately) and by the child (describing each parent). One question, about the likelihood that a parent would dominate the child, was asked of all parents and children; each person rated both parents separately. The other three questions, which dealt with strictness, restrictiveness, and the likelihood that a parent would lay down the law when the child misbehaved, were asked of all parents, and of those children who were still living in their parents' homes at the time of the interview; again, each person rated both parents separately.
High-status parental values
We used items from the 1974 interview to assess the degree to which each parent's values for his or her child resembled the values that differentiate high-SES from low-SES parents (Kohn & Schooler, 1983). The question asked of parents was as follows: "Which three qualities listed on this card would you say are the most desirable, most desirable of all, least desirable, and least desirable of all for a (boy or girl) of (the child's) age to have?" We coded responses for the value characteristics (i.e., having good manners, being neat, having good sense, acting like a boy or girl, being responsible, being obedient, and being interested in how and why things happen) into five categories (1 = the very least desirable; 2 = one of the 3 least, but not the least, desirable; 3 = not chosen; 4 = one of the 3 most desirable, but not the most; and 5 = the very most desirable value).
Socioeconomic status
We used three variables from the 1974 interview to assess SES in 1974 (see the subsequent description of measurement models): (a) family income, from 1 (less than $3,000) to 14 ($75,000 or more); (b) father's education, from 1 (primary grade level) to 9 (advanced degree); and (c) status of father's occupation according to the Hollingshead scale (Hollingshead & Redlich, 1958), from 1 (unskilled) to 7 (managerial or professional).
Background variables
Other background variables from 1974 included mother's age, father's age, and child's age.
Request for support
The central dependent variables were provided by fathers' and mothers' responses to two questions in the 1994 interview:
|
RESULTS |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
). Because there were some missing data, we used Mplus's analysis for missing data. We conducted analyses of men's and women's responses separately. For analyses in which mothers' willingness to ask the target child for help was the dependent variable, all assessments of parental characteristics were of the mother. Similarly, analyses of fathers' willingness to ask the focused-on child for help used assessments of fathers' parental characteristics.
In addition, because we had no reason to examine causal connections between the two conceptually distinct types of help seeking (personal problems, financial problems) that are our dependent variables, we analyzed these separately as well. Consequently, we report the results of four full-information structural equation models. Two of the models used mothers' willingness to request help from the focused-on child as the dependent variables, and two used fathers'. Within each pair, one analysis used willingness to request help with a personal problem as the dependent variable, and one analysis used willingness to request help with a financial problem. Each of these models met the criteria for appropriate fit (Comparative Fit Index > 0.90 and Root Mean Square Error of Approximation < 0.05; see Hu & Bentler, 1990; also see Browne & Cudeck, 1993); model fit measures are given in Table 1.
|
Findings
Results indicated the following. First, a history of greater maternal responsiveness significantly increased the willingness of the mother to ask the child for help with a personal problem; second, a history of greater maternal responsiveness significantly increased the willingness of the mother to ask the child for help with a financial problem; third, a history of greater maternal restrictive dominance significantly decreased the willingness of the mother to ask the child for help with a financial problem. In contrast, the parallel analyses for fathers indicated that neither responsiveness nor restrictive dominance had any significant effect on fathers' willingness to ask for help with either personal or financial problems. A series of supplementary analyses suggested that the findings regarding fathers were due neither to unwillingness to ask for help nor to fathers' relative insensitivity to the nuances of family relationships (analyses available on request).
We found no evidence that either SES or high-status parental values directly affected parents' later willingness to ask for help with a personal or financial problem. However, our models indicated a correlation between high-status parental values and SES. The Pearson product–moment correlations between SES and high-status parental values were statistically significant in all models (mother's personal help model, r =.48, p <.001; mother's financial help model, r =.48, p <.001; father's personal help model, r =.47, p <.001; father's financial help model, r =.45, p <.001). The effects of the ages of the father, mother, and child were all nonsignificant in all models; neither the age of the parent nor the age of the child affected the parent's willingness to ask the focused-on child for either kind of help.2
|
DISCUSSION |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This difference between the determinants of mothers' willingness to ask for financial help as opposed to personal help suggests that willingness to ask for help is not a unitary construct, and that such willingness depends on the nature of the help being requested. It also raises the possibility that, on one hand, the feelings of interpersonal closeness that may be particularly necessary to ask for help with a personal problem are fostered by the personality characteristics and interpersonal relationships that lead to or result from maternal responsiveness. On the other hand, only mothers who did not have a dominant hierarchic relationship (i.e., those whose parenting styles were low in restrictive dominance) with their children may later be able to accept the role reversal involved in asking their children for financial help.
In contrast to these findings, neither responsiveness nor restrictive dominance affected fathers' later willingness to ask for either personal or financial help. Furthermore, none of the analyses we carried out provided strong support for the possibility that this lack of any significant relationships was due either to a generally reduced willingness to ask for such help or to fathers' relative insensitivity to the nuances of family social relationships.
Our results support the predictions based on the attachment literature, at least where maternal behavior is concerned. (Belsky et al., 2001; Lang and Schütze, 2002; Rossi & Rossi, 1990; Whitbeck et al., 1991, 1994). They extend the findings of previous investigators of the effects of earlier parent–child relationships on subsequent social support. Much previous work has demonstrated the importance of earlier parent–child relationships for increasing the adult child's propensity to provide support for an aging parent (e.g., Whitbeck et al., 1994). In contrast, our findings demonstrate a relationship between early parental style and the willingness to request such support. With the available data we cannot determine the direction of the causal relationships between the parents' personal characteristics, their parental behavior, and their willingness to ask an adult child for help. Such a relative readiness to ask for help can be seen as a result of the positive feelings between parent and child to which such childrearing practices would seem to lead, as well as of the personal characteristics of parents that would lead them to adopt such an approach to childrearing to begin with. In addition, recent research suggests that characteristics related to the ones investigated here, such as emotional closeness and parent–child similarity, may affect a mother's willingness to request help of a particular child (Pillemer & Suitor, 2006). Nevertheless, the findings are certainly consistent with the claim that warm, supportive, and responsive parenting, at least on the part of mothers, predicts positive intergenerational relations in later years.
None of our background characteristics predicted the willingness to ask the focused-on child for help. The ages of mothers and fathers, as well as the ages of the focused-on children, were not significant predictors. In addition, family SES did not directly affect the likelihood of being willing to ask for either of these types of help. In addition, although high family SES did correlate positively with the likelihood of parents' holding parental values stressing self-direction rather than conformity (Kohn & Schooler, 1983), holding such values did not affect the later willingness to ask that child for help with either personal or financial problems. Thus, there is no evidence linking SES either directly or indirectly to mothers' willingness to ask for such help.
This pattern of findings may be subject to some limitations. As we have noted, there is some evidence that African Americans, women, and older, unemployed, less educated people may be underrepresented among those interviewed in 1994. It is unlikely, however, that this underrepresentation can account for the pattern of our results. More troubling is our inability to find strong evidence for any plausible explanation of why the earlier parental practices that so strongly affect mothers' willingness to ask for help had so little effect on fathers' willingness to do so. It would seem that although many fathers are clearly willing to ask their adult children for help, this willingness remains unaffected by how they related to their children as parents.
There are obviously related questions we might want to ask in future studies. Principal among these is whether the parents ever actually made such requests for help. Obviously, one would like to know not just whether parenting characteristics predict parents' willingness to request help, but also whether they predict actual requests. Nevertheless, our findings demonstrate that earlier relationships between parents and children do translate into parents' willingness to impose on these children when they are adults. This is particularly the case because the nature of our data permits us to rule out the possibility that a parent's willingness to ask a particular child for help is merely a function of that parent's recollection of his or her parenting practices and relationship to that child, rather than what they actually were.
Our results provide striking evidence of gender differences. They also provide, at least for women, evidence of how behavioral patterns in one stage of life predict behavioral patterns at another, much later, life stage. These effects occur not merely in terms of direct continuity of roles over time, but also in how behavior in one role (i.e., "parent," characteristically marked by giving and controlling behaviors) can affect later behavior in another role (i.e., "elder," characteristically marked by the necessity of getting help from others). Practically, our findings provide social workers and others aiding elderly individuals with some guidance about the characteristics of the past role performance of those individuals that should be considered when assisting them in getting help from their adult children.
|
Acknowledgments |
|---|
We thank the staffs of Cygnus Corporation, Rockville, Maryland, and the National Opinion Research Center for their work on the survey.
|
Footnotes |
|---|
2 Other analyses that also included the gender of the child and the number of sibs as well as latent factors measuring functioning in gross motor, fine motor, cognitive, and self-care activities of daily life (Caplan & Schooler, 2003) produced the same results. An essentially similar pattern of results was also found when analyses were carried out that compared those who asked the focused-on child for help with all other responding parents whether or not they asked anybody else for help.
Decision Editor: Thomas M. Hess, PhD
Received for publication February 14, 2006. Accepted for publication December 5, 2006.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  K. Glaser, C. Tomassini, and R. Stuchbury Differences Over Time in the Relationship Between Partnership Disruptions and Support in Early Old Age in Britain J. Gerontol. B. Psychol. Sci. Soc. Sci., November 1, 2008; 63(6): S359 - S368. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
