| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
RESEARCH ARTICLE |
1 Gerontological Studies, McMaster University, Hamilton, Ontario, Canada.
2 School of Social Work and Family Studies, The University of British Columbia, Vancouver, Canada.
Address correspondence to Lori D. Campbell, PhD, McMaster Centre for Gerontological Studies, McMaster University, 1280 Main Street West, KTH 230a, Hamilton, Ontario, Canada L8S 4M4. E-mail: lcampbe{at}mcmaster.ca
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionREFERENCES |
|---|
Methods. The concepts that guide this research relate to family obligations or motivations to provide care, specifically, commitment to care, legitimate excuses, and caring by default. Data for this research come from the Work and Family Survey (1991–1993) conducted by the Work and Eldercare Research Group of CARNET: The Canadian Aging Research Network.
Results. Although such factors as geographic proximity and sibling network composition predict men's involvement independent of the type of task, the gendered nature of the task is important in how other factors, such as filial obligation, parental status, education, and income influence involvement in care.
Discussion. The findings suggest that, for traditionally male tasks, legitimate excuses or a commitment to care may play a more minor role in influencing men's involvement than is true for traditionally female tasks. Overall, this research demonstrates the importance of examining the gendered nature of the care tasks and highlights the value of the conceptual framework for explaining variations in men's filial care.
RESEARCH on filial caregiving has tended to focus on the care provided by adult daughters. However, adult sons also provide care to older parents (Campbell & Martin-Matthews, 2000a
, 2000b; Chang & White-Means, 1991; Harris, 1998; Kaye & Applegate, 1990; Matthews & Heidorn, 1998). Although this article looks at the relationship between filial caregiving and a variety of sociodemographic factors, our primary goal is to examine the relationship between men's filial caregiving and the gendered nature of care tasks. In particular we want to determine whether the factors influencing the provision of filial care remain the same, or vary, depending on whether the tasks involved are viewed as traditionally male care, traditionally female care, or gender neutral care.
Literature Review
A consistent finding in the literature is that adult daughters are more likely than adult sons to be caregivers to both mothers and fathers (see, e.g., Dwyer & Coward, 1991). Daughters are more likely than sons to provide more domestic assistance and more personal care (traditionally female care), and, when sons do provide care, they tend to perform tasks consistent with traditional or normative roles for men such as home maintenance chores and financial or managerial assistance (traditionally male care; Chang & White-Means, 1991; Stoller, 1990; Young & Kahana, 1989).
Some research (Horowitz, 1985) has found that for tasks that are less structured by traditional gender roles, such as providing transportation, helping with shopping, or running errands (gender neutral care), sons and daughters do not differ in their level of involvement. The literature is mixed on whether emotional support is gender neutral. Some research has found that sons are as likely as daughters to provide emotional support (Horowitz, 1985), whereas other work has suggested that daughters provide more emotional support to elderly mothers than do sons (Houser, Berkman, & Bardsley, 1985). However, there is evidence that a gender bias may be at work, with women overall more willing than men to report emotionally supportive behavior.
Typically, men who are primary care providers are either without siblings, without sisters, or the only geographically available child. Although sons who assume the role of primary caregiver are as extensively involved as daughters in filial care (Horowitz, 1985), daughters are significantly more likely to provide assistance with personal care. Among men and women who provide personal care, gender differences in frequency of care tend to disappear. However, gender differences are related to the specific care tasks performed (Martin Matthews & Campbell, 1995), with daughters more likely than sons to provide the most intimate types of care. Cross-gender taboos involving personal and intimate care are also relevant here. With "strong feelings of gender boundaries" (Rose & Bruce, 1995) surrounding the most intimate kinds of care, it is less expected and less acceptable for men to provide this kind of care, particularly to mothers (Arber & Ginn, 1995).
Some sons may step into care through the principle of substitution (Shanas et al., 1968) when female family members are not able to provide assistance. However, sons are more likely to substitute in some activities than in others, as, again, strong gender norms discourage sons from generally performing traditionally female tasks, even when daughters are not available (Spitze & Logan, 1990). Matthews (1995) suggested that family members, including older parents, adopt cultural assumptions about what constitutes gender-appropriate behavior, such as perceiving women as the "nurturers" in families and men as peripheral to the nurturer role. The gendered nature of these sociocultural assumptions and expectations might also help to explain sons' reduced involvement in certain types of assistance, particularly personal care (Davidson, Arber, & Ginn, 2000).
Our review of the literature identified three particularly salient concepts, each contributing to our understanding of obligating, motivating, and limiting factors that influence family assistance; these form the foundation of the conceptual framework guiding this analysis. These concepts are (a) feelings of filial obligation (Cicirelli, 1983; Lee, 1992) to reflect commitment to care; (b) the availability of legitimate excuses, that is, reasons or circumstances that are perceived as valid for an individual being unable, rather than unwilling, to provide care, such as other family or work obligations (Finch, 1989; Finch & Mason, 1993); and (c) caring by default (Horowitz, 1985) because of the unavailability of other individuals or options for caregiving. We discuss each of these in turn, focusing particularly on the concept of legitimate excuses, because factors associated with this construct dominate the analyses to follow.
Legitimate Excuses
The concept of legitimate excuses has been used in discussions of family obligation (Finch, 1989) and negotiating family responsibilities (Finch & Mason, 1993). Although the term "excuses" implies a kind of illicit avoidance of obligation, we—following Finch and Mason—use it here in a wider and nonjudgmental way to reflect a range of accounts, explanations, and justifications that get constructed when individuals negotiate family obligations and care relationships. Although gender variability is found in some instances of how legitimate excuses are negotiated and deployed in relation to family obligations and support, in the reasons or justifications that are acceptable for limiting assistance to family, and in how readily these reasons are accepted by others, Finch and Mason (1993) concluded overall that a "gender hierarchy does not directly account for what happened" (p. 101). They suggested, rather, that "gendered biographies can mean that women—unlike men—are seen as being in a position which makes them able to juggle various commitments simultaneously without the need to prioritise one over another" (p. 124). Women's biographies, and not men's, put them in a position where the "legitimacy" of other commitments and constraints may hold less weight than the commitments and constraints on men, creating the tendency for men to have their excuses accepted as legitimate more readily than for women. Because the strongest evidence of "a gender dimension" in Finch and Mason's research involved the issue of perceptions of "competence," particularly in the provision of personal care, our research examines the relationship between legitimate excuses and the provision of particular kinds of help.
Although the expectation to provide filial care is not typically placed upon most men, men are in fact involved in different aspects of caregiving—some in extensive and nontraditional ways. This research moves beyond the more typical "gender difference" approach that has dominated the caregiving literature to look at within-gender characteristics or circumstances that influence male involvement in care. When we focus within gender, the circumstances of men's involvement, or justifications for their limited or lack of involvement, are not so clear or unambiguous. Men's family relationships are unique and varied, so that some men will have more or less valid or acceptable reasons for limiting their involvement in caregiving than do other men. Therefore, the concept of legitimate excuses is particularly relevant for examining differences within gender.
Legitimacy of excuses also appears to hinge upon being able to claim an inability, rather than an unwillingness, to provide support. The types of excuses identified by Finch and Mason as acceptable reasons for not being "able" to provide care include other prior commitments, not being available, a lack of skills for the tasks (competence), and geographic distance. Their findings have informed the operationalization of legitimate excuses in this present research on men's caregiving, within the areas of employment constraints, other family commitments, geographic distance, and personal resources.
Obligations and Commitment to Care
Research finds that greater feelings of filial obligation are linked to a greater likelihood of providing instrumental support to parents (Silverstein & Litwak, 1993) or to greater assistance to parents in general (Lee, 1992). Intergenerational affection, intimacy, and altruism appear to be strong motivators for daughters to provide assistance, whereas sons seem to be more motivated by feelings of filial obligation, familiarity, and service, rather than sentiment (Silverstein, Parrott, & Bengtson, 1995).
Finch and Mason (1993, p. 96) discussed responsibility and commitment within family relationships as being uniquely created and shaped by individuals over time, rather than being ascribed, based on gender. However, they also suggested that one's gender may be important in "developing commitments"; that is, "the social conditions under which gender ... [is] lived may help to create conditions which are conducive to the development of certain kinds of communication between individuals." Although individuals have choices related to their obligation or commitment within relationships, gender may also play a role. Further, choices can be constrained by other obligations or commitments (Finch & Mason, 1993). Within this overall context we examine how men's feelings of obligation or commitment influence their involvement in care.
Caring by Default
The third concept, caring by default (Horowitz, 1985), reflects circumstances in which men may be providing care because there are no other individuals to provide care or options for arranging care. Men who are only children, or men who are without sisters, as well as those who live geographically proximate to the older relative in need of care, may find themselves providing care, in part because they are without other viable choices or individuals who can assist with or take responsibility for caregiving, or their proximity and family structure create a situation in which care cannot be easily resisted. The concept relates in some ways to the "principle of substitution" (Shanas et al., 1968), in that sons may be substituting for unavailable daughters and thus providing care by default because of the family structure. There may, however, be limits to the type of care that men will be defaulted to, and this again may be influenced by the gendered nature of the care and the expectations surrounding the type of care, including the expectations of the older parent.
Within the literature on families and aging, there is a recognition of the tensions that can be created between contradictory pressures related to the intersection of gender, individual, and structural characteristics and family care responsibilities; one specific example involves geographical mobility on the one hand, and responsibility for the care of an elderly parent on the other (Finch & Mason, 1993). In order to better understand some of the tensions that may characterize gendered filial care, this article explores the relationship between gender and caregiving responsibilities. By using such concepts as commitment to care, legitimate excuses, and caring by default, we are able to examine the extent and nature of these fundamental tensions in men's experiences of filial care.
Research Questions and Hypotheses
The primary question to be answered in this research is whether factors related to the concepts just identified (commitment to care, legitimate excuses, and caring by default) explain men's involvement in different types of care (traditionally male care, gender neutral care, and traditionally female care). Our expectations, guided by the existing literature, are as follows:
1. Men who report greater feelings of filial obligation are more likely to be involved in each type of care (commitment to care).
2. Men with other family, work, or distance constraints are less likely to be involved in each type of care than are men without these other obligations or constraints (legitimate excuses). We expect this will be particularly true in relation to traditionally female care (as it is considered to be nonnormative for men).
3. Men who are without siblings or sisters, and those who live near their older parent, are more likely to be involved in each type of care than are men with siblings or sisters, or those who live farther away (caring by default).
|
METHODS |
|---|
|
TOPAbstractMethodsResultsDiscussionREFERENCES |
|---|
; Gottlieb, Kelloway, & Fraboni, 1994). Forty-six percent (2,337 respondents) of the original sample responded "yes" to the following question: "Have you provided care or assistance to a relative 65 years of age or older in the previous 6 months?" This secondary analysis utilizes data on the 773 men who indicated that they had provided care or assistance to at least one parent or parent-in-law aged 65 years or older in the previous 6 months. We purposely selected only adult sons who provided some form of filial care or assistance so that we could specifically examine factors that might create differences in men's level of involvement in a range of care tasks. Because of the nature of the data, we were unable to identify those men who had an older parent requiring assistance, but who were not providing care. There may indeed be men in the original sample who had opted out of care overall, just as there were men in our own study who were not involved in all types of care examined. Although this may result in an underestimation of the extent to which sons choose not to provide any care, our sample allows us to examine whether particular sociodemographic and family structure factors do in fact predict men's level of care involvement in different types of tasks. Table 1 presents selected characteristics of our sample.
|
; Coward & Dwyer, 1990; Logan & Spitze, 1996; Montgomery & Kamo, 1989; Spitze & Logan, 1990; Stoller, 1990; Young & Kahana, 1989). For each of the 18 tasks, respondents were asked how often they provided that type of assistance, with response options of 0, never; 1, once a month; 2, several times a month; 3, once a week; 4, several times a week; and 5, daily. Each respondent then had a score (0 to 5) for each task, with his scores summed to create the three caregiving scales. Traditionally male care is measured on a scale that includes the following types of help: managing money (48% of men), completing forms and documents (65%), regular financial assistance (20%), and home maintenance and yard work (56%). The sample size for traditionally male care is 771 men (80 men score zero on the scale, but 691 men provide this type of care). The reliability alpha is.65; range 0–19, X = 2.65, and SD = 2.49.
The scale for gender neutral care includes such help as household chores (44% of men), transportation (59%), assistance with shopping (47%), assistance in getting around (34%), arranging assistance from agencies (33%), dealing with serious memory problems (20%), and dealing with mood swings or extreme behaviors (25%). These tasks are generally those described in Katz's formulation of instrumental activities of daily living (IADLs). The sample size for gender neutral care is 767 (155 men score zero on this scale, with 612 involved in this type of care). The reliability alpha is.81; range 0–28, X = 4.44, and SD = 4.89.
The third scale measures traditionally female care by summing men's help with personal care and other tasks considered nonnormative for men. This help includes dressing and undressing (7% of men); laundry (14%); bathing, washing, and grooming (7%); toileting (4%); feeding–eating (7%); taking medications (15%); and preparing meals (19%). These tasks are described by Katz as activities of daily living (ADLs). The sample size is 756 men (505 men score zero, with 251 men providing this type of care). The reliability alpha is.86; range 0–35, X = 1.46, and SD = 3.77.
Independent variables
Filial obligation: A filial obligation scale was created by summing responses to six statements, including the following: It is my obligation as an adult child to take care of my older parents; as an adult child, settled and with a job, I should help my parents if they need it; I should not allow better financial opportunities to take me away from my parents. There are four response choices for these statements, ranging from 1, disagree a lot, to 4, agree a lot. The reliability alpha is.78; range 6–24, X = 19.96, and SD = 3.11.
Family structure: The measures of legitimate excuses include the family structure variables of respondent's marital status, parental status, and sibling network composition. Marital status includes three categories: married (including common-law couples), which is the reference category (92% of men); formerly married (5%); and never married (4%). Parental status is represented by a dummy coded variable with three categories: any children 12 years of age or younger (the reference category, representing 41% of the sample); all children over 12 years of age (46%); and childless (13%). Sibling network composition, a dummy coded variable, is composed of the four categories of no siblings (reference category and 8% of men), sisters only (19%), brothers only (23%), and brothers and sisters (50%).
Distance constraints: Distance constraints were assessed in terms of "time to travel from respondent's home to relative's home." The mean distance to all relatives is 100.82 min, with a standard deviation of 168.28, and a median and mode of 30 min.
Employment commitments or constraints: Employment variables include number of hours employed in a week (X = 48 hr; SD = 9.4) and whether the respondent does unpaid work after hours for his job. The latter is a categorical variable with "yes" or "no" response options, and it is dummy coded with "no" as the reference category. Fully 76% of the men do extra unpaid work after hours.
The perceived workplace support variable is measured on a three-item scale constructed from responses to statements including the following: I feel that my employer–organization tries to find ways of helping me meet my family responsibilities; my supervisor–manager shows flexibility when I face demands from my family responsibilities. Response choices range from 1, strongly disagree, to 4, strongly agree. The higher the overall score, the more likely the respondent is to perceive his workplace as supportive of his family responsibilities. The reliability alpha is.64; range 3–12, X = 8.30, and SD = 1.48.
Personal resources: The variables used to measure personal resources are the respondent's education and personal income (both continuous variables) and occupation, which is collapsed into two categories: professional–management–administrative (the reference category for the dummy coded variable) and other. The respondents in this sample are relatively well educated (40% have at least a university degree); have a relatively high income (a mean of $50,000–60,000 Canadian); and occupy primarily professional, managerial, or administrative positions (72%). That these men are relatively well off (educationally, financially, and occupationally) limits the generalizability of our findings to the larger population, particularly those men with less education or lower incomes and those unemployed or in nonprofessional–nonmanagerial occupations.
Other variables: Other variables included in the analysis were (a) age of the older relative being cared for (X = 75 years, SD = 6.5, and range 54–95 years); (b) number of relatives cared for (57% assisting one relative, 33% assisting two, 7% assisting three, and 4% assisting four relatives); and (c) length of time caring for a relative (X = 6.2 years; SD = 5.7).
Because the three dependent variables are continuous, ordinary least squares (OLS) multiple regression analysis is used to determine significant predictors of men's care involvement. The somewhat skewed distribution of scores for each dependent variable, and the substantial number of men providing no assistance within each type of care (the large number of zeros), initially raised the issue as to whether OLS multiple regression was the appropriate statistical technique for these data. One significant risk that occurs when OLS multiple regression is used is that a "real" finding might be missed because the violation of linearity could result in underestimating a relationship. However, this will not diminish the significant findings that emerge from these analyses. Further, a tobit analysis approach was also considered, but, on further investigation, it was not judged to be a more appropriate analysis technique for these data. Specifically, as a tobit approach is typically used when one is trying to estimate a regression line where data either are over a particular threshold or are a "negative" (Kennedy, 1998), using tobit would then suggest that we were, in fact, looking to estimate "less than zero" caregiving for our sample of men. Although it is important to acknowledge the potential risks of using OLS analysis, this statistic enables us both to examine the factors that influence each type of care independently and to compare significant predictors across the three types of care.
|
RESULTS |
|---|
|
TOPAbstractMethodsResultsDiscussionREFERENCES |
|---|
.05, **p .01, and ***p .001. The correlation between filial obligation and geographic proximity (-.08*) is not a substantively robust relationship but does indicate that closer proximity to the older relative is related to greater feelings of filial obligation. Men with older children tend to also have older parents (.38***). Higher income is related to having older children (.20***), having older parents (.16***), and living farther away from the elderly relative being cared for (.14***). Among correlations with the three types of care, greater involvement in traditionally male care is associated with having older parents (.13***), living nearer to the older relative (-.13***), and more employment hours (.09*). Greater involvement in gender neutral care is correlated with greater filial obligation (.11**), older children (.15***), older parents (.23***), living closer to the older relative (-.20***), less time providing care (-.07*), and caring for more relatives (.08*). Finally, greater involvement in traditionally female care is related to greater filial obligation (.12***), older children (.13***), older parents (.20***), and living closer to the older relative (-.09*). Many of these specific relationships are explored in the multivariate analyses presented here.
Traditionally Male Care Model
Results for the multiple regression analysis for traditionally male care are reported in Table 2. This model explains approximately 10% of the variance in men's care involvement in traditionally male tasks.
|
Education is significant for this type of care, with higher levels of education correlated with more care involvement. Perceived workplace support is negatively correlated with involvement in care, with men more involved in traditionally male care reporting less workplace support. Age of relative also emerges as significant, with men who have older relatives showing more involvement in care. Filial obligation is not significant here.
Gender Neutral Care Model
Approximately 19% of the variance in men's gender neutral care involvement can be explained by this model (see Table 2). Filial obligation emerges as significant; men who report greater feelings of filial obligation are more involved in gender neutral care. Geographic proximity is also significant, with men who live closer providing more care than those farther away.
Parental status is significant, indicating that men who have older children provide significantly more care than men with younger children. As well, sibling network composition emerges as significant, with men who are without siblings more involved in gender neutral care than men in any of the other sibling networks.
The number of relatives cared for, age of relative, and length of time providing care are all significantly related to care in this model. As expected, the more relatives being cared for, the greater the care involvement, and the older the relatives, the greater the involvement in care. However, the longer care has been provided, the less the involvement in gender neutral care.
Traditionally Female Care Model
Findings for men's involvement in traditionally female care are also presented in Table 2. This model explains 12% of the variance in men's involvement in care. Filial obligation is positively correlated with involvement in traditionally female care. Geographic proximity is also significant, with men who live closer to their parent more involved than men who live farther away.
Parental status is also significant, with men who have older children significantly more involved in care than those with younger children. Sibling network composition is also significant. As with the other models, men without siblings are significantly more involved in traditionally female care than are men with sisters only or with brothers only; unlike the other two models, however, men without siblings are not significantly more involved in traditionally female care than are men with brothers and sisters.
Perceived workplace support is negatively correlated with caregiving in this model. Personal income also emerges as significant here, with men who have a lower income more involved in traditionally female care than are men with a higher income. Finally, the number of relatives cared for and age of relative are both significant: the more relatives cared for, and the older the relatives, the more care involvement.
|
DISCUSSION |
|---|
|
TOPAbstractMethodsResultsDiscussionREFERENCES |
|---|
Legitimate Excuses
Family structure
Men who have older children provide more gender neutral care and traditionally female care (but not more traditionally male care) than do men who have younger children. Because younger children are, potentially or typically, in greater need of parental care than older children, they may be a legitimate reason for not having the time or resources to be more involved in filial care, particularly when that care is less expected from men (traditionally female care). Men in this sample who have younger children also tend to have younger parents. Therefore, it may be a situation of having both younger parents (who typically require less assistance) and younger children (who typically require more care) that together shape less filial care involvement.
That differences do not emerge for traditionally male care suggests that assistance with tasks that are more securely tied to a son's traditional role within the family, such as home repair, yard work, or managerial or financial assistance, may be less influenced by their own child-care circumstances or responsibilities. The expectations surrounding traditional care for men, in essence, may take precedence over or, at least, not be seen as interfering or inconsistent with, the demands of young children in the same way as nontraditional or even gender neutral tasks.
Men without siblings are significantly more involved in all types of care than men in the other sibling groups, with one exception. Men with both sisters and brothers are as involved in traditionally female care as men without siblings. Although the reason for this latter pattern is not readily apparent, it does demonstrate the value of examining the gendered nature of care tasks. The findings overall provide strong support for the hypothesis that men will be more involved in care when they do not have siblings to assist in, or assume, the provision of care. The expectation that men without sisters would be as likely as men without siblings to be more involved in caregiving is, however, not supported by this work. There are several possible explanations: Brothers share the provision of care, or married brothers have spouses (the sisters-in-law) who take on some of the caregiving tasks. Men without siblings must assume that care themselves. Overall, however, this research suggests that being an only child diminishes the legitimacy of other competing demands for all types of caregiving examined.
Marital status is not a significant predictor of men's involvement in any type of care. This suggests that other factors may have a greater influence on care involvement than whether men are with or without a spouse, or even the nature of their unmarried state (divorced vs. never married). Marital status is also not correlated with geographic proximity or filial obligation (in the correlation matrix discussed previously), indicating that never married men in this study are no more likely than married men to live near their older relative or to report greater feelings of filial obligation. In work by Arber and Gilbert (1989), being never married and being proximate (coresiding) are important for greater care involvement. This interweave of marital status, proximity, and commitment to care, however, does not emerge in our study.
Distance constraints
Men who live closer to their relatives are significantly more involved in all types of care than are men who live farther away. Greater distance, therefore, appears to be a legitimate excuse for less involvement in care. However, men who live farther away may have a genuine desire to provide more assistance but are hindered by distance. Therefore, distance may be a legitimate excuse for some, but a barrier to care for others. Further, men who live close to older relatives may choose to do so because of a commitment to provide care.
Employment commitments–constraints
Our expectation that greater employment commitments would be legitimate reasons for less involvement in caregiving is not supported. Further, perceived workplace support is significant for traditionally male care and traditionally female care only, although the relationship is counter to our expectations. Men who are more involved in these two types of care are less likely to perceive their place of employment as supportive of their family responsibilities. Men who are involved in traditionally female care may require more time away from work, or more flexibility from their workplace, to allow them to meet their caregiving responsibilities—accommodations that the workplace may be unable or unwilling to provide. The reason for comparable findings for workplace support in relation to traditionally male care is not immediately apparent. However, that a significant relationship is not found between perceived workplace support and gender neutral care suggests that tasks such as assisting with transportation or shopping, or providing emotional support, are in fact less intrusive in the workplace or just more easily structured into the work–family routine than are other types of assistance, such as traditionally female tasks. Therefore, the issue of workplace support may be less meaningful for gender neutral care tasks as compared with the other types of care.
Personal resources
Men who have a higher income provide significantly less traditionally female care than men with a lower income, as hypothesized. However, income is not a significant predictor of the other types of care. Respondent's education emerges as significant for traditionally male care only, but opposite to our expectation, men with a higher education were more involved in this care. Respondent's occupation is not significant for any dimension of care examined.
These findings, together, show that men with higher education do more traditionally male care and men with higher income do less traditionally female care. In this research, higher education and higher income reflect greater personal resources and, therefore, men who possess more resources may be able to use these resources to take on the role expected for men in care, and to limit or resist care that is more nontraditional care for men. A higher income can provide men with the prestige or power to be less involved in traditionally female tasks, or the financial ability to pay for formal assistance. Further, if one considers tasks that measure traditionally male care, such as managing money, completing forms, and helping with finances, men who are better educated may feel that they are more suited to provide this type of assistance and, because of their educational abilities and skills, they may be viewed by others as most suited for these tasks as well.
There is a belief that higher education is linked to more flexible or progressive gender role ideology and behaviors. However, in this study, men with a higher education are not more involved in traditionally female types of care than are less educated men. It would appear, therefore, that if they do express more progressive beliefs about gender roles, their behaviors do not match these expressions. The relationship between lower income and more traditionally female care, and higher education and more traditionally male care, does provide some support for the hypothesis that educational or financial prestige (or lack of it) is connected to caregiving involvement and to the gendered nature of that care, in important ways.
Caring by Default
The expectation that men who are without siblings and those living closer to their older relative will be more involved in care is generally supported in all types of caregiving examined. However, counter to findings by Horowitz (1985) and to our hypothesis, men without sisters are not as involved in care as are men without siblings. Men who are proximate and men who are without siblings may, in part, be caring by default because of the structural factors of living close by and being without brothers or sisters. However, again, it is possible that men who live closer to their older parent may do so because of a committed relationship and that, as only children, they may have developed a particularly close bond with their parent that influences their desire to provide care.
Other Significant Factors
Because the survey data do not provide an indicator of the older parent's health or need, the age of the older relative was used as a proxy for these more direct measures. Age does emerge as significant for all three types of care, supporting the expectation that men with older parents will be more involved in care, independent of the type of care.
The number of relatives receiving assistance is significantly related to greater help with gender neutral care tasks, but not the other types of tasks. This suggests that gender neutral tasks (such as providing transportation or arranging assistance with outside agencies) can be integrated into one's daily schedule or more easily combined with other activities or commitments. Therefore, this type of help may be more readily provided to more than one relative than can other types of assistance.
This finding may also reflect the unique nature of each type of care. For example, traditionally female tasks generally involve greater face-to-face time, commitment, and frequency. As the number of relatives receiving care increases, such care may therefore be less easily coordinated with other types of assistance or other commitments in men's lives. Further, providing traditionally female care for more than one relative might be even less expected of men (than providing traditionally female care for one relative) and, therefore, less likely undertaken.
The level of assistance to one or multiple relatives can vary not only by the gendered nature of the care tasks but also by the gender and relationship of the older relative(s) receiving care. For example, a son may be more likely to provide traditionally male assistance to his widowed mother living alone than to his married parents or his widowed father-in-law. Ideally, the examination of gender in this context would include in the regression analyses the gender and relationship of each care recipient, rather than grouping all care recipients (mothers, fathers, mothers-in-law, and fathers-in-law) together. However, the data available did not allow for differentiation of care recipients by gender or relationship. The survey questions on level of involvement in specific care tasks combined the care provided to all relatives.
The length of time providing care also emerges as significant in this research, with men who have been providing care longer being less involved in gender neutral care and traditionally female care, but not traditionally male care. The literature is mixed on length of care and care involvement. Some work suggests that sons' involvement in care diminishes over time (Montgomery & Kamo, 1989), whereas other research (Stoller, 1990) finds continuity in sons' caregiving over time. Without longitudinal data, one must be cautious when interpreting our present results. This limitation is relevant to the interpretation of our findings. It may be that for care that is less tied to men's traditional roles in the family, over time, sons relinquish care to other care providers (Montgomery & Kamo, 1989). The needs of the older relative may move beyond the care a son can comfortably or easily provide, or is expected to provide (more personal care), and, therefore, over time, other care arrangements (a female caregiver or more formal assistance) may be needed. For care that is more traditionally male in nature, the length of caregiving may be a less salient issue. Overall, these findings suggest both continuity and change over time in men's involvement related to the gendered nature of the care: continuity in the level of assistance with tasks that are more normative for sons, and a pattern of decreasing involvement for the types of care that may be more representative of the parent's greater level of disability or need.
Conclusions
In sum, findings from this research demonstrate the importance of examining the gendered nature of care tasks to better understand how motivating and obligating factors influence men's involvement in different types of filial care. Although certain competing obligations or family structure factors (e.g., distance and sibling network composition) predict men's involvement in filial care independent of the type of task, the gendered nature of the task is important in whether other factors (e.g., filial obligation, parental status, education, and income) influence men's caregiving. Further, for more traditionally male tasks, legitimate excuses or a commitment to care play a more minor role in influencing men's care involvement than is true for traditionally female tasks. These research findings advance our understanding of the depth and complexity of male caregiving by acknowledging that men's involvement in such caregiving is influenced not just by the gendered context of caregiving per se in the society at large but also by the gendered nature of individual caregiving tasks.
Future research on men's filial caregiving would benefit from further exploration of the relationship history between adult sons and older parents—the relationship that develops across the life course, and how that influences caregiving. Within this focus, there is a need for greater recognition of the older parent as an integral part of the care relationship. Further, research that examines the more subjective dimensions of men's caregiving—how men perceive their caregiving experiences and contributions—is needed. Findings from this present research can be valuable as structural and interpretive guides in this future work.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received for publication May 31, 2002. Accepted for publication April 14, 2003.
|
REFERENCES |
|---|
|
TOPAbstractMethodsResultsDiscussionREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. Sims-Gould and A. Martin-Matthews Themes in Family Care-Giving: Implications for Social Work Practice with Older Adults Br. J. Soc. Work, December 1, 2008; 38(8): 1572 - 1587. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Savla, D. M. Almeida, A. Davey, and S. H. Zarit Routine Assistance to Parents: Effects on Daily Mood and Other Stressors J. Gerontol. B. Psychol. Sci. Soc. Sci., May 1, 2008; 63(3): S154 - S161. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Sims-Gould, A. Martin-Matthews, and M. A. M. Gignac Episodic Crises in the Provision of Care to Elderly Relatives Journal of Applied Gerontology, April 1, 2008; 27(2): 123 - 140. [Abstract] [PDF]
|
|
|
|
 |
|
 M. E. Szinovacz and A. Davey Changes in Adult Child Caregiver Networks Gerontologist, June 1, 2007; 47(3): 280 - 295. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. D. Campbell and M. P. Carroll The Incomplete Revolution: Theorizing Gender When Studying Men Who Provide Care to Aging Parents Men and Masculinities, April 1, 2007; 9(4): 491 - 508. [Abstract] [PDF]
|
|
|
|
 |
|
 A. Martin-Matthews Situating 'Home' at the Nexus of the Public and Private Spheres: Ageing, Gender and Home Support Work in Canada Current Sociology, March 1, 2007; 55(2): 229 - 249. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
