The aging of the population in the United States and elsewhere presents a number of macro-level issues and challenges for families. Declining fertility rates, and perhaps to a greater extent, decreasing mortality rates have influenced the size, age structure, and gender mix of families (Wolf, Reference Wolf, Martin and Preston1994). These demographic trends present care challenges as there are more family members in ages when care needs are relatively common and fewer in care-providing ages (Uhlenberg & Cheuk, Reference Uhlenberg, Cheuk, Szinovacs and Davey2008). At the same time, a host of social trends (serial marriages, co-habitation, delayed childbearing, non-marital parenthood, ubiquity of women in the labor force) and growth in racial, ethnic, and sexual minority populations have added complexity to American family life. In addition, the intensity of demands placed on family caregivers has increased as a result of government efforts to shorten hospital stays and the escalation of expenses associated with the long-term care of older adults with Alzheimer’s disease and other debilitating health conditions. What is not clear is if and how these macro-level population patterns actually mirror changes in micro-level family structures and influence the ways in which contemporary families care for their older members.
In this article, we address the diversity of contemporary family structures and implications for the care of older people in the United States. We begin with a critique of assumptions about conventional family membership and the provision of care for older relatives. We focus primarily on the need for care that occurs when older adults experience changes in their physical, cognitive, or emotional status that require frequent or substantial amounts of assistance with instrumental activities of daily living (IADLs: e.g., managing finances, preparing meals, monitoring medications) and personal care (ADLs: e.g., feeding, bathing, dressing). Our intent is not to provide a comprehensive review of the massive family caregiving literature, but to highlight select findings about the structure and demographic features of family caregivers that are fairly consistent across this body of research. Next, we present common variations in family structures and discuss how these pluralistic family formations influence caregiving roles, responsibilities, and relationships with older family members within the limits of the available research. We conclude with recommendations for further theoretical development to conceptualize the complexity of family structures and methodological strategies to advance research on the diversity within and across family structures and the ways in which contemporary families respond to the care needs of their older members.
Conventional Families and Care for Older Adults
The writings of early theorists and scholars on family structure were predicated on relationships believed to comprise four stages of the normative family life cycle (Glick, Reference Glick1977; Hill & Rodgers, Reference Hill, Rodgers and Christensen1964). Beginning with a heterosexual marriage, the family expanded with the birth of the married couple’s children and then contracted with the marriage of the couple’s biological children. The family life cycle dissolved with the death of a spouse and ended with the death of the surviving spouse. Reflecting Parsons’ (Reference Parsons1964) dichotomous presentation of gendered roles for husbands and wives, men in these conventional families were gainfully employed and provided for the financial and instrumental needs of the family. In turn, women assumed responsibility for the home, reared and cared for the children and other relatives as needed, and maintained relationships with extended kin and others in the community in which the family lived. The focus was on the nuclear family unit, with only peripheral acknowledgement of extended kin and others involved in the daily family life.
Smith (Reference Smith1993) referred to this description of the nuclear family as the “Standard North American Family (SNAF)” (p. 52), an ideological code that influences everyday language, research, advertising, legislation, and more. Reference to normative or conventional families permeated the scholarly family gerontology literature (for reviews, see Allen, Blieszner, & Roberto, Reference Allen, Blieszner and Roberto2000; Blieszner & Bedford, Reference Blieszner and Bedford1995; Brubaker, Reference Brubaker1990; Streib & Beck, Reference Streib and Beck1980; Troll, Reference Troll1971), despite the fact that such family patterns were statistically rare in the history of U.S. families (Coontz, Reference Coontz1992, Reference Coontz1997). For example, the share of the older population that is divorced has increased every decade since 1960 for all age groups and for both men and women (Jacobsen, Kent, Lee, & Mather, Reference Jacobsen, Kent, Lee and Mather2011). This demographic trend is but one contributor to the growth of diverse family structures in late life over the past half-century. Yet, the depiction of the SNAF endures and continues to provide a yardstick by which contemporary marriage, parenthood, and family relationships and responsibilities often are compared (Allen, Blieszner, Roberto, Farnsworth, & Wilcox, Reference Allen, Blieszner, Roberto, Farnsworth and Wilcox1999; Cherlin, Reference Cherlin2010; Coleman & Ganong, Reference Coleman and Ganong2003; Hansen, Reference Hansen2005), and scrutinized (Glenn, Reference Glenn1997; Popenoe, Reference Popenoe1993). Gerstel (Reference Gerstel2011) documented these trends in articles published in one of the most prestigious scholarly journals that concentrate on family life, the Journal of Marriage and Family (JMF). She found 88 per cent of the 537 articles published in JMF between 2004 and 2009 focused on the nuclear family; only 12 per cent focused on extended kin.
In addition, much of the family and caregiving literature reflects the population demographics of the 1980s when the majority of the U.S. population was White and the proportion of racial and ethnic minority group members did not differ much across successive cohorts (< age 20, 26%; ages 20–39, 21%; ages 40–59, 17%; ages 60 and older, 12%; Mather, Reference Mather2007). By 2006, the racial ethnic composition across cohorts differed substantially. Whereas the older groups were still majority White (ages 40–59, 28% minority; ages 60 and older, 20% minority), members of the younger groups were much more likely to be Hispanic, Asian, or multiracial (< age 20, 42%; ages 20–39, 39%).
Not only do the younger generation populations reflect rapid transformation of the racial ethnic composition in the United States, these younger persons are also changing how they are “doing” family. For example, Black and Hispanic adults have lower rates of marriage and higher rates of single motherhood than do non-Hispanic Whites (Carlson, McLanahan, & England, Reference Carlson, McLanahan and England2004), and higher reliance on extended kin throughout adulthood than White adults (Sarkisian, Gerena, & Gerstel, Reference Sarkisian, Gerena and Gerstel2006a). Cherlin (Reference Cherlin2006) has argued that families headed by single mothers exhibit “newer, and less ‘standard’ models [of family] that promise a more balanced reciprocity” (p. 803) within relationships, greater input on who counts as family, and stronger opinions about the rights and obligations associated with family roles. Although yet to be empirically tested, it is likely that these values and beliefs about family inclusion and involvement carry forward into late life and influence expectations for the care of older adults.
Studies focused on nuclear families also typically include mostly middle-class families with the resources to provide for the wants and needs of their members. However, a growing body of literature demonstrates the importance of taking class into account in order to understand how families are constituted (Furstenberg, Reference Furstenberg2009). For example, social class influences family formations (Shafer & James, Reference Shafer and James2013), behaviors (Schoen, Landale, Daniels, & Cheng, Reference Schoen, Landale, Daniels and Cheng2009), and cohesion (Sarkisian, Gerena, & Gerstel, Reference Sarkisian, Gerena and Gerstel2006b). Nevertheless, most studies on aging families treat older adults and their caregivers as a monolithic group and do not consider how class, as well as race, ethnicity, and other intersecting systems influence family relationships and care (Calasanti & Kiecolt, Reference Calasanti, Kiecolt, Blieszner and Bedford2012) and perpetuate the image of the independent nuclear family.
Conventional Family Care for Older Members
One of the earliest conceptualizations of the structure of care for older family members, the hierarchical compensatory model of care (Cantor, Reference Cantor1979, Reference Cantor1991), was predicated on SNAF. It purported that the choice of caregivers by the older adults followed an ordered preference based on the primacy of the relationship between caregivers and care recipients. Spouses were viewed as the preferred and first source of daily care, followed by adult children, and in their absence extended kin, and finally, formal service providers. The core assumptions of the hierarchical compensatory model were that many older adults reside in a conventional household with a living spouse or children nearby to provide care. The model ignored unmarried persons and obscured the involvement of other family members in the care of older adults. It also presumed that the nature of tasks required of family caregivers had little or no influence on the order of preferences for help. Family dynamics that influence the caregiver role also were overlooked. Another underlying premise of the model is that older adults have a positive relationship with their kin and therefore can depend on their family members for care. Finally, the model assumed universality in preferences for care, failing to acknowledge that demographic characteristics and cultural expectations influence who in the family is likely to assume responsibility for the older adult’s care.
The hierarchal model influenced initial beliefs about caregiving and implicitly guided the development and direction of much of the research on caregiving in late life. Studies of family care often address questions about who within the family provides care (Fisher et al., Reference Fisher, Franks, Plassman, Brown, Potter and Llewellyn2011; National Alliance for Caregiving, 2009), relationships between caregivers and older care recipients (Davey & Szinovacz, Reference Davey, Szinovacz, Szinovacz and Davey2008; Stephens & Franks, Reference Stephens, Franks, Qualls and Zarit2009; Stoller & Miklowski, Reference Stoller, Miklowski, Szinovacz and Davey2008), and negative and positive caregiver outcomes (Etters, Goodall, & Harrison, Reference Etters, Goodall and Harrison2008; Pinquart & Sörensen, Reference Pinquart and Sörensen2003, Reference Pinquart and Sörensen2007), but often fail to acknowledge the broader family structure and context that influence elder care roles and responsibilities.
Spouse Caregivers
In the United States, approximately 24 per cent of spouses are the primary caregiver for their husbands or wives (National Alliance for Caregiving, 2009). A meta-analysis of the caregiving literature revealed that spouse caregivers are approximately 70 years of age, White (73.3%), and female (64.79%) (Pinquart & Sörensen, Reference Pinquart and Sörensen2011). They provide the most intensive and extensive care, maintaining the caregiver role longer and tolerating a greater level of disability in the care recipient than would other family caregivers (Biegel, Sales, & Schulz, Reference Biegel, Sales and Schulz1991; Pinquart & Sörensen, Reference Pinquart and Sörensen2011). Research also suggests gender differences in the approach to spousal care (Stoller & Miklowski, Reference Stoller, Miklowski, Szinovacz and Davey2008). In addition to maintaining their routine household and family responsibilities, wives are likely to be solely responsible for the care of their ill husbands, providing assistance with instrumental and other activities of daily living (Hooker, Manoogian-O’Dell, Monahan, Frazier, & Shifren, Reference Hooker, Manoogian-O’Dell, Monahan, Frazier and Shifren2000) as well as assuming such roles as primary decision maker and daily life manager (Roberto, McCann, & Blieszner, Reference Roberto, McCann and Blieszner2013). Husbands, in contrast, are likely to rely on other family members and formal services to assist them with the daily care of their wives (Calasanti & King, Reference Calasanti and King2007).
In one of the few studies of the quality of care provided by older spouses, Williamson and Schaffer (Reference Williamson and Shaffer2001) found that quality of the relationships between spouses prior to one partner’s need for care influenced adaptation to the caregiving role, and quality of care provided. Spouses, regardless of their age, gender, or education level, who described their pre-illness relationship as more communal (e.g., feelings of responsibility for the other’s welfare; responsive to the other’s needs) were less likely to engage in potentially harmful caregiving behavior (i.e., psychological and physical mistreatment) than were those in relationships historically characterized by less frequent communal behaviors. Care recipients were most likely to report potentially harmful caregiver behaviors in situations where they had high levels of functional impairment and their spouse caregivers’ physical, mental, and cognitive health also was compromised (Beach et al., Reference Beach, Schulz, Williamson, Miller, Weiner and Lance2005). These findings highlight the potential for breakdown, under certain conditions, of the typically supportive caregiving situation assumed between spouses.
Adult Children Caregivers
According to the conventional hierarchical model of care, if an older adult is not currently married, or has a spouse who is unable to provide care due to his or her own physical or cognitive limitations, adult children or children-in-law step in as the primary caregivers (National Alliance for Caregiving, 2009). Often motivated by a sense of filial obligation, many adult children assume caregiving responsibilities for their aging parents while actively parenting their own children. In the United States, nearly 47 per cent of adults aged 40 to 59 are providing support to an aging parent while either supporting young or adolescent children or providing assistance to an emerging adult child (Parker & Patten, Reference Parker and Patten2013). Adult children caregivers provide a wide range of care, from providing meals to performing complex medical procedures to providing 24-hour supervision and care.
The pathways adult children and children-in-law follow in becoming caregivers differ across family structures. When there are multiple siblings, the child who lives closest to the parent, the one who is available, or the one perceived by other family members as having the skills and temperament to provide the amount and type of care needed may become the de facto caregiver (Checkovich & Stern, Reference Checkovich and Stern2002). In other families, who assumes the caregiver role for aging parents may be influenced by inherent cultural norms and practices (e.g., wife of oldest son) or the caregiver role may “just happen”, evolving over time as parents experience changes in their functional or cognitive abilities and their need for care intensifies (Roberto, Reference Roberto and Craft-Rosenberg2010). Although often grouped together in research because of assumed filial similarities, there is evidence that adult children and children-in-law caregivers differ with respect to gender and marital status. Fewer adult children caregivers are female, and they are much less likely to be married than children-in-law (Pinquart & Sörensen, Reference Pinquart and Sörensen2011) – two characteristics known to influence who assumes family caregiving roles.
Among adult children caregivers, daughters are more likely to be the primary caregivers for their older parents, spending more hours and engaging in more extensive and intensive care activities, whereas sons often take on a more secondary role (Johnson, Reference Johnson2006). When sons do assume the role of a primary caregiver, they provide care at similar levels to daughter primary caregivers (Zarit & Eggebeen, Reference Zarit, Eggebeen and Bornstein2002). Matthews (Reference Matthews2002) found that gender of one’s siblings makes a difference in parental care responsibilities. Perhaps because of their sisters’ stronger expectations of family participation, men who had sisters provided more care for their parents in coordination with their sisters. Conversely, men who had only brothers tended to respond to their parents’ needs independently and were more likely to take a care management versus a caregiving approach.
Findings from longitudinal analyses of national data, albeit limited, have suggested that caregiving roles assumed by adult children may not be as static as portrayed in the more common cross-sectional studies. Rather, the composition of the sibling care network changes over time (Davey & Szinovacz, Reference Davey, Szinovacz, Szinovacz and Davey2008). The possibility for change is tied to family structure, with greater availability of siblings predicting a greater likelihood of change in the family caregiving network. However, the likelihood of change is less common for the adult child providing the most care. What has yet to be sufficiently addressed is the added-value of multiple children providing care versus having a single adult child or child-in-law assume primary responsibility for the care of their older parents. While sharing the care may result in more-positive outcomes for primary caregivers, it is not clear if having multiple children involved in their care is advantageous for older parents as well.
Extended Kin Caregivers
Bengtson (Reference Bengtson2001) called attention to the increasing importance of multigenerational relationships (involving three or more generations) to individuals and families in American society, hypothesizing that “for many Americans, multigenerational bonds are becoming more important than nuclear family ties for well-being and support over the course of their lives” (p. 5). Intergenerational support patterns ebb and flow over time (Silverstein, Parrott, & Bengtson, Reference Silverstein, Parrott and Bengtson1995) and serve as “latent kin networks” of support (Riley & Riley, Reference Riley, Riley, Bengtson and Achenbaum1993) that often are enacted only in times of crisis. For example, in accordance with the hierarchal model of care, grandchildren may initially serve as secondary caregivers (Hamill, Reference Hamill2012; Tennstedt, McKinlay, & Sullivan, Reference Tennstedt, McKinlay and Sullivan1989) who provide periodic assistance to their grandparents to support the elder’s primary caregivers before eventually becoming full-time caregivers if their parents can no longer maintain that role (Fruhauf, Jarrott, & Allen, Reference Fruhauf, Jarrott and Allen2006; Piercy & Chapman, Reference Piercy and Chapman2004).
Unfortunately, members of extended families often are categorized as the “other” relatives who receive little attention within the context of the studies in which they appear (Gerstel, Reference Gerstel2011). This is particularly true in the caregiving literature where few family members beyond spouses and adult children are the foci of empirical studies. Yet, for families from minority groups and lower socioeconomic classes, who often have fewer economic resources than White, nuclear families, reliance on the extended kinship network for care may be a survival strategy in the face of economic difficulties (Dilworth-Anderson, Williams, & Gibson, Reference Dilworth-Anderson, Williams and Gibson2002; Nichols, Martindale-Adams, Burns, Graney, & Zuber, Reference Nichols, Martindale-Adams, Burns, Graney and Zuber2011).
Despite the plethora of research on caregiving in nuclear families, this structure is found in only a minority of families (Cohen, Reference Cohen2014; Coontz, Reference Coontz1992, Reference Coontz1997). Influences on longevity and changing perceptions about the nature of adult intimate relationships and family life in general, as well as various social movements (civil rights, gender equity), legal and policy changes (marriage and divorce laws, social welfare), and social problems (incarceration trends, poverty) have led to a much broader array of family structures than SNAF. In turn, such variations in family structures can have implications for caregiving provided to older family members.
Pluralistic Families and Care for Older Members
In contemporary society, multiple family structures challenge and diverge from the ideal of the conventional pattern. A convenient term to encompass this myriad of family types is pluralistic families (Allen et al., Reference Allen, Blieszner, Roberto, Farnsworth and Wilcox1999). Variations in family structure by characteristics such as age and generation, marital status, sexual orientation, biolegal ties, parental status, and geographic location may present both opportunities and challenges to pluralistic families caring for older members as compared to caregiving under the conventional family structure. We recognize overlap across various pluralistic structures, yet discuss them separately as exemplars of the diversity found in contemporary families and patterns of care for older adults.
Variations by Age and Generation
Beanpole Families
As life expectancy increases and mortality rates decline, many older adults may be members of three-, four-, and even five-generation families, potentially comprising spouses/partners, children, grandchildren, and siblings (Agree & Hughes, Reference Agree, Hughes, Blieszner and Bedford2012). In families with numerous generations, children have more opportunities to know their aunts, uncles, grandparents, and even their great-aunts/-uncles and great-grandparents, especially their great-aunts and great-grandmothers. This demographic shift has manifested itself in what Bengtson and colleagues termed the beanpole family, a vertical extension of family structure characterized by an increase in the number of living generations within a lineage coupled with a decrease in the number of people within each generation (Bengtson, Lowenstein, Putney, & Gans, Reference Bengtson, Lowenstein, Putney, Gans, Bengtson and Lowenstein2003; Bengtson, Rosenthal, & Burton, Reference Bengtson, Rosenthal, Burton, Binstock and George1990). Despite benefits that may arise from children knowing several generations of older relatives, this vertically extended multigenerational family structure may result in an increased imbalance in caregiving demands for members of the middle generation if they lack siblings or other kin with whom to share care-related tasks (Agree & Hughes, Reference Agree, Hughes, Blieszner and Bedford2012). In addition, individuals who delay childbirth (typically, adults with high education and income levels) are unlikely to experience as many generations (Matthews & Sun, Reference Matthews and Sun2006).
In the United States, the caregiver support ratio, or the number of potential family caregivers between the ages of 45 and 64 for each person aged 80 and older, is declining sharply. In 2010, the family caregiver support ratio was 7:1; by 2030 it is projected to fall to 4:1 (Redfoot, Feinberg, & Houser, Reference Redfoot, Feinberg and Houser2013). These projections lead to several questions related to caregiving: Given that most care for older relatives is provided by a spouse or adult daughter rather than being distributed across other caregivers (National Alliance for Caregiving, 2009), does the declining support ratio necessarily portend neglect of older relatives’ needs? Given that the beanpole family structure implies multiple generations of older adults, will members of younger generations who have opportunities to know several generations of family elders be more inclined to assist them, or will they be more overwhelmed with care needs? If care typically is provided by those in adjacent generations, will members of one older generation be able to care adequately for those in a yet-older generation?
Age-Gapped Families
Among the men and women who delay the birth of their first child, the gap between generations widens. A recurrent, unresolved issue for intergenerational relationships is the long-term effect of postponed childbearing. Some scholars propose that members of such age-gapped families may experience difficulties in developing affective bonds, transferring shared values across generations (Rossi, Reference Rossi, Lancaster, Altman, Rossi and Sherrod1987), and instilling a sense of responsibility to provide care for older family members (George & Gold, Reference George and Gold1991). But Caputo (Reference Caputo1999) found that African American women were more likely than White women both to delay childbirth and to be co-residing in intergenerational families, suggesting that their intergenerational responsibilities might have led to delayed childbirth rather than the reverse. Because age-gapped families also tend to have low fertility, either some older adults may be without family caregivers or the demands of managing both childrearing and caring for older relatives may be heightened for members of the middle generation (Settersten, Reference Settersten2007).
Age-Condensed Families
In contrast, age-condensed families have smaller age distances between each generation. Early fertility results not only in young mothers but also in young grandmothers and great-grandmothers (Burton, Reference Burton1990). Relationships in age-condensed families have a sibling-like quality more than a parent–child quality as parents and children often have similar experiences at the same time (Setterson, Reference Settersten2007). According to Burton, Obeidallah, and Allison (Reference Burton, Obeidallah, Allison, Jessor, Colby and Shweder1996), the overlapping worlds of generations in these families can also contribute to ambiguity in family relationships and care responsibilities. For example, in Caputo’s (Reference Caputo1999) study, caregiving was more likely to be extended downward towards grandchildren than upward towards parents. Ambiguity in care responsibilities and the downward generational focus may mean that family members in different generations who are close in age will be particularly challenged to provide needed care to the oldest relatives and to relatives from multiple generations needing care simultaneously.
Variations by Marital Status
Remarriage
While long-term marriage between heterosexual couples is a common configuration at the center of family structures in late life, second (or additional) unions also are prevalent. Although the proportion of individuals who remarry following divorce or widowhood declines with age, men are more likely to remarry than women (Sweeney, Reference Sweeney2010). Later-life repartnering is about choice and well as constraint. In the United States, a sex-ratio imbalance in older adulthood limits the availability of potential male partners for women and grows markedly with increasing age from 100:114 for 65–69-year-olds to 100:210 for persons aged 85 and older (Kung, Hoyert, & Murphy, Reference Kung, Hoyert, Xu and Murphy2008). Economic resources also influence decisions to repartner, with wealthier older adults more likely to remarry than to stay single (Vespa, Reference Vespa2012). Among partnered older adults with functional limitations, more than 80 per cent received care from their spouse (Lima, Allen, Goldscheider, & Intrator, Reference Lima, Allen, Goldscheider and Intrator2008). Caregiving among remarried couples is as likely as among conventional couples to be influenced by gender, race, ethnicity, and culture (Calasanti & Kiecolt, Reference Calasanti, Kiecolt, Blieszner and Bedford2012; Silverstein, Lendon, & Giarrusso, Reference Silverstein, Lendon, Giarrusso, Blieszner and Bedford2012).
Partnerships
Some single older adults are reluctant to marry or remarry, preferring alternative forms of partnered and family relationships. Romantic co-habitation in late life is on the rise (Brown, Bulanda, & Lee, Reference Brown, Bulanda and Lee2012) and another form of romantic partnering, Living Apart Together (LAT), is gaining visibility among older adults in the United States. Defined as committed, monogamous intimate partnerships between unmarried individuals who live in separate homes but identify themselves as a couple (de Jong Gierveld, Reference de Jong Gierveld2002), older LAT couples describe their relationships as no less committed than happily married couples. Their level of commitment hinges on personal desire alone and is not rooted in any sense of obligation to remain a couple (Reference Benson and WhitbourneBenson, in press). LAT relationships afford older adults safeguards against various forms of loss (e.g., of autonomy or material and financial assets), as well as protection against loneliness and isolation (Upton-Davis, Reference Upton-Davis2012).
An important question is whether or not older adults in LAT relationships can rely on their partners as potential sources of long-term care. When older LAT partners were asked to describe how they would handle the situation of their partner’s becoming seriously ill, those with more conventional beliefs about relationships and those involved in longer LAT relationships showed willingness to provide care. LAT-men demonstrated greater readiness to provide care compared to LAT-women (Karlsson, Johansson, Gerdner, & Borell, Reference Karlsson, Johansson, Gerdner and Borell2007). These women’s responses reflected interest in sustaining their autonomy and reluctance to pursue traditional gendered division of labor roles. In contrast, older LAT partners who are less likely to define their relationships as marriage-like may be reluctant to take on long-term caregiving responsibilities for each other. If the hierarchical compensatory model (Cantor, Reference Cantor1991) of care continues to predominate, we assume when adult children are available and able, they will step in to provide care for LAT partners, or at least for their own parents. Likewise, older adults in LAT partnerships may rely on extended and fictive kin or “chosen family” – close friends who assume family-like roles (MacRae, Reference MacRae1992) – for help if adult children are unavailable or unwilling to support the individual or collective care needs of older LAT couples.
Variations by Sexual Orientation
Single LGBT Persons
Like single heterosexual persons, members of the lesbian, gay, bisexual, and transgendered (LGBT) community do not follow the conventional hierarchical compensatory model for receiving care in old age. Instead, they may rely on siblings and extended kin for needed assistance (if those relatives accept their sexual orientation) as well as fictive kin who volunteer to take on caregiving activities (Croghan, Moone, & Olson, Reference Croghan, Moone and Olson2014). In fact, many older LGBT care recipients rely upon their friends as a safety net against unmet needs for care (Muraco & Fredriksen-Goldsen, Reference Muraco and Fredriksen-Goldsen2011). Caregiving friends acknowledge benefits received from providing care, including feeling good about themselves, perceiving improvement in their self-esteem, and engaging in typical friend activities (Muraco & Fredriksen-Goldsen, Reference Muraco and Fredriksen-Goldsen2014). Still, caregiving can create conflict among friends that may negate these beneficial feelings and such unstructured relationships may be tenuous with respect to meeting long-term caregiving needs (Litwak, Reference Litwak, Adams and Blieszner1989). Even if granted legal authority by an LGBT person, friends may not be perceived by others as having the right to make health care and financial decisions. Lack of support resources and lack of acceptance among formal care providers increase stress among both LGBT persons needing assistance and those attempting to provide it (Croghan et al., Reference Croghan, Moone and Olson2014). LGBT elders worry about discrimination from formal care providers, particularly when their chosen family members lack social or legal sanction, and about not having anyone to help in an emergency (MetLife Mature Market Institute, 2010).
Same-Sex Partners and Spouses
Social, political, and legal changes in the United States have led to greater acceptance of partnered relationships among LGBT older adults. During the 2000s, the LGBT civil rights movement has focused on gaining same-sex marriage or other forms of civil relationship recognition for same-sex couples. A recent qualitative study found that legalizing same-sex marriage brought older couples increased financial and medical security (e.g., with respect to making health care decisions and providing health benefits) (Lannutti, Reference Lannutti2011). Marriage also conveyed greater relational security, even for couples who had already been in long-term committed relationships. They believed that being legally married enhanced their recognition as a legitimate couple, which could increase social support for a bereaved partner upon the death of one of the spouses (Lannutti, Reference Lannutti2011). This recognition might also motivate others to provide needed care for the widowed partner.
Caregiving norms consistent with heterosexual marital relationships appear to govern same-sex partnerships and marriages as well. When the need for care arises, older LGBT adults report that they would first turn to their partners, then to friends and other family members (Cahill, South, & Spade, Reference Cahill, South and Spade2000). Recent findings from a study on long-term care needs in a large nationally representative sample found that both older men and women living with same-sex partners reported significantly greater need for assistance with instrumental and personal care activities than those in different-sex relationships (Hiedemann & Brodoff, Reference Hiedemann and Brodoff2013). In attempting to explain this finding, the authors pointed out that lesbian and bisexual women are more likely to be obese than are heterosexual women, and gay men are more likely to have HIV infection than are heterosexual men. Differences in underlying health conditions as well as variance in social, behavioral, and cultural norms (Bowen, Balsam, & Ender, Reference Bowen, Balsam and Ender2008) may account for the findings. Nevertheless, this study demonstrates the significant caregiving role often assumed by same-sex partners. However, further research is needed on questions such as whether long-term partnered LGBT couples care for each other as married heterosexual couples care for their spouses, and the implications of any differences that may exist. The existing studies do not examine the specific needs of bisexual and transgendered older persons separately. Research has not given sufficient attention to the strengths and resilience skills that older LGBT persons and couples bring to old age and their need for caregiving.
Care by LGBT Adults
To date, there has been little empirical research on family-of-origin caregiving by LGBT adults. Based on available qualitative data and anecdotal accounts, LGBT caregivers’ reasons for providing help (e.g., need; filial responsibility), types of caregiving tasks in which they engage (e.g., emotional and instrumental support), and concerns they have about caregiving (e.g., where to turn for help, how to access services, and what services are available) are very similar to those of heterosexual family caregivers (Cantor, Brennan, & Shippy, Reference Cantor, Brennan and Shippy2004). About half of these caregivers were women, and about two-thirds of lesbian caregivers said that they were either the sole provider or provided most of the care. As with the general population of family caregivers, LGBT caregivers reported significant burden and stress as a result of their caregiver role (e.g., juggling the demands of being employed and giving care, limited time for self and other relationships, and conflicts or disagreements with other family members about the care of the older person). LGBT caregivers may face additional challenges related to strained relationships when the relatives they are trying to help (e.g., their parents) have a negative view of their sexual orientation or choice or partner (Cohen & Murray, Reference Cohen and Murray2007).
Variations beyond Normative Biolegal Ties
Stepfamilies
The social construction of family membership is often altered when either parents or their children divorce and remarry, thus creating reconstituted families. Evidence suggests that in these families, many stepchildren do not define their stepparents as parents or even as relatives, particularly if remarriage occurs when offspring are teenagers or adults. Thus, the filial commitment of adult stepchildren to their aging stepparents is often fragile (Sweeney, Reference Sweeney2010). Differences in norms of obligation and of relationship quality may contribute to adult stepchildren’s being less willing to provide care and assistance for their stepparents than adult children offer their biological parents (Ganong & Coleman, Reference Ganong and Coleman2006; Pezzin, Pollak, & Schone, Reference Pezzin, Pollak and Schone2008; Sweeney, Reference Sweeney2010). Geographic distance may affect care provision as well. Adult stepchildren and their stepmothers are less likely to live together, less likely to live nearby, and less likely to move closer than adult biological children and their mothers (Seltzer, Yahirun, & Bianchi, Reference Seltzer, Yahirun and Bianchi2013), thus making help to stepmothers less convenient. The potential detrimental effects of parental divorce and stepfamily formation on support that adult (step)children provide to elderly adults raises concerns about the adequacy of family caregiving among future cohorts of older adults that experience high rates of divorce and remarriage.
Research on caregiving for older stepparents is not plentiful, and studies of offspring’s caregiving for stepparents in same-sex partnerships are even scarcer. Moreover, as with heteronormative families, not all families headed by same-sex (step)parents are alike. Researchers have not explored similarities and differences in assistance to stepparents from adult stepchildren born before the same-sex partnerships occurred compared to assistance from adult stepchildren born into same-sex unions (Sweeney, Reference Sweeney2010).
Fictive Kin and Impermanent Family Ties
In research on extended family and fictive kin ties in late life (Allen, Blieszner, & Roberto, Reference Allen, Blieszner and Roberto2011), we found that older adults from both mainstream and marginalized families employed kin reinterpretation practices as a means of adapting to impermanence in family ties. They used qualifying language, such as “she is not really my daughter, but she is like a daughter to me”, to convey the meaning of an important relationship that otherwise would have no commonly understood ideological code. We identified family networks that were both expanding (by descendants’ marriages, remarriages, childbearing, adoptions, and the like) and contracting (because of divorce, death, and other losses). The older adults designated the people they considered family by revising the standard structures of kinship (e.g., daughter, sister, son), to promote a more distant relative to a primary or closer kin position, exchange biological or legal ties within the kin hierarchy, convert friends and other nonrelatives to close kin status, retain close kin ties despite changes in family structures (e.g., divorce), or psychologically remove members of their kin network due to death, divorce, or relocation. The purpose of these kin reinterpretation strategies was to normalize and contextualize unanticipated changes in the family life course.
As our study participants described who was in their family and to whom they felt close, they revealed the family members upon whom they actually relied. The meanings associated with qualifying statements such as “He’s not just a brother, he’s more like a twin”, and “‘Step’ is not a good word for it; she’s just my granddaughter”, are ways in which the older adults bridged their familiarity with the modern nuclear family ideal and their actual experiences with complex family situations and relationships. They acknowledged establishing and sustaining a range of ties with non-kin, primary kin, and former kin, and not simply for sentimental reasons. Holding flexible notions of family has the potential to enrich affective bonds and extend the utility of older adults’ kin networks (Johnson, Reference Johnson1999). Any one of these reinterpreted relationships is a potential source of caregiving to the older adults that could substitute for or complement other sources of assistance (Arber, Reference Arber2004; Roberto, Allen, & Blieszner, Reference Roberto, Allen and Blieszner2001). For example, a woman in our sample whose only son had died felt confident that her niece would help her in the future – she “promoted” her niece to the role of daughter. A man whose son married a woman with children from a previous marriage redefined his daughter-in-law’s children as his grandchildren and treated them the same as his biological grandchildren. Having close relationships with the non-biological grandchildren might motivate them to provide help to him in the future. In both cases, these older adults potentially expanded their network of care providers.
Variations by Parental Status
Childlessness
The effect of childlessness on meeting care needs has received relatively scant attention (Dykstra & Hagestad, Reference Dykstra and Hagestad2007). Although a growing body of research confirms that childlessness is not necessarily a negative life course outcome (Dykstra & Hagestad, Reference Dykstra and Hagestad2007), it is still routinely viewed as a risk factor for social isolation and distress in late life. Childless older adults tend to rely heavily on their spouses or partners for social and emotional support (Zhang & Hayward, Reference Zhang and Hayward2001), and they report strong links with extended family members such as siblings, nieces, and nephews, as well as with neighbors, friends, and other community members (Albertini & Kohli, Reference Albertini and Kohli2009; Milardo, Reference Milardo2010). At the same time, they appear to have fewer people to turn to for support when they need care, which may leave them more vulnerable to having unmet needs compared to older parents (Dykstra, Reference Dykstra2006). Thus, those who are childless in late life may have to rely exclusively on formal services to address their care needs (Albertini & Mencarini, Reference Albertini and Mencarini2014), and they have a higher likelihood of requiring institutional care than older adults with children (Grundy & Jitlal, Reference Grundy and Jitlal2007).
The parent versus nonparent classification says little about the diversity of experiences of childlessness in late life (Allen & Wiles, Reference Allen and Wiles2013), and the potential for differential effects on receipt of needed care. Rather than comparing only childless older adults with older parents, it is important to acknowledge various pathways to childlessness in late adulthood. One route involves absence of offspring throughout life; the other involves losing adult children to incarceration or death from illness and accidents. Although never having been a parent is a different experience than losing offspring later in life, whether these two pathways to childlessness have differing impacts on older adults’ receipt of needed care is unclear. It is possible that lifelong childless elders develop strong ties to extended and fictive kin early on, and those long-term close relationships encourage nieces, nephews, cousins, or family-like friends and neighbors to come to their aid. In contrast, if former parents had relied mainly on their offspring for meeting their needs over the years, loss of those children in late life could leave the recently childless elders bereft of family helpers.
Variations by Place
In the conventional family tradition, adult offspring found work and established their families in proximity to their parents, affording opportunities for mutual instrumental, social, and emotional support through the years as well as ease of providing more-extensive care when needed. But educational opportunities, economic pressures, and retirement lifestyle choices have led to geographic dispersion of generations across the United States, at least in families with the financial and other resources to do so. These opportunities for personal development and flexible choices have increased the complexity of caregiving for the oldest generations. Frequently, older parents have at least one adult child living nearby (Johnson, Reference Johnson2006). In all likelihood the older parent, the nearby child, and that child’s siblings all expect the proximal child to provide the majority of needed parent care. At the same time, not all needed assistance requires hands-on proximity, and imaginative division of caregiving tasks across siblings, coupled with effective use of technology to manage certain caregiving responsibilities from a distance (Kinney & Kart, Reference Kinney, Kart, Blieszner and Bedford2012), could ease the burden on the closest sibling and enable all of the children to participate in parent care. The propensity of families to plan for that type of shared caregiving and the overall effects of incorporation of technology on parent care and caregiver burden remain to be investigated.
International and transnational families, in which the dynamics of family life reach across continents and cultures (Skrbiš, Reference Skrbiš2008), represent another dimension of family variation by place. On the one hand, marriage and other partnerships that join nationalities, as well as international adoption, increase within-family cultural diversity. Establishing an international family challenges the members to learn about unfamiliar cultures and how to manage varying familial expectations, including different norms for elder care. On the other hand, transnational families arise when some relatives live at great distances from others, including both temporary relocation and long-term international immigration. These circumstances, too, can require new kinds of kinship and care practices. Filial piety that crosses geographic borders may take on what Sun (2012, p. 1241) deemed new “rules of reciprocity” that encourage traditional ethnic norms of elder care yet allow for parental autonomy, rely on multiple siblings for support rather than just one, and emphasize emotional rather than physical care. This flexibility suggests that, in the context of transnational family caregiving, older adults and their relatives may assume creative and innovative roles while modifying and expanding traditional value systems (Skrbiš, Reference Skrbiš2008; Sun, Reference Sun2012).
Research on care provision to older family members has seldom addressed the implications of international and transnational family structures for meeting care needs of older adults. Expanded cultural experiences and understanding may give rise to families that are more adaptive and resilient in meeting elders’ needs for care than are more homogenous families. It also is likely that the growing acceptance of new technologies will contribute to and help sustain the family’s involvement in long-distant caregiving (Baldassar, Baldrock, & Wilding, Reference Baldassar, Baldock and Wilding2007). In contrast, geographic distance and time-zone differences, lack of formal elder care resources or lack of understanding about how to access them, and clashing norms and expectations might strain international and transnational family members beyond their limits of coping, putting older relatives at risk for inadequate care.
Advancing Research on Family Care for Older Adults
Structural flexibility is likely a more common feature of families than scholars have acknowledged and an understudied aspect of family care. Researchers need to inquire about the potential meaning of all close relationships – not just blood and legal ties, but also anyone who is “like a relative” – in order to understand how individuals associated through family and family-like connections view their filial obligations and responsibilities (Blieszner & Bedford, Reference Blieszner, Bedford, Blieszner and Bedford2012). Investigations should focus not only on individuals’ propensity to provide care for elders currently in their networks, but also on ascertaining their views about supporting their peers and other elders as they grow older together. This type of research requires new conceptualizations of “family” and of caregiving strategies, strong theoretical guidance, and advanced research designs to fully capture and interpret changes in both conventional and pluralistic family structures and the care of older members.
Theoretical Considerations in Understanding Diverse Family Structures
Deeper and broader understanding of the pluralistic structures of aging families requires use of theory that is consonant with the conceptualization of “family”. For example, post-modern approaches such as social constructivism would offer more flexibility and fit in the context of diverse family structures than traditional structural-functionalist approaches. Therefore, it is important that researchers choose the theoretical underpinnings of their studies thoughtfully.
Although no single unifying theory on aging families and caregiving exists, there are opportunities for further development within and greater synthesis across existing paradigms to strengthen the explanatory power of family-focused caregiving research. For example, deriving key individual and family-level variables based on the life course perspective’s core concepts of time, human agency, linked lives, and social forces (Elder, Reference Elder and Lerner1998) would provide historical and cultural evidence on micro, meso, and macro issues (Bronfenbrenner, Reference Bronfenbrenner1979) that influence family relationships and pathways to providing care for older family members. Likewise, incorporating theoretical principles of lifespan human development (Baltes, Reference Baltes1987) would promote attention to neglected research foci such as earlier life antecedents of current care needs and caregiving behaviors, adaptability of family members to changing life circumstances, and the effects of normative and non-normative transitions into the realm of caregiving and care receiving (Roberto, Blieszner, & Allen, Reference Roberto, Blieszner and Allen2006).
Understanding how pluralistic families interact around caregiving matters also requires a close examination of the intersections of social locations based on gender, age, class, race, ethnicity, and sexuality. The use of feminism (Calasanti & Slevin, Reference Calasanti and Slevin2006) and intersectionality (Calasanti & Kiecolt, Reference Calasanti, Kiecolt, Blieszner and Bedford2012) as theoretical approaches brings to the forefront the effects of privileges and disadvantages embedded in social institutions such as aging families and health care settings. The theoretical benefit of feminist and intersectional stances is that, unlike monistic approaches, they account for co-occurring experiential and structural complexity. To date, studies of families and caregiving have focused attention on one or two intersections (e.g., gender and ethnicity or gender, race, and socioeconomic status), but have yet to examine multiple intersections simultaneously (Calasanti & Kiecolt, Reference Calasanti, Kiecolt, Blieszner and Bedford2012). Rather than reducing their effect by statistically controlling for these kinds of variables, investigators should examine their collective impact on the other variables under study.
In addition, theory development and theorizing about relationship dynamics within aging families require further elaboration of existing concepts such as intergenerational ambivalence (Lüscher & Pillemer, Reference Lüscher and Pillemer1998) to identify motivational factors of relatives who are currently peripheral or uninvolved caregivers as well as those of non-traditional caregivers. Finding ways to tap these persons as potential caregivers may better meet needs of older adults than using only formal services. For example, Connidis (Reference Connidis, Blieszner and Bedford2012) endorsed using collective ambivalence (i.e., “mixed feelings across multiple children”; Ward, Reference Ward2008, p. S245) to examine and link individual feelings to family structure and composition. More-complex families have potential for greater collective ambivalence. This concept might shed light on the caregiving motives held by those in multiple types of relationships comprising a particular family.
Finally, increasing diversity of family structures is likely to be accompanied by increased likelihood of partnering with formal service providers to sustain care for older members over time and through their growing need for advanced levels of assistance. Thus, theoretical development is needed to more precisely predict what formal care strategies will be acceptable to older adults and family members under what circumstances and conditions. Andersen’s (Reference Andersen1995) behavioral model of health services is a useful starting place because it considers the constellation of personal characteristics that predispose individuals and families to seek formal help (demographics, health beliefs), enabling resources (personal knowledge, family availability, community services), and need (perceived, evaluated) as predictors of service use and of perceived health outcomes. Although this model has been predominant in the literature on aging service use, its ability to predict outcomes consistently and explain service use or non-use by older adults and their families is limited (Mitchell & Krout, Reference Mitchell and Krout1998). Application of this model to more complex familial and health care situations requires more clearly defined concepts, more valid and reliable measures, and hypothesized outcomes that take into account family context when assessing preference and need for formal care services.
Methodological Enhancements for Understanding Diverse Family Structures
Investigating new aspects of caregiving within the pluralistic family types described previously implies the need for designs that include multiple family structures as well as families that vary by racial ethnic status, class, sexual orientation, and cultural heritage. Studies must also move beyond the individual level of analysis to focus on the family as the unit of analysis. Even though we assume that pluralistic families share some caregiving patterns with conventional ones, such as the likelihood that one family member serves as the primary caregiver, family caregiving research designs that incorporate these features will better reflect the lived experiences of older adults and their relatives than designs that focus on only one person’s responses. Thus, obtaining multiple views of a situation contributes depth and breadth of information about care needs and care provision in the context of diverse family forms.
When divergent perspectives emerge, researchers have opportunities to explore the reasons for and implications of differing perceptions, beliefs, and interpretations of the caregiving situation within families. Further, a family-level design affords chances to fill in missing pieces of the family story and confirm statements, perspectives, and findings across multiple members of the family. Extending the basic family-level data design to multiple occasions of measurement permits examination of similarities and differences over time of all these aspects of caring for older family members. A family-level approach is also important in qualitative research, where rigorous investigations that go beyond descriptive presentations of the family experiences are needed to better understand the dynamics of elder family care in the context of complex family constellations.
Of course, working with a family-level design presents challenges to researchers. Time and money costs are increased when multiple family members are involved in a study. Researchers must carefully evaluate the sample size and number of data collection points needed for the advanced statistical analyses that dependent, nested, and missing data require. They must be prepared to handle missing data due to some family members’ declining to respond to certain questions and missing cases due to inability to recruit all the family members into the study. In the case of longitudinal family-level designs, collecting data from multiple family members over time is more difficult than doing so with just one person, and sample retention strategies become more complicated. Recruiting a diverse sample of families may be particularly challenging if older members have poor health or declining cognitive function, and retaining the oldest members in the sample over time might be especially difficult due to morbidity and mortality.
As pluralistic family structures in the United States become commonplace and other countries around the globe are challenged to assess ways in which families care for older members, it will be important to conduct comparative research from both national and international perspectives. Moving to a cross-national and global level of family data requires not only launching more multi-nation studies, but also merging existing data sets from various cultures. New challenges will emerge related to establishing comparability of family-level concepts and measures and synthesizing results across studies (Hofer & Piccinin, Reference Hofer and Piccinin2010; Piccinin & Hofer, Reference Piccinin, Hofer, Hofer and Alwin2008). Further, the increasing availability of “big data” will represent a powerful opportunity to understand interactions among family members, their informal and formal networks, and the larger communities in which they live. Discovering associations and understanding patterns and trends within the massive amounts of widely available archived and real-time data on whole populations will advance understanding of how societal changes transform family structures and influence the ways in which they care for their older members.
