Intergenerational family structure

Family structure includes factors such as the number of generations, family composition (e.g. size, marital status), geographical proximity, and contact frequency. Many structural aspects of family ties have changed markedly as a result of changes in life expectancy, fertility rates, and divorce. Average life expectancy in the United States of America (USA) was about 49 years in 1900 and is projected to be approximately 83 years in 2050 (Miniño, Heron and Smith Reference Miniño, Heron and Smith2006). The implications of such a change for society and for individuals are profound. The number of people in the USA aged 65 or more years is expected to increase by 135 per cent from 1995 to 2050; with the number of people over 85 years of age expected to increase by more than 400 per cent (Gist and Hetzel Reference Gist and Hetzel2004; US Census Bureau 1997). Given the longer life expectancy of women, the projections differ greatly by gender. The differences in male/female longevity result in an older population that is increasingly female, with approximately 80 males to every 100 females among 65-year-olds and 44 males for every 100 females among 85-year-olds (Gist and Hetzel Reference Gist and Hetzel2004). Thus, intergenerational families are increasingly headed by older women. These figures vary in other parts of the world but are impressive, especially in the developed world (Knodel and Ofstedal Reference Knodel and Ofstedal2003).

In the USA, there are important race and ethnic differences as well, with the proportion of minority elders rapidly increasing. For example, the African American population aged 65 or more years will increase from 3 million (8% of the total) in 2004 to 10.3 million (12%) in 2050, while the number of Latino elders will increase from 2 million (6%) to 15 million (18%) (US Administration on Aging 2006). The Asian American as well as the Central and South American older populations in the USA are substantially smaller but also increasing disproportionately (Federal Interagency Forum on Aging-Related Statistics 2004). The growth in the racial and ethnic diversity of the older population creates a pressing need to consider how culture and racial/ethnic stratification influence intergenerational family experiences.

The increase in longevity has been accompanied by decreased fertility (Antonucci, Blieszner and Denmark Reference Antonucci, Blieszner, Denmark, Landrine and Russo2009). Consequently, the population structure is projected to change from a pyramid with many more younger than older people, to a ‘beanpole’ with similar numbers in each broad age group and many more generations in each family. The historical trend of increased longevity naturally manifests in individuals' lifespans and family lifecourses and structures. As an example, the duration of intergenerational ties is increasing as people live longer, and it is no longer rare for such ties to last four, five or even six decades (Hagestad Reference Hagestad2006; Taylor, Robila and Lee Reference Taylor, Robila and Lee2005). Consequently, multiple transitions are likely to take place within intergenerational ties. As transitions take place across different stages of the lifespan (e.g. from childhood to old age), individuals' intergenerational roles also change, as the child becomes a parent, and the middle-aged grandparent becomes the oldest-old great-grandparent. Additionally, lifecourse experiences as a student, worker or community member continue to change and shape these long-lasting relationships.

Complementing the population changes, new intergenerational family structures are also emerging. Multiple-generation families include members who have experienced divorce, single parenthood, remarriages and blended families, as well as never married co-habiting families, and married couples living apart (de Jong Gierveld Reference de Jong Gierveld2004). The fact is that a very small percentage of the population now live in ‘traditional’ households in first marriages with one wage earner, the father, with a mother not employed outside the home, and with two biologically-related children. A recent report by Williams, Sawyer and Wahlstrom (Reference Williams, Sawyer and Wahlstrom2005) highlighted these changes and indicated that whereas in the USA in 1970, approximately 40 per cent of families were nuclear families, that figure was less than 25 per cent by 2000 and continues to decline. By contrast, the percentage of married people with no children stayed at approximately 30 per cent while the number of single households increased from approximately 15 per cent in 1970 to over 25 per cent in 2000. These various family structures of course reflect the structure of social relations. People are close to parents and children, but also step-parents and step-children. Similarly, people have fewer full siblings but more step- and half-siblings. Multiple serial marriages result in ‘former’ step-siblings or synthetic family members from non-married blended households. Disrupted families with children and fewer resources often rely on grandparents for basic care rather than the traditional honorific role (Hayslip and Patrick Reference Hayslip and Patrick2005). Nonetheless, some grandparents seem to prefer a less hands-on role in their grandchildren's lives. In these cases, grandparents living far away from their children and grandchildren reported that they enjoyed their freedom and were pleased to avoid daily hassles and conflicts with intergenerational family members. They preferred their increased ability to negotiate the timing and duration of family visits (Banks Reference Banks2009).

Despite the fact that these varied structures might be seen as threats to fundamental social relations, the research evidence suggests that families remain committed to their older members (discussed further below). In addition, family members continue to have high levels of contact and to live relatively close to one another. As Shanas and Maddox (Reference Shanas, Maddox, Binstock and Shanas1975) reminded us over a quarter of a century ago, most older people live within a few miles of a son or daughter, and this continues 40 years later. According to the 2002 wave of the US Health and Retirement Study, 51 per cent of parents indicated that at least one adult child lived within ten miles (National Institute on Aging Reference Karp2007: 82). In sum, the structure of the multigenerational family is varied and can be complicated, leading to both predictable and unpredictable support exchanges. These advantage some, e.g. those who acquire grandchildren as a result of a remarriage, and disadvantage others, e.g. those who lose contact with grandchildren who move away with the custodial parent. Although family structure and context have changed a great deal, family members remain an important source of social support.

Support exchange by personal characteristics

Personal characteristics, such as gender, race, socio-economic status and age, influence the direction, quality and quantity of social support. Support exchange appears to occur more often among female family members. One American sample of older mothers aged between 66 and 78 years reported a preference for relying on their daughters rather than their sons for both instrumental and emotional support (Suitor and Pillemer Reference Suitor and Pillemer2006). Similarly, daughters are more likely than sons to provide support to older parents, particularly mothers (Moen, Robison and Fields Reference Moen, Robison and Fields1994). Using data from the US Longitudinal Study of Generations, Silverstein, Gans and Yang (Reference Silverstein, Gans and Yang2006) found that mothers in good health received more support than fathers in poor health. This may be less true in other cultures. In Japan, for example, daughters-in-law rather than daughters are expected to provide direct personal care for both parents-in-law, while first-born sons and their wives are expected to live with and care for his parents until they die (Akiyama, Antonucci and Campbell Reference Akiyama, Antonucci, Campbell and Sokolovsky1997).

In addition to gender, there are variations in social support exchanges by race and income. In the USA, instrumental support, e.g. child care, household chores, is more common in Black and Latino than White families, but socio-economic status is also a factor (Antonucci Reference Antonucci, Birren and Schaie2001; Swartz Reference Swartz2009). Among higher socio-economic status Black and Latino families, financial support is more common relative to instrumental and practical support. Financial support may likely come in the form of co-residence as members across generations live in one household which may facilitate the provision and exchange of support. Schoeni and Ross (Reference Schoeni and Ross2004) found that most parents provided financial support to their children to help pay bills, school tuition, loans and rent. Comparing parents in the upper income quartile with those in the lower income quartile, they found substantial differences, with 90 per cent of the higher-income and 61 per cent of the lower-income parents providing support to their children. There was also a considerable gap in the amount of financial aid received. Among those receiving aid, families in the upper quartile received on average US $17,907, while those in the lower quartile received approximately US $3,548. Racial differences were also evident with African Americans who received financial help from their parents reporting an average transfer of US $5,018, while non-African Americans reported receipt of US $11,765 on average. These findings demonstrate the fundamental, strong and long-lasting contribution of parents to their children but also document significant income and race differences in the amount of intergenerational financial support children receive (Schoeni and Ross Reference Schoeni and Ross2004).

Certain levels and types of social support exchanges appear more common at particular periods of the lifecourse. For example, adolescents may increasingly give and receive emotional support from peers, including siblings, relative to members of their non-peer family members as they seek to establish their independence. On the other hand, the support exchanges of older adults are increasingly family and intergenerationally based. Although there is a general trend for a family's older generations to provide financial support to the younger, the extent of this support varies by specific periods of the lifespan and depends on different lifecourse situational contexts (e.g. school, family, work) (Fuller-Iglesias, Smith and Antonucci Reference Antonucci, Blieszner, Denmark, Landrine and Russo2009; Levitt Reference Levitt2005; Levitt, Guacci and Weber Reference Levitt, Guacci and Weber1992). Using data from the US National Survey of Families and Households, Conney and Uhlenberg (Reference Conney and Uhlenberg1992) found that adult children most frequently reported receiving advice from their parents over the entire lifespan, but the receipt of gifts and money declined once children reached their thirties.

Support exchange by situational characteristics

Situational characteristics including need, health status and culture shape many aspects of social support exchanges. There is evidence to suggest that parents differentiate between their children, giving more to those who are in need but also, perhaps counter-intuitively, to those who are successful (Fingerman et al. Reference Fingerman, Miller, Birditt and Zarit2009). Younger children typically have greater needs and consequently receive more support. Considering intergenerational transfers, Fingerman et al. (Reference Fingerman, Pitzer, Chan, Birditt, Franks and Zarit2010) found that middle-aged adults with adult children and elderly parents in the USA are more likely to give to their children except when the parents experience health declines and limitations in meeting their daily needs. Adult children of divorced parents are just as likely to help their parents as adult children of widowed parents except that divorced fathers receive less help from their adult children (Lin Reference Lin2008).

Caring for an ill individual entails different experiences, norms, and expectations, depending upon situational characteristics. An older parent's need for social support may vary depending on the severity of illness (e.g. acute, chronic, terminal), the number of other family members available to provide support, other demands made on the adult child (e.g. young children to care for and paid work), and the health of the adult child as well as other members of their convoy. Situational factors such as geographical proximity influence contact frequency, which in turn may influence some types of social support exchanges (e.g. providing material goods) more than others (e.g. emotional support).

Support exchange also varies by culture. In the USA, greater support is generally given to younger generations, especially financial support, suggesting a downward flow of intergenerational support (Fingerman et al. Reference Fingerman, Pitzer, Chan, Birditt, Franks and Zarit2011). The same tendency has been found in Europe (Albertini, Kohli and Vogel Reference Albertini, Kohli and Vogel2007; Kohli and Albertini Reference Kohli and Albertini2009), but is less the case in Japan, where children are seen as forever indebted to their parents (Akiyama, Antonucci and Campbell Reference Akiyama, Antonucci, Campbell and Sokolovsky1997). In Taiwan, parents expect and do receive more financial support from their adult sons than daughters, as indicated by the proportionately higher rates of financial exchanges from sons to parents (Chen Reference Chen2006). In a recent study, Akiyama and Antonucci (Reference Akiyama and Antonucci2009) asked regionally-representative samples of adults in the USA and Japan the degree to which they felt that they provided more, received more, or had provided and received equal amounts of support from/with their mother, father and child. Examination of the responses indicated both age and country differences. Overall the Japanese were more likely to say they received more support from their mothers and fathers than Americans, although the majority of people aged 20–50 years in both countries said that they received more support from their mother and father than they provided. The percentage reporting that they provided more support to their mothers and fathers than they received increased with age, with over 20 per cent of those aged 65–74 years in both countries reporting that they provided more than they received. The picture differed for children. Parents in both countries reported that they provided more support to their child than they received.

The difference between the two countries was much smaller for support provided to children than for support provided by adult children to their parents. This pattern was fairly consistent across age groups in both countries, with almost 50 per cent of parents aged 75 or more years reporting that they provided more support to their children than they received. In a study of reciprocity in a regionally-representative sample of people over 65 years of age in south-west France, approximately 35 per cent of the respondents reported that they provided more support than they received (Antonucci Reference Antonucci2008). Of special interest is the fact that people reporting that they provided more support to others were more satisfied with their life six years later than those reporting reciprocal relations or receiving more support than they provided. This finding highlights the positive benefits of providing support to others, especially close family members.

Support exchange and wellbeing

Social support is widely associated with both negative and positive outcomes in terms of wellbeing. This may result from definitional and contextual issues surrounding social support as a multi-dimensional concept. In addition to the types of support, social support is highly contextualised and depends on roles, expectations and norms that may vary by culture, gender, age and time. For example, receiving support may induce psychological distress if it provokes negative feelings such as incompetence, although anticipated support may diminish this association (Liang, Krause and Bennett Reference Liang, Krause and Bennett2001). The anticipation of support may be contingent upon the history of the relationship, such as whether social support exchanges were common and positive in the past. Early relationship characteristics have been shown to influence adult wellbeing. Emotional support received from parents during childhood predicts less depressive symptoms and chronic symptoms in adulthood (Shaw et al. Reference Shaw, Krause, Chatters, Connell and Ingersoll-Dayton2004). Adult children who report spending a great deal of time with their parents during childhood provided more support to their parents later in life (Silverstein et al. Reference Silverstein, Conroy, Wang, Giarrusso and Bengtson2002). This is in keeping with the norm of reciprocity and highlights its implications for wellbeing.

In a cross-national study of Norway, England, Germany, Spain and Israel that examined reciprocity and life satisfaction in the parent–child relationship during late life, parents who provided less support to their adult children relative to the support received reported, on average, lower life satisfaction (Lowenstein, Katz and Gur-Yaish Reference Lowenstein, Katz and Gur-Yaish2007). In contrast, parents who provided more support to their children and received less support in return reported the highest levels of life satisfaction. From further analyses of these data, Katz (Reference Katz2009) found that the effects of intergenerational relations on wellbeing were influenced by multiple factors including nation state, affective solidarity and personal resources.

Support exchange: the context of care-giving

One of the most important types of intergenerational exchanges is care-giving. Provision of informal family care for an older relative is a significant form of intergenerational exchange. In the USA, family members provide 70 per cent of community-based long-term care for older adults (Wolff and Kasper Reference Wolff and Kasper2006). Family care-giving is protective for older adults (Pruchno, Michaels and Potashnik Reference Pruchno, Michaels and Potashnik1990), and current estimates suggest that 44.4 million Americans provide over 37 billion hours of care per year for adult family members and other loved ones with chronic illnesses and disabilities (Gibson and Hauser Reference Gibson and Hauser2008). Younger people report greater expectations that they will provide care than older people expect. The discrepancy has remained consistent over time in the USA. In 1980, Kahn and Antonucci (Reference Kahn and Antonucci1984) found that 90 per cent of US respondents aged 50 or more years agreed that social security benefits should be increased, and 75 per cent felt that older people who could not make it on their own should live with immediate family members rather than in nursing homes. Over 20 years later, people aged 50 or more continued to agree that the family should take the major responsibility for elders (Jackson et al. Reference Jackson, Antonucci, Brown, Daatland, Sellers, Booth, Coulter, Bianchi and Seltzer2008). Fifty-five per cent of this age group believed that families should be responsible for the financial needs of elders who cannot manage on their own; 91 per cent believed families should help with household chores; and 74 per cent believed that families should provide personal care when the elder cannot manage alone. Interestingly, younger people (i.e. adults in their thirties and forties) had the same basic opinions but were more likely to believe that younger people should help their elders. These findings are important because, contrary to the common myths that ‘the family is dead’ or that young people do not expect to care for their elders, it appears that family members do feel close to and wish to support their elders.

While cultural norms and expectations vary with respect to familial or filial obligation (Rossi and Rossi Reference Rossi and Rossi1990), caring for an older family member is often viewed in the context of lifetime exchanges and family solidarity (Bengtson et al. Reference Bengtson, Giarrusso, Mabry and Silverstein2002; Katz et al. Reference Katz, Lowenstein, Phillips, Daatland, Benston, Acock, Allen, Dillworth-Anderson and Klein2005). Fulfilling filial obligations is one of the salient rewards mentioned by family care-givers (Pinquart and Sorenson Reference Pinquart and Sorenson2003; Roff et al. Reference Roff, Burgio, Gitlin, Nichols, Chaplin and Hardin2004). Spouses often mention the fulfilment of marital vows (‘in sickness and in health’), while adult children talk about ‘wanting to give back’ to the parent(s) who cared for them earlier in life.

Care-giving varies by numerous personal characteristics. Both the care recipient's and the care-giver's characteristics shape the experience of intergenerational care exchanges. Among the most frequently documented are gender and race. The vast majority of care-giving continues to be done by women. Wives and adult daughters, and even daughters-in-law, provide significantly more care and more care hours compared to their male counterparts (Johnson and Weiner Reference Johnson and Weiner2006). In addition, men generally assume a more executive style of care-giving, employ formal assistance, and receive more informal support in their role compared to female care-givers (Kramer and Thompson Reference Kramer and Thompson2002).

Race and ethnicity shape intergenerational care-giving expectations and experiences. In the USA, non-White care-givers are less likely to be a spouse and more likely to be an adult child, other relative or friend compared to White care-givers (Connell and Gibson Reference Connell and Gibson1997; Pinquart and Sorensen Reference Pinquart and Sorenson2005). Both African American and Hispanic care-givers are more likely to be co-resident, spend more hours providing care and use less formal support services compared to White and Asian-American care-givers (Cox Reference Cox and Padgett1995). Over 75 per cent of Hispanic care recipients live with an adult daughter, compared to less than 15 per cent of Whites, who are more likely to be in institutional care. African American care-givers are more likely to provide care for both older and younger family members and to be employed outside the home compared to Whites (Connell and Gibson Reference Connell and Gibson1997). While high levels of poverty among older African Americans results in greater financial hardship compared to other care-givers, they were less likely to report care-giver stress and appeared to cope better by praying and/or consulting with spiritual counsellors in comparison to White, Hispanic or Asian-American care-givers (Roff et al. Reference Roff, Burgio, Gitlin, Nichols, Chaplin and Hardin2004).

Situational characteristics such as the family context influence the provision of care-giving support, while family relationship type and quality affect the intergenerational care-giving experience (Steadman, Tremont and Davis Reference Steadman, Tremont and Davis2007). Even within the social relations convoy, there is a natural hierarchy of people from whom an individual prefers to receive support and assistance (Kahn and Antonucci Reference Kahn, Antonucci, Baltes and Brim1980). Older adults prefer to receive support from a spouse, when available, and from adult children, primarily daughters, before turning to friends and neighbours (Cantor Reference Cantor1979). Among siblings, those who are more proximate are perceived to be more available, and those who have a history of more positive relationships are often expected to take on a greater care-giver role (Suitor and Pillemer Reference Suitor and Pillemer1994). Increasingly, adult children care-givers are ‘sandwiched’ between their multiple family and professional roles as they are called upon to provide care for family members across two or three generations (Szinovacz and Davey Reference Szinovacz and Davey2007).

Demographic shifts and higher rates of divorce across the lifespan have stimulated increased interest in how post-divorce and complex family patterns influence intergenerational support in later life (seeGanong and Coleman Reference Ganong, Coleman and Izuhara2009). Post-divorce families are vulnerable to increased tension (Bornat et al. Reference Bornat, Dimmock, Jones and Peace1999; Pasley and Ihinger-Tallman Reference Pasley and Ihinger-Tallman1990), and high levels of intergenerational ambivalence (Luescher and Pillemer Reference Luescher and Pillemer1998; Stewart Reference Stewart2005). Such family dynamics potentially threaten the likelihood and quality of intergenerational care provision. Indeed, a recent study found significant detrimental effects of parental divorce and step relationships on time transfers, probability of co-residence, and likelihood of monetary transfers (Pezzin, Pollack and Schone Reference Pezzin, Pollack and Schone2008). Greater attention is needed to identify and address expectations and experiences of care-giving among minority, gay/lesbian, single adults' and other families that differ in composition from dominant patterns and legal norms (Dilworth-Anderson, Burton and Johnson Reference Dilworth-Anderson, Burton, Johnson, Boss, Doherty, LaRossa, Schumm and Steinmetz1993; Lynch Reference Lynch2000).

While reconfigured stepfamilies and cohabitation have become more prevalent in the USA and elsewhere, care-giving research has not adequately taken these changes in older adults' marital and family histories into account. Elders in such families are vulnerable members of the intergenerational family convoy. One new study of remarried spouse care-givers makes this clear and documents some unique characteristics of dementia care-giving in the remarried and step-family contexts (Sherman Reference Sherman2009; Sherman and Boss Reference Sherman and Boss2007). Remarried care-givers generally reported little to no emotional or instrumental assistance from adult step-children, while a sizeable sub-group also reported considerable tension and conflict with step-children regarding financial and other care-giving-related decisions (Sherman Reference Sherman2009; Sherman and Boss Reference Sherman and Boss2007). The social and care-giving-specific support networks of remarried care-givers comprise biological or adoptive family members and friends, with minimal representation of step-family members. This pattern shifts dramatically for negative (e.g. intrusive or non-support) caregiving networks in which step-family members were nominated most frequently. Such findings of intergenerational non-support patterns among remarried older adults and step-family members raise genuine concern about the potential for additional vulnerability among care-givers in the step-family context.

Extensive research has documented the many adverse relational and health outcomes that can accrue from providing long-term care for an ageing adult who experiences declines in physical and cognitive functioning (Blieszner et al. Reference Blieszner, Roberto, Wilcox, Barham and Winston2007; Gaugler et al. Reference Gaugler, Davey, Pearlin and Zarit2000; Lu and Wykle Reference Lu and Wykle2007; Pinquart and Sorensen Reference Pinquart and Sorenson2003, Reference Pinquart and Sorenson2005; Vitaliano, Zhang and Scanlan Reference Vitaliano, Zhang and Scanlan2003; Zarit Reference Zarit, Qualls and Zarit2009). Women generally report higher levels of depression, anxiety, isolation and burden in their role as care-givers than men. African Americans report higher levels of positive aspects of care-giving, lower anxiety, and lower feelings of bother by the care recipient's behaviour compared to Whites regardless of socio-economic status. African Americans' higher religiosity has been shown to mediate the relationship between race and positive aspects of care (Roff et al. Reference Roff, Burgio, Gitlin, Nichols, Chaplin and Hardin2004). Spouse care-givers, who are older, co-resident and have medical issues of their own, report significantly more physical demands, financial strains and higher levels of relationship and social strain compared to other family carers (Pinquart and Sorensen Reference Pinquart and Sorenson2003). Coping with the challenge of assuming responsibility for their care recipient, they also have to cope with the emotional strain of change and loss in their intimate relationship. Child care-givers report higher relationship quality with the care recipient, greater reward, and less job strain compared to children-in-law, many of whom lack a history of reciprocal support with the care recipient or may feel pressured into the care-giver role by cultural expectations (Pinquart and Sorenson Reference Pinquart and Sorenson2005). Sustaining care-giver health and wellbeing is critical to the provision of long-term care of older adults in an ageing society. As increasing numbers of adults enter later life with complex marital and family histories, the challenge will be to ascertain both their unique vulnerabilities and the strengths of diverse family systems as intergenerational members are called upon to provide long-term care for multiple family members.

Quality of intergenerational relations by personal characteristics

Parent–child relationship qualities vary by personal characteristics including age, gender and race. Research has indicated that parents and children tend to become closer and more positive over time (Rossi and Rossi Reference Rossi and Rossi1990), and that parents and children report lower negative relationship quality over time (Birditt et al. Reference Birditt, Jackey and Antonucci2009). Birditt, Jackey and Antonucci (Reference Birditt, Jackey and Antonucci2009) found that ratings of negativity regarding children decreased over 12 years among young and middle-aged adults but not among older adults. Ambivalence tends to decrease over time, especially as children progress from adolescence to young adulthood (Tighe, Birditt and Antonucci Reference Antonucci, Blieszner, Denmark, Landrine and Russo2009). These findings are similar to those from cross-sectional research which show that older people more than younger people report that the parent–child relationship is closer and less negative (Akiyama et al. Reference Akiyama, Antonucci, Takahashi and Langfahl2003; Umberson Reference Umberson1992). Women are more likely than men to report feeling closer to their children, as well as to be both more positive and more negative about them (Collins and Russell Reference Collins and Russell1991; Fingerman Reference Fingerman2003; Rossi and Rossi Reference Rossi and Rossi1990). For instance, parents of daughters report greater tensions than do parents of sons (Birditt et al. Reference Birditt, Jackey and Antonucci2009). The findings regarding ambivalence are more mixed, with some studies reporting that women feel greater ambivalence (Willson, Shuey and Elder Reference Willson, Shuey and Elder2003) and others finding no gender differences (Fingerman et al. Reference Fingerman, Chen, Hay, Cichy and Lefkowitz2006).

In the USA, because African Americans tend to report greater reliance on their family members for support than other groups (Neighbors Reference Neighbors, Taylor, Jackson and Chatters1997), their parent–child relationships may be a source of support as well as strain and ambivalence (Chatters, Taylor and Neighbors Reference Chatters, Taylor and Neighbors1989; Umberson Reference Umberson1992). Research regarding race differences in the emotional qualities of the parent–child tie has been inconclusive, with some finding that African Americans report greater support and negativity and others finding no differences (Birditt, Rott and Fingerman Reference Birditt, Rott and Fingerman2009; Pillemer et al. Reference Pillemer, Suitor, Mock, Sabir, Pardo and Sechrist2007; Umberson Reference Umberson1992).

Quality of intergenerational relations by situational characteristics

Parents' and children's reports of relationship quality can vary by situational characteristics including social roles, problems, successes, and relationship history. Most parents have invested in having their children achieve roles associated with adulthood. They report more positive relations with young adult children who do not have their own children, who are employed, married, and not co-residing with parents (Belsky et al. Reference Belsky, Jaffee, Caspi, Moffitt and Silva2003). Fingerman et al. (Reference Fingerman, Chen, Hay, Cichy and Lefkowitz2006) found that parents reported greater ambivalence when children had fewer roles (e.g. marriage, employment). Parents report more ambivalence when their children are having problems or are less successful and vice versa. Suitor and Pillemer (Reference Suitor and Pillemer2000) found that older mothers felt closer to children who had involuntary problems (e.g. health problems), whereas they felt least close to children they perceived as having voluntary problems (e.g. trouble with the law). Birditt, Fingerman and Zarit (Reference Birditt, Fingerman and Zarit2010) found that middle-aged men reported greater feelings of ambivalence regarding adult children who had physical and emotional problems and less career success. Middle-aged men and women reported greater ambivalence about children with less relationship success. Children tend to experience greater ambivalence when their parents have poorer health or need care (Peters, Hooker and Zvonkovic Reference Peters, Hooker and Zvonkovic2006; Willson, Shuey and Elder Reference Willson, Shuey and Elder2003). Relationship history also influences the parent–child tie in adulthood. Early parental rejection predicts lower-quality relationships with parents in adulthood (Whitbeck, Hoyt and Huck Reference Whitbeck, Hoyt and Huck1994; Whitbeck, Simons and Conger Reference Whitbeck, Simons and Conger1991). Adult children experience greater ambivalence when they report having had low-quality parent relationships as children (Willson, Shuey and Elder Reference Willson, Shuey and Elder2003).

Quality of intergenerational emotional relations and wellbeing

The emotional qualities of the parent–child tie have important influences on the health and wellbeing of both parents and children. As one might expect, positive aspects of the relationship are associated with better wellbeing among parents and children (Bengtson et al. Reference Bengtson, Giarrusso, Mabry and Silverstein2002; Lowenstein Reference Lowenstein2007). In addition, parents report better wellbeing when their children are more successful (Ryff et al. Reference Ryff, Lee, Essex and Schmutte1994). Silverstein and Bengtson (Reference Silverstein and Bengston1991) examined in a US sample the association between parent's feelings of affective solidarity and parental mortality. While they found no direct effect of affective solidarity on mortality, they did find a buffering effect. Parents who were recently widowed had lower mortality rates if they had greater affective solidarity with offspring. Umberson (Reference Umberson1992) found that greater strain with mothers and fathers associated with greater psychological distress among adult children. Greater ambivalence associates with poorer health and wellbeing among both parents and children. Fingerman et al. (Reference Fingerman, Pitzer, Lefkowitz, Birditt and Mroczek2008) found that adult children reported lower self-rated health when their fathers were more ambivalent about them, whereas mothers reported poorer self-rated health when their children were more ambivalent about them. Ward (Reference Ward2008) found that parents who reported lower collective ambivalence (i.e. lower-quality relationship or less contact with at least some children) reported less depression. Lowenstein (Reference Lowenstein2007) examined the links between how older adults felt about their children (ambivalence, solidarity and conflict) and their quality of life in samples of mothers and fathers aged 75 or more years in England, Norway, Germany, Spain and Israel. She found that ambivalence predicted lower quality of life whereas solidarity predicted greater quality of life in all countries, and that solidarity had a greater impact on quality of life than ambivalence.

Overall, these studies have shown that parents and their children feel both intense love and irritation for one another. Relationship quality includes positive, negative and ambivalent feelings, which vary widely within and between families by personal and situational factors. Parents and children who report greater feelings of ambivalence and lower feelings of positive quality tend to report lower wellbeing. Future studies need to examine these associations over time, however, to understand how these factors operate longitudinally in intergenerational families. To date, it is unclear whether individuals with poorer wellbeing elicit more ambivalence and negativity in their relationships or the reverse. Of course, bi-directionality of influence is very likely. Future researchers should consider using daily diary and longitudinal methods to document this complex relationship and how dyads and family members influence one another over time.