RESEARCH DOSSIER
Later-Life Social Roles and Intergenerational Relations
Abstract. Later-life social roles and intergenerational relations are organized through changing family positions, institutions, obligations, social networks, and reciprocal practices. This fundamental research monograph examines how retirement, grandparenthood, adult parent–child relations, caregiving, widowhood, friendship, volunteering, and civic participation shape psychological experience across later life. It traces the development of disengagement and continuity theories, the convoy model, socioemotional selectivity theory, and multidimensional models of intergenerational solidarity and ambivalence. Original studies and longitudinal family panels are considered alongside systematic reviews of grandparenting, caregiving, loneliness, and interventions to reduce ageism. The evidence synthesis distinguishes network structure from relational quality, symbolic obligations from actual transfers, and voluntary participation from intensive or imposed care. Methodological discussion addresses dyadic reporting, event histories, cohort heterogeneity, cultural and socioeconomic inequality, selection, attrition, and causal identification. The monograph finds that psychological outcomes depend on the conditions and meanings of roles, including access to support, perceived autonomy, reciprocal history, and institutional constraints. It concludes with an evidence-based research agenda on multigenerational panels, inequalities in care, digital mediation, and human–artificial coordination, maintaining a clear distinction between established observations and prospective hypotheses.
Keywords: later-life social roles; intergenerational relations; social gerontology; retirement; grandparenthood; intergenerational solidarity; intergenerational ambivalence; family caregiving; reciprocity; socioemotional selectivity; social participation; life-course psychology; Artificial Era
Historical Theories of Aging and Participation
Modern social gerontology emerged partly through disagreement about the meaning of participation in later life. In Growing Old: The Process of Disengagement, Cumming and Henry (1961) proposed that mutual withdrawal between aging individuals and society constituted a patterned transition. The theory placed later-life role reduction at the center of explanation and was historically consequential because it transformed everyday assumptions about aging into claims open to systematic criticism. Its generalized developmental interpretation subsequently faced evidence of extensive participation, individual differences, and continuing desire for meaningful relationships. Historical significance and contemporary empirical support therefore require separate judgments. The persistence of retirement, widowhood, and changing physical capacity establishes real role transitions; it does not establish that reduced participation is uniformly desired or adaptive. Cumming & Henry (1961) Antonucci et al. (2014)
Continuity theory redirected attention to people maintaining internal preferences and external social patterns through adaptive changes. Atchley (1989) argued that middle-aged and older individuals often use past experiences and identities to manage new circumstances. For example, a person whose professional identity involved teaching might continue mentoring after retirement, retaining a recognizable form of contribution through a different institutional setting. Continuity can occur in goals and relationship functions even when the names of roles change. It should nevertheless be investigated as a tendency within varying constraints, rather than as a universal prescription to preserve earlier routines. Loss of transportation, caregiving obligations, unequal income, and inaccessible community institutions can prevent preferred forms of continuity. Atchley (1989)
The convoy tradition and socioemotional selectivity theory added more differentiated accounts of later-life social patterns. Convoy research organized ties by closeness, function, and changing circumstances; socioemotional selectivity proposed that perceived future time changes social priorities. The traditions illuminate different causal questions. Network composition asks who remains available and what support circulates. Motivational selection asks why particular interactions become preferable. Institutional research asks which relationships and activities a person can actually enter. Empirical work can combine these perspectives without reducing family solidarity to network size or emotionally meaningful contact to mere frequency. The enduring methodological achievement of social gerontology is the replacement of one age-based narrative with testable models of selection, support, opportunity, and longitudinal change. Antonucci & Akiyama (1987) Carstensen et al. (1999) Antonucci et al. (2014)
Retirement and the Reconstruction of Social Position
Retirement changes the arrangement of work obligations, income, time, contacts, and recognized social identity. Its psychological meaning depends on whether work was valued as mastery, economic necessity, status, sociability, or imposed routine, and on whether retirement is anticipated, voluntary, or associated with illness or labor-market exclusion. The loss of a paid occupational role may coincide with entry into care, volunteer, educational, and family roles. Conversely, available leisure may be absorbed by unpaid work. A scientifically useful analysis therefore treats retirement as a sequence of transitions, decisions, and institutional changes, with pretransition characteristics measured explicitly. Simply comparing retirees with workers confounds retirement with selection by age, wealth, health, occupation, and family responsibilities. Atchley (1989) Antonucci et al. (2014)
Role continuity is particularly informative when the observable activity changes but the underlying practice remains recognizable. An experienced worker may provide informal technical assistance to relatives, teach in community organizations, or take part in civic associations. The resulting contribution can sustain competence and social recognition while offering greater discretion over schedules. Other retirees may purposefully reduce demanding commitments and emphasize selected intimate relationships. The latter pattern is compatible with socioemotional selectivity, which links changes in social goals to perceived remaining time. These adaptations can coexist within one person: a smaller network of coworkers, more intense family communication, and increased neighborhood activity may all follow the same retirement event. Carstensen (1992) Carstensen et al. (1999)
Research designs should measure retirement timing relative to health changes, spousal employment, adult children's need for help, and pension eligibility. Panel event studies can follow the same person across multiple pre- and postretirement waves, while qualitative interviews reveal how an occupational identity is reinterpreted. Different outcomes must remain distinct: life satisfaction, depressive symptoms, purpose, loneliness, cognitive engagement, and perceived autonomy are not interchangeable indicators. Associations between volunteering and better well-being may reflect participants' prior health and resources. Likewise, new caregiving after retirement may reflect family need rather than preference. These methodological considerations make the retirement transition central to later-life role analysis without licensing a single conclusion about its benefits. Akhter-Khan et al. (2023) Antonucci et al. (2014)
Continuity, Agency, and Selective Social Investment
Agency in later-life roles refers to the capacity to select, negotiate, modify, or leave social commitments within available opportunities. Agency is expressed through decisions about contact, caregiving, participation, housing, and the distribution of time. Preferences are themselves shaped by prior relationships and social circumstances. A reduction in contact can be an agentic choice, a response to bereavement, a consequence of inaccessible environments, or the cumulative result of others' decisions. Distinguishing these pathways requires information about desired contact, perceived control, actual opportunities, and relationship quality. Atchley's continuity account locates agency in the maintenance of familiar patterns, while the convoy model locates it within social networks that also respond to external events. Atchley (1989) Antonucci et al. (2014)
Socioemotional selectivity theory makes a specific claim: perceived future time shifts the relative value of knowledge-seeking and emotionally meaningful social goals. Carstensen's early empirical work documented selective changes in interaction and emotional closeness, while the later theoretical formulation clarified the role of future-time perspective rather than chronological age alone. The prediction is that situationally limited time horizons can promote selectivity in people of different ages. Later-life roles thus have motivational content: a grandparent may prefer unhurried contact with a few grandchildren to many formal engagements, and an older worker may prioritize mentoring over status-seeking. These examples illustrate theory-derived possibilities; the scientific claim rests on measured goals and interactions, not on idealized descriptions. Carstensen (1992) Carstensen et al. (1999) Carstensen et al. (2003)
The two theories generate complementary but distinguishable tests. Continuity predicts correspondence between earlier identity, preferences, and later role choices, controlling for objective opportunity. Socioemotional selectivity predicts variation in choices with perceived time horizon, even when chronological age is held constant. Network constraints predict contact from mobility, geographic distance, and survival of ties. A longitudinal study that measures all three can ask whether the same person preserves familiar relational functions while changing the size of the network and whether time-perspective change precedes a shift in priorities. This specification prevents a common inferential error: interpreting every contraction of a network as either involuntary loss or deliberate emotional optimization. Antonucci & Akiyama (1987) Atchley (1989) Carstensen et al. (1999)
Life-Course Timing and Linked Generations
Later-life roles develop within connected family biographies. Adult children's education, partnership formation, employment instability, migration, and parenthood affect older parents' opportunities for support and care. At the same time, older adults' health transitions, retirement, and decisions to transfer resources can alter younger generations' prospects. The relevant time scales are multiple: a daily childcare exchange, a year of caregiving, a decade of financial assistance, and transfers accumulated across the family life course. Treating transfers solely as contemporaneous transactions can conceal historical reciprocity and motives that extend across decades. Evidence from British longitudinal data shows that parents in their later working and early retirement years often provide substantial help to adult children, challenging a one-directional dependency model. Grundy (2005) Fingerman et al. (2011a)
Life-course timing gives the same role different consequences. Grandchild care offered occasionally after retirement may provide meaningful contact, whereas full-time custodial care following a family crisis may impose abrupt demands. Support given to an adult child who is entering employment differs from help provided after persistent economic precarity; adult children assisting a healthy parent differ from those responding to disability. An investigation must establish who initiated support, the precipitating circumstances, prior exchanges, and the availability of alternatives. Fingerman and colleagues' interviews with middle-aged adults demonstrated that support to grown offspring and aging parents varies with everyday needs, perceived importance, and parental disability, highlighting the allocation problem facing the middle generation. Fingerman et al. (2011a) Minkler et al. (1997)
Linked lives also make independence relational. An older person's autonomy can depend on assistance that preserves the ability to choose residence, manage daily life, and maintain valued activities. Conversely, accepting assistance can expose disagreement over what constitutes safe or appropriate conduct. Relationship history affects whether help is experienced as solidarity, surveillance, reciprocity, or intrusion. Longitudinal family data increasingly allow researchers to distinguish expectations from realized care. A prospective Amsterdam study found that the number of adult children alone was an inadequate indicator of care received: relationship quality and existing support configurations helped identify whether potential care became actual care. A relational conception of autonomy therefore requires observable opportunities for negotiated assistance. Shen et al. (2025) Lüscher & Pillemer (1998)
Family Structure and Opportunities for Intergenerational Contact
The structural availability of intergenerational relationships depends on family size, survival of relatives, geographic proximity, migration, partnership status, transportation, institutional housing, and communication infrastructure. Structural opportunity is an important dimension of solidarity because contact requires feasible channels, but it cannot stand in for affection, support quality, or normative agreement. The intergenerational solidarity program distinguishes proximity and opportunity from interaction frequency, emotional closeness, beliefs, and functional exchange. A parent may live near adult children and have little intimate contact; another may experience frequent supportive conversation across large distances. Research needs measures of the location of specific relatives and of the extent to which available contact is desired and usable. Bengtson & Roberts (1991) Roberts & Bengtson (1990)
Variation in family structure is both demographic and historical. Families with fewer children, blended families, descendants living across borders, older adults without children, and relationships mediated by step-kin or chosen family offer different possible support constellations. Larger kin networks increase the number of potential helpers without guaranteeing commitment or capacity. Evidence from the Longitudinal Aging Study Amsterdam makes this distinction concrete: the translation of care potential into care receipt varied by partnership status and the support characteristics of adult children. The useful predictor combined opportunities and relationship history. This finding has implications for measurement and services: identifying a living relative is not equivalent to demonstrating that needed help is available at the appropriate time. Shen et al. (2025)
Structural analysis also recognizes the work performed by nonkin and institutions. Friends may provide sustained emotional support, neighbors may assist with transport or monitoring, and health and social-care services may redistribute time-intensive obligations from family members. Institutional access can change whether a family relationship is primarily companionship or crisis management. The convoy model therefore includes the composition and functions of networks, rather than equating a person's social world with a household roster. Studies of later-life contact should collect ties beyond relatives, assess who gives and receives assistance, and separate informal relationships from publicly or privately funded care. The resulting account is capable of identifying isolation risk without prescribing one family form as universally adequate. Antonucci & Akiyama (1987) Antonucci et al. (2014) National Academies (2020)
Multidimensional Intergenerational Solidarity
Intergenerational solidarity describes the patterned connections that link adult family generations through emotional, practical, structural, and normative relations. In the framework developed by Bengtson and Roberts, solidarity comprises multiple dimensions: affection, association, exchange of support, agreement over values, opportunities for interaction, and expectations about obligations. The dimensions provide distinguishable research variables. A parent and adult child may interact weekly because of a care arrangement while reporting limited emotional closeness; another pair may exchange little instrumental help but maintain considerable affection and confidence in future assistance. Solidarity is therefore best understood as a multidimensional configuration, and its components should be measured rather than inferred from a single global rating. The formal analysis of Bengtson and Roberts (1991) and the dimensional test by Roberts and Bengtson (1990) offer an explicit starting point for empirical measurement. Bengtson & Roberts (1991) Roberts & Bengtson (1990)
Multidimensionality matters when studying the later-life transition from giving to receiving assistance. A family relationship can display strong normative obligations and frequent support alongside enduring disagreement about independence. The structure of social opportunities, such as living nearby and having sufficient time to visit, may produce more interaction without changing the underlying emotional orientation. Different dimensions also have different temporal dynamics: affection can remain stable over decades, functional support can change rapidly after an illness, and normative beliefs can shift as younger relatives establish their own households. Cross-sectional data conceal these trajectories if all items are summed into one score. The scientific problem is to identify configurations of dimensions and examine whether they predict actual exchange and psychological outcomes beyond demographic characteristics. Silverstein & Bengtson (1997) Bengtson & Roberts (1991)
Empirical typologies of adult child–parent ties illustrate this agenda. Silverstein and Bengtson (1997) studied the coexistence of different dimensions in American families, emphasizing that adult-child relationships can take qualitatively different forms. Typologies are useful descriptions when the empirical classes are adequately supported, yet they should not be treated as fixed psychological kinds: a distant relationship may become highly involved after a bereavement or caregiving crisis. Measurement should record specific dyads, the direction of transfers, relationship quality, proximity, and contact frequency at each wave. When more than one adult child is present, dyad-level analysis avoids assigning one child's supportive behavior to the family as a whole. The overarching implication is that the quantity of intergenerational contact cannot substitute for the quality, content, or perceived voluntariness of that contact. Silverstein & Bengtson (1997) Roberts & Bengtson (1990) Shen et al. (2025)
Ambivalence, Conflict, and Relational Boundaries
Intergenerational ambivalence concerns the coexistence of positive and negative evaluations, incompatible obligations, or structurally contradictory expectations within family relationships. Lüscher and Pillemer (1998) advanced ambivalence as a framework for examining how affection and resentment can arise together when aging parents and adult children negotiate dependence, authority, property, and care. A relative may feel gratitude for practical help while resisting supervision; an adult child may offer support willingly and experience frustration about the resulting time burden. Ambivalence should be distinguished from a relationship that is merely uniformly conflictual, because simultaneous attachment and tension have particular implications for emotion regulation and boundary negotiation. The concept also discourages coding the family relationship as either supportive or harmful solely on the basis of its average tone. Lüscher & Pillemer (1998)
Relational boundaries are the negotiated limits of access to time, information, financial resources, physical space, and decisions. Such boundaries become salient when an older adult's household routines, medical appointments, or finances involve additional family members. A request for practical assistance can be interpreted as reasonable reliance by one participant and intrusive oversight by another. Later-life parenthood also preserves asymmetries of history: parents may continue to perceive a mentoring obligation while grown children seek recognition as autonomous adults. These interpretations are not directly visible in contact-frequency measures. Interview studies, within-family comparisons of perceptions, and diary reports can locate the situations in which obligations become disputed. The convoy framework predicts that close relationships may exert especially powerful effects because they combine emotional significance with expectations about availability. Antonucci et al. (2014) Fingerman et al. (2011a)
Ambivalence is especially important when examining public policy and professional services that presume familial care. Family members may have divergent ideas about who should provide assistance, which tasks can be delegated, and when institutional support is warranted. A strong norm of family obligation may coexist with economic or geographic constraints that make performance difficult. Researchers should distinguish normative expectation, actual caregiving practice, and experienced relationship strain. Because reports often differ across generations, a single informant can misrepresent the relationship. Dyadic designs can compare an older parent's perception of respect or intrusiveness with the adult child's perception of burden and responsibility. The theoretical claim is thus amenable to falsification: if positive and negative evaluations supply information beyond separate overall satisfaction measures, ambivalence has incremental explanatory value for conflict, distress, and future care decisions. Lüscher & Pillemer (1998) Bengtson & Roberts (1991)
The Distribution of Intergenerational Support
Intergenerational support involves transfers of money, goods, housing, practical services, emotional assistance, advice, and time. These resources differ in scarcity, reversibility, and psychological meaning. A pensioner who provides childcare may be simultaneously receiving transportation assistance from the same adult child. A distant relative may offer regular emotional support without the capacity to provide daily personal care. Evidence on support to both aging parents and grown children shows that the middle generation often distributes resources in more than one direction. Fingerman and colleagues (2011) investigated which family members receive help and why, highlighting perceived needs and relationship considerations as determinants of support. This evidence replaces a simplified model of resource flow from the young to the old with a dynamic account of multidirectional exchanges. Fingerman et al. (2011a) Fingerman et al. (2011b)
The distribution of support reflects available resources and competing obligations. Older parents with relatively secure finances may transfer money to adult children despite having emerging health needs. Adult children facing insecure employment may require assistance while also providing household labor or accompaniment. Transfers therefore depend on both ability and need, and a realized exchange is conditioned on family structure and past interactions. Contemporary prospective research explicitly distinguishes the potential supply of care from the care actually delivered. Shen and colleagues (2025) followed families over time and found that the relationship between adult children and care receipt depended on features of the support network rather than the mere existence of offspring. When interpreting such results, a family-size association is not equivalent to an experimentally demonstrated effect of adding a relative. Shen et al. (2025) Grundy (2005)
Research must separate the decision to seek help from the decision to give it and the practical delivery of assistance. Health status, privacy concerns, cultural expectations, costs, and the availability of community services can change each stage. Transfers are also measured differently across surveys: some ask about money above a threshold, others count hours of care, and others record perceived availability. These outcomes cannot be pooled without defining comparable units and reference periods. Longitudinal data can assess whether a financial transfer is followed by later reciprocal care, but estimates must account for selection into giving, unequal opportunities, and survival. A comprehensive account therefore identifies giver, recipient, type, frequency, intensity, reciprocity, and subjective interpretation for each exchange. Bui et al. (2022) Reyes & Patterson (2025) Fingerman et al. (2011a)
Reciprocity Across Decades
Reciprocity across generations operates on several time scales. Direct reciprocity refers to exchange between the same people; generalized reciprocity can involve support given to one generation in expectation that obligations circulate through family relationships over time. An older parent who helped pay for a child's education may later receive assistance with household tasks. Yet observed sequence does not establish a transactional motive: affection, social norms, economic need, and changes in health may also explain the exchanges. Grundy's (2005) analysis of British families connected socioeconomic conditions and health to patterns of giving and receiving, locating reciprocity within unequal family capacities rather than treating it as a universal exchange rule. Grundy (2005)
Recent longitudinal work investigates the predictive value of historical transfers for subsequent support. Bui, Kim, and Fingerman (2022) examined links between present support and later expectations or receipt of care, while Reyes and Patterson (2025) used life-course exchange data to investigate whether transfers relate to subsequent assistance from adult children. Such studies help establish temporal order and compare directions of transfer, but they remain subject to unobserved family norms and long-standing relationship quality. Parents who are more committed to children may both provide more resources early and receive more assistance later; this joint pattern can arise without a quid-pro-quo agreement. Researchers can improve inference through repeated measurement of preexisting closeness, financial capacity, the timing of unexpected needs, and the set of potential caregivers. Bui et al. (2022) Reyes & Patterson (2025)
Reciprocity also has an emotional dimension. Assistance may communicate continuity, loyalty, and acknowledgement, but a highly unequal exchange may generate guilt, resentment, or pressure to reciprocate beyond one's means. The appropriate empirical outcome is therefore not simply a balance sheet. A transfer of money may be psychologically trivial to a wealthy giver yet decisive for the recipient; several hours of intimate care may demand sustained emotional labor. Studies should record how participants understand the exchange and whether the support fits their stated preferences. This perspective offers a coherent bridge between the solidarity model's functional and normative dimensions and the ambivalence model's focus on conflicting expectations. It also helps distinguish reciprocity as a descriptive historical pattern from reciprocity as an obligation invoked to justify unequal burdens. Bengtson & Roberts (1991) Lüscher & Pillemer (1998) Reyes & Patterson (2025)
Grandparenthood and Its Diverse Forms
Grandparenthood is a family position that may involve emotional closeness, cultural transmission, care work, economic support, and the mediation of relationships between adult children and grandchildren. Its practical expression varies with geographic proximity, grandchildren's ages, the older person's health and employment, custody arrangements, and family history. Some grandparents provide occasional companionship, some perform regular after-school care, some live with grandchildren, and others take full parental responsibility. These arrangements demand separate analytic categories. The systematic review by Danielsbacka, Křenková, and Tanskanen (2022) synthesizes evidence linking grandparenting with health and well-being, while indicating substantial heterogeneity in associations. Researchers should avoid converting an observed benefit in one type of involvement into a prescription for all grandparents. Danielsbacka et al. (2022)
Grandparental participation can sustain valued identities and provide reciprocal learning, particularly when the amount of involvement is compatible with the older person's preferences and available resources. Time with grandchildren can create opportunities for affectionate contact, storytelling, everyday instruction, and continued engagement with family events. At the same time, repeated childcare may constrain paid work, rest, travel, and relationships outside the household. The older adult's role is partly negotiated with the adult parents of the child, who may control access, schedule, and rules for caregiving. Differences in expectations can generate ambivalence even when all participants value the relationship. Measuring hours of care without voluntariness and quality therefore provides an incomplete explanation of psychological outcomes. Danielsbacka et al. (2022) Lüscher & Pillemer (1998)
Intergenerational research also asks whether contact creates benefits for grandchildren or alters family resilience. Plausible mechanisms include stable emotional availability, continuity of family narratives, temporary relief for parents, and support during disruption. Demonstrating those mechanisms requires suitable comparison groups and temporally ordered data. Families with engaged grandparents may differ in household resources, parental employment, and the preexisting quality of relationships. Grandparent involvement may increase precisely when a family experiences adversity, which can produce misleading negative associations if reasons for caregiving are ignored. The longitudinal design should distinguish voluntary recreational contact, routine caregiving, and emergency or custodial arrangements before estimating effects on adults' health or young people's development. Danielsbacka et al. (2022) Minkler et al. (1997) Antonucci et al. (2014)
Custodial Grandparenting and Care Burden
Custodial grandparenting occurs when grandparents assume primary responsibility for a grandchild's upbringing, sometimes following parental illness, disability, substance-related difficulties, death, incarceration, or other family disruption. Its time requirements, authority, and economic implications differ substantially from nonresidential grandparenting. Older caregivers may gain a sense of continuity and commitment while encountering school coordination, legal uncertainty, financial strain, reduced leisure, and changes to employment or retirement plans. The national longitudinal analysis by Minkler and colleagues (1997) investigated depression among grandparents raising grandchildren, establishing the importance of distinguishing this caregiving arrangement when studying psychological health. Its observational design also requires attention to adversity that may have preceded the caregiving role. Minkler et al. (1997)
The later systematic review of grandparenting and well-being found that custodial arrangements and less intensive grandparental involvement cannot be grouped as though they were identical exposures. Danielsbacka and colleagues (2022) reviewed 92 articles reporting 117 studies and described heterogeneity associated with the type of grandparental involvement and research design. An association between custodial care and poorer well-being could involve caregiving demands, socioeconomic disadvantage, or the circumstances that necessitated custody. Conversely, older adults able to provide occasional help may have better baseline health and financial security, creating a positive selection bias. The research question is therefore about mechanisms and role conditions, not a single global effect of grandparenthood. Danielsbacka et al. (2022) Minkler et al. (1997)
Support systems can influence the consequences of custodial care. Access to health services, legal recognition, childcare assistance, flexible work, and emotional support from other adults may affect the burden of responsibilities. Family-centered research can examine whether these resources reduce strain without assuming that grandparents should provide unpaid care regardless of cost. Useful measures include hours of care, responsibility for educational and medical decisions, household financial transfers, perceived control, physical health, social isolation, and the quality of the grandparent–grandchild relationship. A mixed-method design can connect such measures to participants' descriptions of obligation and meaning. The policy consequence is to recognize custodial grandparents as caregivers with specific needs, while the scientific consequence is to model care intensity and precipitating adversity as key determinants. National Academies (2020) Danielsbacka et al. (2022) Minkler et al. (1997)
Adult Children, Parent Identity, and Negotiated Dependence
Parenthood continues after children reach adulthood, although the practices associated with the role evolve. Older parents may remain sources of advice, emotional support, financial help, and family history while adult children establish independent households. With advancing age, additional help can flow toward the parents, sometimes transforming long-standing expectations of who leads decisions. Researchers studying these shifts should distinguish the parent role as an enduring identity from the distribution of particular care tasks. Adults can maintain a parental identity while receiving daily assistance, and grown children can provide substantial assistance without exercising authority over every decision. Fingerman and colleagues' research on support allocations documents how competing needs within multigenerational families shape actual help rather than assuming a fixed direction of dependence. Fingerman et al. (2011a) Fingerman et al. (2011b)
The concept of negotiated dependence makes the quality of decision making central to later-life relationships. Assistance can expand a person's practical autonomy by making independent living possible; it can also narrow perceived control when the helper sets conditions inconsistent with the recipient's preferences. To investigate this distinction, studies need reports from both parties concerning decisions, perceived respect, and whether alternatives were available. Support received is not a direct measure of dependence, because highly autonomous people may coordinate extensive help effectively. Similarly, refusal of help may indicate confidence, privacy preferences, fear of conflict, or lack of trust. Ambivalence theory clarifies why positive intentions can coexist with perceived intrusion and why outcomes should include dignity and relationship tension alongside functional needs. Lüscher & Pillemer (1998) Bui et al. (2022)
Family composition creates further variation: one adult child may provide money, another transportation, and a nonresident relative periodic emotional support. Unequal distributions may generate conflict among siblings as well as between generations. Prospective evidence from the Netherlands suggests that available offspring translate into actual support through relationship-specific processes and circumstances. In a family panel, recording only whether an older parent has any adult child obscures those differences. Analyses should model child-specific proximity, resources, relationship quality, past help, and care involvement, while accounting for the shared older parent. Such models allow testing whether reciprocal history predicts later care beyond current need and prevent family-level generalizations from a single dyad's behavior. Shen et al. (2025) Reyes & Patterson (2025)
Partnership, Widowhood, and Relational Reorganization
Partnership in later life can organize everyday routines, resource sharing, emotional intimacy, social participation, and informal caregiving. A partner's retirement, illness, or death can change several of these arrangements simultaneously. Widowhood is therefore a major role transition: it changes not only marital status but also the distribution of household labor, companionship, practical support, and relationships with extended kin. Some widowed people sustain or expand their friendships, community participation, and family contact; others experience substantial disruption. Longitudinal research is essential because people who become widowed differ in age, prior health, marital experience, and social resources from those who remain partnered. Evaluating change within individuals before and after widowhood reduces, though does not eliminate, these selection problems. Isherwood et al. (2012) Antonucci et al. (2014)
Isherwood, King, and Luszcz (2012) examined social engagement over time in late-life widowhood, an approach suited to the question of how involvement changes with bereavement rather than merely comparing widowed and married groups at one survey wave. The meaning of network change depends on the role previously played by the spouse and the resources now available. A household formerly centered on a partner may develop more direct connections to adult children or neighbors, while a widowed person who cared intensively for a spouse may regain time for other interests. Neither scenario establishes a universal trajectory. A robust design measures caregiving before the loss, the timing of death, social engagement, financial circumstances, and grief-related distress over multiple intervals. Isherwood et al. (2012)
Later-life partnership changes also interact with family boundaries. Adult children may assist a surviving parent with financial administration, transportation, or housing decisions, but the parent may place high value on retaining control over daily life. New intimate relationships can change relations with descendants and step-relatives, especially when questions about inheritance and care expectations arise. Social research should avoid treating kinship as a stable roster after widowhood; family boundaries and the practical importance of particular ties may be renegotiated. The convoy model captures these changes as alterations in social-network composition and function. Combining that perspective with ambivalence theory helps explain how renewed contact can be simultaneously supportive and conflict-producing, and why relationship quality matters more than simple counts of participants in a network. Antonucci & Akiyama (1987) Antonucci et al. (2014) Lüscher & Pillemer (1998)
Friendship Convoys and Nonkin Bonds
Friends and nonkin relationships are integral to later-life social organization. They can provide companionship, emotional disclosure, advice, practical assistance, and social recognition independent of parental and grandparental status. The convoy model characterizes a person as surrounded by relationships varying in closeness and function, with members entering and leaving as circumstances change. This framework allows comparison of family and nonfamily ties without assuming they are substitutes in every domain. A friend may be the most important confidant while an adult child is the primary emergency contact. The appropriate unit is the relationship and its function, not the legal category of the person providing it. Network quality, availability, and subjective satisfaction should be measured separately from the number of connections. Antonucci & Akiyama (1987) Antonucci et al. (2014)
Socioemotional selectivity theory predicts that emotionally meaningful bonds can become prioritized as perceived remaining time changes, a process that may preserve a smaller set of close friends while reducing peripheral contacts. However, selectivity is bounded by survival, geographic accessibility, mobility, and the voluntary nature of relationships. Age-related contraction in the number of friends can result from bereavement or reduced contact opportunities as well as changed preferences. Differentiating mechanisms requires repeated measures of desired interaction, existing network members, environmental barriers, and perceived time horizon. The same older person can report fewer casual acquaintances and greater satisfaction with close relationships, while another with an equally small network experiences acute loneliness. These observations call for independent measures of objective isolation and subjective social experience. Carstensen et al. (1999) National Academies (2020)
Nonkin social bonds also connect older people to communities and institutions. Religious groups, neighborhood associations, educational programs, sports clubs, and online communities may support friendship formation. Social access can depend on transportation, local services, disability accommodation, language, or digital literacy. Participation is therefore an outcome partly of infrastructure and policy. A study examining whether friendship predicts health should account for underlying health, economic security, and the possibility that illness itself reduces participation. Longitudinal transitions, network nominations, and repeated observations of relationship functions can illuminate reciprocal links between social engagement and functioning, while qualitative work investigates the meaning of chosen relationships. A comprehensive theory of later-life social roles must include these forms of affiliation even where family intergenerational relations remain the primary research focus. Antonucci et al. (2014) Holt-Lunstad et al. (2010) National Academies (2020)
Civic Participation and Volunteering
Volunteering and civic participation provide organized roles through which older adults contribute knowledge, labor, care, governance, and companionship. Such activities can maintain recognizable social identities after changes in employment and family roles, and they create ties across age groups outside households. The category must be operationalized carefully: occasional volunteer activity differs from a weekly assignment, formal board membership, political participation, and unpaid caregiving for relatives. Participation can be chosen, expected, or necessary for material reasons. A quantitative study should measure time spent, institutional setting, responsibility, perceived autonomy, and whether the role creates enduring social relations. Continuity theory predicts that a voluntary role may preserve valued activities after occupational retirement, although evidence of continuity requires observing earlier role preferences. Atchley (1989) Akhter-Khan et al. (2023)
A systematic review by Akhter-Khan and colleagues (2023) examined caregiving, volunteering, and loneliness among middle-aged and older adults. Such evidence helps compare engagement categories, but the review's included studies differ in design, sample composition, definitions of engagement, and social contexts. Selection is a major concern: people who are healthier, financially stable, and less lonely may be more capable of volunteering. Volunteer programs themselves may differ in their capacity to establish meaningful rather than superficial contact. Consequently, associations between volunteering and lower loneliness should be presented as suggestive of potentially beneficial social processes, with causal conclusions reserved for designs capable of addressing baseline differences and changes over time. Akhter-Khan et al. (2023)
Intergenerational civic roles offer distinctive psychological functions. Older participants may mentor younger adults, teach practical skills, contribute to community decisions, and receive new perspectives or assistance in return. The reciprocity can challenge assumptions that aging entails a one-way need for care. Contact does not automatically produce mutual respect, however, because organizations may reproduce age hierarchies or assign older volunteers stereotyped tasks. Evaluations should assess the quality and duration of interactions, equality of status, opportunities for shared goals, and whether age-related attitudes change on both sides. Interventions that aim to improve relations across age groups require comparison conditions and follow-up, as emphasized in the broader evidence on reducing ageism. These results are relevant to program design without implying that every older adult must participate in formal volunteer work. Burnes et al. (2019) Antonucci et al. (2014)
Ageism and Intergenerational Contact
Ageism refers to stereotypes, prejudices, and discrimination associated with age. It can shape later-life social roles by narrowing perceived competence, influencing hiring and retirement decisions, limiting leadership opportunities, and altering the way family members interpret an older person's choices. Age-related stereotypes may cast elders as uniformly dependent or unproductive, reducing opportunities for reciprocal engagement. Conversely, expectations that all older adults must be active, healthy, and independent can stigmatize those living with disability or needing assistance. Psychological research should evaluate the specific content, target, and context of attitudes rather than assuming one undifferentiated prejudice. Intergenerational interaction can provide opportunities to revise expectations through sustained experience, but its effects depend on the structure of contact. Burnes et al. (2019)
Burnes and colleagues (2019) systematically reviewed and meta-analyzed interventions intended to reduce ageism against older adults. The evidence base included interventions using education, intergenerational contact, or combined approaches, showing that age attitudes can be studied as outcomes responsive to designed experiences. Such intervention research is especially valuable because it moves beyond correlations between contact and stereotypes. Nevertheless, programs differ in population, outcome measure, exposure duration, and follow-up, and attitudinal improvement does not automatically translate into enduring behavioral change. An appropriate account must distinguish measured stereotypes, self-reported willingness to interact, and actual changes in allocation of resources, social roles, or institutional practice. Burnes et al. (2019)
Intergenerational contact in families differs from contact in educational or civic programs. Family ties have long histories, asymmetric obligations, and established distributions of authority; community programs may create novel relationships under more structured conditions. In either setting, contact quantity alone may fail to predict improved attitudes if participants encounter conflict or unequal treatment. Research should test mechanisms including perspective taking, perceived competence, relationship quality, and meaningful cooperation. Ageism can also intersect with gender, socioeconomic position, ethnicity, disability, and migration status, affecting the roles individuals are offered and the resources that sustain participation. The practical implication is to design inclusive opportunities that preserve autonomy and meaningful contribution while testing outcomes with appropriate controls and diverse samples. Burnes et al. (2019) Antonucci et al. (2014)
Social Isolation, Loneliness, and Health
Social isolation refers to a relative lack of social contact or relationships, while loneliness refers to the subjective experience of a gap between desired and actual social connection. The distinction is essential in later-life research. A person may live alone and maintain fulfilling relationships, or live with adult children and experience loneliness despite frequent contact. National Academies of Sciences, Engineering, and Medicine (2020) reviewed the relevance of isolation and loneliness for older adults and identified consequences for health-care assessment and coordination. Research that uses family size as a proxy for social connection misses both the quality of bonds and the person's own expectations. Effective measurement combines network structure, frequency of contact, perceived social support, and direct questions about loneliness. National Academies (2020)
A widely cited meta-analysis by Holt-Lunstad, Smith, and Layton (2010) synthesized prospective evidence on social relationships and mortality risk. The reported association across diverse indicators supports investigating social relationships as a meaningful health correlate, while variation in measurement and the observational character of much of the evidence require care when interpreting mechanisms. Family solidarity, friendship, and social participation may influence well-being through emotional support, practical assistance, health behavior, and access to services. Health can also shape the network: impairments may reduce mobility and make social engagement more difficult. To clarify direction, studies need repeated health and relationship assessments, plausible confounder adjustment, and attention to timing. Holt-Lunstad et al. (2010) National Academies (2020)
Caregiving and volunteering reveal the complexity of loneliness outcomes. Intense family involvement may involve many interactions but leave little opportunity for chosen companionship; voluntary community engagement may expand contacts yet fail to create close ties. The systematic review by Akhter-Khan and colleagues (2023) examined these associations rather than assuming engagement is uniformly protective. An older adult's experience can reflect caregiving demands, the perceived quality of relationships, reciprocity, and social recognition as much as total contact hours. Clinical or policy interventions should therefore address the specific deficit—access, transportation, relationship quality, support burden, or unwanted isolation—rather than treating every form of contact as equivalent. The evidence warrants concern for isolation and loneliness but should not be converted into individual medical advice or causal promises that any one activity prevents disease. Akhter-Khan et al. (2023) National Academies (2020)
Institutional Care and the Redistribution of Family Responsibilities
Family support and formal care services jointly shape later-life roles. Home care, respite services, accessible transportation, paid leave, health-care coordination, and income support influence which tasks are performed by older adults, adult children, spouses, friends, and professionals. Public policy may change the practical costs of supporting a relative and the meaning of family responsibility. The National Academies report Families Caring for an Aging America (2016) documents the organization and burden of family caregiving in the United States and assesses systems intended to support caregivers. Such institutional evidence is indispensable because a survey of family affection alone cannot explain why equally committed relatives provide very different amounts of care. National Academies (2016)
The allocation of care is particularly consequential when daily support becomes intensive. Adult children may balance work schedules, care for their own children, and tasks related to a parent's health or mobility. Older spouses may become primary caregivers while experiencing their own functional changes. The psychological effect of a demanding role depends partly on the availability and reliability of substitute services. When formal assistance is scarce, family members may assume responsibilities beyond their resources; when services are accessible, assistance from relatives may be directed toward companionship, advocacy, or personally meaningful contact. Research should evaluate both care recipients and caregivers, distinguishing unmet need, service use, role burden, and subjective relationship quality. National Academies (2016) National Academies (2020)
Policy evaluation demands care in causal inference. Regions with extensive formal services often differ from poorly served areas in income, public health, housing, and demography. Cross-national comparisons may therefore misattribute differences in family caregiving to values or norms when institutional opportunity explains part of the variation. Natural policy changes, administrative data linked to surveys, and prospective panels can help evaluate whether service expansion changes care hours, perceived autonomy, family conflict, and employment. Qualitative interviews clarify whether the availability of professional help is viewed as preserving family relationships or displacing a valued responsibility. A scientifically grounded model of intergenerational relations treats family and institutional systems as interacting environments with measurable effects on everyday role performance. National Academies (2016) Grundy (2005)
Which Child Becomes a Caregiver?
The presence of several adult children does not guarantee an equal division of caregiving. Parents may anticipate receiving care from a particular child based on closeness, proximity, family history, or perceived competence. Siblings differ in employment demands, geographic location, their own health, household finances, and obligations to dependent family members. In a within-family study, Pillemer and Suitor (2006) investigated mothers' expectations about which adult child would become a caregiver, showing the value of sampling multiple children within each family. This approach asks how caregivers are selected from a set of potential helpers, rather than treating adult children as exchangeable members of a collective support pool. Pillemer & Suitor (2006)
Caregiver selection also evolves when need becomes immediate. Leopold, Raab, and Engelhardt (2014) analyzed transitions into parental care in the context of siblings' relative costs and commitments, emphasizing the temporal process of role allocation. Expectations expressed before health decline may differ from the practical arrangement adopted after hospitalization, widowhood, or a move. A child who previously provided occasional transportation may assume primary care, while another may offer financial assistance. Selection is also likely to be affected by longstanding relational closeness and obligations. These studies provide evidence on patterns of choice and family constraints, but their observational designs require caution when interpreting how any single family characteristic causes a child to become the helper. Leopold et al. (2014) Pillemer & Suitor (2006)
Measurement must distinguish the identity of the expected caregiver, the person providing help in practice, and the intensity of the tasks performed. The older parent's preference matters alongside each child's capacity and willingness. Bui and colleagues (2022) showed how earlier exchanges, including tangible help and childcare, can be studied as predictors of later expected or received care within one family. Researchers should also attend to adult children who do not become caregivers and to conflicts arising from perceived inequity. A multi-informant longitudinal design can test whether expectations forecast actual support while tracing how responsibilities shift when family circumstances change. The broader implication is that intergenerational solidarity is shaped by differentiated relationships, practical opportunities, and decisions under constraint. Bui et al. (2022) Leopold et al. (2014)
Grandchild Care Intensity and Selection Effects
The intensity of grandchild care is an important variable because different forms of involvement generate different opportunities and responsibilities. A grandparent who provides an occasional afternoon of help may experience the activity as freely chosen, while one responsible for daily childcare may face significant scheduling demands. Nonresident caregiving, co-resident caregiving, and full custodial responsibility also differ in authority, economic costs, and access to parental support. The review by Danielsbacka and colleagues (2022) demonstrates that these distinctions are empirically consequential: its synthesis found divergent associations between well-being and custodial, multigenerational, or non-coresident grandparenting. The methodological lesson is that a binary indicator of grandchild care discards the conditions most likely to explain variation. Danielsbacka et al. (2022)
Di Gessa, Glaser, and Tinker (2016) analyzed health outcomes following different intensities of grandchild care among European older adults using the Survey of Health, Ageing and Retirement in Europe. Their longitudinal design measured care at an earlier wave and health at subsequent follow-ups while adjusting for baseline conditions. This structure offers an important improvement over a cross-sectional association, especially because physical health can determine whether a grandparent is able to provide care. Nevertheless, statistical adjustment cannot guarantee control for every family crisis, preference, or resource that influences both participation and later health. Conclusions therefore should identify particular populations and types of engagement rather than use grandparenthood as a single exposure. Di Gessa et al. (2016)
An informative research program would combine exposure intensity, voluntariness, childcare purpose, family structure, and changing health in the same panel. It should distinguish outcomes such as depressive symptoms, self-rated health, loneliness, life satisfaction, and quality of relationships because they may move in different directions. Interactions with gender, pension status, and formal childcare access can test whether social opportunity modifies the association. Qualitative research can examine the experience of autonomy, intergenerational recognition, and fatigue among caregivers whose hours are similar. These complementary methods would help explain the mechanisms underlying associations without assuming that more childcare necessarily means more meaning or more burden. Di Gessa et al. (2016) Danielsbacka et al. (2022)
Multigenerational Households and Everyday Negotiation
Co-residence places intergenerational relations within a shared domestic space. Grandparents, parents, and children may share resources, meals, transportation, supervision, and care while also negotiating privacy, authority, expenses, and daily routines. The same household arrangement can be driven by cultural preference, housing markets, economic insecurity, illness, or a temporary family transition. Co-residence is therefore a structural descriptor rather than a direct index of solidarity or well-being. It may increase access to help while also increasing the frequency of disagreement. Research should distinguish household membership from the actual quality of interaction and specify who controls financial resources and decisions. The multidimensional solidarity and ambivalence frameworks are especially relevant for these arrangements. Bengtson & Roberts (1991) Lüscher & Pillemer (1998)
Three-generation households involving grandchildren require attention to both older adults' and children's roles. The systematic review by Danielsbacka and colleagues (2022) found mixed associations between grandparental health and three-generation arrangements, with relatively limited evidence compared with other grandparenting forms. Household composition alone may conceal who performs care and whether the older person is a primary provider, occasional helper, or care recipient. Equally important, the arrangement may be initiated by an adult child's housing need or by an older relative's declining capacity. Analyses that fail to specify these circumstances risk treating selection into co-residence as an effect of living together. Multigenerational surveys should collect the transition into the household, prior relationships, task allocations, and whether participants consider the arrangement temporary or enduring. Danielsbacka et al. (2022)
Qualitative methods can uncover the negotiations hidden in quantitative categories. Participants may disagree about acceptable privacy, the discipline of children, household contributions, or who has a duty to provide assistance. Such accounts help generate survey items and mechanisms for longitudinal testing. Observational and diary studies can estimate fluctuations in conflict and mutual help during routine tasks, while administrative or housing data may contextualize why co-residence occurred. A life-course perspective identifies how household membership changes when grandchildren mature or when older relatives experience illness or bereavement. The central scientific question concerns how resources, authority, intimacy, and agency are coordinated across generations sharing an environment, with attention to the institutional conditions that made the arrangement possible. Lüscher & Pillemer (1998) Antonucci et al. (2014)
Relational Mechanisms and Psychological Well-Being
The relation between social roles and psychological well-being may operate through several pathways. Meaningful participation can provide continuity of identity, recognition by others, opportunities to exercise skills, and experiences of reciprocity. Close relationships can provide emotional reassurance and practical assistance, while conflict and excessive obligations may increase distress or restrict activities. These processes should be treated as distinct hypotheses rather than blended into an all-purpose claim that social contact is beneficial. A study of mentoring should measure perceived contribution and autonomy; a study of caregiving should measure responsibility and strain; and a study of family contact should measure perceived responsiveness alongside frequency. The convoy and solidarity traditions provide variables for specifying the mediating relational conditions. Antonucci et al. (2014) Bengtson & Roberts (1991)
The outcome domain matters. Social engagement may be related to life satisfaction without similarly predicting depressive symptoms, physical functioning, or loneliness. Holt-Lunstad and colleagues' (2010) meta-analysis of social relationships and mortality demonstrates the significance of network indicators for prospective health research, while the National Academies' 2020 review documents the distinction between isolation and loneliness. Both bodies of evidence concern multiple forms of social connection rather than a direct experimental test of an individual later-life role. Extrapolations to retirement, grandparenthood, or volunteering thus require direct measurement and careful confounder control. The systematic review of volunteering and caregiving further cautions that participation categories contain heterogeneous activities and intentions. Holt-Lunstad et al. (2010) National Academies (2020) Akhter-Khan et al. (2023)
Role strain and role meaning can coexist within the same person, making average effects difficult to interpret. A custodial grandparent may describe strong affection and purpose alongside exhaustion; an older adult receiving help may be grateful while frustrated by reduced control. Longitudinal mediation analyses should establish the temporal order of role changes, proposed mechanisms, and outcomes, and should consider common causes of each. Within-family data can separate shared circumstances from differences among relatives, although unmeasured change remains possible. From a clinical and community standpoint, support should address particular difficulties—including overwhelming tasks, limited choices, unmet care needs, or loneliness—rather than prescribe a single normative pathway through later life. The scientific objective is to explain under which conditions roles support psychological functioning and when the same nominal role imposes costs. Danielsbacka et al. (2022) Minkler et al. (1997) Lüscher & Pillemer (1998)
Ethical Autonomy and Intergenerational Responsibility
Later-life intergenerational relations raise ethical questions about respecting preferences, distributing responsibilities, and protecting people whose circumstances create dependence. Psychological evidence can clarify the conditions under which assistance is experienced as supportive, controlling, or burdensome. It cannot itself establish a universal moral ranking of family obligations. Adults differ in their legitimate preferences for privacy, shared decision making, kin participation, and professional services. Older adults receiving care remain participants in decisions affecting their own lives wherever decision-making capacity permits. Research that records only the caregiver's account risks misrepresenting the recipient's priorities. The literature on intergenerational ambivalence makes explicit why a close family tie can contain both commitment and a dispute about authority. Lüscher & Pillemer (1998) National Academies (2016)
Responsibility also extends beyond the immediate dyad. Intensive family caregiving can affect siblings, partners, grandchildren, employment, and the caregiver's own health. Institutions shape the distribution of these obligations through service availability and financial policy. The National Academies report on family caregiving identifies the public importance of caregiver support and coordinated services, while empirical studies of caregiver selection demonstrate that family burden is rarely distributed equally. Respect for older adults and caregivers therefore requires studying both the quality of consent and the real alternatives available to participants. A stated preference for unpaid family care may have a different meaning when publicly supported services are accessible than when no viable alternative exists. National Academies (2016) Pillemer & Suitor (2006)
Research ethics extends to the collection and interpretation of sensitive family data. Dyadic studies can expose disputes, financial transfers, or health information that one relative would not wish to share with another. Researchers should protect confidentiality across informants, obtain appropriate consent, and specify whether participants are reporting their own activities or judging another family member. Technology-mediated care introduces further questions about access to messages, automated decisions, and control over data. The empirical agenda is to assess whether such practices strengthen autonomy, equitable assistance, and trust under observable conditions. Ethical design supports the scientific integrity of intergenerational research because coerced reporting, incomplete privacy protection, and unexamined power imbalances can distort precisely the relationships the study aims to understand. National Academies (2016) National Academies (2020)
Gender, Culture, and Socioeconomic Stratification
Later-life roles are differentiated by social positions accumulated across the life course. Employment histories, pension entitlements, health insurance, property ownership, and access to transportation constrain the roles available after retirement. Gendered patterns of paid work and unpaid caregiving can create unequal financial security and expectations about who should provide care. Cultural and family norms influence how people interpret reciprocity, filial responsibility, co-residence, and the acceptability of formal assistance. These conditions shape role opportunities even when older adults have similar ages or apparent family structures. Research in the United States on support to aging parents and grown children has identified differences in the amount and organization of assistance across family groups, underscoring the need to study social position rather than interpret age as a sufficient explanatory variable. Fingerman et al. (2011b) Grundy (2005)
Cross-cultural comparisons require explicit measurement equivalence. A survey item asking whether adult children 'should' care for parents may refer to personal affection, moral obligation, economic necessity, or the lack of viable public services, depending on the setting. High rates of co-residence can similarly indicate preference, housing costs, migration patterns, or multigenerational property arrangements. Researchers should measure institutions and subjective motivations independently and avoid attributing population differences to immutable cultural values. Even within one country, variation by household resources, region, and access to services may be substantial. The convoy perspective accommodates this variation because social networks are situated within historical and structural opportunities; intergenerational solidarity models then provide more precise constructs for what those relationships accomplish. Antonucci et al. (2014) Bengtson & Roberts (1991)
Inequality also affects the measurement and consequences of social participation. Older adults with substantial time and economic security may pursue voluntary community roles; those who face economic pressure may continue employment or provide unpaid care. The same number of caregiving hours can represent different burdens depending on health, support availability, and household income. Ageist practices can further restrict opportunities for work and participation, and interventions to reduce stereotypes may operate differently across settings. Analytically, stratified longitudinal models can test whether transitions have heterogeneous associations with psychological outcomes, while qualitative interviews elucidate why participants interpret similar arrangements differently. Such evidence supports context-sensitive explanation and limits the transportability of a finding from affluent, urban, or highly educated samples to populations facing different constraints. Burnes et al. (2019) Akhter-Khan et al. (2023) National Academies (2020)
Migration, Distance, and Digital Mediation
Geographic separation can alter the feasible forms of intergenerational support without necessarily weakening emotional bonds. Migration for employment, family reunification, education, or protection may redistribute kin across cities and countries. Families can sustain frequent conversation and economic transfers over long distances while remaining unable to provide immediate hands-on care. Measuring support therefore requires distinguishing communication, financial remittances, travel, emergency assistance, and daily caregiving. Structural solidarity concerns the opportunity to interact, whereas functional solidarity concerns the help actually given or received. This distinction is especially important when interpreting seemingly low face-to-face contact as a sign of deficient family commitment. Bengtson & Roberts (1991) Roberts & Bengtson (1990)
Digital communication makes it possible to participate in some family practices despite distance. Video calls, messaging, shared photo archives, online scheduling, and remote coordination can help older adults retain roles as parents, grandparents, friends, and caregivers. Their psychological effects depend on usability, trust, privacy, social preferences, and access to reliable infrastructure. Technology can facilitate meaningful communication but does not remove the practical demands of physical care, transportation, or housing support. A rigorous study would compare communication frequency, perceived intimacy, actual help during crises, and the older person's preferences before and after a technological change. It would also examine digital exclusion: inaccessible interfaces or unaffordable connectivity can make technological mediation an additional source of inequality. The explanatory constructs remain role continuity, contact opportunity, subjective meaning, and functional exchange rather than a presumption that digital contact is identical to in-person care. Antonucci et al. (2014) National Academies (2020)
The Artificial Era introduces an additional institutional context in which artificial systems may help coordinate intergenerational communication, manage reminders, translate messages, or support access to services. Such functions can be operationalized as task completion, user control, error rates, privacy protection, and changes in the distribution of family work. Whether people perceive a system as supportive is an empirical question, and its technical reliability requires separate assessment. An older adult using automated tools to organize appointments may gain practical independence while family members retain responsibility for emotionally significant decisions. New research should test whether these arrangements facilitate autonomy and equitable reciprocity, especially when technology changes who has access to household information. This inquiry extends established studies of social organization into human–artificial settings while preserving a clear boundary between documented family psychology and hypotheses about emerging technological mediation. Antonucci et al. (2014) National Academies (2020)
Measurement of Roles and Intergenerational Relations
Measurement begins with specifying whether the target is a social status, an activity, an expectation, a relationship quality, or a psychological outcome. Retirement status cannot stand in for freedom from work demands; being a grandparent does not establish childcare involvement; the presence of adult children does not establish usable care. Surveys should collect role identities and actual task involvement, then distinguish formal membership, intensity, voluntariness, subjective centrality, and perceived recognition. For intergenerational relations, the solidarity tradition offers domains including affection, association, function, consensus, norms, and structural opportunity. Roberts and Bengtson's (1990) dimensional analysis provides an important reminder that a latent solidarity score needs empirical support before use. Roberts & Bengtson (1990) Bengtson & Roberts (1991)
Measurement units can be individuals, family dyads, households, or social networks, and the choice determines which hypotheses are testable. A parent may describe the relationship with each adult child differently; a child's report of care may disagree with the parent's report of receiving it. Data from several children within one family are statistically dependent and call for multilevel or related designs. Network methods can map tie types and functions, while event histories record when role transitions occur. Repeated diary surveys may capture short-term fluctuations in contact and tension; retrospective interviews can reconstruct role meaning but require attention to memory and narrative reconstruction. Recent studies of support exchanges illustrate the value of linking reports of prior giving to later care outcomes within families rather than measuring one generic family-support construct. Bui et al. (2022) Shen et al. (2025)
Construct validity also requires attention to cultural and temporal invariance. A caregiving item may change meaning as formal services expand or as employment conditions alter what family help entails. Loneliness cannot be inferred solely from network size, and engagement cannot be inferred solely from hours of volunteering. Researchers should predefine primary outcomes and distinguish reflective indicators of subjective experience from formative inventories of different activities. Reporting should include sampling rules, item wording, reliability evidence, missing-data patterns, and the implications of attrition. Where differences between groups are interpreted, researchers should test whether measurement operates comparably across groups. Such practices permit meaningful comparison of age cohorts and national contexts without collapsing role structure into a single index with unclear psychological content. National Academies (2020) Akhter-Khan et al. (2023) Bengtson & Roberts (1991)
Longitudinal Designs and Causal Inference
Later-life roles unfold through transitions, so longitudinal evidence is crucial. Repeated panels can observe support, health, retirement, and relationship quality before and after a change. Event-history methods can represent retirement, widowhood, caregiving onset, or moves between households as dated occurrences; within-person models can distinguish changes in an individual's experience from enduring between-person differences. However, temporal sequence alone does not remove confounding. Worsening health can both cause a transition into receiving care and predict lower well-being, so a naive comparison may attribute an existing health trajectory to care receipt. Designs should consider time-varying confounders, informative attrition, cohort effects, and the dependence of outcomes within families. Isherwood et al. (2012) Shen et al. (2025)
Prospective exchange studies illustrate both the promise and the limits of family-panel methods. Bui and colleagues (2022) related earlier support patterns to later expected and actual caregiving within families, while Shen and colleagues (2025) examined when potential adult-child support translated into real assistance. These approaches improve on one-time reports because prior exchanges are observed before later care outcomes, but common underlying attachment, economic advantages, and family norms may still explain both. Causal interpretation can be strengthened by measuring baseline trajectories, specifying plausible confounder sets, and examining sensitivity to unmeasured influences. Natural experiments based on policy changes can contribute complementary evidence when the policy genuinely alters role opportunities and the comparison assumptions can be justified. Bui et al. (2022) Shen et al. (2025)
Evidence synthesis must also recognize study-design heterogeneity. Danielsbacka and colleagues' (2022) grandparenting review spans custodial and noncustodial arrangements with differing selection mechanisms, and Akhter-Khan and colleagues' (2023) review examines varied forms of unpaid productive participation. A pooled association across incomparable exposures would produce a deceptively simple answer. Research should specify eligibility, design, exposure definition, outcome domain, and whether the estimate addresses selection or an intervention. Qualitative findings can generate mechanisms and identify missing constructs, while longitudinal quantitative tests estimate associations with known uncertainty. Independently replicating a result across populations and methods improves confidence when the same core construct is measured adequately. The scientific aim is explanatory integration through explicit assumptions, not the declaration of causality based on statistically significant coefficients. Danielsbacka et al. (2022) Akhter-Khan et al. (2023) Reyes & Patterson (2025)
Scientific Convergences, Contradictions, and Research Agenda
Several conclusions recur across distinct research traditions. Later-life role trajectories are diverse; social networks change through selection, loss, and opportunity; intergenerational ties contain multiple dimensions; support can flow in both directions; and grandparenting and caregiving outcomes vary with responsibility and context. The convoy model organizes changing networks, continuity theory explains adaptation through familiar identities, socioemotional selectivity provides testable motivational predictions, and solidarity and ambivalence theories clarify the relational content of family exchanges. These frameworks become most informative when their explanatory roles are kept explicit. An observed reduction in contact could be predicted from changed time horizons, altered opportunity, or a bereavement; distinguishing the hypotheses requires more than adding the theories together. Antonucci et al. (2014) Atchley (1989) Carstensen et al. (1999) Bengtson & Roberts (1991) Lüscher & Pillemer (1998)
Empirical disagreements often emerge from different exposures or comparison groups. An association between voluntary grandparental contact and well-being may coexist with evidence of distress in intensive custodial care. Volunteering may be linked to lower loneliness in one context while providing little relief when roles are demanding or contact lacks intimacy. Strong family norms may support help but increase guilt when obligations cannot be fulfilled. Researchers should treat these patterns as prompts to investigate intensity, voluntariness, relational quality, prior adversity, and material resources. Review evidence on caregiving, grandparenting, and loneliness provides support for this differentiated agenda, but the field still needs designs that compare mechanisms across contexts rather than relying on a universal benefit of participation. Danielsbacka et al. (2022) Akhter-Khan et al. (2023) Minkler et al. (1997)
Priorities for future research follow from these gaps. Large multigenerational panels should collect parallel parent and child reports, documented transfers, role meaning, perceived choice, and local institutional conditions. Preregistered hypotheses can compare solidarity, ambivalence, continuity, and motivational models using the same observations. Policy natural experiments may clarify how pension rules, caregiving services, accessible transport, and affordable communication affect the distribution of obligations. Research should include older adults without children, migrants, people living with disabilities, rural communities, and economically vulnerable populations. New technological forms of coordination warrant evaluation against substantive outcomes—autonomy, relationship quality, task reliability, privacy, and distribution of care—rather than adoption rates alone. Together these priorities permit a cumulative, inclusive science of social roles that explains how relationships are sustained and reorganized across the later life course. National Academies (2020) Burnes et al. (2019) Shen et al. (2025)
Conclusion
Later-life social roles and intergenerational relations form a changing system of social positions, obligations, opportunities, and meanings through which people organize participation and mutual assistance. Retirement, grandparenthood, partnership changes, caregiving, and civic involvement can each restructure the distribution of time and resources, but their psychological consequences depend on prior biographies, health, material circumstances, relationship histories, and perceived choice. Intergenerational relations persist through multiple dimensions of solidarity, reciprocity, and ambivalence. A complete account therefore identifies what a role requires, how it is negotiated, who is involved, what resources circulate, and how its meaning changes over time. Antonucci et al. (2014) Bengtson & Roberts (1991) Lüscher & Pillemer (1998)
The strongest synthesis comes from triangulating complementary traditions. Continuity theory locates adaptive change within enduring identity and practices; socioemotional selectivity specifies motivational consequences of perceived time horizons; convoy theory examines the structure and functions of changing social networks; solidarity and ambivalence models make the content of family ties measurable. Longitudinal studies show how giving and receiving help are intertwined, while research reviews document that the apparent benefits or costs of caregiving and participation vary substantially by context. These findings support differentiated explanations and caution against assigning a uniform developmental purpose to withdrawal, activity, grandparenting, or family assistance. Atchley (1989) Carstensen et al. (1999) Shen et al. (2025) Danielsbacka et al. (2022)
A mature research program must combine explicit constructs, multigenerational evidence, reliable measurement, and careful causal design. It should investigate how institutions and technological infrastructures enable or constrain meaningful participation, how unequal resources affect the ability to choose roles, and how negotiated support preserves agency as needs change. The Artificial Era makes these questions especially timely as communication and coordination increasingly involve artificial systems, creating new possibilities and responsibilities that require direct empirical study. The enduring scientific object remains the organization of relationships among generations and the conditions under which those relationships sustain psychological well-being, autonomy, reciprocity, and participation over time. National Academies (2020) Antonucci et al. (2014)
References
- Akhter-Khan, S. C., Hofmann, V., Warncke, M., Tamimi, N., Mayston, R., & Prina, M. A. (2023). Caregiving, volunteering, and loneliness in middle-aged and older adults: A systematic review. Aging & Mental Health, 27(7), 1233–1245. https://doi.org/10.1080/13607863.2022.2144130
- Antonucci, T. C., & Akiyama, H. (1987). Social networks in adult life and a preliminary examination of the convoy model. Journal of Gerontology, 42(5), 519–527. https://doi.org/10.1093/geronj/42.5.519
- Antonucci, T. C., Ajrouch, K. J., & Birditt, K. S. (2014). The convoy model: Explaining social relations from a multidisciplinary perspective. The Gerontologist, 54(1), 82–92. https://doi.org/10.1093/geront/gnt118
- Atchley, R. C. (1989). A continuity theory of normal aging. The Gerontologist, 29(2), 183–190. https://doi.org/10.1093/geront/29.2.183
- Bengtson, V. L., & Roberts, R. E. L. (1991). Intergenerational solidarity in aging families: An example of formal theory construction. Journal of Marriage and the Family, 53(4), 856–870. https://doi.org/10.2307/352993
- Bui, C. N., Kim, K., & Fingerman, K. L. (2022). Support now to care later: Intergenerational support exchanges and older parents' care receipt and expectations. The Journals of Gerontology: Series B, 77(7), 1315–1324. https://doi.org/10.1093/geronb/gbac059
- Burnes, D., Sheppard, C., Henderson, C. R., Jr., Wassel, M., Cope, R., Barber, C., & Pillemer, K. (2019). Interventions to reduce ageism against older adults: A systematic review and meta-analysis. American Journal of Public Health, 109(8), e1–e9. https://doi.org/10.2105/AJPH.2019.305123
- Carstensen, L. L. (1992). Social and emotional patterns in adulthood: Support for socioemotional selectivity theory. Psychology and Aging, 7(3), 331–338. https://doi.org/10.1037/0882-7974.7.3.331
- Carstensen, L. L., Fung, H. H., & Charles, S. T. (2003). Socioemotional selectivity theory and the regulation of emotion in the second half of life. Motivation and Emotion, 27(2), 103–123. https://doi.org/10.1023/A:1024569803230
- Carstensen, L. L., Isaacowitz, D. M., & Charles, S. T. (1999). Taking time seriously: A theory of socioemotional selectivity. American Psychologist, 54(3), 165–181. https://doi.org/10.1037/0003-066X.54.3.165
- Cumming, E., & Henry, W. E. (1961). Growing old: The process of disengagement. Basic Books. https://search.worldcat.org/title/Growing-old-%3A-the-process-of-disengagement/oclc/1078826751
- Danielsbacka, M., Křenková, L., & Tanskanen, A. O. (2022). Grandparenting, health, and well-being: A systematic literature review. European Journal of Ageing, 19(3), 341–368. https://doi.org/10.1007/s10433-021-00674-y
- Di Gessa, G., Glaser, K., & Tinker, A. (2016). The health impact of intensive and nonintensive grandchild care in Europe: New evidence from SHARE. The Journals of Gerontology: Series B, 71(5), 867–879. https://doi.org/10.1093/geronb/gbv055
- Fingerman, K. L., Pitzer, L. M., Chan, W., Birditt, K. S., Franks, M. M., & Zarit, S. H. (2011). Who gets what and why? Help middle-aged adults provide to parents and grown children. The Journals of Gerontology: Series B, 66(1), 87–98. https://doi.org/10.1093/geronb/gbq009
- Fingerman, K. L., VanderDrift, L. E., Dotterer, A. M., Birditt, K. S., & Zarit, S. H. (2011). Support to aging parents and grown children in Black and White families. The Gerontologist, 51(4), 441–452. https://doi.org/10.1093/geront/gnq114
- Grundy, E. (2005). Reciprocity in relationships: Socio-economic and health influences on intergenerational exchanges between Third Age parents and their adult children in Great Britain. The British Journal of Sociology, 56(2), 233–255. https://doi.org/10.1111/j.1468-4446.2005.00057.x
- Holt-Lunstad, J., Smith, T. B., & Layton, J. B. (2010). Social relationships and mortality risk: A meta-analytic review. PLoS Medicine, 7(7), e1000316. https://doi.org/10.1371/journal.pmed.1000316
- Isherwood, L. M., King, D. S., & Luszcz, M. A. (2012). A longitudinal analysis of social engagement in late-life widowhood. International Journal of Aging and Human Development, 74(3), 211–229. https://doi.org/10.2190/AG.74.3.c
- Leopold, T., Raab, M., & Engelhardt, H. (2014). The transition to parent care: Costs, commitments, and caregiver selection among children. Journal of Marriage and Family, 76(2), 300–318. https://doi.org/10.1111/jomf.12099
- Lüscher, K., & Pillemer, K. (1998). Intergenerational ambivalence: A new approach to the study of parent–child relations in later life. Journal of Marriage and the Family, 60(2), 413–425. https://eric.ed.gov/?id=EJ579056
- Minkler, M., Fuller-Thomson, E., Miller, D., & Driver, D. (1997). Depression in grandparents raising grandchildren: Results of a national longitudinal study. Archives of Family Medicine, 6(5), 445–452. https://doi.org/10.1001/archfami.6.5.445
- National Academies of Sciences, Engineering, and Medicine. (2016). Families caring for an aging America. National Academies Press. https://doi.org/10.17226/23606
- National Academies of Sciences, Engineering, and Medicine. (2020). Social isolation and loneliness in older adults: Opportunities for the health care system. National Academies Press. https://doi.org/10.17226/25663
- Pillemer, K., & Suitor, J. J. (2006). Making choices: A within-family study of caregiver selection. The Gerontologist, 46(4), 439–448. https://doi.org/10.1093/geront/46.4.439
- Reyes, A. M., & Patterson, S. E. (2025). Dynamics of family exchanges: How life course transfers influence support from adult children. Research on Aging, 47(7–8), 346–360. https://doi.org/10.1177/01640275251321485
- Roberts, R. E. L., & Bengtson, V. L. (1990). Is intergenerational solidarity a unidimensional construct? A second test of a formal model. Journal of Gerontology, 45(1), S12–S20. https://doi.org/10.1093/geronj/45.1.S12
- Shen, Y., van Tilburg, T. G., & van der Horst, M. (2025). The translation of intergenerational care potential into care receipt of older parents: A prospective study. Research on Aging, 47(7–8), 392–403. https://doi.org/10.1177/01640275251326507
- Silverstein, M., & Bengtson, V. L. (1997). Intergenerational solidarity and the structure of adult child–parent relationships in American families. American Journal of Sociology, 103(2), 429–460. https://doi.org/10.1086/231213
Social Roles as Life-Course Structures
Later-life social roles are socially recognized positions, expectations, practices, and relationships through which people participate in families, communities, institutions, and intergenerational exchanges as they age. A role may describe employment, marriage, parenthood, grandparenthood, caregiving, friendship, voluntary service, civic participation, or the receipt of care. Roles have an institutional component because access and obligations are shaped by pensions, labor markets, housing, public services, and family law. They also have a psychological component because individuals interpret expectations, identify with roles, negotiate obligations, and experience their continuity or revision as meaningful. Intergenerational relations encompass sustained ties and transfers among persons occupying different familial or social generations, including contact, affection, material exchange, perceived obligations, conflict, and shared activity. The question investigated here is how these roles are reorganized in later life and how reciprocal intergenerational relations shape adaptation, autonomy, and well-being. Antonucci et al. (2014) Bengtson & Roberts (1991)
Chronological age does not determine a single pattern of roles. An employed grandparent caring for a spouse, a childless retired volunteer, an older migrant maintaining distant family ties, and an older adult receiving formal assistance all occupy distinct configurations of opportunities and obligations. Retirement can be gradual, partial, involuntary, or absent; grandparenthood is neither universal nor socially homogeneous; and kinship can coexist with friendship and relationships formed through organizations. Research therefore needs a person-by-role-by-context unit of analysis rather than assuming one average transition. This principle follows the convoy model's emphasis on persons embedded in dynamically changing networks and the life-course emphasis on timing, cumulative circumstances, and linked biographies. It also changes operationalization: the number of roles, their subjective centrality, time demands, reciprocity, and perceived freedom of choice must be measured separately. Antonucci & Akiyama (1987) Antonucci et al. (2014)
The substantive scope of this monograph concerns retirement and later-life participation, reorganization of adult parent–child relationships, grandparental and caregiving roles, bereavement, intergenerational exchange and solidarity, contact across age groups, social inequality, and the methods required to identify change. It treats psychological well-being as a consequential outcome rather than the definition of successful aging. A socially demanding role may provide purpose while producing exhaustion; limited contact may reflect deliberate selection rather than exclusion; and frequent assistance may signal need, affection, obligation, or conflict. The central analytical distinction is between a role's formal designation, its actual practices, and its meaning to participants. These dimensions can diverge and generate different predictions for subjective experience, relational quality, and observable behavior. Atchley (1989) Lüscher & Pillemer (1998)