Suppose that your parents divorced when you were 10 years old and your mother moved far away from you and your father. Suppose also that you and she have not been close for years. Would that affect how you would feel about helping her if she broke a hip when she was 75? What if she were your stepmother rather than your mother? Would that make a difference to you? What if your stepmother had raised you since you were 11? What if it was your aging father who broke a hip? Your stepfather? We have been asking these types of questions, using vignettes, for two decades in studies examining normative beliefs about intergenerational responsibilities to older (and younger) family members when the family had experienced a divorce or remarriage.

Most Americans agree that "adult children should take care of their parents when they get old" (e.g., Lee, Netzer, & Coward, 1994), a belief so widespread that most states have filial responsibility laws defining younger family members' obligations to provide care for aging parents and grandparents (Bulcroft et al., 1989). Critics argue that these social policies are based on outdated assumptions about families (e.g., families are emotionally close and loving, family members are able and willing to assist older kin, family membership is stable; (Hooyman & Gonyea, 1995). These assumptions do not reflect the experiences of many current families because decades of structural changes due to divorce, remarriage, and cohabitation have made connections among kin more tenuous than in the past.

Although the divorce rate has plateaued (Kreider, 2005), families continue to be affected by divorce and subsequent family transitions. An increasing proportion of older adults have been divorced and most of them have remarried, as have many widowed people (Kreider, 2005). Nearly half of all marriages are remarriages for one or both partners (U.S. Census Bureau, 2000), and often there are children from prior relationships. A recent Pew Center national poll reported that more than 40% of adults have at least one steprelationship. About 40% of multiple-generation families in the United States contain a stepgrandparent.

Policymakers are, therefore, faced with laws and social policies designed for a mid-twentieth century extended family at a time when multigenerational family structures are becoming increasingly more complex. Our intergenerational obligation research was stimulated by the idea that even though most Americans generally agreed that younger family members should support older ones, researchers were reporting far less agreement about intergenerational assistance when individuals were asked about specific contexts (Rossi & Rossi, 1990). We were interested in how divorce and remarriage affected beliefs about intergenerational assistance.

What people do in relationships is based partly on personal beliefs about appropriate actions between kin and partly on normative beliefs about what should be done regarding intergenerational responsibilities (Finch & Mason, 1993; Ganong & Coleman, 1999). Normative beliefs about intergenerational responsibilities also are important to understand because they influence the development and application of public policy.

Motives for intergenerational assistance

Scholars have offered multiple explanations for intergenerational assistance: the norm of family obligation (helping is a duty required because of shared kinship), altruism (a genetic predisposition to help kin or because helping kin makes the donor happier than helping others), reciprocity (children must repay older parents for help they gave when children were young), gratitude (children help parents because they appreciate parents' sacrifices for them), moral duty (help older kin because that is what good people should do), emotional attachments (help is more likely if relationships are emotionally close), and family solidarity (a combination of several of the above explanations). Most explanations for intergenerational assistance are based on the assumption that parents take good care of children when they are young, which elicits aid from the children when the older generation needs help.

Evidence suggests that parental divorce when children are minors may have long-term consequences on intergenerational assistance to parents later in life (e.g., Amato, Rezac, & Booth, 1995). Most of the common explanations are potentially affected by divorce (such as family obligation norms, reciprocity, gratitude, emotional attachment, family solidarity, and even altruism). For instance, if nonresidential parents do not continue to be invested in their children and stay in contact with them, their children may no longer see them as deserving assistance in later life.

Parental remarriages when children are young also may affect familial responsibilities later in life. Ties to parents may be affected in unknown ways by parental remarriages and bonds with stepparents may be ambiguous (Cherlin, 1978). Many of the explanations for intergenerational aid do not automatically apply to stepparents. As with divorced parents in later life, remarried parents and stepparents are likely to be seen as worthy recipients of intergenerational support based on what transpired as children were raised. For stepparents, whether or not they are seen as family may be particularly relevant.

Results from Our Intergenerational Family Obligations Studies

What We Know

In general, genetic bonds represent greater intergenerational obligations than do affinal bonds. All other factors being equal, people believe that parents should receive more assistance than stepparents. Similarly, the obligation to aid children is greater than the obligation to stepchildren.

Genetic kinship is not enough to justify intergenerational assistance. In our studies, kinship was immutable for only about 25% of the study participants. For most people kinship alone (norm of family obligations) was inadequate justification for providing intergenerational aid. For the majority, other relational characteristics were more important.

Relationship quality is important in attributing obligations to assist, regardless of relationship type (genetic or step). Parents and adult children were thought to be much more obligated to help each other when relationships were emotionally close. When relationships were distant or hostile, providing help was discretionary and more limited (Ganong & Coleman, 2006a). Stepparents and stepchildren who had close relationships were perceived to be as obligated to help each other as were parents and adult children with close bonds (Coleman et al., 2005; Ganong & Coleman, 1999).

Beliefs about helping kin are stronger when kin have previously provided help. Most people think adult children are obligated to help parents only if the parents had fulfilled parental responsibilities to children when they were young (Ganong & Coleman, 2006b). Genetic kinship had significance, but without past histories of mutual help it was as if the special loyalties and responsibilities attendant to kinship were unimportant (Coleman et al., 1997; Ganong & Coleman, 2006b). Without a history of financial, tangible, and emotional support from divorced parents, adult children were seen as having a lesser debt to repay than if parents had maintained contact and continued to provide for them. Divorced and remarried parents who were perceived to have broken the reciprocity "contract" lost their "rights" to receive help from adult children.

Stepparents and stepchildren who had helped each other in the past (i.e., the stepparent helped raise the stepchild or they mutually assisted each other as adults) were perceived as obligated to assist each other at levels similar, but not quite equal, to older parents and adult children who had reciprocal exchanges (Coleman et al., 2005; Ganong & Coleman, 1999). Stepparents who helped raise stepchildren thus can "earn" help later in life.

Intergenerational assistance is limited for relationships formed later in life. Steprelationships formed in later life have less time to build close bonds and exchange resources, which reduced the likelihood that relatives would be perceived as kin and have responsibilities to help each other (Ganong & Coleman, 2006a; Ganong et al., 1998). Although direct reciprocity norms may not apply in later-life steprelationships, some people thought that an adult stepchild should repay an older stepparent for help the stepparent provided stepchild's parent or as an indirect way to repay the parent by helping the new spouse or partner.

For middle-generation adults, helping children takes precedence over helping elders. There is a hierarchy of intergenerational assistance, with children at the top as targets of aid.

Intergenerational assistance is conditional for most Americans. For most of the thousands of respondents in our studies, intergenerational obligations were not automatic, but were related to a number of interpersonal and intrapersonal factors. The interpersonal factors of relationship closeness, quality, and reciprocity have been mentioned, but people also considered resource availability and other demands on family members' resources.

Older people think differently than younger people do about intergenerational assistance. The most consistent demographic difference we found (more than gender, race, ethnicity, or social class) was age. Adults older than 60 differed from middle-aged and young adults in their beliefs across a broad spectrum of tasks.

What We Think We Know

Kinship definitions are important in making judgments about intergenerational aid. Social scientists have argued that kinship definitions have become more flexible. Instead of limiting family membership to individuals related by genetic and legal bonds (Schneider, 1980), postmodern families rely on mutual affection and shared interests to denote kinship (Scanzoni & Marsiglio, 1993). We found that divorce and remarriage or repartnering can change how individuals define kin membership through reduced contact, lower exchange of resources, emotional distance, and greater interpersonal hostility.

However, stepkin may become family members psychologically and emotionally (Schmeekle, Giarrusso, Feng, & Bengtson, 2005; Widmer, 2006). How steprelationships are defined is an important factor in understanding and predicting resource exchanges between older stepparents and adult stepchildren. Steprelationships characterized by past mutual exchanges of resources and strong emotional bonds generally were seen as indicators that stepkin were thought of as family. In those cases, family obligation norms applied as if there were genetic ties among them (Coleman et al., 2005; Ganong & Coleman, 1998a).

Gender may be important in predicting the ways family members support kin, but men and women hold similar beliefs about intergenerational assistance. We found few gender differences in our studies-men and women held similar attitudes and beliefs about intergenerational responsibilities.

Normative obligation beliefs are significantly related to personally held, felt obligations, and both types of obligation beliefs are significantly related to family behaviors. Beliefs about how family members should interact with each other are predictive of what individuals believe they should do in their own families.

What do we still need or want to know?

We need to understand better how kinship definitions function in making judgments about intergenerational assistance. We need to either confirm or disconfirm our model of the importance of kinship definitions by looking at more family contexts and by asking people more specifically about their kin-defining cognitions.

We need to know more about how individuals gain or lose family membership. For instance, how does a divorced parent maintain family membership standing? What is the process by which stepgrandparents become family? How do members of cohabiting households, adoptive parents and adopted children, and in-laws gain or lose family membership status?

We need to know more about the hierarchy of intergenerational assistance. In particular, where do stepchildren, cohabiting adults and their children, and others fit into a hierarchy of recipients of aid? Where do adopted children and in-laws fit in this hierarchy?

We need to know if there is a hierarchy of tasks or types of intergenerational transfers. Do individuals prioritize the ways in which they help older kin? Are the conditions for helping based partly on the type of aid needed?

We need to know how fault (blame) influences judgments about intergenerational assistance. If a family member is at fault for his or her problems, does that affect whether or not the person should be helped by other family members? Research on friends and strangers indicates that being at fault for problems affects judgments about helping-does this apply to helping family members as well?

We need to know more about how policies affect and are affected by normative beliefs about intergenerational obligations.

Finally, we need to know if there are cross-cultural differences in beliefs about intergenerational assistance and how those beliefs are affected by divorce and remarriage.

Although 20 years of studies examining normative beliefs about intergenerational responsibilities among family members after divorce or remarriage have resulted in some certainties and probabilities, many questions remain. Exploration of cross-cultural differences, kinship definitions, and influences on public policy, among other questions, await further research.

Editor's note: For information about the references, please contact one of the authors: ColemanMa@missouri.edu and ganongl@missouri.edu