Linked Lives and Cumulative Inequality: A Multigenerational Family Life Course Framework

Abstract

Growing social and economic inequalities in the United States have been accompanied by shifts in family structure. Social and demographic changes may make multigenerational family ties (e.g., between grandparents, parents, and children) an even more important contributor than ever in perpetuating inequality. The family life course framework, which considers how dimensions of individual (age), sociohistorical (period, cohort), and processual (generation) time intersect, provides a useful structure for thinking about how multigenerational families matter for inequality today. We integrate these dimensions of time with the ideas of linked lives and cumulative inequality to propose ways in which advantage and disadvantage are transmitted and grow across multiple generations. In particular, we propose an integrative model of linked lives and cumulative inequality that extends the family life course framework into a multigenerational perspective. We conclude by identifying data sources and methodologies useful for family scholars interested in pursuing a multigenerational family approach to inequality.

A central focus of family scholarship is the role that family plays in aspects of inequality. Across various social and economic dimensions, many societies—including the United States—are more unequal now than in the past (e.g., Duncan & Murnane, 2011; Meara, Richards, & Cutler, 2008; Piketty, 2014; Torche, 2015). This increasing social and economic inequality has been accompanied by changes in families. In particular, aspects of the second demographic transition—including postponement and abstention from marriage; increases in divorce, alternative union forms, and nonmarital childbearing; and decreases in remarriage and fertility (Lesthaeghe, 1995, 2010)—have altered the configuration of families. One of the most striking consequences of these changes is a shift in the structure of families from a pyramid-like structure with multiple individuals at the base to a longer, thinner structure that more closely resembles a beanpole (Bengtson, Rosenthal, & Burton, 1990). As a result, there has been a move from an emphasis on horizontal ties within generations (e.g., spouses, siblings) to vertical ties between generations (e.g., parents and children, grandparents and grandchildren). Further, because these changes in family behaviors have been accompanied by increases in life expectancy and health improvements (Arias, 2015), new opportunities for multigenerational relationships (e.g., those between noncontiguous generations) exist.

A large body of research shows that generations of families (generally parents and children) tend to resemble one another in terms of markers of advantage, especially socioeconomic status (e.g., Charles & Hurst, 2003; Torche, 2015). However, current economic, social, and demographic conditions may make various types of family connections more important contributors to the perpetuation of inequality now than ever (Swartz, 2008). For example, in recent decades, parental spending for children has increased in magnitude in real dollars and extends into (at least) the third decade of life (Kornrich & Furstenberg, 2013), which reflects an extended transition to adulthood (Swartz, 2008).

Further, multigenerational family ties may be particularly important. For example, because college costs have reached unprecedented levels (College Board, 2010), family financial resources are more influential in college graduation and student debt now than in the past. Because parent financial resources are a primary way that individuals pay for college, and parents may be better able to assist their children with college costs because their own parents helped them obtain a college degree, the behavior of these students’ grandparents in the past may have consequences for their grandchildren indirectly via parental education. As another example, as more aging adults are unmarried (Brown & Lin, 2012), their children—and grandchildren—may need to assume caregiving responsibilities formerly assumed by spouses (Dellmann-Jenkins, Blankemeyer, & Pinkard, 2000).

In addition, this growing diversity in family structure has the potential to increase inequality between families and to heighten the importance of various family connections in how advantage is transmitted across generations (Tach, 2015). For example, educational assortative mating in marriage (Schwartz & Mare, 2005) has increased over the past several decades. This means that the advantage of growing up with two married parents is more likely now than in the past to be consolidated with the advantage of having (two) highly educated parents. As such, the gap between the most advantaged children and the least advantaged children has grown. In addition, because multiple types of advantage and disadvantage tend to cluster together, material and nonmaterial resources and family structure may mutually reinforce each other to amplify the transmission of advantage and disadvantage (Hout, 2015). For example, single mothers spend less time with their children than married mothers as a result of differences in social structural factors such as employment and education (Kendig & Bianchi, 2008). Single mothers are also more likely than married mothers to experience financial insecurity throughout their lives (Johnson & Favreault, 2004). Further, a large literature documents that grandparents—especially grandmothers—play a critical role in many single-parent families (e.g., Bengtson, 2001; Hayslip & Kaminski, 2005; McLanahan & Sandefur, 1994). Because aspects of socioeconomic status and health are intertwined throughout the life course (e.g., Hayward & Gorman, 2004), single mothers are more likely than married mothers to experience poor health in later life, which may result in additional caregiving responsibilities for their children.

A multigenerational framework may be especially important for understanding Black–White racial disparities in a variety of outcomes, because of deeply rooted institutional racism in the United States, including exclusion of African Americans from mainstream credit markets (Conley, 1999; Oliver & Shapiro, 2006), school segregation (Reardon & Owens, 2014), and unethical medical practices such as the Tuskegee Study (Roberts, 2014; Washington, 2006). African Americans may continue to be affected by the experiences of their parents, grandparents, and more distant ancestors through multigenerational family processes. For example, African Americans have experienced economic disenfranchisement tied to government policies such as Social Security and the Servicemen’s Readjustment Act, as well as the bank lending practices of redlining and housing covenants (Coates, 2014; Conley, 1999; Oliver & Shapiro, 2006). These policies have left generations of African American parents with lower levels of wealth, leaving them less able to assist their children financially (Charles & Hurst, 2002). Because family financial resources are critical for many individuals who are buying their own homes (Charles & Hurst, 2002), previous racist economic policies may continue to have an impact on African Americans today via multigenerational family processes related to family financial support (Charles, Roscigno, & Torres, 2007).

In this article, we integrate the perspectives of linked lives and cumulative inequality into a multigenerational family life course framework. Linked lives and cumulative inequality are useful theoretical concepts for understanding the role of families in inequality. However, these concepts are also limited in that they typically focus on the relationships between one generation of parents and their children. As a result, these concepts may fail to consider how advantage or disadvantage in one generation may affect other generations beyond the parent–child tie. Just as early-life conditions have implications for later-life outcomes, the experiences of earlier generations may have consequences for later generations, and vice versa. We advocate for the integration of cumulative inequality and linked lives concepts into considerations of how families reproduce and amplify inequality across generations. Families transmit socioeconomic status—and other aspects of advantage—through a variety of pathways. The fact that family members’ lives are linked means that transmission is facilitated by the existence of spillover effects of various types of resources (e.g., money, time, knowledge, social networks) within multiple generations of the same family. We propose an integrative model of linked lives and cumulative inequality that extends the family life course into a multigenerational perspective. Specifically, we discuss how the family life course framework’s expanded concept of time—one that incorporates multiple family generations—facilitates a more comprehensive examination of inequality.

We begin by reviewing the family life course framework and the dimensions of time considered by family scholars: age, period, cohort, and generation. We then review the theoretical and empirical literatures on linked lives and cumulative inequality and discuss how these two concepts could be integrated to examine the transmission of advantage and disadvantage across multiple generations. We conclude by identifying data sources and methodologies useful for family scholars interested in pursuing a multigenerational approach to inequality.

Time and the Family Life Course Framework: Age, Period, Cohort, and Generation

In the early 1990s, scholars integrated the concepts of the individual life course and family development into the family life course framework (e.g., Aldous, 1990; Bengtson & Allen, 1993). This framework draws from both sociological theories of social change and psychological theories of individual and family development and highlights the importance of historical and social contexts and individual time and development (Elder, 1995; Settersten, 2003, 2005). This framework draws on the life course perspective, which has often been utilized to examine individuals as they age (Elder, 1994; Elder, Johnson, & Crosnoe, 2003). However, from its inception, the life course perspective has been embedded in the family context and in the analysis of social inequality. Indeed, Elder’s (1974) prominent work Children of the Great Depression: Social Change in Life Experience described how economic downturn differentially affected children’s outcomes depending on the family context in which they were situated (e.g., parental financial resources) and their relative position within the family (e.g., birth order, gender). The family life course framework articulates how the lives of individual family members are interconnected and how families transmit advantage.

Time is central to the family life course framework and involves concepts from the life course perspective (age, period, and cohort; Elder, 1994; O’Rand, 1996) as well as the family development perspective (processual time, often measured as “generation”; Rodgers, 1973). Time operates not only at the individual and sociohistorical level (Elder, 1975, 1994; Riley, 1973; Rodgers, 1973) but also at the family level (Aldous, 1990; Rodgers, 1973). At the individual level, time is usually conceptualized and operationalized as chronological age to capture biological, psychological, and social changes that are age graded (Baltes, Lindenberger, & Staudinger, 2006). The life course perspective acknowledges that development is a lifelong process and that all developmental stages should be considered (i.e., infancy through adulthood). The social stratification literature recognizes the centrality of age-related processes in inequality through its focus on the social origins of opportunity, which are rooted in early life (Hout, 2015).

The life course concept of sociohistorical time acknowledges that individuals are situated within particular social and historical contexts, with varying geographical, political, and economic environments. To understand the influence of these contexts, sociohistorical time is usually conceptualized as cohort and period effects. The term cohort extends the life course concept of age by grouping individuals together on the basis of the historical period in which they were born or experience a certain major life event (e.g., college matriculation) (Ryder, 1965). Period effects refer to historical events that have wide-reaching impacts across individuals (e.g., the Great Recession). Social stratification scholars also have acknowledged the importance of the intersection of period and cohort. For Hout (2015), “Life chances can even depend on the year in which a person was born” (p. 31). For example, changes in the Vietnam War draft policy in the early 1970s meant that men born after certain dates would never be drafted, which potentially had consequences for their subsequent earnings (Angrist, 1990).

Just as time progresses at the individual and sociohistorical levels in terms of age, period, and cohort, it also moves forward at the family level in terms of processual time (Aldous, 1990; Rodgers, 1973). Processual time captures family members’ relational roles and positions in terms of the family of origin and the family of procreation. In particular, processual time captures not only progression across roles (e.g., from being childless to being a parent) but also stages within a particular role (e.g., being the parent of an infant versus of a middle-aged adult). Further, the birth of a first child has implications for the new parent and also marks the beginning of a new role for the older family member as a grandparent. In this way, processual time captures how a biological event (e.g., the birth of a child) has effects that reverberate beyond the family of procreation through the creation of extended family ties.

Processual time is frequently denoted in the notion of generation. Bengtson (1993) defines generation as the “rank-descent ordering of individuals within families” (p. 11). The concept of generation is typically captured by terms such as grandparent, parent, child, and grandchild. Further, “movement into these lineage social positions” is related to—but far from perfectly correlated with—individual time” (Bengtson & Allen, 1993, p. 481). For example, a woman may become a grandmother at age 30 or at age 60 (Burton & Bengtson, 1985). Sometimes the terms generation and cohort are used interchangeably to denote cohort effects (as in the discussion of generational differences between baby boomers and millennials). However, it is important to retain a definition of generation that explicitly references family position; as we show below, the implications of occupying a certain generational position are independent of other dimensions of time such as age.

The term generation refers to an individual’s position in their family relative to both ancestors and descendants (i.e., grandparent, parent, and child). Scholars who examine intergenerational family relations commonly refer to the oldest living generation as G1, the middle generation as G2, the next generation as G3, and so on (e.g., Birditt, Tighe, Fingerman, & Zarit, 2012). The concept of generation acknowledges that the interactions between family members are largely affected by generational position (Bengtson & Allen, 1993). Just as individuals pass through different stages of development, they also pass through different generational positions as older generations pass away and/or new generations are born. In contrast to the individual life course, in which the individual can occupy only one developmental stage at a given moment, within the family life course, individuals can occupy multiple generational positions simultaneously. For example, a woman in her 30s may be a mother (of a young child), an adult child (of a parent in their 60s), and an adult grandchild (of a grandparent in their 90s).

Among the most prominent examples of the utility of the concept of generation as distinct from age or cohort is the intergenerational stake hypothesis (Bengtson & Kuypers, 1971). According to this hypothesis, individuals in higher generational positions (i.e., parents or grandparents) are more invested in their relationships with successive generations (i.e., children or grandchildren) than younger generations are in their relationships with older generations. Bengtson and Kuypers’ (1971) rationale for this hypothesis is that older generations are more invested in—and have more favorable perceptions of—relationships with younger generations because older generations consider younger generations their legacy. Younger generations are less invested in relationships with older generations because they are seeking a sense of autonomy and independence. The intergenerational stake hypothesis has been tested in a variety of two-generational models including parents and their children and grandparents and grandchildren (Giarrusso, Feng, & Bengtson, 2004). Recent work by Birditt and colleagues tested this hypothesis in a three-generational model and found that individuals reported more closeness and less conflict with their children than they did with their parents (Birditt, Hartnett, Fingerman, Zarit, & Antonucci, 2015). These models have been tested on individuals in a variety of age ranges, which suggests that these findings more likely reflect generational position than chronological age (Giarrusso et al., 2004).

The intergenerational stake hypothesis predicts that older generations will transfer more resources to younger generations than younger generations transfer to them, which may result in increasing inequality across generations. For example, emerging research suggests that in today’s high-cost college environment, parents may be borrowing from their own retirement accounts or taking out additional loans to finance their children’s college education (Walsemann & Ailshire, 2016). This behavior may put parents’ own financial futures at risk by leaving them less prepared for their own retirement; although one can acquire loans to pay for college, there are fewer options to finance retirement. Further, a potential unintended consequence of parents’ financing of their children’s educations is that parents may need to rely more on their own children in later life because of their diminished financial resources.

Fingerman’s concept of developmental schisms provides a framework for understanding how discrepancies in ages between the generations have an impact on intergenerational relationships (Fingerman, 1996). In particular, each developmental stage has specific tasks associated with it, and these tasks have implications for intergenerational interactions. Developmental schisms occur when generations are competing for the same resources (e.g., time, finances, social support). For example, in later stages of the life course, parents may experience physical and cognitive declines that result in their becoming more reliant on their adult children. As a result, these adult children often find themselves juggling care of their parents with their other work and family responsibilities. According to Fingerman and colleagues, differences in developmental tasks between the generations can often be a source of intergenerational tension (Birditt, Miller, Fingerman, & Lefkowitz, 2009; Fingerman, Miller, & Seidel, 2009).

The idea of developmental schisms also helps us to understand how inequality may accumulate across generations. Families in which older generations have fewer financial resources may be less likely to provide financial and instrumental support for younger generations. In addition, older generations in these families may be more likely to need support from younger generations because neither generation has the financial resources to pay for alternative care. The provision of this additional care can have financial consequences for the younger generation by putting its members in vulnerable positions as they enter later life themselves. As such, financial inequality may result in greater developmental schisms in disadvantaged families (Lee, Tang, Kim, & Albert, 2015).

Integrating these various conceptions of time—age, period, cohort, and generation—facilitates an understanding of the role that the family plays in contemporary inequality. For example, period effects related to the Great Recession affected not only individuals but also family processes tied to coresidence. One way that the cohort of recent college graduates responded to a lack of employment opportunities was by continuing to live with parents or even returning to the parental home (i.e., “boomerang kids”) (Fry, 2013; Newman, 2012). Further, there was also an increase in multigenerational coresidence (e.g., grandparents living with grandchildren—either with or without parents present), especially among African American families (Fry, 2013). These shifts in family coresidence demonstrate how multigenerational family ties may operate as an informal safety net in response to economic hardship (Newman, 2012). However, the provision of this safety net is not without potential costs. For example, prior research shows that grandmothers caring for grandchildren are more likely to live in poverty (Fry, 2013). This may reflect both causal processes related to the increased economic strain of supporting additional family members and structural factors related to previous economic vulnerability, both of which increase the likelihood that grandmothers may care for grandchildren.

Linked Lives

The perspective of linked lives emphasizes the generational dimension of time in recognizing that an individual’s life is embedded within the lives of their family members, including those from other generations (Elder, 1994; Macmillan & Copher, 2005). Because of the linked, or interdependent, lives of individuals within a family system, the events, trajectories, or transitions occurring within one family member’s life may have reverberating effects on the lives of the other members (Bengtson, Elder, & Putney, 2005). Children’s, parents’, and grandparents’ lives are intricately and dependently intertwined across each generation’s life course. Researchers using a linked lives perspective frequently identify an event within the family (e.g., parental job loss) and the immediate outcome it has on another family member (e.g., child’s behavioral outcome). This generational approach has been critical in understanding the connections between generational dyads. For example, the Family Transitions Project (formerly known as the Iowa Youth and Families Project) demonstrated how the Farm Crisis of the 1980s had an impact not only on the stress levels of parents but also on the quality of family relationships between parents and children (Conger & Elder, 1994). Further, longitudinal examination of these families has shown that this early-life economic downturn had long-term impacts for children’s outcomes (Conger & Conger, 2002; Conger, Conger, & Martin, 2010).

Family scholars studying young families often use the concept of linked lives to understand the impact of a parent’s life event on their children’s internalizing and externalizing behaviors (e.g., Dearing, McCartney, & Taylor, 2006; Yuan, 2008). Further, life events can damage the parent–child relationship, potentially having long-term developmental implications for children (Ge et al., 1992). Thus, children’s well-being is closely intertwined with the lives of their parents such that parental events can have serious, detrimental effects on children. It is not necessarily the event that directly affects the child but rather the event’s impact on the parent (Elder, van Nguyen, & Caspi, 1985; Ge et al., 1992; Thornberry, Freeman-Gallant, Lizotte, Krohn, & Smith, 2003; Yuan, 2008). Strain or stress affects the parent’s own well-being, as well as their parenting, which in turn disrupts the child’s development. In other words, it is parental stress that mediates the relationship between an adverse event and children’s outcomes (Yuan, 2008).

Researchers are beginning to recognize that the reciprocal associations between parents and children have implications for one another’s advantage and disadvantage across the life course (Polenick, DePasquale, Eggebeen, Zarit, & Fingerman, 2016; Umberson, Pudrovska, & Reczek, 2010). Just as parents have an impact on children’s socioeconomic status, children can also have an impact on their parents’ socioeconomic status. For example, a recent study (Houle & Berger, 2017) found that parents who have a child with an early-life disabling health condition experience greater debt burden in the years following the child’s birth than do parents of children without these conditions. In addition, similar to the literature that has demonstrated the deleterious effects of parents’ economic distress on children’s outcomes in early life (Conger, Ge, Elder, Lorenz, & Simons, 1994; Elder, Conger, Foster, & Ardelt, 1992), Gilligan and colleagues found that adult children’s employment problems during the Great Recession had negative consequences for relationships with their older mothers (Gilligan, Suitor, & Pillemer, 2013). Further, additional work has shown that problems in adult children’s lives have implications for older parents’ psychological well-being (Fingerman, Cheng, Birditt, & Zarit, 2012). Filial and parental obligations may change as children and parents age. For example, many adult children continue to receive a variety of types of support from their parents well into midlife (Suitor, Gilligan, & Pillemer, 2015), whereas in the later life stages, adult children provide more financial, instrumental, and emotional support to their aging parents (Suitor et al., 2015). Changes in support dynamics can strain both the parent–child relationship and the well-being of each generation.

Younger generations’ characteristics can also have a positive impact on older generations. For example, adult children’s education has an inverse association with parental mortality (Friedman & Mare, 2014), net of parents’ own socioeconomic characteristics. Further, the relationship between children’s education and parental mortality is stronger for causes of death in which behavioral factors play a larger role, which highlights the importance of education as a family-level resource. Taken together, the literature has shown that members of both generations can have positive and negative impacts on one another throughout life.

Figure 1 depicts linked lives. Specifically, it shows the bidirectional effects of parent and child generations on one another. It also emphasizes the generational dimension of time as expressed by the vertical positioning of the older generation (parent) above younger generation (child).

Figure 1.

Linked Lives.

Cumulative Inequality

Explorations of cumulative inequality typically consider how resources accumulate across time (usually measured by age) within an individual’s life course. This is in contrast to the concept of linked lives, which in general has focused on how the experiences of family members—including those of different generations—influence one another at one point in sociohistorical time. The concept of cumulative inequality (Ferraro & Shippee, 2009), and the related notion of cumulative (dis)advantage (Crystal & Shea, 1990; Dannefer, 2003; DiPrete & Eirich, 2006), stems from Robert Merton’s (1968) description of how small initial advantages in scholarly productivity beget increased reputation and resources, which in turn beget even greater productivity, reputation, and resources across a scholar’s career. This phenomenon has also been referred to as the “Matthew effect” for its similarity to the parable of the talents found in the Gospel of Matthew: “For unto every one that hath shall be given, and he shall have abundance: but from him that hath not shall be taken away even that which he hath” (Matthew 25:29, King James Version). Thus, cumulative inequality describes the phenomenon by which inequalities grow over time as initial advantages beget additional advantages. A key facet of the process of cumulative inequality is that early events set in motion trajectories that influence opportunities for future advancement. This accumulation of advantages over time also parallels the idea of compound interest, in which interest is applied to financial principal over time, yielding exponential returns as accrued interest becomes part of the sum to which interest is applied.

Within the family context, scholars in a variety of disciplines have applied cumulative inequality theory to describe how early events in an individuals’ life—even in utero—may have long-lasting effects on individual outcomes by setting individuals on paths of either accumulating advantage or disadvantage (e.g., Barker, 1995; Crystal & Shea, 1990; Dannefer, 2003; DiPrete & Eirich, 2006; Ferraro & Shippee, 2009). The focus on the early life of an individual is consistent with the emphasis within cumulative inequality on the importance of the timing of advantages (O’Rand, 1996). Families of origin are critical because they structure children’s access to resources as well as their exposure to stressors, which affect children’s trajectories across various domains. In particular, a large body of research documents how socioeconomic adversity early in life (e.g., limited parental financial resources) plays an important role in health and mortality (e.g., Hayward & Gorman, 2004; Pudrovska, 2014). Hayward and Gorman (2004) have stated that early-life events are important because “childhood conditions potentially set in motion a range of cascading events (e.g., lifestyle preferences, exposure to adverse work conditions, and income and wealth trajectories) that can have more temporally proximate effects on mortality” (p. 88).

Figure 2 depicts cumulative inequality and shows that factors related to the family context in childhood have impacts that may endure into later life. Cumulative inequality emphasizes the chronological dimension of time as indicated by the horizontal positioning of the individual at an earlier point (e.g., childhood) shown on the far left side and the individual’s later-life outcomes (e.g., adulthood) shown on the far right side.

Figure 2.

Cumulative Inequality.

Integrating Linked Lives and Cumulative Inequality in a Multigenerational Family Framework

The concepts of linked lives and cumulative inequality emphasize different dimensions of time. Linked lives stresses the importance of connections between individuals in different generations, whereas cumulative inequality stresses how earlier experiences affect individuals as they age. Just as the fates of parents and children are connected across the life course, so, too, may be those of more distant generations across processual time (McGoldrick, Gerson, & Petry, 2008; McGoldrick, Gerson, & Shellenberger, 1999). The concepts of both linked lives and cumulative inequality emphasize relationships between parents and children; however, neither of these perspectives generally considers the influence of more distant generations (e.g., grandparents and grandchildren). Therefore, as have other scholars (e.g., Bengtson, 2001; Mare, 2011, 2014; Pfeffer, 2014), we call for researchers to expand the family life course model to a multigenerational framework.

Extending the idea of linked lives beyond a two-generation paradigm into a multigenerational framework is important because the ties between two adjacent generations (G1–G2) have potential implications for the ties between the next two adjacent generations (G2–G3). The notion of cumulative inequality can likewise be expanded to incorporate multiple generations. Just as previous experiences set individuals on pathways that generate further advantage and disadvantage across the individual life course, advantage (or disadvantage) may be amplified across generations. Not only do generations affect one another; advantage or disadvantage may grow across generations (McGoldrick et al., 1999; McGoldrick et al., 2008). For example, in the case of parental financial support, a three-generation model is critical because what parents transmit to their own children may be affected by what parents received from their own parents (Steelman & Powell, 1991). If G1 parents help G2 children pay for college, then G2 children are more likely to graduate from college and have better-paying jobs, which leaves the G2 generation better positioned to financially assist their own children (G3) with paying for college.

An examination of three-generation (or more) models can contribute to our understanding of how families transmit values and behaviors (Blau & Duncan, 1967; Bowen, 1978; Lareau, 1993; McGoldrick et al., 1999; McGoldrick et al., 2008; Schoeni, 1997). Returning to the example of parental financial support for a college education enables a more complete investigation of the intergenerational transmission of the importance of a college education. The symbolic meaning of this support may be important for a variety of outcomes beyond absolute levels of financial support. For example, when children grow up in families in which there is a history of parental financial support for college across prior generations, the knowledge of this legacy of support may promote college aspirations independent of the magnitude of the financial transfer itself. Further, receipt of support may make parents more inclined to provide support to their own children in order to be consistent with family norms regarding parenting behaviors.

A consideration of multiple dimensions of time may help us understand the accumulation of inequality within a multigenerational family context. First, sociohistorical contexts when one generation is at a particular age (e.g., growing up in the Great Depression or Great Recession) may influence not only that generation’s later lives but also the lives of subsequent generations. For example, current cohorts of college students are graduating with unprecedented debt (Houle, 2014). This debt has implications for their family formation and future parenting decisions (Addo, 2014). In particular, these young adults may find themselves pulled in two directions by the competing demands of paying off their own student loans and saving for their own retirement and their children’s education. Although considerable attention has been paid to rising college costs for current and future cohorts of young adults, far less discussion has considered how college costs today may have an impact on the children and grandchildren of recent graduates.

These examples explicate how G1 influences G3 via G2. However, G1 may also have a direct effect on G3, for example, if grandparents pay directly for grandchildren’s college. It is also important to note that G1 does not necessarily have to be alive to have an impact on G3. Bequests are perhaps the most striking example of grandparents having an impact on grandchildren after grandparents are deceased (McGarry, 2001).

Recent work also considers how family structure modifies the multigenerational transmission of advantage. Using data from three generations of the Panel Study of Income Dynamics, Song (2016) found that the direct effect of grandparents’ educational attainment on grandchildren’s educational attainment (i.e., independent of parent’s educational attainment) is stronger among grandchildren who grew up in two-parent families. The author reports, however, that the family structure of parents (the middle generation) does not modify the grandparent effect. Heterogeneity in the multigenerational transmission of advantage by family structure—and stronger transmission of status among two-parent families specifically—may reflect the fact that the lives of multiple generations are more tightly linked in two-parent families because, in part, family structure influences the closeness of ties between grandparents and grandchildren.

In addition, generations have reciprocal impacts on each other, and these impacts have lasting effects across the family life course. These reciprocal effects can also operate in a multigenerational context. A phenomenon that has gained attention in recent years is skipped-generation families, in which grandparents raise their grandchildren (e.g., Mills, Gomez-Smith, & De Leon, 2005). Often grandparents in these families reside with and provide extensive support to grandchildren because parents are absent. In the case of skipped-generation families, this is evidenced by research showing the impacts of the provision of care for grandparents themselves. Although work has found that, overall, grandparenting does not have negative consequences on grandparents’ health and well-being, there is evidence that grandmothers in skipped generation families who live with their grandchildren experience declines in physical and psychological health (Hughes, Waite, LaPierre, & Luo, 2007). Further, African American women are more likely to provide care for grandchildren (Caputo, 2000), which highlights that the likelihood of exposure to these reciprocal family dynamics intersects with social structural characteristics including gender and race (Collins, 1990).

Figure 3 depicts our proposed three-generation model that demonstrates the bidirectional effects that multiple generations can have on one another. First, grandparents and grandchildren can affect each other indirectly via parents. Second, there can be direct impacts between grandparents and grandchildren. Figure 3 also combines linked lives’ time dimension of generation with cumulative inequality’s time dimension of age. Extending the family life course beyond two-generation parent–child paradigms will facilitate the understanding of how families may consolidate advantage and perpetuate inequality across multiple generations.

Figure 3.

Linked Lives and Cumulative Inequality in a Multigenerational Perspective.

Data and Methodological Approaches to Studying Multigenerational Families

Recent developments in data and methodology offer promising new directions for scholars interested in exploring aspects of inequality within a multigenerational family life course framework. Just as longitudinal data has enabled scholars to study the individual life course (Elder et al., 2003), several data sources facilitate the examination of the family life course across generations. In particular, a growing number of data sets contain independent reports from three (or more) generations of family members. As such, scholars have begun to examine how continuity and discontinuity of family characteristics and behaviors across generations may lead to diverging trajectories between families.

For example, the Panel Study of Income Dynamics (PSID), begun in 1968, is the world’s longest, nationally representative household panel survey. As of the 2013 wave of data, the PSID contains more than 1,300 grandparent (G1), parent (G2), and child (G3) triads. The Longitudinal Study of Generations (LSOG), which began in 1971, contains information on approximately 300 three-generation families. Since 1989, the Family Transitions Project (FTP) has collected data from more than 500 individuals, their parents, and their oldest child. The PSID, LSOG, and FTP are especially valuable data sources because in addition to their inclusion of multiple generations of family members, they have also followed these families over several decades. As a result, these data can facilitate the examination of the family life course across multiple developmental stages of individual family members within particular sociohistorical contexts. More recently begun, the Family Exchanges Study (FES) interviewed more than 600 three-generation families between 2008 and 2013. Taken together, the PSID, LSOG, FTP, and FES have facilitated the examination of a wide range of multigenerational topics, such as class mobility (Hertel & Groh-Samberg, 2014), the transmission of health and health behaviors (Davis, McGonagle, Schoeni, & Stafford, 2008), intergenerational relationship quality (Birditt et al., 2012), the transmission of beliefs and values (Bengtson, Copen, Putney, & Silverstein, 2009), and the role of genetic, environmental, and epigenetic factors (Conger, Belsky, & Capaldi, 2009).

These data sets have begun to show that examining only parents and children may lead to underestimating the extent of social inequality—as well as the family’s role in its transmission. This is because more distant generations have direct impacts on one another and do not just operate through intermediate generations (e.g., grandparents have impacts on grandchildren beyond the impact of grandparents on parents) (Mare, 2011, 2014). As multigenerational data sets become more readily available, scholars should continue to explore how families reproduce and amplify inequality across generations in changing political, economic, and social environments.

Further, methodological developments that facilitate the analysis of multiple family members enable scholars to address research questions involving multiple generations. Multilevel models have grown in prominence in the past 10 years, allowing for scholars to account for multiple siblings and cousins nested within families (Teachman & Crowder, 2002). Further, actor–partner interaction models (APIM) enable scholars to account for the reciprocal effects of family members on one another’s outcomes. Typically, APIM models have been used to examine how parents and children affect each other’s well-being (Fingerman, Pitzer, Lefkowitz, Birditt, & Mroczek, 2008; Polenick et al., 2016). We propose that these models can be extended to examine how grandparents and grandchildren influence one another.

In addition to the quantitative methodologies mentioned above, scholars should also utilize qualitative data to examine advantage and disadvantage across multiple generations. Combining quantitative and qualitative data has become increasingly common in research on families across the past decade (Creswell & Plano-Clark, 2010; Neal, Hammer, & Morgan, 2006; Plano-Clark, Huddleston-Casas, Churchill, Green, & Garrett, 2008). Although quantitative analyses can identify patterns of relationships among variables, such analyses are less fruitful for pursuing the processes underlying statistical relationships (Morgan, 2007; Neal et al., 2006; Umberson & Montez, 2010). Further, scholars should consider collecting qualitative data from members of more than one generation to capture multiple family members’ experiences (Reczek, 2014). These data sources and methodological approaches provide promising opportunities to extend the family life course framework to consider how the linked lives of multiple generations are a mechanism through which inequality accumulates.