Marriage, Cohabitation, and Divorce in Later Life

Older adults are at the forefront of family change as a declining share experiences lifelong marriage and rates of cohabitation and divorce in later life continue to rise. The goal of this article is to review recent scholarship on marriage, cohabitation, and divorce among older adults and identify directions for future research. The varied family experiences characterizing the later life course demonstrate the importance of moving beyond marital status to capture additional dimensions of the marital biography, including transitions, timing, duration, and sequencing. Cohabitation operates as an alternative to marriage for older adults and is increasingly replacing remarriage following divorce or widowhood. The gray divorce rate has doubled in recent decades as older adults abandon marriage in favor of unmarried partnerships or singlehood. The retreat from marriage among older adults raises important questions about the ramifications of family change for health and well-being as well as access to caregivers given that spouses historically have been the primary source of care.

Our review indicates that a growing segment of older adults may be at risk for poorer health outcomes and at the same time have fewer informal sources of support, necessitating additional institutional mechanisms for ensuring the health and well-being of today’s older population.

U.S. family life is characterized by marked demographic change. Recent decades have witnessed a retreat from marriage, sustained high levels of divorce, and a rapid acceleration in unmarried cohabitation ( Cherlin, 2010; Kennedy & Ruggles, 2014). Older adults have not been immune to family change. In fact, some of the most dramatic shifts in family life are occurring among adults aged 50 years and older ( Cooney & Dunne, 2001). Today’s baby boomers (born 1946–1964), for example, were the generation that as young adults popularized premarital cohabitation and experienced the divorce revolution. Now entering older adulthood, boomers remain at the vanguard of family change, eschewing marriage, and embracing unmarried partnerships such as cohabitation. They are also driving the gray divorce revolution, which is largely a reverberation of the initial run-up in divorce decades ago. Many of the boomers who first divorced as young adults got remarried and are divorcing yet again ( Brown & Lin, 2012).

Our goal is to review the recent literature on older adult (which we define as aged 50 years and older) marriage, cohabitation, and divorce. Later life couplehood is no longer confined to the boundaries of marriage. Older adults are taking advantage of the flexibility afforded by unmarried partnerships, including cohabitation ( Calasanti & Kiecolt, 2007). A growing share does not seem to feel compelled to remain coupled. Long-term marriages are increasingly ending through divorce and most individuals who call it quits are not repartnering ( Brown, Lin, Hammersmith, & Wright, 2016). Using Census data, we establish how levels of marriage, cohabitation, and divorce have shifted over time among older adults, documenting the drops in marriage and widowhood and the increases in divorce and cohabitation for men and women. Additionally, we outline the theoretical and conceptual explanations for these current patterns and consider the ramifications of these changes for individual health and well-being. There are many well-established explanations for the benefits associated with marriage, such as the resource perspective, crisis perspective, and cumulative disadvantage theory, but theorizing on cohabitation and divorce in later life is limited. Finally, we conclude with a discussion of directions for future theoretical and empirical research on family change in later life.

Marriage

Marital Status

The proportion of U.S. adults who are currently married is at a historic low ( Cherlin, 2010), and this retreat from marriage is apparent among older adults. Over the past quarter century, the share of men ages 50 years and older who are married has declined from 78% in 1990 to 67.3% in 2015 (see Table 1) (data from 1990 decennial census and 2015 American Community Survey). For older women, the percentage married has stagnated, hovering at 52.6% in 1990 and 52.7% in 2015. This stability reflects a corresponding decline in widowhood as women’s husbands are living longer these days. Widowhood fell slightly among men from 7.5% in 1990 to 5.7% in 2015. The decline was sharper for women, whose levels of widowhood plummeted from 31.6% to 18.9%. Meanwhile, being divorced is now more prevalent among both men and women. In 1990, 8.1% of men and 10.1% of women were divorced. In 2015, figures stood at 14.3% for men and 18.1% for women. Likewise, the shares of never-married and cohabiting older adults have risen over the past 25 years. Among men, 5% were never-married in 1990 versus 9.1% in 2015. For women, the growth in the never-married has been more modest, increasing from 4.9% in 1990 to 7.7% in 2015. Cohabitation levels more than doubled among men from 1.5% to 3.6% and from less than 1% to 2.6% between 1990 and 2015 for women. Underscoring the growing diversity of marital statuses in later life, these patterns signal that traditional lifelong marriage that eventuates in spousal loss is decreasingly characteristic of the older adult family life course.

Percentage Distribution of Marital Status for Men and Women, 1990 and 2015

Note: The figures for 1990 come from the decennial census data and the 2015 figures are from the American Community Survey. Calculations by the authors.

Percentage Distribution of Marital Status for Men and Women, 1990 and 2015

Note: The figures for 1990 come from the decennial census data and the 2015 figures are from the American Community Survey. Calculations by the authors.

This declining prevalence of marriage during the second half of life is driven largely by the baby boomer generation. As of 2015, more than one in three boomers (37%) was unmarried (authors’ calculation using the 2015 American Community Survey). This share will grow in the coming years as more boomers experience marital dissolution through either gray divorce or widowhood and do not subsequently remarry. Unmarried boomers are disadvantaged compared with married boomers. Despite similar education levels, poverty is four times higher among unmarried than married boomers, and disability is twice as high ( Lin & Brown, 2012). Economic disadvantage combined with potentially fewer sources of social support leave unmarried older adults particularly vulnerable in the event of a health crisis ( Zhang, Liu, & Yu, 2016). With fewer boomers having a spouse, it is unclear who will provide care to them as they experience health declines.

The Marital Biography

As the family life course experiences of older adults become more varied, it is important to move beyond current marital status. Other factors, such as prior marital transitions, their timing, the duration of time spent in particular marital statuses, and the sequencing of these transitions combine to shape health and well-being ( Cooney & Dunne 2001; Hughes & Waite, 2009; Reczek, Pudrovska, Carr, Thomeer, & Umberson, 2016; Zhang et al., 2016). Marital dissolution, for instance, has enduring negative effects on individual health in midlife, even for those who remarry. Remarriage offsets only some of the health disadvantage linked to marital disruption, whether through divorce or spousal loss. Time spent in either the divorced or widowed state is related to worse health outcomes, including chronic conditions and mobility limitations ( Hughes & Waite, 2009), although not to cardiovascular disease ( Zhang & Hayward, 2006). Still, marital disruption itself is associated positively with cardiovascular disease ( Zhang & Hayward, 2006). Timing also matters: the detrimental health outcomes associated with divorce attenuate whereas the negative effects of widowhood intensify with age for women ( Liu, 2012). Consistent with the cumulative disadvantage perspective, dissolutions appear to have additive negative effects on health, as individuals who experience two divorces fare worse, on average, than those who only divorce once ( Dupre, Beck, & Meadows, 2009; Zhang, 2006). Moreover, the negative health effects of divorce are not necessarily immediately apparent and can emerge years later ( Hughes & Waite, 2009), reinforcing the stress model perspective that stipulates marital dissolution is a stressful life event that often involves enduring, chronic strains which take a toll on health ( Zhang et al., 2016).

Marriage and Well-being

In fact, researchers have challenged the conventional finding that marriage is advantageous for well-being, arguing instead that the apparent gains to marriage are actually due to the detrimental influences of disruption on health ( Williams & Umberson, 2004). Ultimately, the health benefits of marriage are less apparent today than a generation ago. The gap between the married and never-married has shrunk for men and the negative health outcomes associated with marital disruption are more severe, particularly among women ( Liu & Umberson, 2008). These findings challenge the marital resources model which stipulates that marriage provides spouses with psychological, economic, and social benefits that should enhance well-being ( Zhang et al., 2016) and longevity ( Dupre et al., 2009). Likewise, there are notable differentials by gender and race in the benefits of marriage with men and Whites typically enjoying more advantages than women and non-Whites, although the gender differential may be attenuating ( Carr & Springer, 2010). One way to shed new light on gender dynamics and marriage is to consider same-sex couples ( Umberson & Kroeger, 2016). Now that same-sex marriage is legal across the United States, researchers can investigate whether same-sex couples realize benefits from marriage akin to different-sex couples.

Marital benefits are contingent on marital quality with the greatest gains accruing to those with the happiest marriages. In poorer quality marriages, the health benefits are often negligible or even negative compared to the alternative of getting divorced ( Zhang et al., 2016). Marital strain exacerbates the decline in self-rated health that typically occurs over time, and this effect is larger at older ages ( Umberson, Williams, Powers, Liu, & Needham, 2006). For individuals with a disability or functional limitations, a high quality marriage helps to minimize the psychological burdens related to quality of life whereas a low quality marriage diminishes mental health and quality of life ( Bookwala, 2011). The association between marital quality and health increases with age and the linkage is more pronounced for women than men ( Liu & Waite, 2014).

Cohabitation

The Growth in Cohabitation

Cohabitation is now growing more rapidly among older than younger adults. As depicted in Figure 1, the number of cohabitors aged 50 years and older has more than quadrupled since 2000, rising from roughly 951,000 to over 4 million in 2016. In the past decade alone, the number of individuals aged 50 years and older who were cohabiting surged 85% from 2.3 to 4 million ( Stepler, 2017b). One reason for the rise of cohabitation in later life is because fewer older adults are married, meaning a larger share is eligible to cohabit. Several demographic trends have contributed to growth in unmarried older adults. First, there has been a slight increase in people who never marry, especially for men ( Lin & Brown, 2012). Second, the rise in gray divorce (i.e., among those aged 50 years and older) results in newly single individuals who increasingly form cohabiting unions rather than remarriages ( Brown et al., 2016). Third, remarriage rates have declined 60% in recent decades and have stalled among older adults ( Brown & Lin, 2013; Sweeney, 2010). Together, these factors signal an increase in the number of adults who could cohabit. But the older adult cohabitation rate also has risen. Since 2000, the share of unmarried adults who are cohabiting has doubled from 7% to 14% (authors’ calculations using the 2000–2016 Current Population Survey). Cohort replacement has contributed to a rise in favorable attitudes towards cohabitation among older adults. Baby boomers are especially likely to be supportive of cohabitation compared with older cohorts ( Brown & Wright, 2016).

years and older, 2000–2016." />

Number of cohabiting individuals aged 50 years and older, 2000–2016.

Why Cohabit?

The rising popularity of older adult cohabitation was first documented more than two decades ago ( Chevan, 1996; Hatch, 1995). This early research articulated numerous economic and social benefits of cohabitation in later life. The economies of scale traditionally confined to marriage also can be achieved through cohabitation and without the legal obligations marriage involves. Couples can live together in a close, intimate partnership and pool their resources to the extent that it works for them. By remaining unmarried, they are not legally responsible for the partner’s medical expenses nor do the partners have any claims to each other’s assets. Cohabitation enables couples to preserve their financial autonomy, ensuring their wealth transfers to their offspring rather than their partner. Likewise, unmarried couples can continue to receive Social Security and pension benefits that may terminate upon remarriage. Granted, in some cases marriage holds unique advantages, such as when one partner does not have access to health insurance or when marriage would provide a larger Social Security benefit ( Chevan, 1996).

There are also social reasons to cohabit in later life. An incomplete institution in which the norms and expectations for partners’ roles lack clear definition, cohabitation requires couples to actively construct their relationship dynamics (cf., Nock, 1995). This process can create conflict and disagreement, but it is also an opportunity for couples to carve out alternative relationship scripts that do not hew to traditional marital expectations ( Vespa, 2013). Men may find cohabitation desirable because it gives them access to a resident partner who provides social support ( de Jong Gierveld, 2002). Women may favor cohabitation because it is not predicated on the gendered caregiving obligations typifying marriage, allowing them to preserve some of their autonomy ( McWilliams & Barrett, 2014; Talbott, 1998; Watson & Stelle, 2011). Older cohabitors are less likely to provide care to their partners than are older married spouses ( Noël-Miller, 2011).

The Role of Cohabitation in Later Life

The meaning or purpose of later life cohabitation is unique. Whereas cohabitation among young adults tends to operate as a prelude to marriage or an alternative to singlehood, culminating in either marriage or separation within a year or two of its inception, cohabitation among older adults functions as a long-term alternative to marriage ( King & Scott, 2005). The relationship quality and stability of older cohabitors exceeds that of younger cohabitors, even though older cohabitors are relatively unlikely to report plans to marry their partners ( King & Scott, 2005). Indeed, cohabitation in later life tends to be quite stable, with an average duration of nearly ten years ( Brown, Bulanda, & Lee, 2012; Brown & Kawamura, 2010). Only a minority of older cohabiting couples wed or break up. Rather, the most common union outcome for older cohabitors is dissolution resulting from the death of the partner ( Brown et al., 2012). The relationship dynamics of later life cohabitation are akin to remarriage. Older cohabitors and remarried individuals report comparable levels of emotional satisfaction, openness, pleasure, interaction, criticism, and demands, although cohabitors are less likely than remarried individuals to say their relationships are very happy ( Brown & Kawamura, 2010).

Profile of Older Cohabitors

As for demographic profiles, older adult cohabitors are distinct from both older remarried and unpartnered individuals. Table 2 provides a portrait of the previously married, differentiating among individuals aged 50 years and older who are cohabiting, remarried, or unpartnered using the 2015 American Community Survey. Nearly all (89%) older adult cohabitors are previously married ( Brown, Lee, & Bulanda, 2006). The majority of cohabiting and remarried older adults are men, whereas over two-thirds of unpartnereds are women. The median age of cohabitors (60) is younger than both remarrieds (63) and singles (68). Over 80% of remarrieds are White, compared to just over three-quarters of cohabitors and 70% of unpartnereds. The majority of both cohabitors (85%) and unpartnereds (56%) are divorced. Remarried individuals have more education than either cohabitors or unpartnereds, on average. Over one-quarter of remarried older adults have at least a college degree, whereas just over one-fifth of cohabitors and one-fifth of unpartnereds have a college degree or more. Cohabitors are the most likely to be working (62%). Over half of remarried respondents report being employed, and just 37% of unpartnereds are working. The high employment level of cohabitors does not yield the economic returns that remarried individuals enjoy. Remarried individuals have the highest median household income at $101,027, followed by cohabitors with $88,829, and $55,519 among unpartnered persons. Over one-fifth of cohabitors (21%) and 17% of unpartnereds report being poor compared with less than 5% of remarrieds. More than one-third of unpartnered older adults have a disability versus about one-fifth of cohabitors and remarried individuals. Finally, approximately 10% of older cohabitors have no health insurance, whereas only 6% of unpartnereds and 4% of remarried individuals are uninsured.

Percentage Distributions of Demographic, Economic, and Health Characteristics of Previously Married Adults Aged 50 and Older, by Union Status, 2015

Note: Data come from the 2015 American Community Survey. Calculations by the authors. NA = not applicable.

Percentage Distributions of Demographic, Economic, and Health Characteristics of Previously Married Adults Aged 50 and Older, by Union Status, 2015

Note: Data come from the 2015 American Community Survey. Calculations by the authors. NA = not applicable.

This national portrait echoes earlier research showing that older cohabitors tend to have fewer economic resources, including wealth and homeownership, than their remarried counterparts despite having largely comparable education and employment levels ( Brown et al., 2006). Nonetheless, research on later life union formation shows that wealthier individuals are not more likely to remarry than to cohabit ( Vespa, 2012). The economic advantages accruing to cohabitors versus unpartnered older adults ( Brown et al., 2006) align with work showing wealth is positively associated with forming a cohabiting (or marital) union in later life ( Vespa, 2012). Cohabitors typically report the weakest social ties to friends and family ( Brown et al., 2006). For cohabiting women, having friends and family close by is associated with a lower likelihood of marrying and a greater chance of breaking up with the partner ( Vespa, 2013), which suggests that women with larger support networks may be less committed to their cohabiting partners because they have alternative sources of social support. Cohabiting women who receive entitlement income are also less likely to marry ( Vespa, 2013), reinforcing the notion that cohabitation allows individuals, especially women, to maintain financial independence. The transition to marriage among older cohabiting couples, while unusual, appears to follow a gendered pattern of exchange in which men are most likely to marry when they are in poor health and have considerable wealth whereas women’s marriage entry is highest when they have little wealth and excellent health ( Vespa, 2013). In other words, men exchange economic security for women’s caregiving and vitality.

Cohabitation and Health Outcomes

Given that most cohabiting unions are quite stable and operate as an alternative to marriage in later life, it is possible that older cohabitors enjoy health benefits that are on par with those of older married individuals. There is limited research on the well-being of older cohabitors. An early cross-sectional study indicated that the levels of depressive symptoms did not differ for women by union type but that married men reported fewer symptoms, on average, than did cohabiting men. Cohabiting men’s psychological well-being was comparable to that of married and cohabiting women ( Brown, Bulanda, & Lee, 2005). A more recent, longitudinal examination came to a different conclusion about men, namely, that the psychological well-being of cohabitors is similar to or even better than that of marrieds whereas women’s psychological well-being did not vary by union type ( Wright & Brown, 2017). The physical health benefits of cohabitation are largely unexplored. There is no mortality advantage of marriage versus cohabitation for Blacks ( Liu & Reczek, 2012). Among Whites, cohabitation is associated with higher mortality than marriage but this differential diminishes with age ( Liu & Reczek, 2012), perhaps reflecting the unique role of cohabitation as an alternative to marriage in later life.

Same-Sex Cohabitation

Research on same-sex cohabiting older adults is slim. Same-sex cohabiting older adults are more socioeconomically advantaged than different-sex cohabitors and appear more comparable to different-sex married older adults ( Baumle, 2014; Manning & Brown, 2015). Same-sex male cohabitors are largely similar to different-sex married men in terms of physical health but experience more psychological distress. Same-sex female cohabitors report poor mental and physical health and more functional limitations than different-sex married women ( Gonzales & Henning-Smith, 2015). A similar pattern occurs when comparing individuals in same-sex couples to those in different-sex couples: men have equivalent health outcomes whereas women’s health is worse, on average, in same-sex than different-sex cohabiting couples ( Baumle, 2014). These differentials emerge despite evidence that same-sex couples monitor and encourage healthy behaviors for their partners ( Reczek, 2012). The health advantages experienced by same-sex cohabitors, when they exist, are largely due to their high socioeconomic status, otherwise their health outcomes are similar to unpartnereds ( Liu, Reczek, & Brown, 2013). A key task for future research is to address whether those in same-sex or different-sex cohabiting or marital unions experience similar health outcomes and whether these outcomes vary by gender ( Fredriksen-Goldsen & Muraco, 2010).

Divorce

The Gray Divorce Revolution

The fragility of later life marriages is at an all-time high. Since 1990, the gray divorce rate has doubled, rising from 4.9 divorced persons per 1,000 married persons to 10 per 1,000 in 2015 ( Brown & Lin, 2012; Stepler, 2017a). In 2010, more than one-quarter of individuals who divorced were over age 50, compared to just 1 in 10 in 1990 ( Brown & Lin, 2012). The scope of the gray divorce revolution will intensify in the coming years with the aging of the population. Even if the gray divorce rate remains unchanged, by 2030 the number of persons experiencing gray divorce is estimated to grow by one-third merely due to the larger size of the older adult population ( Brown & Lin, 2012). The rise in gray divorce is remarkable considering that the overall divorce rate has been stable since 1990 and is falling among younger adults, reflecting the growing selectivity of marriage for this age group ( Kennedy & Ruggles, 2014).

More than three decades ago, scholars identified key social and demographic trends foretelling a rise in later life divorce ( Berardo, 1982; Uhlenberg & Myers, 1981). First, Uhlenberg and Myers (1981) noted that widespread divorce created new norms about the acceptability of calling it quits. As individuals experienced divorce either first hand or within their social networks, the stigma attached to divorce diminished. Thus, exposure (direct or indirect) to divorce could encourage couples to seek a divorce when they are dissatisfied with their marriage. Second, the increase in remarriage that accompanied the divorce revolution also portended a rise in subsequent divorce as remarriages are at higher risk of divorce than first marriages. Indeed, the gray divorce rate is 2.5 times higher for those in a remarriage than a first marriage ( Brown & Lin, 2012). Remarried couples tend to be less homogenous and this heightens their chances of divorce. Also, remarriage frequently results in stepfamilies, which present considerable challenges for couples as they blend children from prior relationships. The obstacles associated with stepfamily formation are so formidable that it can take five to seven years for families to reach equilibrium. Yet, many couples divorce within a few years of remarrying. Individuals who have experienced divorce are more likely to divorce again in the event of an unsatisfactory marriage ( Amato, 2010). One reason why the divorce rate is lower in first marriages is because some fraction of them are unhappy but refuse to get divorced. A third factor is women’s employment. The dramatic increase in wives’ labor force participation when these older people were at their prime changed the marital bargain by making wives less dependent on their husbands ( Schoen, Astone, Kim, Rothert, & Standish, 2002). Many wives now have sufficient financial autonomy that they can afford to get divorced. Finally, lengthening life expectancies have changed the calculus about divorce. Individuals who survive to age 65 can expect to live another 20 years, which could be a long time to spend with someone from whom one has grown apart. Adults are living healthier longer, which could nudge them to make a significant life change like gray divorce. The centrality of marriage has receded in modern society and living alone or with an unmarried partner are now viable alternatives ( Cherlin, 2004).

The gray divorce revolution is unfolding in a larger social context in which the meaning of marriage (and divorce) has shifted dramatically in recent decades ( Wu & Schimmele, 2007). The prevailing framework of individualized marriage, marked by self-fulfillment, flexible roles, and open communication, pervades across the generations ( Cherlin, 2004). Like their younger counterparts, older adults hold marriage in high esteem but also have lofty expectations for what constitutes a good marriage. When one’s marriage fails to live up this standard, divorce is viewed as an acceptable solution. Qualitative research on gray divorce reveals that growing apart is a common reason why older couples call it quits. After raising children and having careers, many couples retire only to find that they do not enjoy spending time together ( Bair, 2007).

Predictors of Gray Divorce

From a life course perspective, it is plausible that key turning points such as an empty nest, retirement, or failing health could prompt couples to reflect on their marriage and decide to get divorced. These turning points are invoked in the narratives of individuals who experience gray divorce ( Bair, 2007). But empirical research reveals they are not associated with a couple’s risk of gray divorce. Rather, the same factors that are associated with divorce earlier in the adult life course are most salient for gray divorce, too. Marital duration is inversely associated with divorce and remarriages tend to be of shorter duration than first marriages. Interracial couples are more likely to experience gray divorce than same race couples. Marital quality is negatively associated with divorce. And couples with fewer economic resources, namely wealth, are at greater risk of gray divorce ( Lin, Brown, Wright, & Hammersmith, 2016).

Consequences of Later Life Divorce

Divorce is among the most stressful life events and it can take years for individuals to recover psychologically, socially, and financially. There is little work on the consequences of gray divorce ( Carr & Pudrovska, 2012) but it seems likely that the range of outcomes for older adults is more varied than for younger adults. On the one hand, older adults who wanted to get divorced, are financially secure, and in good health may experience few or no downsides to calling it quits. Their quality of life could actually improve following divorce. On the other hand, individuals who are vulnerable due to financial hardship or poor health could be devastated by a gray divorce. Unlike their younger counterparts, they do not have decades remaining in the labor force to make up for the financial losses associated with divorce. Poor health could impede their ability to work, compounding financial difficulties. Navigating health declines without the support and care of a spouse may pose significant challenges to gray divorced individuals, diminishing their well-being. For example, a recent study by Karraker and Latham (2015) suggests that healthy midlife married couples are at risk of gray divorce with the onset of wife’s heart problems, but not when the husband’s health declines.

One-third of first later life marital dissolutions now occur through gray divorce rather than widowhood, making it vital that researchers broaden their scope to encompass both dissolution pathways ( Brown et al., 2016). Gray divorce results in two individuals eligible to repartner and they are much more likely to form a new union than those who experience dissolution through spousal death. Thus, we can expect later life repartnering to climb in the coming years. Still, as shown in Figure 2, few gray divorced women form either a remarriage (15%) or a cohabitation (9%). The levels of repartnering are somewhat higher for gray divorced men at 28% for remarriage and 15% for cohabitation, but most remain single ( Brown et al., 2016).

Repartnership status by dissolution type and gender. Note: Statistics are from Table 2 of Brown et al. (2016) and reflect the 2010 repartnership status of individuals who had experienced divorce or widowhood at age 50 years or older.

An important task for future research is to evaluate whether the outcomes associated with gray divorce are similar to widowhood as well as whether repartnering reduces the negative effects of disruption. From a financial standpoint, it seems gray divorce and widowhood may be largely equivalent for men, but for women, gray divorce is often a bigger economic shock. Among those who are age-eligible for Social Security, 27% of gray divorced women are in poverty compared with just 13% of widowed women. For men, the share is about 13% regardless of dissolution type. Those who have repartnered are unlikely to be poor at only about 4% ( Lin, Brown, & Hammersmith, 2017). Divorce also shapes the relationships between fathers and their adult children. Older men receive less support from their adult children if they are divorced from the children’s mother ( Lin, 2008). Later life divorce is also tied to decreased contact with adult children, especially for fathers ( Kalmijn, 2013). In turn, repartnering following divorce further weakens men’s relationships to their children ( Kalmijn, 2013; Noël-Miller, 2013).

Discussion

The past few decades have witnessed rapid change in the family formation and dissolution patterns of older adults. Declining shares of older adults are either married or widowed, and rising proportions are cohabiting, divorced, or never-married. The changing marital status composition of older adults foregrounds the salience of the larger marital biography, encompassing not merely current marital status but also transitions and their key features, including timing, duration, and sequencing. Multiple transitions, especially the experience of marital disruption, can be detrimental to health and well-being and these negative outcomes often persist over time and even after repartnering occurs ( Hughes & Waite, 2009; Zhang et al., 2016).

The varied marital biographies of today’s older adults raise a host of questions about the diverse trajectories of the family life course after age 50. Here, we reviewed recent research that focuses on marriage, cohabitation, and divorce in later life. But family pathways are not restricted to marriage or even to coresidential relationships. Non-coresidential partnerships, including dating and living apart together (LAT) relationships, are arguably more common than is cohabitation in later life but they remain understudied ( Brown & Shinohara, 2013; Connidis, Borell, & Karlsson, 2017; Lewin, 2016). Dating relationships are concentrated among the most advantaged unmarried older adults, with those who have higher levels of education and are in better health the most likely to be dating ( Brown & Shinohara, 2013). LAT relationships, which can be conceptualized as long-term dating relationships that are unlikely to eventuate in either cohabitation or marriage, offer unprecedented flexibility and autonomy by allowing couples to define their obligations and responsibilities to one another within a framework of a high commitment relationship ( Benson & Coleman, 2016; Connidis et al., 2017; Duncan & Phillips, 2011; Upton-Davis, 2012). Older adults in LAT relationships report less happiness than do cohabitors and married individuals, but also less relationship strain, which aligns with the notion that LAT couples can establish the relationship expectations and norms that work for them ( Lewin, 2016).

In short, there are arrays of relationship options for older adults that merit consideration in future research. Remarkably little is known about the basic levels and patterns of emergent relationship types, such as LAT, let alone whether and how these relationships affect the health and well-being of older adults. Greater attention to how marital biographies and current relationship type (including dating or LAT) are linked to well-being in later life is sorely needed. Theory development on nonmarital relationships is also vital as the motivations for dating or cohabitation are unlike those that prevail earlier in the life course. Bulcroft and Bulcroft’s (1991) conclusion more than a quarter century ago that explanations for dating in young adulthood do not readily apply to older adult dating remains true and extends to other relationship types such as cohabitation. Likewise, the costs and benefits of gray divorce are arguably distinctive, yet the predictors appear to be largely the same as those identified for younger adults ( Lin et al., 2016). This paradox merits further conceptual and empirical attention.

It is also essential to address how these partnership dynamics impinge on other family ties, namely between parents and their children. Couples often pursue LAT relationships rather than cohabit or marry because they have resident children ( de Jong Gierveld & Merz, 2013). Similarly, many couples choose cohabitation over remarriage at the urging of their adult children ( Bildtgård & Öberg, 2017). When an older adult experiences a health decline does the partner step in to help or is it the adult child who serves as the caregiver? Cohabitors cannot count on their partner like married spouses do ( Noël-Miller, 2011). Probably LAT and dating partners are even less likely to provide care than cohabiting partners, but this question remains unexplored. If partners and children are less willing to be caregivers, then the burden increasingly falls on institutions and society to manage the care of frail elders which could have significant public policy implications.

Aging is a global phenomenon with far-reaching ramifications for societies. Yet, comparative research on partnerships and unions in later life is slim. In fact, to the best of our knowledge, the only available information on older adult marital status distributions in other countries is now somewhat dated. The Survey of Health, Ageing and Retirement in Europe (SHARE) provides some basic insights. In 2004, individuals aged 50 and older living in ten European countries were typically in partnered relationships. About 77% of older men and 56% of older women were married. Only 1% of older men and just 0.4% of older women were cohabiting, levels that are remarkably lower than in the United States. The percentages currently divorced among older European adults, which stood at 6% and 7% for men and women, respectively, are also considerably lower than in the United States. Nearly 8% of men and 8% women were never married. Widowhood was much less common among men at 8% than women at 29% ( Kohli, Kunemund, & Ludicke, 2005). Of course, these overall figures belie considerable variation across European nations. In Sweden, for example, about 10% of men and 6% of women were cohabiting in 2004 ( Kohli et al., 2005). Globally, the proportions of older men and women who are married has grown modestly and the proportions widowed have fallen in recent decades, with both trends mainly reflecting gains in life expectancy ( Kinsella & Phillips, 2005). Childlessness is on the rise for older adults internationally, and the proportions divorced are also expected to increase in the coming years, reflecting family patterns established earlier in the life course and raising new questions about the availability of family support and caregiving in later life ( Kinsella & Phillips, 2005). Future research should pay greater attention not only to the diverse family demographic trends marking older adulthood but also how these patterns align with cross-national economic and social policies, which may provide incentives to form (or dissolve) various types of unions. And, here again, cross-national information on non-coresidential unions, such as dating and LAT relationships, appears to be lacking.

To ensure researchers can capture the richness of the family life course experiences of older adults, major national data collections on older adults may benefit from expanding beyond the narrow focus on marital status to include non-coresidential relationships such as dating and LAT. Marital biographies are now diverse, so collecting more detailed marital and cohabitation histories for same-sex and different-sex relationships is warranted to ensure researchers can identify the components of the marital biography that are most closely tied to well-being in later life ( Umberson, Thomeer, Kroeger, Lodge, & Xu, 2015). There is growing recognition that couple-level data offer much richer insights into aging but there are a lot of unanswered questions about how linked lives shape health outcomes. In particular, the ways in which changes in spousal health may shape one’s own outcomes are poorly understood ( Cooney & Dunne, 2001; Zhang et al., 2016). Even less is known about how nonmarital partners influence each other’s health.

Dramatic family changes are occurring during the second half of life. Today’s older adults have complex marital biographies, reflecting their varied experiences of cohabitation, marriage, divorce, and remarriage. Some have shunned marriage altogether whereas others are calling it quits later in life. New relationship paradigms offer attractive alternatives to marriage and even cohabitation. Gerontologists and family scholars are only beginning to investigate the patterns and consequences of these new frontiers in later life couple relationships.

Funding

This research was supported in part by the Center for Family and Demographic Research, Bowling Green State University, which has core funding from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (P2CHD050959).