The Gender‐as‐Relational Approach for Theorizing About Romantic Relationships of Sexual and Gender Minority Mid‐ to Later‐Life Adults

Abstract

We deploy the “gender-as-relational” (GAR) approach to enhance the study of the long-term romantic relationships of sexual and gender minority mid- to later-life adults. The GAR approach states that gender within relationships is shaped by three key factors: own gender, partner’s gender, and the gendered relational context. This approach highlights that the relationship dynamics of men, women, and gender nonconforming people are highly diverse, reflecting that gender is a social construct formed within interactions and institutions. We explicate how GAR can reorganize the study of sexual and gender diversity in three research areas related to aging and relationships—caregiving, marital health benefits, and intimacy—and discuss theory-driven methods appropriate for a GAR research agenda. A GAR framework reorients research by complicating taken-for-granted assumptions about how gender operates within mid- to later-life romantic relationships and queering understandings of aging and romantic relationships to include experiences outside of heteronormative and cisnormative categories.

Long-term romantic relationships (e.g., marriage, cohabitation, dating) are key sources of intimacy, support, and identity in mid- to later-life (Waite & Das, 2010). Compared to their unpartnered peers, married and cohabiting people have better mental and physical health, more health-promoting behaviors, and greater longevity (Umberson & Thomeer, 2020). Romantic relationships may be particularly important for mid- to later-life adults who often have smaller social networks than younger adults due to the loss of coworker ties through retirement, adult children leaving their household, and deaths of parents, siblings, and friends (Waite & Das, 2010). As a result, older adults’ social networks are more often spouse- or partner-centered (Cornwell, Laumann, & Schumm, 2008). Relationships in mid- to later-life are also more likely to be long-term relationships, meaning that couples have experienced an accumulation of benefits (or costs) from that relationship and developed systematic patterns of gendered relationship dynamics (e.g., division of household labor, conflict style; Umberson, Thomeer, & Lodge, 2015). Finally, as people age, they become increasingly likely to experience chronic conditions and health limitations, and romantic partners play a key role in caregiving and influencing health in those situations (Revenson et al., 2016).

Research on long-term romantic relationships in mid- to later-life has primarily focused on different-sex couples. This research shows that romantic ties are highly gendered, with women and men experiencing their relationship differently (Thomeer, Umberson, & Pudrovska, 2013; Waller & McLanahan, 2005). Yet, calls to “queer” our understandings of aging within families highlight the key limitations of past approaches which tend to overlook experiences outside of heteronormative and cisnormative categories, including the general exclusion of sexual and gender minority (SGM) people in romantic relationships (Brown, 2009; Fish & Russell, 2018). (See Table 1 for definitions of terms used in this article.) According to recent U.S. estimates, among adults 65 and older, there are about 217,050 transgender adults, and about 1.5 million lesbian, gay, and bisexual adults (Flores et al., 2016; LGBT MAP & SAGE, 2010). These are likely underestimates, given that some people may not disclose sexual and gender information on a survey (Fredriksen-Goldsen & Kim, 2015). These populations are expected to increase significantly—perhaps even doubling—over the next decade (LGBT MAP & SAGE, 2010). Overlooking gender dynamics in SGM people’s relationships prevents a full understanding of gender within romantic partnerships in mid- to later-life within an increasingly sexual and gender diverse population.

Table 1:

Definitions of Key Terms

Key Terms	Definitions
Gender Minority*	A person whose gender identity does not align with hegemonic gender norms (e.g., transgender, gender nonconforming)
Sexual Minority*	A person whose sexual identity, attraction, or behavior does not align with heterosexuality (e.g., lesbian, asexual)
Cisgender	A person whose assigned sex at birth aligns with their gender identity and expression
Transgender	A person whose gender identity is other than their sex assigned at birth, and/or someone who identifies as a transgender identity including transwoman, transman, and transgender
Gender nonconforming	A person whose gender identity and expression is outside of the gender categories of man and woman
Asexual	A person who does not have sexual attraction to or sexual interest in other people, and/or identifies as asexual
Heterosexual	A person who is primarily attracted to or in romantic relationships with people of a different sex or gender, and/or identifies as heterosexual
Bisexual	A person who is attracted to or in romantic relationships with both men and women; a person who is attracted to or in romantic relationships with people of any gender; and/or a person who identifies as bisexual
Gay/Lesbian	A person who is attracted to or in romantic relationships with people of their same sex or gender, and/or identifies as gay or lesbian

^*Minority status reflects both a statistical minority and the reduced access to resources and power as a result of this status

In addition to including older SGM-member couples in empirical scholarship, we also must develop and deploy theoretical frames that attend to the experiences of sexual and gender diverse relationships across a range of SGM-member relationship types. In this article, we make a case for using the “gender-as-relational” (GAR) approach as a tool for theorizing about and analyzing the gendered experiences of romantic relationships in mid- to later-life. In addition to discussing the primary tenets of this framework and some empirical applications, we also discuss opportunities for diversity within the GAR approach and suggest several appropriate theory-driven methodological approaches to best move forward a GAR research agenda. A GAR approach draws attention to the fact that within romantic relationships, people enact gender not only as individuals but also in relation to each other, namely each other’s gender (Reczek & Umberson, 2016; Umberson, Thomeer, & Lodge, 2015). People engage with and understand their own gender differently depending on whether they are interacting with a man, a woman, or a gender nonconforming person. Thus, because SGM-member romantic relationships consist of two men, two women, or other gender combinations, the dynamics of SGM-member relationships are unique depending on their sexual and gender diversity.

In this article, we focus on long-term romantic relationships from mid- to later-life within the U.S. (ages 40 and older). We limit our discussion to long-term relationships (generally defined as 7 or more years; Umberson, Thomeer, & Lodge, 2015) because research on health and aging shows that long-term romantic relationships are most strongly associated with health benefits in aging populations (Umberson et al., 2006). In mid- to later-life, people are also increasingly likely to experience health conditions (e.g., increased cancer risk, development of physical symptoms; Levine & Crimmins, 2018), leading to potential change in gender relationship dynamics. Finally, we focus specifically on the U.S. context as this is where the bulk of research using a GAR frame has occurred. Although not the focus of our discussion and examples, GAR can be used to organize the way scholars think about, theorize, and study sexual and gender diversity in aging populations across diverse cultural contexts, life stages, and relationship durations.

The Case for a Gender-As-Relational Framework

The Limits of Focusing on Cisgender Heterosexual Romantic Relationships

Most past research on mid- to later-life romantic relationships is built on heteronormative assumptions, meaning scholars rely on binary gender constructions (i.e., men and women as the only gender identities) and heterosexual orientation (i.e., man partnered with woman as the only relationship type) (Allen & Mendez, 2018). Existing theoretical paradigms generally fall short in addressing the complexities of SGM-member romantic relationships because these paradigms are typically rooted in the assumption that romantic relationships refer only to a heterosexual cisgender man with a heterosexual cisgender woman. For example, over fifty years ago, Jessie Bernard (1982) argued that within every marriage there were two distinct relationships: his and hers. As Bernard noted—and other feminist research supported, in different-sex relationships women did more unpaid work around the home (e.g., emotion work, household work, care work), and there were high levels of disagreement between men and women regarding the state of their marriage (Thomeer, Reczek, & Umberson, 2015b; Umberson & Williams, 2005; Waller & McLanahan, 2005). In subsequent studies, scholars pointed to women’s lower levels of happiness, greater likelihood to initiate divorce, lower desire to get remarried after a marriage dissolved, and fewer health and well-being benefits from marriage as evidence that relationship dynamics were strongly gendered, with more costs for women (Amato & Previti, 2003; Brown et al., 2016; Umberson & Williams, 2005). These past studies—even within feminist gerontology (Calasanti & King, 2007)—generally relied on theoretical perspectives that assumed a heteronormative context of gender differences in relationships. These perspectives include classic family theories (e.g., functionalism, exchange theory) as well as feminist gerontological theories that focus on gender inequities in power and resources between men and women in a couple (see Ferree, 2010 for more detailed discussion of these theories).

Historically, men and women within different-sex relationships have been constructed and understood as “opposite,” especially in regard to their relationship desires and tendencies (Ridgeway, 2009). The discourse that men and women are opposites undergirds gendered experiences of marriage that are evident in many domains central to mid- to later-life romantic relationships, including sex, intimacy, and caregiving (Thomeer, Reczek, & Umberson, 2015b). Scholars have held that these gender differences are accentuated at older ages, both due to more traditional gender dynamics within the current cohort of older adults and the increased demands for unpaid care work (Pinquart & Sorensen, 2006; Shu & Meagher, 2017). While patriarchal culture and related gender dynamics may be changing, particularly for younger cohorts, older generations grew up in and sustained their relationships through a more gender-traditionalist time, which may mean that their cultural lexicon is rooted more strongly in heteronormative systems (Scarborough, Sin, & Risman, 2019).

In light of these shifts, recent scholarship has challenged dominant heteronormative paradigms and called for queering of the family in ways that reflect the sexual and gender diversity of contemporary romantic relationships (Allen & Mendez, 2018; Fish & Russell, 2018). Queering theory and research on the family allows dominant notions of “the couple” to be deconstructed and highlights gender and sexual diversity within family life. Scholars also champion queering theory and research around aging, calling attention to the reality that most aging research is heteronormative and cisnormative, rendering invisible the romantic experiences of older SGM adults (Brown, 2009; Fabbre, 2016). Although we know a great deal about how gender matters in relationships involving older cismen and ciswomen in different-sex unions, we know less about gender dynamics in romantic relationships outside of these heteronormative and cisnormative contexts.

Theoretical Foundations of the Gender-As-Relational (GAR) Framework

By challenging taken-for-granted assumptions about mid- to later-life romantic relationships based on different-sex couples with cisgender partners, we can expand our understandings of gender dynamics. We advocate for a Gender-as-Relational (GAR) framework to advance this goal. A GAR framework aligns with conceptualizations of gender as formed and maintained through social interactions with others, including within romantic relationships (West & Zimmerman, 1987). Gender scholars emphasize that there are multiple forms of masculinities and femininities that must be understood in relation to each other, including in terms of power differentials (e.g., relational, interactional, hierarchical; Connell, 2005; Ferree. 2010). The positions of men, women, and other genders are inherently tied together and cannot be independently understood, especially in the context of romantic relationships (Calasanti, 2004). For example, when a husband works to understand his wife (and vice versa), he draws on broader cultural discourses of gender, and these discourses do not emphasize similarities between men and women but rather their differences (e.g., men are rational, women are emotional). Even as men and women’s individual experiences within a relationship may challenge these assumptions (e.g., a wife desires sex often, a husband cries easily), culturally-motivated gender discourses are powerful frames through which partners view their relationship, especially as these discourses are often institutionalized (i.e., involving socially established and legally sustained patterns of norms, values, and behaviors that provide guides to roles and obligations; Cherlin, 1978)—at least within different-sex marriages (Ferree, 2010).

The GAR perspective emerged from feminist scholarship (Calasanti, 2004; Connell, 2005) to specifically advance theory regarding gender relationship dynamics beyond the different-sex couples with cisgender partners’ context (Reczek & Umberson, 2016; Umberson, Thomeer, & Lodge, 2015). A GAR approach builds on gender theories which argue that gender should not be understood as static and binary but as “dynamic and situational, [with] attention to differences among women and among men” (Springer et al., 2012, 1661). Relationship dynamics may differ in important ways within couples of different gender compositions. Gender enactments in cisgender-member different-sex couples—in which men and women are culturally viewed as “opposites”—often differ from gender enactments in SGM-member couples—in which essentialized gender differences may be muted.

This is not to say that there are not gender differences within or across same-sex or other SGM-member relationships, especially as related to diversity in gender presentation, performance, expression, and/or ideology. Importantly, although most previous studies using a GAR approach have focused on gender identity (discussed in more detail below), the GAR approach allows researchers to consider both within- and between-person aspects of gender-related attitudes, presentation, and beliefs. For example, previous studies of mid- and later-life different-sex couples consider the gender role beliefs of each spouse, comparing couples in which both partners hold traditional gender beliefs, both hold egalitarian beliefs, and one partner holds traditional beliefs and the other egalitarian (Mickelson, Claffey, & Williams, 2006; Walker-O’Neal & Futris, 2011) Concordance and discordance in these beliefs is associated with the division of household labor, marital quality, and emotional spousal support (Mickelson et al., 2006). GAR is particularly useful in moving away from a focus on gender as binary to gender as a more nuanced, multi-faceted, and dynamic construct.

A GAR framework also takes into account intersectionality to address the interlocking systems of oppression and privilege that shape experiences (Collins, 2000). An intersectional perspective emphasizes the layered oppressions of ageism with sexism, heterosexism, and transphobia that are experienced differently depending on intersecting social positions such as gender and race (Wight et al., 2015; Witten, 2016). Scholars have called for more attention to intersectionality within aging and family research (Cronin & King, 2010; Witten, 2016). Centering SGM older adults within studies of aging romantic relationships requires recognition of intersectional identities and the specific historical, cultural, and structural positions of those identities (Calasanti & Kiecolt, 2007). SGM adults experience sexual minority stressors (i.e., stressors that are uniquely experienced by people who face the systematic disadvantage of stigma; Meyer, 1995) that shape their romantic relationship dynamics over their life course, and these experiences vary across intersecting social positions (e.g., the experience of an SGM identity is not the same for white people and people of color; Grollman, 2014).

Applying a Gender-As-Relational Framework in Research on Aging Romantic Relationships

A new wave of research has begun to examine whether mid- to later-life SGM-member romantic relationships enact gender relationship dynamics in ways that parallel heteronormative and cisnormative scripts in mid- to later-life, or whether SGM-member relationships queer relationship dynamics by drawing on alternative scripts of masculinity, femininity, and/or androgyny. A GAR framework is ideal for advancing this research. Scholarship on sexual and gender diversity often overlooks aging populations, as younger romantic relationships tend to be viewed as the harbingers of the gender revolution (Brown, 2009). This renders research on same-sex couples and other SGM-member families in mid- to later-life an area—as Brown (2009) argues—of “rhetorical silence,” effectively limiting how we think about gender relationship dynamics in aging populations. Calasanti and others draw attention to these issues in their development of feminist gerontology, a perspective that focuses on gender and power relations across the life course (Calasanti, 2004; Calasanti & Slevin, 2001; Hooyman et al., 2002). Similarly, Fredriksen-Goldsen, Brown, Fabbre, and others highlight the unique positions of older SGM adults, arguing that their life course experiences have given them both obstacles and resiliencies not seen among heterosexual and cisgender older adults and calling for more inclusion of sexual and gender diversity in aging studies (Brown, 2009; Fabbre, 2016; Fredriksen-Goldsen et al., 2013). These scholars further suggest that “[SGM older adults] have been excluded from queer theory in much the same way as they are excluded from… gerontology” (Brown, 2009, 65). Although we focus on queering aging within the context of romantic relationships through a GAR framework, we direct readers to comprehensive overviews on queering aging by Reczek (2018), Brown (2009), and Fabbre (2016).

To illustrate how a GAR frame can provide insight into gender relationship dynamics of mid- to later-life couples, we focus below on three main areas that are especially salient for aging romantic relationships and that have already received some research attention: caregiving, health benefits of marriage, and intimacy. GAR can also be applied to many other wide-ranging issues of aging romantic relationships, including end-of-life planning, conflict, adjustment to the death of a spouse, caring for aging parents, and division of household work. Below, we first highlight research on different-sex and same-sex couples with cisgender partners, and then we expand these arguments to consider other gender contexts as well as theory-driven methods for approaching this research agenda.

Focusing on three areas (caregiving, marital health benefits, intimacy) central to studies of mid- to later-life same-sex and different-sex cisgender-member couples allows us to illustrate two major implications of the GAR frame. First, the GAR approach points to three key factors: own gender, partner gender, and gendered relational contexts (e.g., same-sex compared to different-sex dyads). Thus, not just personal experiences of gender shape the dynamics of the relationship but also the interaction between each partner’s gendered experiences. Gender dynamics are co-constructed by partners within a relationship. Second, and as an extension of this first point, this approach highlights that relationship dynamics of cisgender men, cisgender women, and gender minority people are highly diverse, both across and within couples, reflecting how dynamic and complex gender is as a social construction formed within interactions and institutions.

Caregiving

When an older partner or spouse has illness or infirmity, women typically provide more informal caregiving than do men in the context of different-sex cisgender-member couples (Pinquart & Sorensen, 2006). In different-sex romantic relationships, compared to men, women tend to be more intensive in their caregiving (e.g., more multitasking involved, more inflexible tasks, more emotion work) and do more of their care work in isolation (e.g., without adult children helping or support of paid caregivers) and are less likely to be explicitly appreciated by the care recipient (Pinquart & Sorensen, 2006; Thomeer, Reczek, & Umberson, 2015b). Our recent research that includes same-sex couples goes further to show that, compared to men, women provide more caregiving—including more intensive caregiving and more caregiving to support emotional needs—to their partner whether they are married to a woman or a man and even when ill themselves (Umberson et al., 2016b).

Why do we see such stark gender differences in caregiving regardless of the gender composition of the couple? GAR points to broader gender discourses regarding masculinity and femininity (e.g., men as providers, women as nurturers) and structural constraints (e.g., women paid less in formal work force so specialize in care work at home) as connected to gendered patterns of caregiving. These discourses include the construction of gender strategies and family myths to explain and justify unequal divisions of caregiving (Pfeffer, 2010). In our study of mid- to later-life different-sex couples, one woman described how the responsibility of caring for her husband was passed from his mother to her when she told his dying mother, “I will take care of [your son]. And you don’t have to worry about him” (Thomeer, Reczek, & Umberson, 2015b, 18). For women in different-sex couples, caring for men is seen as the primary responsibility of women, but a GAR perspective also suggests that caring generally—whether in a same- or different-sex union—is central to the way that women approach their romantic relationships.

A GAR frame pushes beyond a heteronormative context and provides a vantage point to see not only gender difference as central to how men and women approach caregiving, but also that caring for a same- or different-sex partner shapes the overall illness and caregiving experience. In our study of caregiving in same-sex and different-sex couples, we found that most men in the sample viewed both their own and their partner’s illness as something to be dealt with alone as an individual, whereas most women conceptualized illness as better dealt with together as partners (Umberson et al., 2016b). These different ways of making sense of illness experiences meant women both provided and desired care, but men were less likely to provide or desire it. Gendered caregiving dynamics are thus a result of dissimilar constructions of illness for men and women. Same-sex spouses describe similar constructions of illness much more so than different-sex spouses, and—as a consequence—same-sex spouses describe less illness-related disagreement and stress.

The second major tenet of GAR—that the relationship dynamics of cisgender men, cisgender women, and gender minority people are highly diverse due to gender being interactive and dynamic—is also evident in older couples’ caregiving. Men and gender nonconforming people provide substantial care to their partners, breaking out of the theoretical conceptualization of caregiving as only “women’s work.” Feminist gerontologists have emphasized for years that men provide caregiving, especially at older ages and within the context of romantic relationships (Calasanti & Bowen, 2006; Calasanti & King, 2007). In most of these studies, scholars focus on men in different-sex relationships, finding that men perform this caregiving within dominant masculinity discourses. For example, one man said of caring for his wife, “You just have to pick it up like you do a trade. Like laying brick or finishing concrete” (Calasanti & King, 2007, 520). Other studies highlight caregiving by gay men and—much less studied—transgender, asexual, and bisexual men and gender nonconforming people within romantic relationships (Pfeffer, 2010; Reczek & Umberson, 2016; Umberson et al., 2016b). A GAR approach emphasizes that there is not just one way to do (or not do) caregiving related to gender, and that gendered provisions of caregiving within romantic relationships are co-constructed, with important implications for the balance of costs and rewards associated with caregiving across diverse populations.

Gendered Health Benefits of Marriage

Previous studies find that men experience more of a health benefit from different-sex marriage than women, and—going even further—that women experience health costs from different-sex marriage (Liu & Umberson, 2008). This gender gap is likely greater at older ages within current cohorts, given the more gender conservative contexts of older age cohorts (Shu & Meagher, 2017). Yet these conclusions are complicated by a GAR approach, which emphasizes that the intersection between own gender and partner’s gender shapes the extent to which marriage benefits health, leading to potential differences for same-sex compared to different-sex couples and to a high-level of variation in who benefits from marriage across and within couples.

To the first point, a GAR approach expands beyond the different-sex older couple context to examine the health impact of marriage in SGM-member couples, finding that same-sex couples, much like different-sex couples, experience health benefits from cohabitation and marriage (Liu, Reczek, & Brown, 2013; Wight, LeBlanc, & Badgett, 2013). Overall, these effects appear consistent across couples with two men or two women, suggesting that women and men in same-sex relationships receive similar benefits from these relationships (Liu, Reczek, & Brown, 2013). There are reasons to expect that same-sex couples may experience even more benefits from marriage than different-sex couples, given our research showing that same-sex couples are less likely to experience gender inequalities in who provides care, support, or other types of work or resources within the household due to gender similarity and a greater commitment to equality (Reczek & Umberson, 2012; Thomeer, Reczek, & Umberson, 2015a; Umberson et al., 2016a). However, this finding has not been demonstrated with population-level data reflecting the absence of nationally representative longitudinal studies of same-sex and different-sex couples (Reczek, 2020).

To the second point, studies find that it is not just one’s personal experience of marriage that impacts what marriage means for health, but the interaction between each partner’s experiences—and that this is, in turn, shaped by each partner’s gender. Through an emphasis on dyadic methods, GAR suggests an important avenue of research is examining discordance between partners on their ratings of relationship quality in driving gender differences in the health impacts of marriage as couples age. Women in different-sex relationships tend to rate their relationships as lower quality than men in different-sex relationships, even though these men and women are in a relationship with each other (Robles et al., 2014; Umberson et al., 2006). Thus, to the extent that women in different-sex relationships report worse relationship quality than men, this translates to sustained disadvantage for the health of married women at mid- to later-life (Umberson & Williams, 2005; Umberson et al., 2006).

Yet research within a GAR approach allows us to examine relationship dynamics in multiple contexts, comparing couples with two men or two women and testing whether his and hers marriages might extend to these contexts. These studies show that there are gender differences when comparing relationships between two men and two women in the determinants of relationship quality (Cooper et al., 2017; Pollitt, Robinson, & Umberson, 2018). For example, Pollitt and colleagues (2018) in their study of mid-life couples find that greater gender conformity is associated with more shared power and also greater marital quality in same-sex men couples and different-sex couples, but not same-sex women couples. Yet aside from Kurdek’s early work showing that same-sex couples report similar relationship quality regardless of gender (2004), few studies consider whether men in same-sex couples have better or worse relationship quality compared to women in same-sex couples, how relationship quality differs in comparison to different-sex couples, or whether the impact of relationship quality on health is moderated by gender in same-sex couples. Scientific knowledge of how gender matters for the health benefits of romantic relationships is still in progress, and a GAR approach is needed to continue to push this forward, especially in studies of aging romantic relationships. As inclusion of same-sex couples in studies of relationships becomes more common, studies on gender that include only different-sex spouses should be seen as incomplete in documenting gender relationship experiences and outcomes.

Intimacy

The third topic we highlight is intimacy in romantic relationships, with intimacy referring to feelings of closeness and support and the sharing of physical space, touch, emotions, and information with another. Previous research demonstrates that intimacy is gendered within mid- to later-life couples, yet our knowledge of this is limited without a GAR approach that brings to center stage the fact that intimacy is co-constructed within couples. The co-construction of intimacy often results in tensions and disagreements around intimacy not only within different-sex but also within same-sex couples. Like other aging research in a heteronormative contexts, men and women are viewed as inhabiting distinct realms for how they understand and experience intimacy. Within the dominant paradigm, women express more desire to talk and share emotions with their partner, and see this as evidence of intimacy. But, for older men, intimacy is often evidenced by whether or not the couple is having sexual intercourse (Elliott & Umberson, 2008). Constructions of men and women as opposite—especially in terms of how they view and experience intimacy—impact relationship dynamics, including conflict and disagreement around intimacy and sex (Elliott & Umberson, 2008).

Studying gender discourses around intimacy becomes especially important at older ages when—largely due to health and functional changes—frequency of sex often decreases, sometimes leading to changing notions of intimacy (Elliott & Umberson, 2008; Lindau et al., 2007). A GAR approach, which queers understandings of gender and intimacy, fits well with a focus on aging romantic relationships where experiences of intimacy may shift from passionate to companionate, even while—as our research has shown—partners continuing to view their relationship as fundamentally intimate and often sexual (Umberson, Thomeer, & Lodge, 2015). Through a GAR lens, scholars can consider how men and women co-construct notions of intimacy together, within a couple. For example, Elliott and Umberson (2008), in their study of aging different-sex couples, find that aging men and women draw on cultural ideas about gender when explaining their own changing desires for sex and intimacy, and that they also are influenced by—and sometimes conform to—the desires of their spouse. Further, women in different-sex relationships often take on extra labor (which the authors call “performing desire”) to bring each partner’s view of intimacy into alignment. Our qualitative study of same- and different-sex mid-life couples finds that previously accounted for sexual desire differences lessen with age, even in the most gender traditional couples, and that couples draw more on their similarities than their differences as they age (Umberson, Thomeer, & Lodge, 2015).

What does intimacy look like within same-sex couples? To some extent, research bears out what we would expect based on a GAR approach—that same-sex couples are more likely to share notions of intimacy than are different-sex couples. Same-sex women couples, for example, are more likely to work to minimize emotional boundaries within their relationship, cultivating an environment of openly sharing feelings and viewing this as integral to intimacy; but men in same-sex relationships do more work to maintain emotional boundaries and are much less likely to view the sharing of emotions as essential (Umberson, Thomeer, & Lodge 2015). While important, this research may oversimplify gender in same-sex couples, and a GAR framework highlights that tensions and disagreements around sex and intimacy also occur within same-sex couples and illuminates diversity both across and within couples. An overemphasis on gender sameness is as untenable as an overemphasis on gender difference and overlooks considerable hetereogeneity.

A GAR approach shows that gender is enacted in myriad ways, especially within romantic relationships. Thus, not all same-sex couples are driven by a discourse of “gender sameness” just as not all different-sex couples are driven by discourses of “gender difference.” As such, scholars should not expect all same-sex couples to be conflict-free around sex and intimacy, nor should they expect all different-sex couples to have conflict in gendered ways. Gender discourses are dynamic over time, especially in long-term aging couples. For example, in our study of same-sex and different-sex couples, one partner in a same-sex couple described why he had changed over the course of his relationship, “I might be the kind of person that would tend to say what is on my mind. [My partner] would give me pause to stop and think before I opened my mouth” (Umberson, Thomeer, and Lodge, 2015, 549–550). The gender sameness around maximizing gender boundaries developed over the relationship as both partners became more attuned to each other. Thus, we might expect the impact of gender to lessen at older ages, as the dynamics of the relationship become more salient than general gender ideologies, but future studies would need to test this possibility.

Diversity beyond Same-sex Cisgender-Member Couples

The GAR frame has primarily been applied to cisgender-member same-sex and different-sex couples, yet a GAR perspective has tremendous potential for expanding scholarship on other romantic relationships with sexual and gender diversity. Below, we highlight the burgeoning research on sexual and gender diversity, romantic relationships, and aging within four areas: 1) transgender and gender nonconforming people in relationships, 2) the role of gender diversity beyond gender identity, 3) non-traditional relationship types, and 4) mixed-sexual identity relationships. We end this section by discussing the need to bring intersectionality perspectives as well as the broader socio-historical context into discussions of gender dynamics within mid- to later-life relationships.

Transgender and Gender Nonconforming People in Relationships

The inclusion of older gender minority populations (e.g., transgender people, gender nonconforming people, genderfluid people) is essential for theoretical development concerning romantic relationships, gender, and aging. The above discussion largely centered on a static and binary construction of gender, comparing cisgender men and women in different-sex and same-sex relationships. It is important for GAR studies to not only ask whether someone is in a relationship with a man or a woman, but to consider multiple gender identities and the complex ways these configurations might impact romantic relationship dynamics. Older gender minority adults typically transitioned (or identified as gender nonconforming or another gender diverse identity) at later ages compared to younger cohorts (Porter et al., 2016), and thus likely followed historically-specific relationship biography trajectories as they aged. This older age of transitioning or coming out may reflect high levels of transphobia and rigid conceptualizations of gender within the historical context of this cohort, or it may reflect changing preferences and notions of the self within a developmental understanding of aging (Bockting et al., 2016). What this means from a GAR perspective is that gender minority older adults in relationships demonstrate that gender is not binary but is dynamic, and experiences of older gender minority adults in romantic relationships need to be incorporated into our theories of how gender is enacted at older ages.

Gender Diversity beyond Gender Identity

A GAR perspective also highlights that gender diversity matters for couples even beyond gender identity, although few studies examine this. Older adults have a range of gender ideologies, gender presentations, and gender expressions—even within the same gender identity categories. Studies of mid- to later-life couples suggest that each of these elements of gender matter for couple dynamics; for example, a survey of mid-life men and women in same-sex relationships found that women with more stereotypically masculine traits did less housework, whereas men with more stereotypically feminine traits did more (Civettini, 2016). Yet few of these studies rely on well-validated measures of gender presentation, expression, or ideology, specifically for older adults—and even less so for older SGM adults. Developing new measures with multiple components of gender will advance our knowledge of how gender is relational within older couples (Springer & Mouzon, 2019). Recognizing how multiple masculinities and femininities are enacted in mid- to later-life relationships is key to moving research forward. As people in all relationship types, including different-sex couples, may enact gender in many ways, we must move beyond just talking about gender as being a man or woman and begin to discuss negotiations of masculinities, femininities, and androgynies.

Non-traditional Relationship Types

Given the diversity of identities and relationship types which have been present throughout history, it is imperative to not confine a GAR perspective to traditional dyads (e.g., committed cohabiting couples, married couples) (Schippers, 2016). Non-monogamous relationships (e.g., polyamorous relationships, open relationships) may have more complex gender configurations, but have been largely ignored in gerontological research. We do not have estimates of the number of non-monogamous relationships at older ages, yet gender discourses within these couples—especially polyamorous relationships—potentially differ from those of monogamous couples. Schippers (2016) argues that sociologists should turn to studying polyamorous relationships in order to better understand and develop new conceptual frames regarding gender, and we suggest this is especially lacking in mid- to later-life. Moving beyond a dyadic understanding of romantic relationships by including non-monogamous older couples can expand ideas about how gender is co-constructed in older ages, and how people engage in and sustain intimacy in relationships as they age.

Mixed-sexual Identity Relationships

Finally, a GAR frame does not assume that relationships composed of a man and a woman necessarily include a heterosexual man and a heterosexual woman, nor that same-sex relationships are a lesbian woman with a lesbian woman or a gay man with a gay man. There is a range of sexual orientations within same-sex and different-sex unions, including bisexual, queer, and asexual orientations. The current cohort of older adults likely has a relatively high number of mid- to later-life different-sex couples composed of sexual minority adults, in part because sexual minority status was highly stigmatized and some sexual minority older adults entered different-sex marriages in order to escape some of the stigma. Experiences of bisexual and pansexual people in relationships are especially overlooked, as these populations are typically rendered invisible within a relationship context. Yet there is evidence that bisexual people report lower relationship quality and happiness, especially if currently in different-sex unions (Thomeer & Reczek, 2016). A GAR perspective highlights relationship histories of bisexual people and other sexual minority and even heterosexual people, pointing to a need for a life course approach to understand how gender dynamics shift throughout different relationships with people of different genders.

The gender experiences of aging asexual people in relationships are also important to include. According to a recent report, 1.7 percent of sexual minority adults identify as asexual, and people who identify as asexual are just as likely to report being in a romantic relationship as non-asexual adults (Rothblum et al., 2019). Hammack, Frost, and Hughes (2019), in their argument for queering understandings of intimacy, call attention to the fact that intimacy can occur in the absence or limited experience of sexual or romantic desire, pushing scholars to look beyond conventional notions of romantic relationships and include a wider range of experiences. This queering intimacy frame acknowledges the many people who identify as asexual within relationships—recognizing that their relationships may not fit within traditional understandings yet still have important gender dynamics to unpack. This is especially important for older couples, who may see their own sexuality and sexual desires shift over time—even if not identifying as asexual—in ways that matter for their relational discourses and gender enactments (Elliott & Umberson, 2008; Lindau et al., 2007). These wide-ranging mixed-orientation relationships across the life course illustrate that various sexualities shape gender experiences and dynamics, and mixed-orientation older adult relationships are an important site in which to understand the GAR frame.

Intersectionality

As discussed above, an intersectionality perspective highlights the interlocking oppressions of ageism, sexism, heterosexism, and transphobia that shape the life chances and experiences of mid- to later-life couples. Gender is not the only important identity within a relationship, and a GAR approach may run the risk of overemphasizing gender and ignoring the role of age, sexuality, social class, race/ethnicity, ability, immigration status, and other relevant contexts. Each of these social locations—independently and in interaction—impacts the relationship dynamics of aging couples. Experiences of discrimination and stigma generally negatively shape the relationship quality of SGM-member couples, contributing to more conflict and worse relationship functioning (Petruzzella, Feinstein, & Lavner, 2019). At the same time, intimate partners can serve as a key resource for SGM people experiencing stigma and discrimination (Donnelly, Robinson, & Umberson, 2019; Rosenthal et al., 2019).

Van Eeden-Moorefield’s introduction to the 2018 special issues of Family Relations dedicated to intersectional variations in the experiences of queer families criticized that few studies expand beyond homogenous understandings of queer families, with most research only considering couples that are middle- or upper-class, college educated, cohabiting or married, gay or lesbian, cisgender, without children, and with disposable income. With a few notable exceptions, these queer couples have dominated family science journals, as well as most of the studies we have highlighted. These couples are more likely to conform to—or even reproduce—the gender dynamics of different-sex relationships. Thus, much like research on different-sex couples, research with older same-sex couples may be limited because they are built on heteronormative assumptions simply transferred to a same-sex context. This has been a long-standing critique of queer theory perspectives. Further, researchers, by using theories and assumptions based on studies of middle- and upper-class White cisgender couples and due often to their own positionality, may overlook important points of diversity.

Socio-historical contexts.

The historical and cultural context is especially important for current cohorts of mid- to later-life SGM-member couples. For example, among older couples born before and during the Baby Boom cohort, there is more age hetereogeneity within same-sex couples than different-sex couples, with the gap largest for same-sex men couples (Manning, Brown, & Stykes, 2016). This age gap likely interacts with gender and gender expectations. Aging couples may expect the younger partner to provide care for the older partner, regardless of gender, with the younger partner generally being the woman in different-sex couples and this age gap being even larger—and thus expectations perhaps more intense—for same-sex men couples.

Scholarship on mid- to later-life SGM people’s relationships must also take into account cumulative exposures to various discriminatory laws, policies, and attitudes (Frost et al., 2017; Rosenthal et al., 2019). People in older SGM cohorts grew up in and formed their relationships in distinct historical contexts such as living through the HIV/AIDS crisis and higher degrees of closeted-ness. Although the past few decades have seen federal recognition of same-sex marriage and increased visibility and acceptance of SGM persons and their relationships, this did not characterize the majority of the life course for older SGM people. As recently as 2004, 60 percent of adults in the U.S. disapproved of same-sex marriage (Pew Center, 2019). SGM people experienced decades of institutionalized discrimination if their SGM status was revealed in certain contexts, including high risk of job loss, loss of custody of children, denial of professional certifications, eviction, refusal of medical services, and harassment and arrests from law enforcement (Meyer 2015). These legal challenges have lessened, but their cumulative effects resonate within the lives of SGM older adults. Aging brings new legal issues; for example, “Religious Freedom” laws that allow some hospices to deny services to SGM people. Romantic relationships serve a critically important role in the lives of SGM people, in part because they provide social support in the midst of these lifelong experiences of discrimination and stigma (Donnelly, Robinson, & Umberson, 2019).

The research we have presented has been centered on the U.S., yet national contexts differ in terms of their gender power structures, attitudes towards SGM people, and attitudes towards mid-life and older adults (Hu & Scott, 2016; Lee, Tufiş, & Alwin, 2010; Smith, Son, & Kim, 2014). Each of these factors, as well as specific public policy and institutional structures (Collins, 2019; Sjöberg, 2004), affect enactments of gender and gender relations for mid- to later-life couples. Many aspects of the GAR framework could be applied across cultures—given that gender is important in all societies even if the enactments of gender and specific gender relations vary significantly. Thus, a complete understanding of the GAR framework will require careful attention to variation across national contexts.

GAR-Driven Methods

Finally, just as a comprehensive study of gender dynamics within aging couples requires a strong theoretical foundation, this research agenda also benefits by allowing theory to drive the choice of research methods. A GAR approach is ideally addressed through dyadic methods (i.e., collecting and analyzing data from both partners within a relationship)—or similar methods in the case of relationships with more than two partners—and methods that allow researchers to study change over time. Regarding dyadic methods, given that a GAR approach emphasizes dynamics within and between partners, having data from only one partner’s perspective is limiting, providing a biased picture of how gender operates within the couple. Dyadic data allow researchers to look within relationships to compare partners’ behaviors, reports, and perceptions across a variety of outcomes (Umberson et al., 2015). These data can be collected in multiple ways, including through surveys (Rothblum, Balsam, & Solomon, 2011a), in-depth interviews (Reczek & Umberson, 2012), ethnographies (Moore, 2008), narrative analysis (Rothblum, Balsam, & Solomon, 2011b), and relationship timelines (Thomeer et al., 2018). Dyadic analysis can be complemented by approaches that address variability within and across couples (e.g., person-centered and variable-centered approaches; Muthén & Muthén, 2000).

Qualitative dyadic approaches have been readily used, and are ideal in conjunction with a GAR approach. Depending on the goals of the study, researchers diverge in whether data are collected from both partners simultaneously, for example, through joint interviews, and thus analyze the partners’ collective responses as well as their interactions, or from data collected from each partner separately, for example, through separately administered surveys or interviews (Reczek, 2014). Regardless of data collection strategy, researchers can then analyze dyadic data to develop insight into gender dynamics within couples by considering both overlap and points of difference between partners’ accounts and outcomes. Some qualitative data analysis software allows researchers to compare coding reports from more than one person in a relationship, which is ideal for qualitative dyadic analysis. Further, it may also be important to collect qualitative data not only from dyad members, but other network members—either other romantic partners in a polyamorous relationship or others who contribute to relationship dynamics (e.g., children, parents).

The primary statistical method for analyzing dyadic survey data involves estimation of Actor-Partner Interdependence Models (APIM). When this approach is applied to the study of different-sex couples, models typically distinguish dyad members based on the gender of the partner, looking at “husband effects” and “wife effects.” APIM can also be applied to same-sex couples, even though members of these dyads are indistinguishable on gender identity. West and colleagues (2008) have addressed this challenge with the development of the factorial method, allowing the estimation of gender effects in multilevel models that include both same-sex and different-sex couples (see also Kroeger and Powers, 2019). For example, recent studies using the factorial method found that some of the specific processes around health behaviors within a relationship depend on the gender of the respondent/actor but others depend primarily on whether one is married to a man or a woman or whether in a same-sex or different-sex relationship (Holway, Umberson, & Donnelly, 2018; Umberson, Donnelly, & Politt, 2018).

In addition to dyadic methods, a GAR approach benefits from methods that allow researchers to capture data indicating dynamic—rather than static—processes, as the impact of gender within relationships shifts over time, especially within aging couples confronting health challenges and major life course transitions. We highlight three key methodological strategies: life course methods, daily diary methods, and longitudinal multi-year data. Life course methods, such as relationship timelines (de Vries et al., 2017), collect detailed information about the entire course of the relationship, often from when partners first met and including major relationship milestones (e.g., marriage, major illness), and analyze this data quantitatively and qualitatively (Frost et al., 2017; Thomeer et al., 2018). This approach allows researchers to examine how gender shapes couple-level experiences over time, placing events within historical contexts and individual biographies. Daily diary methods involve the collection of data, typically via short daily questionnaires, over a period of days or weeks (Bolger & Laurenceau, 2013). This approach is ideal for examining relationship dynamics that unfold over short periods of time (e.g., the effect of daily stress levels on relationship conflict) and has been used to examine gender differences in relationship experiences and consequences (Pollitt, Robinson, & Umberson, 2018). Finally, longitudinal multi-year datasets are rare in studies of SGM-member couples, but have been paramount in studies of different-sex cisgender-member couples in order to understand how gender dynamics unfold over time within aging romantic relationships as well as the gendered consequences of relationship dissolution. For example, longitudinal multi-year datasets provide insight into how the health consequences of widowhood are more severe for men than women by allowing comparison of the health of individuals before and after the death of their spouse (Lee & DeMaris, 2007).

There are methodological challenges to a GAR approach, especially if the goal is longitudinal analysis of nationally representative data. A key concern is achieving sufficient statistical power in order to examine multiple identities and relationship configurations. In order to expand research in this area within a GAR framework, data collection efforts must oversample SGM-member couples, ensuring that there are enough members of that subgroup in the sample for reliable estimates. Oversampling is an intentional sampling technique in which researchers include more people within specific communities that often have low prevalence in the general population or are generally hard to reach; research must then adjust for oversampling in the analysis with sample weights (See Anderssen & Malterud, 2017 and Vaughan, 2017 for more details). Respondent-driven sampling (RDS), a method that draws on social networks, also can be a useful strategy for producing unbiased population estimates, depending on whether the underlying assumptions can be met (Meyer & Wilson, 2009). Deploying this method online may be an effective strategy to reach a larger portion of a hidden or hard-to-reach population (Tourangeau et al., 2014). Datasets also need to add measures of sexual and gender identity, as well as other gender and sexuality measures, as this allows researchers to identify SGM respondents already included within existing longitudinal datasets (Morabia, 2017). Some SGM people are not visible because many surveys have historically not asked about sexual identity and rarely asked about gender identity beyond binary man/woman measures, especially among older adults (Fredriksen-Goldsen & Kim, 2015). For example, a transman with a cisgender woman would not be identified as a SGM-member couple in many surveys given the absence of measures of transgender status. Without these innovations, a GAR-driven analysis using nationally representative datasets will remain stalled.

Conclusion

This article describes a theoretical frame that can be used across disciplines, methodological approaches, and relationship types to further scientific knowledge of gender relationship dynamics of sexual and gender diverse populations in mid- to later-life. Given the rapidly shifting legal and social landscape concerning SGM populations in the U.S., gerontologists and family scholars have a unique opportunity to broaden and develop theoretical understandings of gender relationship dynamics, moving beyond the decades of research on different-sex couples with cisgender partners to include and even center the experiences of aging SGM-member couples. Even outside of the SGM context, mid- to later-life couples are experiencing rapid change due to increased rates of divorce in aging populations and more short-term relationships—including dating and cohabitation (Brown et al., 2016). These romantic relationship shifts may be accompanied by changing gender dynamics that structure the long-term romantic relationship context. A focus on a GAR theoretical perspective in aging populations can guide future data collections and empirical research aiming to more comprehensively understand romantic relationship dynamics of diverse couples. Our goal is to move towards a multi-dimensional picture of gender within mid- to later-life couples and highlight how gender is relational and dynamic as couples age. The unique socio-historical contexts of mid- to later-life SGM-member couples highlights the limitations of simply transferring what we know about romantic relationships based on different-sex couples with cisgender partners who have not faced the same historical and structural constraints as SGM-member couples or gender relational experiences, and to continue a legacy of research on SGM-member couples.