Gay Men and Lesbians: The Family Context

Lawrence A Kurdek. Handbook of Contemporary Families. Editor: Marilyn Coleman & Lawrence H Ganong. Sage Publication. 2004.

Because of the stigma associated with being identified as gay or lesbian, there are no reliable data on the number of gay and lesbian Americans. Further, whatever prevalence data are available vary according to whether “homosexuality” is defined in terms of behavior, attraction, or identification. Perhaps the most reliable data come from Laumann, Gagnon, Michael, and Michaels’s (1994) National Health and Social Life Survey of 3,432 Americans who were interviewed face to face in 1992 on the basis of probability sampling. Laumann et al. found that 4.9% of men and 4.1% of women reported same-gender sex partners since the age of 18, 6.2% of men and 4.4% of women reported some attraction to same-sex persons, and 2.8% of men and 1.4% of women reported identifying with a label denoting same-gender sexuality. Hewitt (1998) reports similar prevalence rates from other surveys. Even though these estimates are well below the 10% figure usually associated with the Kinsey data (Kinsey, Wardell, & Martin, 1948; Kinsey, Wardell, Martin, & Gebhard, 1953) and were obtained from data collected in 1992, it is clear that a sizable number of Americans and their family members deal with issues regarding being gay or lesbian.

Although there is currently active scholarly interest in the multidimensional character and assessment of homosexuality (e.g., Peplau, 2001), the developmental process of constructing a gay or lesbian identity (e.g., Dube, 2000), attitudes and prejudice toward gay men and lesbians (e.g., Chen & Kenrick, 2002), the effects of HIV-positive status on relationships (e.g., Hatala, Baack, & Parmenter, 1998), validation of existing measures (e.g., Greenfield & Thelen, 1997) or findings (e.g., Buunk & Dijkstra, 2001) with gay and lesbian samples, and use of data from gay and lesbian individuals to test hypotheses derived from theories of evolutionary psychology (e.g., Buunk & Dijkstra, 2001), the focus of this chapter is on the intersection between the study of families and the study of gay men and lesbians. In keeping with this theme and the space limitations of this volume, I review selectively what is known and what needs to be known in two content areas pertinent to this general area of study. The first deals with gay and lesbian adolescents and the second with gay and lesbian couples.

Gay and Lesbian Adolescents and their Families

What is Known

Perhaps the first time family members directly confront issues relevant to homosexuality is when another family member freely discloses his or her identity as a gay man or lesbian or has such an identity discovered. The best-studied scenario is one in which an adolescent’s sexual orientation is made known to his or her family members. Although adolescence is a developmental period during which many aspects of one’s identity are in flux (Graber & Archibald, 2001), an increasing number of adolescents are claiming a gay or lesbian identity (Savin-Williams, 1998). This cohort effect may be due to general social-cultural changes in which attitudes toward homosexuality have become relatively less negative. Nonetheless, it is sobering to note that data from the 1998 General Social Survey indicate that more than 50% of the 25,668 respondents thought that sexual relations between two adults of the same sex are always wrong (Butler, 2000) and that data from a randomly selected nationally representative sample collected in the year 2000 indicate that 38% of the 2,283 respondents completely agreed that homosexual behavior is morally wrong (Kaiser Family Foundation, 2001).

As one might expect from problems in determining the number of gay and lesbian adults, there are also no conclusive data on the number of gay and lesbian adolescents. Remafedi, Resnick, Blum, and Harris (1992) found in a survey of 34,706 Minnesota students from Grades 7 through 12 (49.8% male and 50.2% female) that 0.7% of the males and 0.2% of the females described themselves as mostly or totally homosexual and that 0.8% of the males and 0.9% of the females described themselves as bisexual. In a national study of 12- to 19-year-old high school students (5,758 males and 6,182 females), Russell, Seif, and Truong (2001) found that 0.7% of the males and 1.5% of the females indicated exclusive same-sex attraction and that 6.5% of the males and 3.7% of the females indicated bisexual attractions. Because the sexual identity of young adolescents is unlikely to be stable (Diamond, 2000), it is unclear how many of these adolescents will retain a gay or lesbian identity into adulthood.

The “coming out” process of adolescents has been of interest to family scholars because it poses issues for parent-child relationships, because gay and lesbian youth are at risk for negative consequences associated with the verbal and physical abuse they are likely to receive, and because heterosexual family members confront their own issues regarding identifying themselves as being related to a gay or lesbian individual. With regard to parent-child relationships, although relevant data are not extensive (see Rotheram-Borus & Langabeer, 2001, and Savin-Williams, 2001b, for reviews), the initial reactions that parents have to their child’s disclosing a gay or lesbian identity are quite diverse including disbelief, denial, shock, anger, and lack of surprise but are typically neither extremely positive (immediate acceptance) nor extremely negative (dismissal from the home or physical abuse). There is also evidence (Maguen, Floyd, Bakeman, & Armistead, 2002; Savin-Williams, 2001b) that gay and lesbian adolescents are more likely to disclose their sexual orientation to friends than to parents, that if adolescents do disclose their orientation to a family member it is more likely to be to mothers than to fathers, and that mothers react more positively than fathers do.

Regarding the at-risk status of gay and lesbian youth, Rivers and D’Augelli (2001) noted that gay and lesbian adolescents deal with unique stressors in addition to those experienced by adolescents in general. These include feeling different, perceiving little family support, and viewing HIV and AIDS as gay diseases. Among the most unfortunate consequences of these stressors is the likelihood that gay and lesbian adolescents will be the targets of verbal abuse (Thurlow, 2001) and physical abuse (Rivers & D’Augelli, 2001; Savin-Williams, 1994). Such abuse, in turn, may lead gay and lesbian youth to be at risk for school problems, running away, substance abuse, prostitution, and attempts at suicide (Russell et al., 2001; Savin-Williams, 2001c).

Finally, although coming out is usually used to describe the process of one’s coming to terms with one’s own gay or lesbian sexual identity, it is increasingly also being used to describe public acknowledgment of being related to a gay or lesbian individual. Crosbie-Burnett, Foster, Murray, and Bowen (1996) developed a social-cognitive behavioral model to integrate the multiple aspects of how families adjust to the news of having a gay/lesbian family member. This model describes the coming out process with regard to subsystems in a family that include a mother, a father, a gay/lesbian child, and a heterosexual child (e.g., the gay child-mother/father subsystems, the heterosexual child-mother/father subsystems, the marital subsystem, and the sibling subsystem) and links the family of origin to the extended family (such as cousins, aunts, uncles, and grandparents), the community (such as school and church), and other social contexts (such as the workplace).

What Needs to be Known

Because empirical research on gay and lesbian adolescents is relatively new, many specific topics need further study (see Savin-Williams, 2001b, for a detailed research agenda). Three are highlighted here.

The Antecedents of a Gay or Lesbian Identity

The fact that many adolescents claim a gay, lesbian, or bisexual identity indicates that the factors determining sexual orientation are likely to operate fairly early in development. Further, whatever factors are responsible for sexual orientation are also likely to be complex. For example, although there is evidence that prenatal estrogen contributes to the development of a homosexual orientation in women (Meyer-Bahlburg et al., 1995), that gay adolescents have a greater than average number of male siblings and a later than average birth order than nongay counterparts (Blanchard, Zucker, Bradley, & Hume, 1995), and that childhood cross-sex-typed behavior is predictive of adult homosexual orientation for both men and women (Bailey & Zucker, 1995), the relevant effects are not large and, consequently, argue against any single factor determining sexual orientation. Simply put, some children or adolescents with all of the relevant characteristics just listed will not develop into adults who claim a gay, lesbian, or bisexual orientation. Further, some individuals may recognize same-sex attractions but not engage in same-sex behaviors, and some individuals may engage in same-sex behaviors but not incorporate that dimension into their sexual identity (Savin-Williams, 2001a). Following the lead of scholars such as Horowitz and Newcomb (2001) and Peplau (2001), one avenue for future research is to adopt the perspective that human sexuality is fluid, malleable, and dependent on context. This perspective would be a useful conceptual basis for prospective studies attempting to describe the complex pathways by which sexual attractions, behaviors, and identities are shaped.

The Ecology of the Coming Out Process

Despite the comprehensive nature of Crosbie-Burnett et al.’s (1996) model, surprisingly few researchers have used it to guide the study of how a family member’s coming out affects other family members and the social contexts within which the family is embedded (however, see Oswald, 2000). The Crosbie Burnett et al. model provides a rich framework for future studies examining how the identification of a gay or lesbian family member leads other family members to revise schemas about both themselves (e.g., I am the father of a gay son and may not be a grandparent) and the gay/lesbian person (e.g., My son will not get married and probably will not raise children) as well as how issues within the family (such as the secret of having a gay or lesbian member; see Vangelisti & Caughlin, 1997) spill over to the wider social context. The Crosbie-Burnett model can also be extended to include the study of older individuals who experience the coming out process after adolescence (e.g., women who explore lesbian identities after having been married; Morris, Balsam, & Rothblum, 2002).

Resilience in Gay and Lesbian Youth

Although much of the work done with adolescents has addressed the ways in which gay and lesbian youth are at risk for a variety of problems, there has also been a call (e.g., Oswald, 2002; Savin-Williams, 2001c) for future researchers to examine how gay and lesbian youth, despite the psychological and social stressors they experience, manage to negotiate the developmental tasks of adolescence and young adulthood and achieve a stable identity in adulthood. Although personality factors (such as psychological hardiness) may be important, coping mechanisms and social support are also likely to be highly relevant.

Gay and Lesbian Cohabiting Couples

What we Know

Estimating the number of American gay and lesbian couples is even more problematic than estimating the number of gay/lesbian individuals because presenting oneself publicly as part of a gay or lesbian couple may open the door for discrimination, abuse, and even violence (Bryant & Demian, 1994; Kaiser Family Foundation, 2001). Thus, current estimates regarding the number of gay and lesbian couples are likely to be underestimates of the actual number. Still, on the basis of data from the 2000 U.S. Census (U.S. Bureau of the Census, 2002), 301,026 households were headed by a male householder and a male partner, and 293,365 households were headed by a female householder and a female partner. Contrary to popular myths (Herek, 1991) that lesbians and gay men are not capable of forming durable cohabiting relationships, surveys indicate that between 45% and 80% of lesbians and between 40% and 60% of gay men are currently involved in a romantic relationship (Bradford, Ryan, & Rothblum, 1994; Falkner & Garber, 2002; Morris et al., 2002). More importantly, several authors (Blumstein & Schwartz, 1983; Bryant & Demian, 1994; Falkner & Garber, 2002; Kurdek, 2003; “The Advocate Sex Poll,” 2002) report that between 8% and 21% of lesbian couples and between 18% and 28% of gay couples have lived together 10 or more years. Clearly, despite a general social climate of prejudice against gay men and lesbians, being part of a couple is integral to the lives of many lesbians and gay men.

The limited information we have regarding lesbian and gay couples can be used to answer four general questions: (a) What is the ethnography of gay and lesbian relationships? (b) Do gay and lesbian couples differ from heterosexual couples? (c) Do gay and lesbian couples differ from each other? and (d) How does the experience of parenting affect couple-level functioning? Each of these questions is addressed in turn.

A General Ethnography of Lesbian and Gay Couples: What are Gay and Lesbian Couples Like?

Because of the social stigma surrounding homosexuality, lesbians and gay men receive little if any information regarding lesbian and gay couples in the course of their socialization. As a result, lesbian and gay close relationships develop without consensual norms (Laird, 1993). It should be underscored at the outset that most of the available descriptive data on lesbian and gay couples come from relatively young, white, and well-educated volunteer respondents. Thus, the “typical couple” profiles presented below probably apply to only a select group of lesbian and gay couples. The most detailed descriptive account of lesbian and gay couples comes from Bryant and Demian’s (1994) national survey of 1,749 individuals who represented 706 lesbian couples and 560 gay couples. Selected findings of the study are presented by survey topic. Unless indicated otherwise, the data come from the Bryant and Demian survey.

Previous Relationships. Do members of gay and lesbian couples have extensive previous experience with heterosexual relationships? Many of the gay and lesbian respondents (38% and 32%, respectively) indicated that their current relationship was their first major lesbian/gay relationship. Lesbians were more likely than gay men to have been involved in a heterosexual marriage (27% vs. 19%, respectively). Consistent with the latter finding, Blumstein and Schwartz (1983) reported that 22% of their lesbian respondents and 15% of their gay respondents had been previously married, and Morris et al. (2002) reported that 25% of their sample of lesbian women had been married. However, it is unclear how previous experiences with heterosexual relationships and marriage affect the development of gay and lesbian couples.

Name Used to Identify Partner. Given that gay and lesbian relationships are not socially legitimized, one might expect that there would be little consensus on how members of these couples refer to each other. Most gay men (40%) called their partner “lover,” whereas most lesbians (35%) referred to their partner as “partner” or “life partner.” McWhirter and Mattison (1984) also reported that most members of gay couples referred to each other as “lover.” It appears that no studies have addressed how the simple act of naming one’s partner affects relationship functioning, although heterosexual couples typically go through a stagelike sequence of changing the way they refer to their partner as the relationship progresses (e.g., “date” to “boyfriend/girlfriend” to “fiancé/fiancée” to “husband/wife”).

How Partner Was Met. Because one’s sexual identity is often kept a secret and is not readily detected, how do possible partners meet each other? Lesbians were most likely to have met their partner through friends (28%), at work (21%), or at a social event (16%). Gay men were most likely to have met their partner at a bar (22%), through friends (19%), or at a social event (13%). McWhirter and Mattison (1984) also mentioned that the first contact between members of a gay couple was likely to occur in a gay establishment.

Relationship Rituals. Because gay and lesbian couples cannot legally get married in the United States, they do not have readily accessible strategies for formalizing their relationships. Nonetheless, 57% of the lesbians and 36% of the gay men reported that they wore a ring or some other symbol to represent their relationship. Nineteen percent of the lesbians and 11% of the gay men had held some type of commitment ceremony. The current trend toward reporting the union of same-sex couples in “marriage” announcements suggests that gay and lesbian couples may be receiving stronger social validation now than in the past, though not without controversy.

Residence. Cohabitation may be a significant event in the course of any relationship because it marks a milestone in the extent to which each partner identifies with the relationship. Most of the couples (82% male, 75% female) had lived together during the previous year of the relationship. Thirty-six percent of the male couples and 32% of the female couples jointly owned their residence, whereas 27% and 33%, respectively, jointly rented or leased their residence. Housing discrimination was reported by 15% of the renters and 9% of the homeowners. McWhirter and Mattison (1984) reported that most of the male couples they studied moved in together after about 1 month of acquaintance.

Finances. Although living together can certainly be seen as a public statement of being a couple, how partners arrange finances can also be seen as a statement of the extent to which the partners are invested in the longevity of the relationship. Eighty-two percent of the gay male couples and 75% of the lesbian couples shared all or part of their incomes. Both Blumstein and Schwartz (1983) and McWhirter and Mattison (1984) reported that the probability of a couple’s pooling finances increased with the length of time living together.

Division of Household Labor. One clear indication that members of gay and lesbian couples construct their own relationships is that, unlike members of heterosexual couples who may use gender as a way of determining who does what household tasks, members of gay and lesbian couples begin their relationships with no preset ideas of how household labor will be divided. In their study of gay couples, McWhirter and Mattison (1984) noted that the handling of household chores varied by stage of the relationship. In the first year of the relationship, partners shared almost all chores. Later, however, routines got established as chores were assigned primarily on the basis of each partner’s skill and work schedule. In instances where each partner was skilled, partners willingly unlearned previous skills to create complementarity and a sense of balance in the relationship. Blumstein and Schwartz (1983) and Kurdek (1993) noted that lesbian couples at any stage of the relationship were particularly likely to divide household labor equally. Blumstein and Schwartz speculate that lesbians may avoid task specialization in the area of household work because of the low status traditionally associated with the women who do it. This speculation is consistent with other reports that lesbian couples are more likely than either gay or heterosexual couples to follow an ethic of equality (Chan, Brooks, Raboy, & Patterson, 1998; Clunis & Green, 1988; Peplau & Cochran, 1990).

Conflict. One consequence of the interdependence experienced by relationship partners is that conflict is almost certain to arise.

Most of the lesbian and gay male respondents (28% and 26%, respectively) reported two small arguments per month. The percentage of lesbians reporting verbal abuse and physical abuse from their partners was 17% and 3%, respectively. Corresponding values for gay men were 15% and 3%. With regard to the content of conflict, Kurdek (1994a) found that managing finances, driving style, affection/sex, being overly critical, and household tasks were troublesome areas for both gay and lesbian couples.

Social Support. Given the stigma associated with being gay or lesbian, traditional sources of support for members of a couple (such as parents) may not exist. Indeed, both gay men and lesbians indicated that the major source of social support for their relationships came from other gay/lesbian friends, followed in order by siblings, mother, and father. Most gay men (43%) and most lesbians (47%) were also most likely to turn to friends for help with their relationships. Similar findings were reported by Kurdek (1988).

Legal Arrangements. Because gay and lesbian couples enjoy no “default” legal privileges, members of these couples often protect their relationships in a proactive manner. Thirty-two percent of the lesbian respondents and 39% of the gay male respondents had executed a will, and 28% of the lesbian respondents and 27% of the gay respondents had made arrangements for power of attorney.

Stages of Relationship Development. In recognition of the fact that gay and lesbian couples, just like heterosexual couples, undergo rather systematic changes in their relationships, McWhirter and Mattison (1984) have described the development of gay couples, and Clunis and Green (1988) have described the development of lesbian couples. McWhirter and Mattison derived their stage model from a cross-sectional study of 156 predominantly white, well educated male couples. Clunis and Green modified and expanded the McWhirter and Mattison stages on the basis of their experiences as therapists. Both sets of authors noted that not all couples fit the stage model, and neither set tested their model with longitudinal data.

McWhirter and Mattison proposed that gay male couples develop in a six-stage sequence. The Blending Stage occurs in the first year and is characterized by merging, limerance (e.g., intense preoccupation with and longing for the partner), shared activity, and high sexual activity. Years 2 and 3 (Nesting) are marked by homemaking, finding compatibility, the decline of limerance, and ambivalence about the relationship. Recognizing individual needs, expressing dissatisfactions, dealing with conflict, and establishing traditions occur during the Maintaining Stage (Years 4 and 5). The Building Stage happens during Years 6 through 10 and includes collaborating, increasing individual productivity, establishing independence and individual habits, and acknowledging the dependability of partners. Years 11 through 20 are described as a Releasing Stage that involves trusting, merging of money and possessions, midlife evaluations of priorities, and taking each other for granted. The last stage, Renewing, occurs after 20 or more years of cohabitation and is characterized by achieving financial and emotional security; shifting perspectives regarding time, health, and loss; restoring romance in the partnership; and remembering events in the relationship history.

Unlike McWhirter and Mattison (1984), Clunis and Green (1988) begin the description of their developmental stage model with a Prerelationship Stage, which refers to the time during which partners decide whether to invest time and energy in getting to know each other better. This is followed by a Romance Stage that has many of the characteristics of McWhirter and Mattison’s Blending Stage and then a Conflict Stage that is similar to McWhirter and Mattison’s Maintaining Stage. With the Acceptance Stage comes a sense of stability and an awareness of the faults and shortcomings of each partner. This is followed by the Commitment Stage, in which partners work on balancing opposing needs while accepting each other as trustworthy. In the Collaboration Stage, the couple works to create something together (e.g., a baby, a business venture) in the world outside the relationship that enhances the relationship.

In a test of the McWhirter and Mattison’s (1984) stage model, Kurdek and Schmitt (1986) used cross-sectional data to examine differences in multiple assessments of relationship quality for gay, lesbian, and heterosexual couples from the first three stages of the model (Blending, Nesting, and Maintaining). Regardless of type of couple, findings were consistent with McWhirter and Mattison’s prediction that the second and third years of the relationship were likely to involve stress and disillusionment. Relative to Blending couples and Maintaining couples, Nesting couples reported the lowest satisfaction with affection and sex, the lowest amount of shared activity, and the most frequent number of dysfunctional beliefs regarding sexual perfection.

Sexual Behavior. Although the defining feature of a gay or lesbian relationship is some type of sexual behavior with another member of one’s own sex, the extant literature on the sexual behavior of gay and lesbian couples is surprisingly sparse. Blumstein and Schwartz (1983) and McWhirter and Mattison (1984) found no evidence that lesbian partners and gay partners regularly assume “active” and “passive” roles in sexual interactions. However, there is limited evidence regarding the specific types of sexual activity engaged in by gay and lesbian couples. McWhirter and Mattison reported that over 90% of the gay couples they interviewed engaged in kissing and hugging, body rubbing and kissing, tongue kissing, fellatio, being fellated, mutual fellatio, and mutual masturbation. Seventy-one percent of the respondents reported engaging in anal intercourse, and 41% reported engaging in analingus. Eighty-three percent of the respondents reported they were satisfied with their sexual relationship, and 91% reported that their level of satisfaction had increased since the beginning of the relationship. Of the 81 respondents who reported sexual problems, the most common (43%) was erectile failure.

Loulan (1987) reported the results from survey data gathered from 1,566 lesbians, 62% of whom were coupled. At least 80% of the respondents reported that they did the following activities to their partners: touching breasts, kissing breasts, licking breasts, and putting fingers in vagina. Seventy-one percent of the respondents reported that they performed oral sex on their partner; 56% said they put their tongue in their partner’s vagina; and 55% said they masturbated their partner. Diamant, Lever, and Schuster (2000) reported similar findings regarding sexual activities in their sample of 6,935 lesbians.

Do Gay/Lesbian Couples Differ from Heterosexual Couples?

Despite an increased scientific interest in gay and lesbian couples (e.g., see review by Peplau & Spaulding, 2000), systematic comparisons of members of both gay and lesbian couples to members of heterosexual couples (as well as comparisons of members of gay couples to members of lesbian couples) have been characterized by several methodological problems. These include studying only one partner from the couple (e.g., Peplau, Cochran, & Mays, 1997); averaging individual scores from both partners (e.g., Schreurs & Buunk, 1996); using measures with unknown psychometric properties (e.g., Blumstein & Schwartz, 1983); performing qualitative analyses of in-depth interviews that although providing a rich description of the lives of some gay men and lesbians make it difficult to determine the size and generality of any effects (e.g., Carrington, 1999); comparing couples without first ensuring that the members of these couples were equivalent on demographic characteristics such as age, education, income, and length of relationship (e.g., Haas & Stafford, 1998); treating members of the couples as independent units of analysis (e.g., Mackey, Diemer, & O’Brien, 2000); and recognizing the interdependence of scores derived from the same couple but performing statistical analyses that did not take this interdependence into account (e.g., Gaines & Henderson, 2002).

Kurdek (2001) addressed these methodological limitations by collecting longitudinal data from both members of gay, lesbian, and heterosexual married couples with psychometrically sound measures, providing estimates of effect sizes, using demographic variables (e.g., age, education, income, and months living together) as control variables, and employing statistical analyses (multilevel modeling; see Kenny, Mannetti, Pierro, Livi, & Kashy, 2002) that took partner interdependence into account. Because children are known to affect marital functioning (e.g., Erel & Burman, 1995) and because the gay and lesbian couples studied did not have children, members of both gay and lesbian couples were compared to members of married couples without children on sets of individual differences, relationship-related attitude, conflict resolution, and social support variables known to be linked to relationship quality (Huston, 2000; Karney & Bradbury, 1995) as well as on global appraisals of the relationship.

Overall, few differences between members of gay/lesbian couples and members of non-parent couples were found, and most of these effects were small. Nonetheless, some specific differences were notable. With regard to individual-differences variables, only 9 of the 36 effects (25%) were significant. Relative to members of married couples, members of both gay and lesbian couples had higher private self-consciousness (focus on their own thoughts and feelings, perhaps due to their stigmatized status), were more comfortable with closeness (perhaps because many were in long-term relationships without much institutional or social support), and were higher in openness (proclivity toward variety, intellectual curiosity, and aesthetic sensitivity, perhaps due to a propensity to explore an alternative lifestyle). With regard to relationship-related attitudes, only 6 of the 22 effects (27%) were significant, with members of both gay and lesbian couples reporting more autonomy than members of married couples (perhaps reflecting high levels of agency and instrumentality for economically advantaged gay and lesbian partners) and in the one of the strongest effects in the studywith members of lesbian couples reporting higher levels of equality than members of married couples (perhaps due to lesbians having encountered inequities more frequently in their past relationships).

Although members of the three types of couples were relatively equivalent on the set of conflict resolution variables (only 1 of the 16 effects [6%] was significant), consistent effects were obtained for the set of social support variables (for which 6 of the 10 effects [60%] were significant), with members of gay and lesbian couples reporting less support from family members than members of married couples (probably due to the stigma of being gay or lesbian) and members of lesbian partners reporting more support from friends than members of married couples (perhaps due to lesbians’ living within close geographical proximity to other lesbians). Finally, with regard to global appraisals of the relationship, only two of the eight effects (25%) were significant, with members of lesbian couples reporting higher satisfaction than members of married couples (perhaps the result of two “doses” of women’s relationship-enhancing socialization experiences) and members of gay couples reporting less commitment than members of married couples (perhaps the result of two “doses” of men’s self-enhancing socialization experiences).

In sum, although some significant effects were obtained and interpretable, the bulk of the evidence clearly indicated that members of gay/lesbian couples were more similar to than different from members of married couples without children, with only 24 of the total of 92 effects (26%) being significant. Additional evidence of the general similarity among members of gay/lesbian couples to members of heterosexual couples comes from reports of no differences between members of gay/lesbian couples and members of married couples in the areas of conflict and conflict resolution styles (Kurdek, 1994a, 1994b), the rate of change in relationship satisfaction (Kurdek, 1998), and adjustment to relationship dissolution (Kurdek, 1997a) as well as in the predictors of relationship quality (Kurdek, 1997a, 1997b, 1997c, 2000), change in relationship quality (Kurdek, 1998), and dissolution (Kurdek, 1998). However, it should be noted that these findings are not independent of each other because they are based on the same longitudinal samples.

Do Gay and Lesbian Couples Differ from Each Other?

Using almost the same sets of variables as those used in comparing members of gay/lesbian couples to members of married couples, Kurdek (2003) compared members of gay couples to members of lesbian couples at multiple points over time. Again, demographic variables were used as control variables, comparisons used statistical analyses that took partner interdependence into account, and effect sizes were calculated. Overall, even fewer differences were found between members of gay couples and members of lesbian couples than were found between members of gay/lesbian couples and members of nonparent, married couples, and what differences were found were mostly small.

Only 1 of the 22 effects (4%) for the set of individual differences variables was significant, and no effects were found for the sets of eight conflict resolution variables and the set of five social support variables. However, three of the eight effects (37%) for the set of relationship-related attitude variables were significant, with members of lesbian couples reporting stronger liking, trust, and equality than members of gay couples. These differences are consistent with the view that lesbians create relationships in which women’s relationship-enhancing influences are activated and potentiated (Ferree, 1990) and that lesbian partners as women are likely to redress the imbalances in power and influence that they are likely to have experienced in their relationship histories (Fox & Murry, 2000). As additional evidence of the general similarity between members of gay couples and members of lesbian couples, no differences between members of these two types of couples have been found in level of similarity between members within the couple (assortative mating or homogamy; Kurdek, 2003), areas of conflict and conflict resolution styles (Kurdek, 1994a, 1994b), the rate of change in relationship quality (Kurdek, 1996, 2003), the predictors of relationship quality (Kurdek, 1996, 2003), the prevalence of dissolution (19% for gay couples and 23% for lesbian couples, Kurdek, 2003), and adjustment to relationship dissolution (Kurdek, 1997a). However, it should be noted that these findings are not independent of each other because they are based on the same longitudinal samples.

How Does Parenting Affect Couple-Level Functioning?

Consistent with the questionable reliability of other prevalence data already presented, there are no reliable data on the number of children living part time or full time with gay and lesbian couples. In the Bryant and Demian (1994) sample, 21% of the lesbians and 9% of the gay men reported “caring” for children, although it was not clear whether these children resided with the couple (20% of Morris et al.’s 2002 lesbian sample also had children). These children were offspring from a previous marriage for 74% of the lesbians and for 79% of the gay men. Thirteen percent of the lesbians were impregnated through donor insemination. Ten percent of the lesbians and 4% of the gay men planned to have children. Blumstein and Schwartz (1983) reported that 7% of their lesbian couples had children living with them more than 6 months per year. Unfortunately, gay couples were not asked about children. Eleven percent of the respondents from the Kaiser Family Foundation (2001) study reported having children under age 18 living in the household, but separate numbers were not given for gay and lesbian respondents.

Much of the work with regard to gay and lesbian parents has focused on the adjustment of children growing up with gay or lesbian parents, and the relevant studies have almost exclusively involved lesbian mothers (see reviews by Patterson, 2000, and Stacey & Biblarz, 2001). Two kinds of families have been studied: one in which children were born before the mother began a relationship with another woman (e.g., in the context of marriage which was ended through divorce) and the other in which children were born during that relationship (usually by means of donor insemination).

From a parenting transition perspective (Capaldi & Patterson, 1991), the first kind of family has much in common with families in which children experience both parental divorce and the remarriage of at least one parent. However, few studies (e.g., Lynch, 2000) have compared children in the two forms of “remarried” families. Instead, most studies involve comparing children of custodial lesbian mothers to children of custodial heterosexual mothers. In her review of these studies, Patterson (2000) noted that the majority of studies have found no differences between children with the two types of mothers on measures of sexual identity, personal development, and social relationships. Stacey and Biblarz (2001), however, pointed to a pattern in some of these studies (e.g., Green, Mandel, Hotvedt, Gray, & Smith, 1986) that shows that children living with lesbian mothers are likely to internalize non-traditional gender roles. Patterson also noted that few studies have revealed differences in the adjustment of the two types of mothers.

The second kind of family has been of particular interest because children in these families cannot be expected to suffer any of the negative consequences (e.g., Capaldi & Patterson, 1991) associated with marital conflict, divorce, repartnering, and, possibly, the mother’s coming out process. Stacey and Biblarz (2001) noted, however, that because donor insemination is an expensive process, it may be available to only economically privileged couples. Consequently, the generalizability of relevant findings can be questioned. Nonetheless, as with findings for the first kind of family, those for this type of family show no adverse effects associated with living with a lesbian couple. In fact, there is some evidence that the lesbian co-parent (i.e., the member of the couple who does not give birth) is more skilled and invested in parenting than fathers typically are (e.g., Brewaeys, Ponjaert, Van Hall, & Golombok, 1997). To the extent that lesbian couples represent a “double dose” of female-linked socialization experiences and expectations, this pattern of enhanced parenting for lesbian couples is similar to the pattern of enhanced relationship quality for lesbian couples reported earlier.

What we Need to Know

Although there are many gaps in our knowledge of lesbian and gay couples, only eight will be highlighted.

Recruiting Nationally Representative Samples

One major advance in the study of gay and lesbian couples is the recruitment of fairly large samples of gay and lesbian individuals, some of whom are in relationships. The most common strategies used in recruitment have been to distribute questionnaires to visitors to gay and lesbian bookstores, community centers, and political/social groups (e.g., Morris et al., 2002); to send questionnaires to individuals on mailing lists for gay/lesbian political, social, and health care organizations (e.g., Peplau et al., 1997); to request participants from notices placed in gay/lesbian periodicals (e.g., Diamant et al., 2000) or gay/lesbian online sites (Falkner & Garber, 2002); and to use snowball sampling from friendship networks (e.g., Kurdek, 2003). Although these recruitment strategies result in national samples, they are unlikely to be representative samples. However, because many members of the larger gay/lesbian population are hidden and because information about sexual orientation is not obtained in government-sponsored surveys, it is extremely difficult to assess whether any sample of gay and lesbian persons is representative. This dilemma might be addressed if census forms make it easier for same-sex couples to identify themselves, if antigay/lesbian sentiments continue their gradual decline, and if innovative sampling strategies (e.g., the adaptive sampling approach described by Blair, 1999) continue to be devised.

Roles, Scripts, and Structure

In the absence of conventional partner roles (such as husband and wife), it is unclear how members of lesbian and gay couples create predictability in their relationships (Klinkenberg & Rose, 1994). Do they seek out “established” couples for advice? Do they construct roles on a trial-and-error basis or on the basis of which tasks need to be performed (McWhirter & Mattison, 1984)? One way to obtain information on these issues would be to conduct a longitudinal study of lesbian and gay couples that starts when partners first begin dating and to document how roles in the relationship get established and change over time. As a beginning step in this direction, Klinkenberg and Rose (1994) showed that gay and lesbian individuals followed well-defined scripts for their first dates. The scripts for gay men tended to include sexual behavior more frequently than those of lesbians, whereas the scripts of lesbians included intimacy and sharing of emotions more frequently than those of gay men. Relative to the scripts for heterosexual women, those for lesbians did not include telling parents about the date and monitoring the date with regard to safety concerns.

The Construction of Commitment

Work with heterosexual couples has indicated that partners employ several psychological strategies to enhance commitment to maintaining their relationships. These include a tendency to accommodate and to forgive rather than to retaliate in response to a partner’s transgression, to sacrifice desirable activities when the partner does not want to engage in these activities, to derogate tempting alternatives to the relationship, and to produce relationship-enhancing illusions (e.g., Finkel, Rusbult, Kumashiro, & Hannon, 2002). Given the striking similarities found in the functioning of gay/lesbian and heterosexual couples summarized above, there is no reason to expect that these strategies would not also be important for gay and lesbian couples. However, because commitment can also be enhanced by forces outside the relationship (“constraint commitment”; Kurdek, 2000), it would be of interest to see if relationship outcomes for gay and lesbian couples are affected by the degree to which members of social systems that include one or both partners of the couple explicitly support the relationship (Oswald, 2002). These systems might include families of origin, friends, co-workers, neighbors, and members of the immediate heterosexual and lesbian/gay community. Because gay and lesbian partners experience little institutionalized external support for their relationships, it would also be of interest to see if the link between psychological maintenance strategies and commitment is moderated by the degree of perceived external constraints to leaving the relationship. It is possible that this link is especially strong when few external constraints are perceived.

Relationship Tensions

Because resolving conflict constructively is critical to relationship satisfaction and relationship stability (Karney & Bradbury, 1995), it would be useful to examine dyadic processes in conflict resolution for gay and lesbian couples using methods other than self-report ones (Kurdek, 1994b). Kollock, Blumstein, and Schwartz (1985) audiotaped conversations of members from gay, lesbian, and heterosexual couples in which members were to arrive at a joint decision. They found that regardless of type of couple, the partner with the greater power (the one who had more influence in daily decisions) was more likely to interrupt the speech of the less powerful partner. In what appears to be the only published study using videotaped observations of conflict interaction, Gottman et al. (in press) found that initiators of conflict discussion from gay and lesbian couples were less belligerent and less domineering than those from married couples, that gay and lesbian partners were more positive in how they received conflict-related information than married partners, that gay partners had trouble repairing conflict when their partners become too negative, and that lesbian partners were more expressive of both positive and negative affect than gay partners. These intriguing findings need to be replicated and extended. Also in need of empirical support is Patterson and Schwartz’s (1994) speculation that gay and lesbian couples are more likely than heterosexual couples to control consciously the level of conflict in their interactions in order to ensure the survival of their relationships.

Sexual Behavior

Sexual behavior in gay and lesbian couples has not been widely studied. The limited data on this topic are of note because of all of the areas of couplehood studied by Blumstein and Schwartz (1983), the largest differences between lesbian and gay couples occurred in this area. Blumstein and Schwartz found that gay couples were more sexually active than lesbian couples in the early years of the relationship. After 10 years together, however, gay couples had sex less frequently and often devised explicit arrangements for sexual activity outside the relationship. Gay couples’ acceptance of sexual nonexclusivity is one of the most distinctive features of their relationships (“The Advocate Sex Poll,” 2002), yet we know little regarding how nonexclusivity affects the relationships of gay men in light of the HIV/AIDS epidemic (Hatala et al., 1998).


Because lesbian couples have only recently chosen donor insemination as a strategy for becoming parents, most of the children studied in these families are fairly young. Longitudinal studies that track the development of these children through adolescence and young adulthood would be of interest because of speculation (e.g., Patterson, 1992) that adolescents who are negotiating their own sexual identities and are prone to the positive and negative effects of peer contact may have some difficulty openly acknowledging that they have a lesbian parent. Given that some lesbian couples are likely to separate (Kurdek, 2003), information is also needed on how children are affected by the conflict that is likely to precede the separation, how children adjust to the separation, and how children are affected by the possible repartnering of both the birth mother and the lesbian co-parent. To the extent that gay and lesbian couples have been used to test tenets of evolutionary psychological theories (Buunk & Dijkstra, 2001), “divorcing” lesbian couples are of particular interest with regard to child custody and visitation issues because a couple with two women which represents a possible double dose of evolutionarily based parenting investment may have stronger competing claims about parenting issues than a couple with a man and woman. Finally, although the number of gay couples raising children full time is small, a larger number of gay couples may include one partner who has children from a previous marriage. These children are likely to live primarily with their mothers but may have regular contact with their gay fathers and their father’s partner. We know little about how these families function and change with the development of the child involved and how these children come to understand the nature of their father’s relationship.

Diversity within Diversity

To date, the study of lesbian and gay couples has primarily attended to the type-of-couple variable as a moderator variable (e.g., is the link between Relationship Process X and Relationship Outcome Y as strong for lesbian and gay couples as for heterosexual couples?). Although these studies are important for documenting the robustness of relationship process across diverse types of couples, they necessarily involve treating each type of couple as a homogeneous group. This is unfortunate in light of evidence that the initiation, development, and dissolution of relationships are affected by the personal-social-cultural-historical context within which members of a couple develop (Huston, 2000). But how does one identify what kinds of diversity are important within each type of couple? For gay and lesbian couples, there is evidence (see Friend, 1991; Greene, 1997) that older lesbians and gay men as well as lesbians and gay men of color are likely to experience rejection and forms of discrimination from other gay men and lesbians as well as from the heterosexual majority. In addition, geographical location is likely to affect relationship development inasmuch as some areas of the country are more repressive or more liberal than others. For instance, large cities are more likely than small towns or rural areas to have lesbian and gay organizations and establishments in which lesbians and gay men can meet prospective partners and take advantage of available support programs and local ordinances protecting gay and lesbian individuals. It remains for future researchers to determine the extent to which these factors affect the development of gay and lesbian relationships.

Methodologies and Theories

Although the past decade has seen an enormous increase in scientific information about gay and lesbian couples, future work would benefit from new methodologies. At the descriptive level, much could be learned from diary-type data that would provide information about change in relationship functioning over time at the micro level (see Huston, 2000). At the statistical level, the issue of nonindependence in partners’ scores is no longer a problem with the advent of hierarchical linear modeling (e.g., Kenny et al., 2002; Kurdek, 2000) and the extension of work with the intraclass correlation coefficient (Gonzalez & Griffin, 2001).

Although there are several well-done reviews of the literature on gay and lesbian couples (e.g., Peplau & Spalding, 2000), overall, the level of theoretical analysis in this area is not very advanced (see, however, Peplau, 2001). One might argue that theoretical advances need to wait until descriptive data are available from samples of gay and lesbian couples that are not predominantly white and privileged. Still, conceptual work at two levels seems promising and complementary. At the first level, specific content-based models of couples with particular characteristics (e.g., gay couples, lesbian couples, gay/lesbian couples of color, aging gay/lesbian couples) can systematically describe how diverse couples flesh out the content of their relationships. Diary-type data and qualitative interviews (e.g., Carrington, 1999) may be especially useful in this regard. At the second, more abstract level, the development of general function-based theories of close relationships (such as interdependence theory; see Finkel et al., 2002) would be enhanced if data from gay and lesbian couples were routinely used to test such theories. Large-scale data collections and the use of sophisticated data analytic techniques (such as hierarchical linear modeling) to test for the robustness of findings across diverse types of couples would be especially useful in this regard. Despite the tremendous diversity in how two people create the form of their relationship, there is likely to be equally impressive uniformity in the structure of how any couple begins, endures, and dissolves.