Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Aug 18.
Published in final edited form as: J Marriage Fam. 2020 May 11;82(4):1141–1158. doi: 10.1111/jomf.12686

The Psychological Toll of Emotion Work in Same-Sex and Different-Sex Marital Dyads

Debra Umberson 1, Mieke Beth Thomeer 2,*, Amanda M Pollitt 3,**, Sara E Mernitz 4,***
PMCID: PMC9387901  NIHMSID: NIHMS1783135  PMID: 35992218

Abstract

Objective:

This study considers how the provision of daily emotion work may affect the psychological well-being of the emotion worker, and how this linkage may vary for men and women in same- and different-sex marriages.

Background:

Emotion work–work intended to bolster a spouse’s well-being by reading and managing the spouse’s emotional needs—is common within marital relationships and often gendered, with women more aware of and concerned with emotion work than men. Yet, the psychological cost of performing emotion work is largely unexplored.

Method:

This study relies on 10 days of daily experiences data from spouses in same- and different-sex marriages (n = 756 individuals). Mixed effects multilevel regression modeling is used to examine how the provision of emotion work is associated with the emotion workers’ psychological well-being.

Results:

Providing emotion work is inversely associated with emotion workers’ psychological well-being, especially when provided for a spouse with elevated depressive symptoms. These estimated effects are generally similar for men and women but greater for those married to a man than for those married to a woman, whether in a same- or different-sex marriage.

Conclusion:

Emotion work appears to adversely affect the worker’s own psychological well-being, especially when a spouse has elevated depressive symptoms and when one’s spouse is a man. These results point to the importance of dyadic approaches and consideration of gendered relationship dynamics of same- as well as different-sex couples in studies of emotion work and other marital processes.

Keywords: couples, emotions, gender, marital relations, same-sex marriage, well-being

Emotion work, defined as work intended to bolster another person’s emotional well-being by reading and managing that person’s emotional needs, is a common dynamic in intimate relationships (Erickson, 2005; Hochschild, 2003; Pfeffer, 2010) that helps foster intimacy and closeness between spouses (Minnotte, 2017; Umberson, Thomeer, & Lodge, 2015). Although emotion work is positively associated with marital quality for the couple and for the well-being of the individual spouse receiving emotion work (Curran, McDaniel, Pollitt, & Totenhagen, 2015; Horne & Johnson, 2018; Minnotte, Pedersen, & Mannon, 2010), the provision of emotion work may be stressful and psychologically costly for the spouse who provides that work (Duncombe & Marsden, 1993, 1995; Umberson et al., 2015). Moreover, because research on heterosexual couples generally suggests that emotion work is a more immersive experience for women than men and that women are more concerned with the importance of emotion work within their relationship (Elliott & Umberson, 2008; Erickson, 2005; Rao, 2017; Wong, 2017), emotion work may more strongly undermine psychological well-being for women than for men.

Recent qualitative research suggests that gender differences in emotion work in different-sex couples may be paralleled in same-sex marriages with women married to women doing more emotion work and valuing it more than men married to men, yet spousal inequalities in emotion work may be less pronounced in same-sex couples (Umberson et al., 2015). Thus, women, and especially women partnered with men, may be more vulnerable to psychological costs related to providing emotion work than are men. The negative impact of providing emotion work may be further intensified in cases where spouses are perceived as having a greater need for emotion work, such as when a spouse has elevated depressive symptoms. Prior studies show that having a spouse with depressive symptoms is stressful (Rehman, Gollan, & Mortimer, 2008), perhaps partly because providing emotion work in this context is challenging. However, prior studies have not considered how the provision of emotion work affects the psychological well-being of the emotion worker, especially when the spouse providing emotion work may already be vulnerable because they have a spouse with depressive symptoms, nor has prior research considered gender and same-sex compared to different-sex couple variation in these dynamics.

To address these research gaps, we analyze 10 days of daily experiences data (based on questionnaires completed daily for 10 days) from both spouses in same-sex and different-sex marriages (n = 756 individuals, 378 married couples) and examine the relationship between the provision of emotion work and the psychological well-being of the emotion worker. We work from a gender-as-relational perspective to further consider how this linkage may differ for men and women in same-sex compared with different-sex marriages and to address how having a spouse with elevated depression symptoms may affect this link. A gender-as-relational perspective draws attention to the fact that within intimate relationships people enact gender not only as individuals but as individuals in relation to each other, making it especially important to consider the gender composition of the dyad (Springer, Hankivsky, & Bates, 2012; Thomeer, Umberson, & Reczek, 2020; Umberson et al., 2015). Dyadic daily experiences data collected over a 10-day period allow us to consider average daily fluctuation in associations between emotion work and psychological well-being for the emotion worker, and to consider respondents’ reports about their own emotion work and well-being in relation to their spouse’s reports about their own depressive symptoms. Our data include cisgender men and women in both same-sex and different-sex marriages, thus are uniquely suited to address the following questions: (a) Is the performance of emotion work associated with the psychological well-being of the emotion worker? (b) Does this association vary for men and women, and/or in the context of same-sex compared with different-sex marriages? and (c) Does emotion work undermine well-being more when one’s spouse has elevated depressive symptoms, and in different ways for men and women in same-compared with different-sex marriages?

These questions are important because emotion work may have costs for the psychological well-being of the emotion worker, even if emotion work benefits the recipient of that work. Moreover, this psychological toll may not be shouldered equally for men and women, particularly in social and relational contexts often overlooked in family studies, such as same-sex compared with different-sex marriages or having a spouse with depressive symptoms. The present study contributes to our understanding of emotion work and marital dynamics by extending the focus to psychological consequences of emotion work for the emotion worker, by broadening the study of gendered emotion work processes to include same-sex as well as different-sex married couples, and by paying attention to the mental health context of emotion work provision.

Background

Emotion work is a common type of informal work in intimate relationships that has the goal of managing and enhancing a partner’s emotional well-being (Erickson, 2005; Hochschild, 2003; Pfeffer, 2010). Such work can take multiple forms but typically includes assessment of a spouse’s emotional needs and intentional efforts to improve and manage their emotional state (Erickson, 2005; Minnotte, 2017; Pfeffer, 2010; Thomeer, Reczek, & Umberson, 2015). For instance, the provision of emotion work often requires an individual to first recognize a partner’s emotional needs (e.g., partner needs encouragement) and then act to manage those needs (e.g., offers encouragement to partner). In this way, our focus on emotion work differs from research on “emotional labor” in that it does not occur in paid or public sectors of the market economy but represents a separate—albeit related—construct (Hochschild, 1989). The study of emotion work within intimate relationships has a rich legacy within family studies, signifying that it is a key component of family and marital dynamics (Erickson, 2005; Horne & Johnson, 2018; Pfeffer, 2010; Rao, 2017; Umberson et al., 2015; Wong, 2017). Emotion work in intimate relationships seems to benefit the psychological well-being of the recipient of that work (Thomeer, Umberson, & Pudrovska, 2013), but we know very little about the implications of emotion work for the emotion worker’s psychological well-being.

Although the explicit goal of emotion work is to improve the emotional state of the emotion work recipient through managing their emotional needs, a feminist perspective (Bernard, 1987; Thomeer et al., 2013) suggests that the provision of emotion work may itself be a stressor, perhaps especially if this work goes unacknowledged, unappreciated, or unreciprocated (Duncombe & Marsden, 1993, 1995). In-depth interview studies support the idea that the provision of emotion work might be associated with lower levels of well-being: spouses who provide emotion work on a regular basis tend to describe this work as “draining,” because they perform it continually or feel that they are always monitoring their partner’s emotional needs and must be poised to meet those needs (Thomeer et al., 2013; Umberson et al., 2015). Thus, based on prior research on emotion work in intimate relationships, we test the following hypothesis:

Hypothesis 1: The provision of emotion work will be inversely associated with the psychological well-being of the emotion worker.

Gendered Experiences of Emotion Work

Most research on emotion work in intimate relationships is based on data from different-sex couples. Past findings on gender differences in amount of emotion work provided to spouses are inconsistent, but a general theme of past research is that women’s emotion work is more intensive and immersive, and that their emotion work efforts are less recognized by their spouse compared to men (Erickson, 2005; Umberson et al., 2015). Some qualitative studies suggest that, in different-sex marriages, wives provide more emotion work than husbands even when the wife is physically ill or depressed (Thomeer et al., 2013; Thomeer et al., 2015). These findings align with conceptualizations of gender as formed and maintained through social interactions with others (Thomeer et al., 2020; West & Zimmerman, 1987). From a gender-as-relational perspective, the gendered division of emotion work fits within the dominant gender discourse that casts women as naturally more nurturing and emotionally expressive (e.g., “emotion experts”) in relation to men who are cast as more stoic and emotionally detached (Connell, 2005; Lutz, 1988). Women are also more likely to be viewed as “relationship keepers,” meaning they are expected to be more attuned to relationship dynamics and emotional nuance, recognizing when emotion work may be helpful in maintaining the relationship and improving their spouse’s well-being (Erickson, 2005; Rao, 2017; Wong, 2017).

Past research emphasizes how the external constraints and demands on women to care for others—especially for other people’s emotional well-being—alongside internalized messaging about women’s responsibility for the well-being of others impose pressures that may undermine psychological well-being (i.e., “cost of caring” hypothesis; Kessler, McLeod, & Wethington, 1985; Taylor, 2015). Although relationship dynamics of same-sex couples may diverge from those of different-sex couples, dominant discourses about gender and relationship orientations apply to women broadly. Thus, we expect that women in both same- and different-sex relationships will be more adversely affected than men by doing this work, and we test the following hypothesis:

Hypothesis 2: The provision of emotion work will be associated with lower psychological well-being for women than men.

As articulated with the gender-as-relational perspective, the dynamics of same-sex relationships, which consist of two men or two women, may be unique compared with different-sex couples because people enact gender differently depending on whether they are interacting with a man or a woman even when they are influenced by broader societal constraints and opportunities associated with gender (Ferree, 2010; Thomeer et al., 2020). A key limitation of past studies on emotion work is that this research has focused on different-sex unions wherein gender differences and inequality in care work are central (Curran et al., 2015; Horne & Johnson, 2018; Minnotte et al., 2010). Although most research on different-sex couples finds that women are more focused than men on emotion work in their intimate relationships (Rao, 2017; Thomeer et al., 2015; Wong, 2017), a fuller understanding of gendered emotion work processes requires that we also examine emotion work and well-being linkages for men compared to women in same-sex marriages.

A growing body of evidence using a gender-as-relational perspective suggests that intimate partner dynamics, including emotion work, may unfold differently in same-sex compared with different-sex relationships (Behler, Donnelly, & Umberson, 2019; Umberson et al., 2015). Same-sex relationships are often characterized by more equality and reciprocity (Balsam, Beauchaine, Rothblum, & Solomon, 2008; Gotta et al., 2011); thus, because emotion work is more likely to be shared, same-sex partners may experience less harm to their own well-being when providing emotion work. Indeed, a recent qualitative study shows that, although women in both same- and different-sex relationships performed more emotion work than men, women married to women generally understood their emotion work as mutual and expressed greater appreciation for each other’s emotion work than other couples (Umberson et al., 2015). Moreover, men in same-sex relationships described providing more emotion work to their partner than did men in different-sex relationships, but they did not generally view this work as stressful—especially compared with how women in same- and different-sex relationships viewed their emotion work. These findings lend support for the expectation that poorer well-being related to greater provision of emotion work by women compared with men may be attenuated in same-sex unions. Thus, we test the following hypothesis:

Hypothesis 3: The association of emotion work with psychological well-being will have less impact on the well-being of emotion workers, particularly women, in same-sex compared with different-sex unions.

Emotion Work and Mental Health of One’s Partner

It is also important to consider the mental health context in which emotion work is provided, specifically whether or not this work is provided to a partner with depressive symptoms, particularly when the partner meets criteria suggesting clinically significant levels of depressive symptoms. Although emotion work is routine within intimate relationships (Erickson, 2005; Hochschild, 2003; Pfeffer, 2010), conditions may shift when one’s spouse experiences significant depressive symptoms (Thomeer et al., 2013). Depressive symptoms may drive the provision of more emotion work within the relationship. At the same time, a spouse with depressive symptoms may be less capable of responding to their partner’s emotion work. Indeed, having a spouse with depressive symptoms places significant strain on a relationship (Sharabi, Delaney, & Knobloch, 2016) and undermines the non-depressed spouse’s well-being (Behler et al., 2019). Providing emotion work to a spouse with depressive symptoms may also be difficult because so much is riding on alleviating those symptoms with the emotion worker possibly feeling responsible for helping the spouse with depressive symptoms and frustrated when that is not the case—a common dynamic within couples wherein one spouse is experiencing depressive symptoms (Reczek, Thomeer, Gebhardt-Kram, & Umberson, 2020). A spouse with depressive symptoms may also be less capable of feeling or expressing appreciation for the emotion work they receive, leading to more stress for the emotion worker. Thus, having a spouse with depressive symptoms may exacerbate the impact of emotion work provision on the emotion worker’s psychological well-being, a premise we test with Hypothesis 4:

Hypothesis 4: Emotion work will have a stronger negative effect on the emotion worker’s psychological well-being under the condition of having a spouse with depression.

As discussed above, a gender-as-relational perspective further suggests that the impact of a spouse’s depressive symptoms on emotion work and well-being processes would differ for men and women, and for men and women in same- and different-sex marriages. Qualitative analysis of interviews with men and women in different-sex couples in which one spouse had depressive symptoms concluded that women draw on femininity scripts that women are emotion experts and become deeply involved with their husband’s depression, often to their own emotional detriment (Thomeer et al., 2013). This process was less apparent for men when their wife had depressive symptoms—likely because scripts of masculinity position men as less capable and responsible for dealing with mental health issues and emotions. Qualitative research also suggests that women in same-sex relationships similarly become deeply involved in their partner’s depression to the detriment of their own well-being (Thomeer et al., 2015). However, partners in same-sex relationships may feel better equipped to help with one another’s mental health issues, and thus be more involved in emotion work around depressive symptoms because they draw on a script of “gender same-ness” rather than difference (Reczek et al., 2020). Thus, we propose our final hypothesis:

Hypothesis 5. The negative impact of providing emotion work to a spouse with depression will be stronger for women compared with men, and more so for different-sex compared with same-sex spouses.

Method

Participants and Data

We use dyadic data from a study on same-sex and different-sex marriages and health in midlife couples that included a baseline survey as well as 10 days of daily experiences questionnaires in 2015. Data include 157 lesbian couples (314 women), 106 gay couples (212 men), and 230 heterosexual couples (115 women, 115 men). The sample was recruited in a systematic way to ensure that different-sex and same-sex couples were comparable; participants were matched on relationship duration, age, and geographic location. Relationship duration was measured based on the total number of years cohabiting and married due to past legal restrictions on marriage for same-sex couples. Massachusetts was chosen as the study site because it was the first U.S. state to legalize marriage for same-sex couples in 2004, but couples who married in Massachusetts and resided in other states were also invited to participate.

Using the Massachusetts Registry of Vital Records, same-sex couples who married between 2004 and 2015 and met the midlife age requirements (ages 35–65) were invited to participate in letters mailed to their address by the lead author. Approximately 70% of same-sex couples were recruited with this strategy. These couples were asked to refer both same-sex and different-sex couples who met the study requirements. The remaining 30% of same-sex couples were recruited via these referrals. Different-sex couples were recruited from Massachusetts public records on households in each zip code with records including addresses and demographic information for city residents; we recruited these couples from zip codes corresponding to neighborhoods of the enlisted same-sex couple study participants (approximately 40% recruited via this method). The remaining 60% of different-sex couples were obtained through snowball sampling referrals from enlisted same-sex and different-sex study participants—a strategy used in other studies of same- and different-sex couples (Balsam et al., 2008). Out of all couples recruited, 55% of male same-sex couples, 62% of female same-sex couples, and 51% of different-sex couples lived outside of Massachusetts at the time of initial recruitment in 2014–2015. All participants were cisgender men and women, aged 35 to 65, legally married, and living together for at least 3 years at the time of the study. The demographic characteristics of the sample were consistent with nationally representative data comparing same-sex and different-sex spouses in midlife on many characteristics, such as age, relationship duration, place of residence, and education (Gates, 2014, 2015), but this sample is not representative of the U.S. population.

Both spouses in each marital dyad participated in the study. Each spouse completed questionnaires online and separately from one another. They first completed a 45-minute survey to capture baseline information on their relationship and health history and current status. Participants also completed a daily experiences questionnaire that took 5 to 10 minutes to complete at the end of each day for 10 days. Both spouses must have completed at least 6 of the 10 daily experiences questionnaires to have their data retained in the study; 90 percent completed all daily questionnaires, with no differences in number of days completed by gay, lesbian, and heterosexual dyads. Data are publicly available at the Inter-University Consortium for Political and Social Research (https://doi.org/10.3886/ICPSR37404.v1).

Measures

Psychological Well-Being of the Emotion Worker.

We assessed daily psychological well-being with 9 items of positive and negative affect based on measures used in previous research (Almeida & Kessler, 1998; Mroczek & Kolarz, 1998; Watson, Clark, & Tellegen, 1988). Well-being is assessed with 9 items that tap into daily moods on a scale from 1 (not at all) to 5 (extremely). Items asked to what extent, in the past 24 hours, participants felt calm, happy, frustrated, worried, tired, sad, irritable, angry, and upset. Each item is coded so that higher scores indicate better psychological well-being (α = 0.87).

Depressive Symptoms of Spouse.

We created a binary measure indicating whether spouses scored above a clinically significant cut-point based on the spouse’s responses to the 11-item Center for Epidemiological Studies Depression Scale (CES-D; Irwin, Artin, & Oxman, 1999; Radloff, 1977; Zauszniewski & Bekhet, 2009) reported at the baseline interview (α = 0.84). Participants were asked to indicate how often they experienced certain negative (e.g., “During the past week, I felt depressed.”) and positive (e.g., “During the past week, I was happy.”) feelings during the past week. Response options were 0 (rarely/none of the time), 1 (some of the time), 2 (occasionally), and 3 (most of the time); questions pertaining to positive emotions were reverse coded. Spouses were categorized as having clinically significant levels of depressive symptoms if their summed score met the clinical cut-point for depression (≥10); though we do not know whether participants have been formally diagnosed with depression, previous research shows that this cut-point predicts clinical diagnoses of major depressive disorder with 97% accuracy (Andresen, Malmgren, Carter, & Patrick, 1994). We refer to respondents meeting this threshold as spouses with “elevated depressive symptoms” throughout the results.

Emotion Work.

Drawing on prior measures of emotion work developed by Erickson (1993, 2005), as well as the theoretical and conceptual literature on emotion work (Thomeer et al., 2015; Umberson et al., 2015), we created a scale to measure daily emotion work. This scale included the following 3 items that tap into assessing a spouse’s emotional state and engaging in efforts to manage the spouse’s emotions: (a) sense that your spouse is bothered about something, (b) suggest solutions to your spouse’s problems, and (c) try to improve your spouse’s mood. Response options for each question ranged from 1 (not at all) to 5 (a great deal). The within- and between-person reliabilities of this measure were 0.70 and 0.71, respectively, indicating that this emotion work measure is consistent over time and between people (see Bolger & Laurenceau, 2013, pp. 127–141, for details on calculating within-person reliability).

Other Covariates.

We controlled for a number of factors that may be associated with daily psychological well-being including day of daily questionnaire completion (Day 1–10), participants’ own depressive symptoms at baseline (using the CES-D; grand mean = 6.04), whether there were children living in their household (0 “no”, 1 “yes”), relationship duration (years living together including years married; grand mean = 15.08), whether participants graduated from college or not (0 “no”, 1 “yes”), race (0 “non-White”, 1 “White”), employment status (referent: full-time employed outside the home), and personal income (0 = no personal income to 6 = $150,000 or more).

Analytic Plan

We conducted analyses in Stata 15. First, we examined the descriptive statistics, comparing study variables for men married to men, men married to women, women married to men, and women married to women using one-way ANOVAs with a priori planned contrasts for the continuous measures and chi-square tests for the categorical variables. We present means and standard deviations, cell sizes and percentages, standardized mean differences, chi-square test statistics, and difference test p-values for descriptive data comparisons in Table 1. Second, to test all hypotheses and specifically consider three different “gender effects” in the model—respondent gender (men compared with women), spouse gender (respondents married to men compared with those married to women), and the interaction of respondent and spouse gender (same-sex compared with different-sex dyads), we employed the factorial method (West, Popp, & Kenny, 2008), an extension of the Actor Partner Interdependence Model (Cook & Kenny, 2005), via mixed effects multilevel regression modeling (to account for interdependence between dyad members; Kroeger & Powers, 2019). This method allows us to directly compare differences in associations between men with men, men with women, women with men, and women with women in the same analytic model. We specify daily variables (Level 1) as nested within respondents (Level 2) and respondents as nested within couples (Level 3). However, there is no variability at Level 3 because there are only two respondents per dyad; thus, these models cannot be estimated as true three-level models. To account for this lack of variability and to model interdependence, and because we have both distinguishable (different-sex couples) and indistinguishable (same-sex couples) dyads, we treat all couples as indistinguishable by specifying one shared error variance and one covariance per couple at Level 2. We centered emotion work at the grand mean of all respondents across all diary days such that results are interpreted as the effect of above sample average daily emotion work on daily psychological well-being over time. We present standardized results (in standard deviation units) alongside unstandardized results for all models; measures were z-scored prior to estimation to reduce bias in standardized coefficients that can occur in multilevel models. We assessed model fit using likelihood ratio tests of the fit of the nested models. Statistically significant likelihood ratio statistics between nested models indicate improved model fit for models with additional parameters; when the test is not significant, it indicates that the additional parameters are not improving fit to the data.

Table 1.

Descriptive Results of Study Variables by Gendered Relationship Context (N = 756)

Men with men, mean (SD)/N (%) Men with women, mean (SD)/N (%) Women with men, mean (SD)/N (%) Women with women, mean (SD)/N (%) Omnibus F/χ2
Baseline variables
 Depressive Symptoms 5.57 (5.10) 6.02 (5.40) 6.92 (5.23) 6.09 (5.37) 1.64
 Spouse Has Depressive Symptoms 42 (19.81) 28 (24.35) 25 (21.74) 72 (22.93) 0.63
 Children in Household (couple-level) 27 (12.7)b,c,d 82 (71.3)a 82 (71.3)a,d 127 (40.4)a,c 155.86
 Years Living Together (couple-level) 16.23 (0.54)d 15.89 (0.73)d 15.89 (0.73)d 13.73 (0.44)a,b,c 8.10
 College Graduate 170 (80.19)b 80 (69.57)a,d 87 (75.65)d 271 (86.31)b,c 17.18
 White 185 (87.26) 95 (82.61) 93 (80.87)d 278 (88.54)c 5.60
 Not Employed 36 (16.98) 23 (20.00) 27 (23.48) 54 (17.20) 2.73
 Part-time Employed 14 (6.60)c 4 (3.48)c 23 (20.00)a,b,d 28 (8.92)c 22.46
 Full-time Employed 162 (76.42)c 87 (75.65)c 65 (56.52)a,b,d 232 (73.89)c 17.10
 Personal Income 3.60 (1.64)b,c,d 2.85 (1.50)a,c 2.24 (1.43)a,b,d 3.11 (1.42)a,c 21.61
Daily variables
 Emotion Work 1.99 (0.82)b,c 2.08 (0.82)c 1.82 (0.76)a,b,d 2.02 (0.86)c 21.67
 Psychological Well-being 38.52 (4.98)b,c,d 37.73 (5.01)a,c 36.53 (5.71)a,b,d 37.48 (5.41)a,c 36.86
Open in a new tab
a

Significantly different (p < .05) from men with men.

b

Significantly different (p < .05) from men with women.

c

Significantly different (p < .05) from women with men.

d

Significantly different (p < .05) from women with women.

Results

Descriptive Results

As shown in Table 1, there were several key differences in comparisons of men married to men, men married to women, women married to women, and women married to men. In terms of the main outcome and predictor variables, women married to men reported performing less emotion work than men married to men (d = −0.21, p < .001), men married to women (d = −0.31, p < .001), and women married to women (d = −0.24, p < .001). Men married to men reported performing less daily emotion work than men married to women (d = −0.10, p = .04). There were no other significant differences in emotion work between groups. Men married to men reported higher average daily psychological well-being compared with men married to women (d = 0.16, p < .001), women married to men (d = 0.37, p < .001), and women married to women (d = 0.20, p < .001). Women married to men reported lower average daily psychological well-being compared with men married to women (d = −0.23, p < .001) and women married to women (d = −0.17, p < .001). Men married to women and women married to women did not differ from one another on average daily psychological well-being.

Regarding the control variables, in these descriptive statistics, men married to men were less likely to have children in the household than men and women in different-sex couples (χ2 = 154.21, p < .001) and women married to women (χ2 = 46.93, p < .001). Women married to women were also less likely to have children in the household than different-sex couples (χ2 = 50.82, p < .001). Women married to women had shorter relationship durations than men married to men (d = −0.32, p = .001) and men and women in different-sex couples (d = −0.27, p = .003). Men married to men and women married to women were more likely to have a college degree than men married to women (χ2 = 4.67, p = .03; χ2 = 32.08, p < .001, respectively); women married to women were also more likely to have a college degree (χ2 = 6.92, p = .01) and identify as white (χ2 = 4.33, p = .04) than women married to men. Women married to men were less likely to be employed part-time or full-time compared with men married to men (χ2 = 13.33, p < .001; χ2 = 13.90, p < .001, respectively), men married to women (χ2 = 15.15, p < .001; χ2 = 9.39, p = .002), and women married to women (χ2 = 9.87, p = .002; χ2 = 11.91, p = .001). Men married to men reported higher personal income than men married to women (d = 0.48, p < .001), women married to men (d = 0.88, p < .001), and women married to women (d = 0.32, p = .001). Women married to men had lower personal income than men married to women (d = 0.39, p < .001) and women married to women (d = −0.63, p < .001).

Hypothesis 1: Emotion Work and Well-Being of the Emotion Worker.

First, we considered how daily emotion work is associated with daily psychological well-being (shown in Table 2). We found support for Hypothesis 1: Respondents who reported higher levels of emotion work averaged across 10 days scored lower on daily psychological well-being (b* = −0.07). That is, we found a small, negative effect such that a one standard deviation increase in daily emotion work is associated with a 0.07 standard deviation decrease in daily psychological well-being.

Table 2.

Mixed-Effects Multilevel Regression Models Predicting Average Daily Psychological Well-Being (N = 756)

Daily emotion work Daily emotion work: Gender Daily emotion work: Couple type
b SE b* p b SE b* p b SE b* p
Intercept 40.89 0.58 0.18 < .001 41.58 0.63 0.20 < .001 41.65 0.65 0.22 < .001
Day 0.12 0.02 0.02 < .001 0.12 0.02 0.02 < .001 0.12 0.02 0.02 < .001
Depressive Symptoms −0.38 0.02 −0.38 < .001 −0.38 0.02 −0.38 < .001 −0.38 0.02 −0.38 < .001
Children in Household −0.98 0.26 −0.18 < .001 −0.97 0.26 −0.18 < .001 −0.82 0.29 −0.16 < .001
Relationship Duration 0.02 0.02 0.03 .26 0.02 0.02 0.02 .33 0.02 0.02 0.03 .26
College Graduate −1.12 0.28 −0.21 < .001 −1.03 0.29 −0.19 < .001 −1.06 0.29 −0.20 < .001
White −0.47 0.32 −0.09 .13 −0.49 0.31 −0.09 .12 −0.51 0.32 −0.10 .11
Not Employed 0.57 0.31 0.11 .07 0.57 0.31 0.11 .06 0.56 0.31 0.11 .07
Part-time Employed 0.01 0.40 0.00 .99 0.13 0.40 0.03 .74 0.13 0.40 0.02 .75
Personal Income 0.20 0.08 0.06 .02 0.18 0.08 0.05 .04 0.16 0.08 0.05 .06
Daily Emotion Work −0.43 0.07 −0.07 < .001 −0.66 0.12 −0.10 < .001 −0.64 0.13 −0.10 < .001
Woman — — — — −0.57 0.37 −0.11 .13 −0.80 0.59 −0.19 .18
Woman × Daily Emotion Work — — — — −0.01 0.15 −0.00 .93 −0.10 0.23 −0.02 .67
Spouse is a Woman — — — — −0.51 0.37 0.05 .17 −0.77 0.60 −0.03 .20
Spouse is a Woman × Daily Emotion Work — — — — 0.38 0.15 0.06 .01 0.31 0.22 0.05 .15
Woman × Spouse is a Woman — — — — — — — — 0.44 0.85 0.13 .61
Woman × Spouse is a Woman × Daily Emotion Work — — — — — — — — 0.14 0.30 0.02 .65
Model loglikelihood χ 2 p Model loglikelihood χ 2 p Model loglikelihood χ 2 p
Loglikelihood Ratio Test −20,920.11 401.94 < .001 −20,913.02 14.18 .007 −20,912.24 1.57 .46
Open in a new tab

Notes. Standardized results shown. Significant p-values are bolded. Statistically significant loglikelihood ratio tests indicate significant improvement in model fit from prior model. The daily emotion work model is compared to a null model with only random effects (loglikelihood = −21,121.08).

Hypothesis 2: Gender Differences in the Effects of Emotion Work on Well-Being.

We considered the possibility of gender differences in the link between daily emotion work and average daily well-being in Table 2. We found no interaction between respondent gender and emotion work on average daily psychological well-being, indicating that men and women who provided emotion work were similarly affected, thus, Hypothesis 2 was not supported. However, there was a significant interaction between spouse gender and emotion work. The negative association between emotion work and average daily psychological well-being was stronger for respondents married to a man (b* = −0.10) compared with respondents married to a woman (b* = −0.04).

Hypothesis 3: Differences in the Effects of Emotion Work on Well-Being for Same-Sex Compared with Different-Sex Marriages.

We found partial support for Hypothesis 3. The three-way interaction between respondent gender, spouse gender, and emotion work was not significant (Table 2), suggesting that the negative association between emotion work and average daily psychological well-being did not differ for respondents in same-sex compared with different-sex marriages. However, significant group differences in the slope of the association between emotion work and psychological well-being could be masked by the reference groups (i.e., men married to men); rather than fitting the models with new reference groups, we calculated the adjusted predicted slopes for each group and tested for slope differences using Wald tests. The Wald tests indicated that this association was weaker for women married to women (b* = −0.04) compared with women married to men (b* = −0.11, z = 4.68, p = .03) and men married to men (b* = −0.10, z = 4.58, p = .03). There were no other significant differences between groups. We showed these group differences in the association between daily emotion work and daily psychological well-being in Figure 1.

Figure 1.

Figure 1.

Open in a new tab

Gendered Marital Dyad Differences in the Association Between Daily Emotion Work and Daily Psychological Well-Being.

Hypothesis 4: Emotion Work and Well-Being under the Condition of a Spouse’s Elevated Depressive Symptoms.

Next, we considered whether daily emotion work might be more strongly linked to daily psychological well-being when one is providing emotion work to a spouse with elevated depressive symptoms. In Table 3, we showed results from the model from Hypothesis 1, including the main effect of having a spouse with elevated depressive symptoms and the interaction between emotion work and having a spouse with elevated depressive symptoms. We found support for Hypothesis 4: The negative association of daily emotion work with daily psychological well-being of the emotion worker was greater for those whose spouse has elevated depressive symptoms (b* = −0.11) than for respondents whose spouse does not (b* = −0.05).

Table 3.

Mixed-Effects Multilevel Regression Models Predicting Average Daily Psychological Well-Being by Gender (N = 756)

Daily emotion work: Spouse with depression Daily emotion work: Gender and spouse with depression Daily emotion work: Couple type and spouse with depression
b SE b* p b SE b* p b SE b* p
Intercept 40.93 0.59 0.21 <.001 41.31 0.65 0.22 <.001 41.46 0.67 0.24 <.001
Day 0.12 0.02 0.02 <.001 0.12 0.02 0.02 <.001 0.12 0.02 0.02 <.001
Depressive Symptoms (Baseline) −0.39 0.02 −0.38 <.001 −0.39 0.02 −0.38 <.001 −0.39 0.02 −0.38 <.001
Children in Household −0.96 0.25 −0.18 <.001 −0.96 0.26 −0.18 <.001 −0.81 0.28 −0.15 <.001
Relationship Duration 0.02 0.02 0.03 .29 0.01 0.02 0.02 .39 0.02 0.02 0.02 .32
College Graduate −1.19 0.28 −0.22 <.001 −1.14 0.29 −0.21 <.001 −1.17 0.29 −0.22 <.001
White −0.46 0.31 −0.09 .15 −0.44 0.31 −0.08 .16 −0.46 0.31 −0.09 .14
Not Employed 0.51 0.31 0.10 .10 0.53 0.31 0.10 .08 0.51 0.31 0.10 .10
Part-time Employed −0.04 0.40 −0.01 .92 0.08 0.40 0.01 .85 0.07 0.40 0.01 .87
Personal Income 0.20 0.08 0.06 .02 0.17 0.08 0.05 .04 0.16 0.08 0.05 .06
Daily Emotion Work −0.34 0.08 −0.05 <.001 −0.45 0.14 −0.07 <.001 −0.47 0.15 −0.07 <.001
Spouse with Depression 0.22 0.40 −0.10 .58 1.71 0.73 −0.01 .02 1.53 0.82 0.02 .06
Daily Emotion Work × Spouse with Depressive Symptoms −0.38 0.15 −0.06 .01 −0.89 0.27 −0.14 <.001 −0.73 0.31 −0.11 .02
Woman −0.26 0.43 −0.08 .55 −0.69 0.65 −0.14 .29
Woman × Daily Emotion Work −0.09 0.17 −0.01 .60 −0.02 0.26 −0.00 .95
Woman × Spouse with Depressive Symptoms −1.38 0.87 −0.12 .11 −0.93 1.32 −0.21 .48
Daily Emotion Work × Spouse with Depressive Symptoms × Woman 0.37 0.33 0.06 .27 −0.08 0.51 −0.01 .87
Spouse is a Woman −0.27 0.42 0.04 .51 −0.72 0.68 −0.01 .28
Spouse is a Woman × Daily Emotion Work 0.26 0.17 0.04 .14 0.33 0.25 0.05 .19
Spouse is a Woman × Spouse with Depressive Symptoms −1.03 0.88 −0.02 .24 −0.57 1.31 −0.09 .66
Daily Emotion Work × Spouse with Depressive Symptoms × Woman Partner 0.46 0.34 0.07 .17 0.04 0.49 0.01 .94
Woman × Spouse is a Woman 0.80 0.95 0.10 .40
Woman × Spouse is a Woman × Daily Emotion Work −0.14 0.35 −0.02 .69
Woman × Spouse is a Woman × Spouse with Depressive Symptoms −0.81 1.78 0.15 .65
Daily Emotion Work × Woman × Spouse is a Woman × Spouse with Depressive Symptoms 0.80 0.68 0.13 .24
Model loglikelihood χ 2 p Model loglikelihood χ 2 p Model loglikelihood χ 2 p
Loglikelihood Ratio Test −20,914.80 412.56 <.001 −20,904.55 20.50 .009 −20,902.82 3.47 .48
Open in a new tab

Notes. Standardized results shown. Significant p-values are bolded. Statistically significant loglikelihood ratio tests indicate significant improvement in model fit from prior model. The daily emotion work: spouse with depression is compared to a null model with only random effects (loglikelihood = −21,121.08).

Hypothesis 5: Gender Differences in the Effects of Emotion Work on Well-Being under the Condition of a Spouse with Elevated Depressive Symptoms, for Same-Sex Compared with Different-Sex Marriages.

In the test for Hypothesis 5 (shown in Table 3), we included a three-way interaction between emotion work, having a spouse with elevated depressive symptoms, and respondent gender as well as a three-way interaction between emotion work, having a spouse with elevated depressive symptoms, and spouse gender. The reference groups for these interactions were men whose spouse did not have elevated depressive symptoms and respondents married to men without elevated depressive symptoms, respectively. There was no significant three-way interaction between emotion work, having a spouse with elevated depressive symptoms, and respondent gender, and the Wald tests revealed no significant group differences in slopes among those whose spouses with elevated depressive symptoms. However, we found a significant three-way interaction between emotion work, having a spouse with elevated depressive symptoms, and spouse gender (see Figure 2). Wald tests revealed that there were significant differences in the slope of the association between emotion work and psychological well-being for respondents married to men with elevated depressive symptoms (b* = −0.18) compared with respondents married to women with elevated depressive symptoms (b* = −0.07, z = 6.10, p = .01). That is, the detrimental effect of emotion work on average daily psychological well-being among those whose spouse has elevated depressive symptoms was stronger for those married to a man compared with those married to a woman.

Figure 2.

Figure 2.

Open in a new tab

Partner Gender Differences in the Association Between Daily Emotion Work and Daily Psychological Well-Being Based on Spouse Depressive Symptoms.

We found mixed support for Hypothesis 5 that, under the condition of having a spouse with elevated depressive symptoms, the negative association of average daily emotion work and average daily psychological well-being would differ in same-sex compared with different-sex marriages: The four-way interaction between emotion work, respondent gender, spouse gender, and having a spouse with elevated depressive symptoms was not significant (Table 3). The reference group for this model was men who are married to a man without elevated depressive symptoms; however, because we were interested in gendered marital dyad differences in the strength of this association among those with spouses with elevated depressive symptoms, we conducted Wald tests further comparing these groups. This association was stronger for men married to a man with elevated depressive symptoms (b* = −0.18) compared with women married to a woman with elevated depressive symptoms (b* = −0.04, z = 7.87, p = .005) and for women married to a man with elevated depressive symptoms (b* = −0.20) compared with women married to a woman with elevated depressive symptoms (z = 6.74, p = .009). We illustrated these differences visually in Figure 3. This figure showed that the negative association between the provision of emotion work and average daily psychological well-being when married to someone with elevated depressive symptoms was more pronounced for respondents—whether men or women—married to men than it was for women married to women.

Figure 3.

Figure 3.

Open in a new tab

Gendered Marital Dyad Differences in the Association Between Daily Emotion Work and Daily Psychological Well-Being Among Respondents Who Have a Spouse with Depressive Symptoms.

Discussion

Among married couples, men and women are more likely to name their spouse than anyone else as their most important source of emotional support (Taylor, 2011). Emotion work is likely central to this dynamic. Spouses often perform emotion work by devoting effort toward assessing and managing each other’s emotions in an effort to bolster emotional well-being (Erickson, 2005). Whereas emotion work is generally understood as benefiting the well-being of the recipient (Thomeer et al., 2013), the provision of emotion work can be stressful (Erickson, 2005) which might undermine the emotion worker’s psychological well-being. Moreover, past research suggests that emotion work is a gendered process whereby women provide more emotion work to their intimate partners than do men, receive less appreciation for that emotion work, and perhaps experience more stress from that emotion work, with this depending not only on one’s own gender but also the gender of one’s spouse (Umberson et al., 2015). Prior research also suggests that emotion work under certain conditions may be more stressful, such as when a spouse has depressive symptoms (Thomeer et al., 2013). The present study focuses specifically on the impact of emotion work on the emotion worker’s psychological well-being, whether having a spouse with elevated depressive symptoms exacerbates that impact, and possible variation by gender as well as whether one is in a same-sex or different-sex marriage.

Our primary findings both support and modify expectations based on prior research. In line with a gender-as-relational theoretical frame, we find it is too simplistic to view our results through a lens of gender similarity or gender difference. Rather, interpretations depend on whether the focus is on one’s own gender, the gender of one’s spouse, or the interaction between the two. Here we highlight three important themes. First, the performance of emotion work seems to come with some psychological cost to the well-being of the emotion worker, supporting the notion that emotion work is often stressful for the emotion worker. Based on prior work on emotion and mental health “contagion” (Kouros & Cummings, 2010; Rosenquist, Fowler, & Christakis, 2011), the emotion worker’s diminished psychological well-being could plausibly have a negative impact of both spouses’ well-being over time—pointing to the need for longitudinal research as well as more attention to the dyadic experiences of spouses’ emotions and well-being. Future research should further consider the possibility that emotion work negatively impacts marital quality over time, with psychological well-being (of one or both partners) mediating this linkage. The general finding of the psychological cost of emotion work also points to the importance of designing policy and clinical strategies that do not further burden spouses who provide care to their partners, and to the need for strategies that reduce burden on emotion workers and attend to their well-being.

Second, we find both similarities and differences by gender and across same- and different-sex dyadic contexts in the link between emotion work and well-being. Generally speaking, the provision of emotion work is associated with lower levels of well-being for men and women, and in same-sex compared with different-sex marriages. This finding raises the possibility that we may have underestimated the impact of emotion work on men in past studies (Erickson, 2005; Thomeer et al., 2015). However, there are two important exceptions to patterns of emotion work and well-being similarity for men and women across couples—a finding made possible only by including men and women in both same- and different-sex couples in the study design. In the first instance, we find that the gender of one’s spouse shapes the degree to which emotion work is associated with the emotion worker’s well-being, such that the inverse association between emotion work and average daily psychological well-being is stronger when one is married to a man than when married to a woman, regardless of one’s own gender or whether one is in a same- or different-sex couple. This suggests the provision of emotion work may be more psychologically taxing when one’s partner is a man, perhaps because men are less likely to be aware of, or to reciprocate, emotion work exchanges as suggested by recent qualitative research (Pfeffer, 2010; Thomeer et al., 2013; Wong, 2017).

In the second instance, compared with women married to men, the well-being of women married to women seems to be less affected by emotion work. Prior qualitative research suggests this finding may occur because there is greater reciprocity in, and appreciation for, emotion work provided by one’s partner in same-sex marriages (Umberson et al., 2015). Thus, approaches comparing same-sex couples to different-sex couples, or only comparing men to women (with this often restricted to different-sex couples), may miss key gendered variation in intimate relationship dynamics. The present findings underscore the need to pay close attention to not only who provides emotion work but who that work is provided to, especially in terms of their gender and—as discussed below—their mental health status. These findings further illustrate the salience of the gender-as-relational perspective (Springer et al., 2012; Umberson et al., 2015); gendered relationship dynamics reflect one’s own gender in relation to whether one is interacting with a man or a woman. Future theory and research should seek to replicate the current findings and to identify other marital dynamics that may be experienced in similar and/or different ways across same-sex and different-sex marriages and depending on the gender of the spouse as well as the respondent. This approach can help advance scientific understanding of marital dynamics that influence the psychological well-being of both partners in a relationship, with potentially different costs and benefits for men and women in same- and different-sex relationships—information needed to ground effective policies and interventions to support couples and their mental health.

The third major theme is that the provision of emotion work may undermine psychological well-being more when one is married to a spouse with elevated depressive symptoms. Emotion work is likely more stressful when one is married to a spouse with elevated depressive symptoms because it is perceived that one’s emotion work is critical to helping relieve their spouse’s symptoms, and because emotion work efforts may be less effective when dealing with a spouse with elevated depressive symptoms whose negative emotions are more intractable than those of a spouse without depressive symptoms (Reczek et al., 2020). This points to the necessity of paying attention to the mental health context of marital dynamics, and conceptualizing mental health as a contextual factor rather than solely as an outcome, as is the case in many studies.

Moreover, we find that the psychological cost of providing emotion work to a spouse with depressive symptomology is greater when the spouse is a man rather than a woman, whether in a same-sex or a different-sex marriage. These findings suggest that not only is it more challenging to provide emotion work to a spouse with elevated depressive symptoms or to provide emotion work to a spouse who is a man (as described above), but it is particularly more challenging when that spouse is a man with elevated depressive symptoms. Indeed, a previous study concluded that women with elevated depressive symptoms are more likely than men with elevated depressive symptoms to be concerned about the impact of their depression on their spouse and to attempt to shield their spouse from this impact (Thomeer et al., 2013). This finding likely also reflects the greater stigma men experience concerning their own depressive symptoms, with men less likely to seek help when they experience depressive symptoms (Johnson, Oliffe, Kelly, Galdas, & Ogrodniczuk, 2012; Yousaf, Grunfeld, & Hunter, 2015). This hesitance to seek external help may intensify the burden experienced by the spouse who provides emotion work (Reczek et al., 2020)—an important possibility for future research to consider. Dominant masculinity discourses also encourage men to downplay their emotional distress (Flynn, Hollenstein, & Mackey, 2010); thus, it may be harder for partners to determine the effectiveness of their emotion work when men have elevated depressive symptoms. Notably, however, our findings do not support prior theoretical work suggesting that caring for a spouse with elevated depressive symptoms is more emotionally difficult for women than men, as men married to men with elevated depressive symptoms have similarly negative outcomes as women married to men with elevated depressive symptoms (Rehman et al., 2008).

Although this study provides one of the first examinations into how the provision of emotion work affects the psychological well-being of the emotion worker, we must acknowledge limitations. First, the generalizability of our results is limited due to study design. A significant limitation is the use of inferential statistics with data that are not from a random sample. For example, 60% of the different-sex couples were recruited through snowball sampling with referrals from enlisted study participants but 70% of the same-sex couples were recruited through vital records, and although we matched couples on key sociodemographic characteristics in the recruitment process, there may be unmeasured differences that drive some findings. Additionally, the sample includes highly educated, currently married, midlife adults and results likely differ by age, class, and relationship type, as well as more inclusive measures of gender. The sample includes only cisgender respondents, although prior research demonstrates that transgender respondents may experience different emotion work dynamics within a couple (Pfeffer, 2010) and gender non-binary and nonconforming people are largely overlooked in family studies. Future research on gendered dynamics in intimate relationships should broaden inclusion based on gender diversity, as well as age, race, and socioeconomic diversity (van Eeden-Moorefield, 2018). Second, the measure of emotion work is based on self-reports and may not objectively capture emotion work provision. The emotion work measure focuses on assessing and managing the emotions of one’s spouse, but future research—in line with past studies (Erickson, 2005; Umberson et al., 2015)—should also incorporate measures regarding management of one’s own emotions and testing how this is associated with psychological well-being. Third, we distinguish partners based on gender and gendered dyadic contexts but future research might also distinguish partners based on the amount of emotion work provided by each spouse. Fourth, in analyses predicting psychological well-being based on emotion work provided to a spouse with elevated depressive symptoms, we do not know if spouses were formally diagnosed or receiving treatment for depression. The stress of emotion work provision on the emotion worker may vary depending on whether their partner is in treatment for depression—or other mental health issues. And finally, it is not possible with the present data to isolate emotion work as the cause of respondent well-being when a spouse has elevated depressive symptoms, and so we caution against a causal interpretation of our findings, given that it may be that the spouse’s depressive symptoms are the driver of the respondent’s increased emotion work as well as lower psychological well-being.

Future research should address the psychological costs of emotion work for emotion workers and the social conditions that exacerbate or ameliorate the psychological costs of emotion work. The strain of emotion work may also take a toll on marital quality as well as the health behaviors and health of both partners, and these possibilities should be explored in future work. Marital dynamics that increase distress for either partner may undermine overall marital quality (Williams, Frech, & Carlson, 2010), as well as the health and well-being of others in their family or friend groups. Marital quality is in turn strongly associated with long-term health, well-being, and mortality risk (Proulx, Helms, & Buehler, 2007; Robles, Slatcher, Trombello, & McGinn, 2014). Despite the importance of emotion work for the well-being of a couple and the emotion work recipient, special consideration should be given to the well-being of the emotion worker across diverse marital dyads and to how processes of emotion work may serve to exacerbate existing inequalities in relationships.

Acknowledgments

This research was supported by grant R21AG044585 from the National Institute on Aging (awarded to Umberson) and grant P2CHD042849, Population Research Center, and T32HD007081, Training Program in Population Studies, both awarded to the Population Research Center at The University of Texas at Austin by the Eunice Kennedy Shriver National Institute of Child Health and Human Development. This study was also funded in part by the National Institute on Alcohol Abuse and Alcoholism grant number F32AA025814 (awarded to Pollitt). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Contributor Information

Debra Umberson, University of Texas at Austin.

Mieke Beth Thomeer, The University of Alabama at Birmingham.

Amanda M. Pollitt, University of Texas at Austin.

Sara E. Mernitz, University of Texas at Austin.

References

  1. Almeida DM, & Kessler RC (1998). Everyday stressors and gender differences in daily distress. Journal of Personality and Social Psychology, 75, 670–680. 10.1037/0022-3514.75.3.670 [DOI] [PubMed] [Google Scholar]
  2. Andresen EM, Malmgren JA, Carter WB, & Patrick DL (1994). Screening for depression in well older adults: Evaluation of a short form of the CES-D. American Journal of Preventive Medicine, 10, 77–84. 10.1016/S0749-3797(18)30622-6 [DOI] [PubMed] [Google Scholar]
  3. Balsam KF, Beauchaine TP, Rothblum ED, & Solomon SE (2008). Three-year follow-up of same-sex couples who had civil unions in Vermont, same-sex couples not in civil unions, and heterosexual married couples. Developmental Psychology, 44, 102–116. 10.1037/0012-1649.44.1.102 [DOI] [PubMed] [Google Scholar]
  4. Behler R, Donnelly R, & Umberson D (2019). Psychological distress transmission in same-sex and different-sex marriages. Journal of Health and Social Behavior, 60, 18–35. 10.1177/0022146518813097 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bernard J (1987). The future of marriage. New Haven, CT: Yale University Press. [Google Scholar]
  6. Bolger N, & Laurenceau JP (2013). Intensive longitudinal methods: An introduction to diary and experience sampling research. New York, NY: Guilford Press. [Google Scholar]
  7. Connell RW (2005). Masculinities. Berkeley, CA: University of California Press. [Google Scholar]
  8. Cook W, & Kenny D (2005). The actor-partner interdependence model: A model of bidirectional effects in developmental studies. International Journal of Behavioral Development, 29, 101–109. 10.1080/01650250444000405 [DOI] [Google Scholar]
  9. Curran MA, McDaniel BT, Pollitt AM, & Totenhagen CJ (2015). Gender, emotion work, and relationship quality: A daily diary study. Sex Roles, 73, 157–173. 10.1007/s11199-015-0495-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Duncombe J, & Marsden D (1993). Love and intimacy: The gender division of emotion and emotion work: A neglected aspect of sociological discussion of heterosexual relationships. Sociology, 27, 221–241. 10.1177/0038038593027002003 [DOI] [Google Scholar]
  11. Duncombe J, & Marsden D (1995). “Workaholics” and “whingeing women”: Theorising intimacy and emotion work—The last frontier of gender inequality? The Sociological Review, 43, 150–169. 10.1111/j.1467-954X.1995.tb02482.x| [DOI] [Google Scholar]
  12. Elliott S, & Umberson D (2008). The performance of desire: Gender and sexual negotiation in long-term marriages. Journal of Marriage and Family, 70, 391–406. 10.1111/j.1741-3737.2008.00489.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Erickson RJ (1993). Reconceptualizing family work: The effect of emotion work on perceptions of marital quality. Journal of Marriage and the Family, 55, 888–900. 10.2307/352770 [DOI] [Google Scholar]
  14. Erickson RJ (2005). Why emotion work matters: Sex, gender, and the division of household labor. Journal of Marriage and Family, 67, 337–351. 10.1111/j.0022-2445.2005.00120.x [DOI] [Google Scholar]
  15. Ferree MM (2010). Filling the glass: Gender perspectives on families. Journal of Marriage and the Family, 72, 420–439. 10.1111/j.1741-3737.2010.00711.x [DOI] [Google Scholar]
  16. Flynn JJ, Hollenstein T, & Mackey A (2010). The effect of suppressing and not accepting emotions on depressive symptoms: Is suppression different for men and women? Personality and Individual Differences, 49, 582–586. 10.1016/j.paid.2010.05.022 [DOI] [Google Scholar]
  17. Gates GJ (2014). LGB families and relationships: Analyses of the 2013 National Health Interview Survey. Los Angeles: The Williams Institute, UCLA School of Law. [Google Scholar]
  18. Gates GJ (2015). Demographics of married and unmarried same-sex couples: Analyses of the 2013 American Community Survey. Los Angeles: The Williams Institute, UCLA School of Law. [Google Scholar]
  19. Gotta G, Green RJ, Rothblum E, Solomon S, Balsam K, & Schwartz P (2011). Heterosexual, lesbian, and gay male relationships: A comparison of couples in 1975 and 2000. Family Process, 50, 353–376. 10.1111/j.1545-5300.2011.01365.x [DOI] [PubMed] [Google Scholar]
  20. Hochschild AR (1989). The second shift: Working parents and the revolution at home. New York: Viking. [Google Scholar]
  21. Hochschild AR (2003). The managed heart: Commercialization of human feeling. Berkeley, CA: University of California Press. [Google Scholar]
  22. Horne RM, & Johnson MD (2018). A labor of love? Emotion work in intimate relationships. Journal of Social and Personal Relationships, 36, 1190–1209. 10.1177/0265407518756779 [DOI] [Google Scholar]
  23. Irwin M, Artin KH, & Oxman MN (1999). Screening for depression in the older adult: Criterion validity of the 10-item Center for Epidemiological Studies Depression Scale (CES-D). Archives of Internal Medicine, 159, 1701–1704. 10.1001/archinte.159.15.1701 [DOI] [PubMed] [Google Scholar]
  24. Johnson JL, Oliffe JL, Kelly MT, Galdas P, & Ogrodniczuk JS (2012). Men’s discourses of help-seeking in the context of depression. Sociology of Health & Illness, 34, 345–361. 10.1111/j.1467-9566.2011.01372.x [DOI] [PubMed] [Google Scholar]
  25. Kessler RC, McLeod JD, & Wethington E (1985). The costs of caring: A perspective on the relationship between sex and psychological distress. In Social support: Theory, research and applications (pp. 491–506). Dordrecht, The Netherlands: Springer. [Google Scholar]
  26. Kouros CD, & Cummings EM (2010). Longitudinal associations between husbands’ and wives’ depressive symptoms. Journal of Marriage and Family, 72, 135–147. 10.1111/j.1741-3737.2009.00688.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kroeger RA, & Powers DA (2019). Examining same-sex couples using dyadic data methods. In Schoen R (Ed.), Analytical family demography: The Springer series on demographic methods and population analysis (pp. 157–186). Cham, Switzerland: Springer. 10.1007/978-3-319-93227-9_7 [DOI] [Google Scholar]
  28. Lutz CA (1988). Unnatural emotions. Chicago: University of Chicago Press. 10.7208/chicago/9780226219783.001.0001 [DOI] [Google Scholar]
  29. Minnotte KL (2017). Integrative and masking emotion work: Marital outcomes among dual-earner couples. Marriage & Family Review, 53, 88–104. 10.1080/01494929.2016.1157563 [DOI] [Google Scholar]
  30. Minnotte KL, Pedersen D, & Mannon SE (2010). The emotional terrain of parenting and marriage: Emotion work and marital satisfaction. The Social Science Journal, 47, 747–761. 10.1016/j.soscij.2010.07.011 [DOI] [Google Scholar]
  31. Mroczek DK, & Kolarz CM (1998). The effect of age on positive and negative affect: A developmental perspective on happiness. Journal of Personality and Social Psychology, 75, 1333–1349. 10.1037/0022-3514.75.5.1333 [DOI] [PubMed] [Google Scholar]
  32. Pfeffer CA (2010). “Women’s work”? Women partners of transgender men doing housework and emotion work. Journal of Marriage and Family, 72, 165–183. 10.1111/j.1741-3737.2009.00690.x [DOI] [Google Scholar]
  33. Proulx CM, Helms HM, & Buehler C (2007). Marital quality and personal well-being: A meta-analysis. Journal of Marriage and Family, 69, 576–593. 10.1111/j.1741-3737.2007.00393.x [DOI] [Google Scholar]
  34. Radloff LS (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385–401. 10.1177/014662167700100306 [DOI] [Google Scholar]
  35. Rao AH (2017). Stand by your man: Wives’ emotion work during men’s unemployment. Journal of Marriage and Family, 79, 636–656. 10.1111/jomf.12385 [DOI] [Google Scholar]
  36. Reczek C, Thomeer MB, Gebhardt-Kram L, & Umberson D (2020). “Go see somebody”: How spouses promote mental health care. Society and Mental Health, 10, 80–96. 10.1177/2156869319834335 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Rehman US, Gollan J, & Mortimer AR (2008). The marital context of depression: Research, limitations, and new directions. Clinical Psychology Review, 28, 179–198. 10.1016/j.cpr.2007.04.007 [DOI] [PubMed] [Google Scholar]
  38. Robles TF, Slatcher RB, Trombello JM, & McGinn MM (2014). Marital quality and health: A meta-analytic review. Psychological Bulletin, 140, 140–187. 10.1037/a0031859 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Rosenquist JN, Fowler JH, & Christakis NA (2011). Social network determinants of depression. Molecular Psychiatry, 16, 273–281. 10.1038/mp.2010.13 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Sharabi LL, Delaney AL, & Knobloch LK (2016). In their own words: How clinical depression affects romantic relationships. Journal of Social and Personal Relationships, 33, 421–448. 10.1177/0265407515578820 [DOI] [Google Scholar]
  41. Springer KW, Hankivsky O, & Bates LM (2012). Gender and health: Relational, intersectional, and biosocial approaches. Social Science & Medicine, 74, 1661–1666. 10.1016/j.socscimed.2012.03.001 [DOI] [PubMed] [Google Scholar]
  42. Taylor J (2015). Gender orientation and the cost of caring for others. Society and Mental Health, 5, 49–65. [Google Scholar]
  43. Taylor S (2011). Social support: A review. In Friedman HS (Ed.), The Oxford handbook of Health Psychology (pp. 189–214). New York: Oxford University Press. [Google Scholar]
  44. Thomeer MB, Reczek C, & Umberson D (2015). Gendered emotion work around physical health problems in mid-and later-life marriages. Journal of Aging Studies, 32, 12–22. 10.1016/j.jaging.2014.12.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Thomeer MB, Umberson D, & Pudrovska T (2013). Marital processes around depression: A gendered and relational perspective. Society and Mental Health, 3, 151–169. 10.1177/2156869313487224 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Thomeer MB, Umberson D, & Reczek C (2020). The gender-as-relational approach for theorizing about romantic relationships of sexual and gender diverse mid- to later-life adults. Journal of Family Theory & Review. 10.1111/jftr.12368 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Umberson D, Thomeer MB, & Lodge AC (2015). Intimacy and emotion work in lesbian, gay, and heterosexual relationships. Journal of Marriage and Family, 77, 542–556. 10.1111/jomf.12178 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. van Eeden-Moorefield B (2018). Introduction to the special issue: Intersectional variations in the experiences of queer families. Family Relations, 67, 7–11. 10.1111/fare.12305 [DOI] [Google Scholar]
  49. Watson D, Clark LA, & Tellegen A (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54, 1063–1070. 10.1037/0022-3514.54.6.1063 [DOI] [PubMed] [Google Scholar]
  50. West C, & Zimmerman DH (1987). Doing gender. Gender & Society, 1, 125–151. 10.1177/0891243287001002002 [DOI] [Google Scholar]
  51. West TV, Popp D, & Kenny DA (2008). A guide for the estimation of gender and sexual orientation effects in dyadic data: An actor-partner interdependence model approach. Personality and Social Psychology Bulletin, 34, 321–336. 10.1177/0146167207311199 [DOI] [PubMed] [Google Scholar]
  52. Williams K, Frech A, & Carlson DL (2010). Marital status and mental health. In Scheid T & Brown T (Eds.), A handbook for the study of mental health: Social contexts, theories, and systems (pp. 306–320). New York: Cambridge University Press. [Google Scholar]
  53. Wong JS (2017). Competing desires: How young adult couples negotiate moving for career opportunities. Gender & Society, 31, 171–196. 10.1177/0891243217695520 [DOI] [Google Scholar]
  54. Yousaf O, Grunfeld EA, & Hunter MS (2015). A systematic review of the factors associated with delays in medical and psychological help-seeking among men. Health Psychology Review, 9, 264–276. 10.1080/17437199.2013.840954 [DOI] [PubMed] [Google Scholar]
  55. Zauszniewski JA, & Bekhet AK (2009). Depressive symptoms in elderly women with chronic conditions: Measurement issues. Aging and Mental Health, 13, 64–72. 10.1080/13607860802154481 [DOI] [PubMed] [Google Scholar]

RESOURCES