A labor of love? Emotion work in intimate relationships

Abstract

Drawing on relational developmental systems and gender relations perspectives, this study analyzed data from 1,932 heterosexual couples from Waves 1 and 2 of the German Family Panel to answer three questions: (1) What are the longitudinal associations between male and female partners’ emotion work provision and relationship satisfaction? (2) Are there gender differences in associations between emotion work and relationship satisfaction? (3) Does autonomy moderate associations among these focal variables? An actor–partner interdependence model revealed emotion work was linked to heightened future relationship satisfaction, and female partners’ emotion work was the strongest predictor of both partners’ relationship satisfaction. Latent variable interactions demonstrated male partners’ emotion work was linked to female partners’ heightened relationship satisfaction only when men also reported high levels of autonomy. Emotion work may be a “labor of love” that builds future relationship satisfaction while under the differential “management” of autonomous self-representation and gender norms of affective care.

Keywords

Autonomy developmental theory dyadic data emotion work gender intimate relationships

It is commonly believed that “relationships are a lot of work,” but what kind of “work” sustains an intimate relationship? In response to this question, intimate relationship scholars have explored the concept of emotion work, which refers to supportive behaviors that foster a partner’s positive emotions (e.g., listening attentively to a partner’s thoughts and acknowledging his or her emotions; Erickson, 2005). Although prior research found emotion work was associated with higher relationship satisfaction and stronger commitment (Curran, McDaniel, Pollitt, & Totenhagen, 2015; Minnotte, Pederson, & Mannon, 2010), habitually focusing one’s efforts on enhancing a partner’s emotional well-being was also linked to relationship conflict and psychological distress, particularly for women (Strazdins & Broom, 2004; Umberson, Thomeer, & Lodge, 2015). Further, the greater discrepancy between one’s sense of self, beliefs, and goals and emotion work enacted for a partner may prove especially detrimental to personal and relationship well-being. In short, the literature lacks consensus about whether emotion work is helpful or harmful for relationship quality, especially over time, and little research has explored how intrapersonal characteristics (e.g., one’s ability to stay autonomous in the face of influence from others; Noom, Dekovic, & Meeus, 2001) may shape this association for men and women.

The present study combines relational developmental systems (RDS) theory (Lerner, Johnson, & Buckingham, 2015) with a gender relations perspective (Ferree, 2010) and draws on data from couples in the German Family Panel (pairfam) study (Brüderl et al., 2015) to answer three main questions: (1) What are the longitudinal associations between male and female partners’ emotion work provision and each partner’s relationship satisfaction? (2) Are there gender differences in associations between emotion work provision and relationship satisfaction? (3) Does autonomy moderate associations among these focal variables?

Background

Theoretical framework

The present study is guided by a combined RDS and gender relations theoretical perspective. The RDS meta-model frames human development as a relational process involving continual, bidirectional transactions between an individual and his or her intrapersonal (e.g., cognitions), interpersonal (e.g., family), and sociocultural (e.g., gender norms) contexts over the life course (Lerner et al., 2015). Grounded in the worldview of relationism (paradigm that focuses on processes within, change throughout, and relations among developmental systems) and the principle of holism (the need to consider context to derive meaning), an RDS perspective suggests person and context mutually constitute one another and evolve together across time (Overton, 2013). As such, an individual’s cognitive and behavioral trajectories will inevitably shape and be shaped by his or her relational environments (Lerner et al., 2015). One’s intimate union is certainly a salient and proximal relational environment that influences individual development. An RDS perspective considers how individuals’ attributes, relationship cognitions, and partnership dynamics may be influenced by their own experiences, their partners’ characteristics, and their multiple contexts.

Applied to the present study, an RDS perspective necessitates the use of longitudinal data from couples to explore interrelations among behavioral (emotion work) and cognitive (relationship satisfaction) characteristics over time. This study represents, to our knowledge, the first attempt at investigating emotion work processes in relationships with multiple waves of data, addressing calls for such a study over two decades ago (Erickson, 1993). The coevolution of individual and relational contexts under an RDS framework also leads us to consider both partners’ emotion work and relationship satisfaction. Indeed, providing emotion work may be a delightful or woeful task with consequences for one’s own relationship satisfaction, but it may also influence how the receiving partner appraises the quality of his or her relationship.

We are also informed by a gender relations perspective that focuses on the salience of gender across micro-, meso-, and macro-environments (Ferree, 2010). Grounded in feminist theory, this perspective defines gender as a “socially constructed stratification system” (Risman, 2004, p. 430) that structures individual behavior and interpersonal relations based on assumed differences between men and women. Although men and women are generally more similar than different on many intra- and interpersonal qualities (e.g., Carothers & Reis, 2013), a gender relations perspective suggests gendered behaviors are continually (re)produced to maintain inequalities between men and women (Ferree, 2010). The division of family work (i.e., housework, child care, and emotion work) is one way gender differences are preserved, with women held more responsible for this work than men given cultural constructions of women as nurturers and emotional supporters (and men as independent and stoic; Curran et al., 2015; Thomeer, Reczek, & Umberson, 2015). Individuals may feel compelled to engage in or disengage from family work to uphold these gendered identities, which may result in different—even divergent—relationship satisfaction appraisals for men and women.

A gender relations perspective requires an examination of emotion work on the basis of gender. Stronger associations may emerge between women’s emotion work and both partners’ relationship satisfaction, as women’s emotional support may be expected and preferred because it aligns with traditionally feminine behavior. Although men’s emotion work may still impact both partners’ relationship appraisals, this link may be weaker because emotion work is antithetical to conventional masculinity and deemed less central to positive relationship functioning (e.g., Duncombe & Marsden, 1993; Holmes, 2015). We compare the relative strength of partners’ pathways between emotion work and relationship satisfaction to test whether associations differ based on gender.

Emotion work, relationship quality, and gender

Although Hochschild (1979) originally conceptualized emotion work as an intrapersonal process of altering one’s emotions to produce a certain bodily display, she argued emotion work can also be performed “by the self upon others” (p. 562). As such, more recent relationship research has explored this interpersonal form of emotion work and whether it is beneficial for relationship functioning. Moreover, while most studies have aligned their conceptualizations of emotion work with Erickson’s (1993) pioneering study—defining it as “the enhancement of others’ emotional well-being and the provision of emotional support” (p. 888)—they utilize varying operationalizations of the construct.

Several studies document the benefits of such emotion work for relationship well-being, leading some to conclude “the more emotion work performed, the better” (Erickson, 1993, p. 896). Erickson (1993) analyzed cross-sectional data from 205 married women and found spouses’ emotion work (e.g., sharing thoughts and feelings, expressing concern for a partner’s well-being) was positively associated with women’s marital stability and satisfaction and negatively associated with their marital burnout, and Wilcox and Nock (2006) found men’s emotion work (operationalized as displays of affection and understanding) was the strongest predictor of women’s marital happiness. Other work found men and women were most satisfied with their relationship when they provided similar levels of emotion work (e.g., attempting to relieve a partner of his or her negative mood; Holm, Werner-Wilson, Cook, & Berger, 2001). Providing emotion work also predicted higher daily relationship satisfaction, love, commitment, and closeness for men and women in a week-long diary study that utilized Erickson’s original emotion work items (Curran et al., 2015).

These findings have been corroborated by various studies drawing on data from a family and work study in the Western U.S. that assessed cross-partner reports of emotion work performance (items inquired about self-disclosure, reducing negative moods, expressing praise, providing solutions, initiating problem talks, and sensing a partner’s distress) and relationship outcomes. For example, receiving emotion work from a partner was associated with higher marital satisfaction and lower marital burnout for men and women (Minnotte et al., 2010; Pederson, Minnotte, Mannon, & Kiger, 2011). Findings are mixed on gender differences in providing emotion work: Minnotte et al. (2010) found emotion work performance was positively linked to men and women’s self-reported marital satisfaction, but Pederson et al. (2011) only found this association for women. These studies suggest receiving emotion work is linked to more satisfying intimate ties, and providing emotion work is also linked to enhanced marital well-being, particularly for women.

Other research suggests, however, that emotion work is not always beneficial to intimate partnerships. Curran, McDaniel, Pollitt, and Totenhagen (2015) found men with partners who provided frequent emotion work had more fluctuating daily reports of love and commitment. Further, female partners who provided more emotion work than they received (operationalized as understanding affectual needs, providing emotional and instrumental support, and working to improve one’s relationship) reported feeling less loved, higher marital conflict, and more depressive symptoms in a study of couples raising young children (Strazdins & Broom, 2004). Moreover, heterosexual couples experienced more conflict around emotion work performance compared to same-sex couples (Umberson et al., 2015). These studies suggest emotion work may be regulated by norms that specify who should (and should not) provide emotion-based support in intimate unions, which may have differential consequences for men’s and women’s well-being.

Taken together, the literature on emotion work, relationship satisfaction, and gender reveals inconsistent findings, necessitating further exploration into associations among these variables. The aforementioned studies all had cross-sectional designs (with the exception of Curran et al.’s (2015) 7-day study) and many continue to rely on only one partner’s perception of emotion work in his or her relationship (e.g., Erickson, 2005; Wilcox & Nock, 2006). We use both partners’ reports of their emotion work provision and relationship satisfaction at two time points to address these limitations, testing for actor (e.g., intrapersonal) and partner (e.g., interpersonal) effects in how these variables are connected over a 1-year time span. In line with most emotion work studies, relationship satisfaction is our outcome variable of interest because it is one of the most salient indicators of relationship quality (Finkel, Simpson, & Eastwick, 2017) and is theoretically consistent with emotion work efforts. Given that emotions and satisfaction are distinct (but related) dimensions of subjective well-being (Lucas, Diener, & Suh, 1996), it is theorized that the emotions partners experience or elicit from one another serves as the affectual material that informs cognitive appraisals of general relationship satisfaction.

Autonomy as a potential moderator

To understand the conditions under which emotion work is constructive or destructive for relationship well-being, Curran et al. (2015) suggested exploring factors that moderate associations between these constructs; we examine one’s autonomy as a moderator of the emotion work-relationship satisfaction link. Autonomy broadly refers to self-governance or feeling a sense of control over one’s thoughts and actions and has been conceptualized as a fundamental human need that optimizes well-being (Ryan & Deci, 2000). More precisely, autonomy involves a level of self-assurance in personal beliefs, decisions, and goals, which is characterized by the ability to maintain one’s sense of self and resist external influence from others (Noom et al., 2001). Autonomy may impact the emotion work process because these efforts may be more successful if an individual is confident in (and can communicate), their genuine desires to meet a partner’s emotive needs, ultimately boosting each partner’s relationship satisfaction. Conversely, the emotion work of individuals who are overly influenced by others (i.e., have low autonomy) might seem forced or feigned, which may strain relational well-being.

Several studies provide empirical support for the conceptual link between emotion work, autonomy, and relationship quality. For example, those who felt autonomous in their relationships (rather than coerced or pressured) also reported stronger cohesion with their partners and higher relationship satisfaction (Blais, Sabourin, Boucher, & Vallerand, 1990). Furthermore, individuals with autonomous motivations engaged in more positive coping strategies during stressful relationship events (e.g., expressed emotions and attempted to understand partners’ emotions; Knee, Patrick, Vietor, Nanayakkara, & Neighbors, 2002). In contrast, failing to express true emotions and elevating others’ needs above one’s own is associated with more negative emotions, difficultly sharing thoughts and feelings, and discomfort with receiving support (Helgeson & Fritz, 1998). As such, we examine the potential moderating role of autonomy in the link between emotion work and relationship satisfaction.

The present study

Drawing on RDS (Lerner et al., 2015) and gender relations (Ferree, 2010) perspectives and data from 1,932 couples, the present study examines three questions. First, what are the longitudinal associations between each partner’s emotion work provision (e.g., “I listen to my partner and give him/her the chance to express himself/herself”) and relationship satisfaction? Second, do associations between emotion work and relationship satisfaction differ between men and women? Third, does male and female partner autonomy (e.g., “I often agree with others, even if I’m not sure” [reverse coded]) moderate these associations? This study represents the first effort to explore longitudinal links between emotion work and relationship satisfaction from a national sample of couples, enabling a rigorous exploration of individual and cross-partner effects.

Based on the literature reviewed, we offer three hypotheses. First—and perhaps most tentatively given mixed evidence from past research—we expect that emotion work provision will generally be associated with higher relationship satisfaction in the future (Hypothesis 1). Next, we predict that women’s emotion work provision will demonstrate stronger associations with male and female partners’ relationship satisfaction compared to men’s emotion work (Hypothesis 2). Finally, we hypothesize there will be a positive association between emotion work provision and both partners’ relationship satisfaction for individuals with high autonomy, but a negative association between emotion work provision and both partners’ relationship satisfaction for individuals with low autonomy (Hypothesis 3).

Method

Procedures

Data from Waves 1 (2008) and 2 (2009) of the German Family Panel (pairfam) study were analyzed to answer the research questions (Brüderl et al., 2015). Pairfam is a longitudinal, multi-informant study funded by the German Research Foundation that commenced in 2008 and is scheduled to culminate in 2022 (Huinink et al., 2011). The aim of pairfam is to explore couple and family development over the life course from psychological, sociological, and ecological perspectives.

In 2008, 12,402 focal participants (referred to as anchors) who belonged to three age cohorts were recruited into pairfam: adolescents 15–17 years old (born between 1991 and 1993), young adults 25–27 years old (born between 1981 and 1983), and adults approaching midlife 35–37 years old (born between 1971 and 1973). Just over half of the 12,402 anchors recruited were in intimate relationships and asked for permission to contact their partners to participate in the study. A total of 3,743 intimate partners consented and returned surveys, resulting in a subsample of 3,743 couples. Further details are included in the study’s concept paper (Huinink et al., 2011) and website (http://www.pairfam.de/en/study.html).

Sample

We first filtered the sample (from the initial 3,743 intimate partner pairs) to only include those individuals in continuing partnerships from Waves 1 to 2 (n = 2,302). We then excluded those in the adolescent cohort (n = 336) given the instability of relationships in that age range (e.g., Carver, Joyner, & Udry, 2003), as well as those who reported being in same-sex partnerships (n = 34) because there were too few individuals in this subgroup for analysis, resulting in our final subsample of 1,932 heterosexual couples used in this study.

Most couples (n = 1197; 62.0%) were nearing midlife (female partner age: M = 34.96, SD = 3.45; male partner age: M = 37.81, SD = 3.71), and the remaining couples (n = 735; 38.0%) were in young adulthood (female partner age: M = 25.54, SD = 2.29; male partner age: M = 28.21, SD = 4.02). In terms of ethnicity, the majority of participants (78.5% of females and 82.1% of males) were German natives, while the rest were non-German (8.3% of females and 5.0% of males), half German (6.6% of females and 5.2% of males), ethnic-German immigrants (4.6% of females and 4.5% of males), or Turkish (2.1% of females and 3.2% of males). Over three-quarters of participants (80.4%) lived in Western Germany. Most participants (35.5% of females and 33.0% of males) completed a university degree, and the median monthly net household income in the sample was €2,600 (M = €2,797; SD = €1,333). Partners were together for 9.27 years (SD = 5.61) on average, with 66.1% married, 26.2% cohabiting, and 7.7% living apart together. Over one-third of couples (35.8%) had no children, 24.5% had one child, 28.9% had two children, and 10.8% had three or more children.

Measures

Emotion work

An emotion work scale was developed from 6 items assessed at Wave 1 that measure emotionally supportive behaviors in intimate unions. Parceling was utilized to generate three groupings of items based on conceptual similarity. Parcel 1 was comprised of 3 items from Bodenmann’s (2000) Dyadic Coping Questionnaire that measured supportive behaviors toward partners during times of stress. Participants were asked: “When your partner is stressed out, how often do you react in the following ways?” Items were: “I let my partner know that I understand him/her,” “I listen to my partner and give him/her the chance to express himself/herself,” and “I support my partner in concrete ways when he/she has a problem.” Parcel 2 consisted of 2 items from the Network of Relationships Inventory (Furman & Buhrmester, 1985) that assessed frequency of self-disclosure to a partner: “How often do you tell your partner what you’re thinking?” and “How often do you share your secrets and private feelings with your partner?” The response scale for parcels 1 and 2 ranged from 1 = never to 5 = always, and mean scores for items in each parcel were computed. Parcel 3 consisted of 1 item adapted from the Positive Problem Solving Behavior scale of Bodenmann’s (2000), Marital Communication Questionnaire and assessed participants’ constructive behaviors during conflicts in the past 6 months. Participants were asked: “When you have a disagreement with your partner, how often do you engage in the following behaviors?” The item was: “Listen to and ask questions of your partner in order to understand better.” Responses ranged from 1 = almost never or never to 5 = very frequently. These parcels were indicators of the latent emotion work construct.

Relationship satisfaction

Male and female partners’ satisfaction with their intimate unions was assessed with 1 item adapted from the Relationship Assessment Scale (Hendrick, Dicke, & Hendrick, 1998) at Waves 1 and 2: “All in all, how satisfied are you with your relationship?” Responses ranged from 0 = very dissatisfied to 10 = very satisfied. This measure is consistent with several global relationship satisfaction items from validated scales (e.g., Funk & Rogge, 2007), and psychometric evidence from research on life satisfaction—a related construct—suggests single-item measures (“All things considered, how satisfied are you with your life as a whole?”) were as valid as multi-item measures (Cheung & Lucas, 2014).

Autonomy

The mean of 3 items adapted from the Autonomy Scale (Noom et al., 2001) assessed participants’ levels of autonomy at Wave 1: “I often change my mind after hearing what others think,” “I strongly tend to follow the wishes of others,” and “I often agree with others, even if I’m not sure.” Responses ranged from 1 = not at all to 5 = absolutely, and items were reverse coded so that higher values reflected higher levels of autonomy. Cronbach’s α for these items was .67 for males and .68 for females.

Control variables

Relationship duration and satisfaction, number of children, gender role ideology, division of family work (housework and child care), and living in Eastern Germany are included as covariates in our analysis. Support provision (Johnson, Horne, & Galovan, 2016) and marital happiness (VanLaningham, Johnson, & Amato, 2001) tend to decline over time, and baseline relationship satisfaction must be held constant to ensure links between emotion work and future satisfaction are not due to unexamined continuity in satisfaction. Couples with children experienced stronger declines in supportive behaviors over time (Johnson et al., 2016). In terms of gender ideology, traditional gender role attitudes were linked to men’s lower emotion work provision (Erickson, 2005). Men’s housework and child care involvement is positively linked to both partners’ emotion work and women’s marital well-being (Erickson, 1993). Historically, Eastern Germany enacted more egalitarian policies and norms on work and family life than the West, which may influence family work practices (Cooke, 2007).

At baseline, participants were asked to report on their relationship duration in months (recoded into years) and their relationship satisfaction. Participants also specified the number of children in their household, as well as whether they resided in Eastern Germany. The mean of 2 items adapted from Hill and Arránz Becker (2004) measured the extent to which participants had traditional gender role ideologies: “Women should be more concerned about their family than about their career,” and “A child aged under six will suffer from having a working mother.” Responses ranged from 1 = disagree completely to 5 = agree completely. Relative housework and child care involvement was assessed with the following question adapted from Baxter (2002): “To what extent do you and your partner share duties in the following domains?” The domains were “housework (washing, cooking, cleaning)” and “taking care of the children.” Available responses for each domain were: 1 = (almost) completely my partner, 2 = for the most part my partner, 3 = split about 50/50, 4 = for the most part me, 5 = (almost) completely me, 6 = another person, and 7 = doesn’t apply to our situation. Responses were recoded as missing if they indicated someone else performed housework (.2% for females and .2% for males) and child care (.1% for females and 0% for males) tasks or if these tasks were irrelevant (housework: 1.5% for females and .8% for males; child care: 14.5% for females and 11.6% for males), with resulting responses ranging from 1 to 5. The tasks included in the housework measure are representative of the most time-consuming and routine household tasks assessed in existing research that uses composite housework measures (e.g., Erickson, 1993; Horne, Johnson, Galambos, & Krahn, 2017). The child care measure is consistent with other studies that have used single-item assessments of this variable (e.g., Pederson et al., 2011) and is further strengthened by the inclusion of raising children as a covariate, which will aid in the estimation of accurate parameter estimates for couples who reported child care does not apply to their situations (this absence was because they had no children).

Analytic plan

Missing data on our focal variables were quite low in this study (ranging from .10% on female partner emotion work to 2.6% for female partner Wave 1 relationship satisfaction) and were handled through full information maximum likelihood estimation. Analyses were computed in Mplus 7.4 (Muthén & Muthén, 1998–2012). We evaluated model fit on the basis of local fit testing (inspection of residuals) and global fit indices. Good fit was evident by a nonsignificant χ² test, values greater than .95 for the comparative fit index (CFI) and Tucker–Lewis index (TLI), and smaller than .06 and .08 for the root mean square error of approximation (RMSEA) and the standardized root mean square residual (SRMR; Little, 2013).

We first computed descriptive statistics and bivariate correlations for all study variables, followed by confirmatory factor analyses of the latent constructs (i.e., emotion work and autonomy). We then computed an actor–partner interdependence model (APIM; Kenny, Kashy, & Cook, 2006) to examine associations among male and female partners’ emotion work and relationship satisfaction. Finally, latent variable interactions were computed to test the moderating influence of autonomy on the emotion work/relationship satisfaction pathway following procedures outlined by Maslowsky, Jager, and Hemken (2015).

Results

Descriptive statistics and bivariate correlations

Table 1 contains descriptive statistics and bivariate correlations for the study variables. Mean levels of these variables revealed that participants reported performing emotion work quite frequently, were in highly satisfying relationships, and tended to be autonomous in their interpersonal interactions. Turning to the correlations, higher emotion work was moderately associated with higher relationship satisfaction for male and female partners. There were also small associations between emotion work and relationship satisfaction with higher autonomy for both partners. Relationship duration and raising more children was linked to less emotion work. Although higher emotion work was linked to men’s increased relative housework and child care involvement, emotion work was linked to women’s lower relative housework involvement and not to their child care. Associations among emotion work, having a traditional gender role attitude, and living in East Germany were nonsignificant for both partners. These correlations suggest emotion work, relationship satisfaction, and autonomy may be positively interrelated for male and female partners, and gender differences in emotion work may exist based on personal, familial, and sociodemographic characteristics (reaffirming the inclusion of these covariates).

Table 1.

Correlations and descriptive statistics for females above and males below the diagonal (n = 1,932 couples).

Variable	1	2	3	4	5	6	7	8	9	10	M	SD	Range
1. Emotion work	.21**	.30**	.15**	.35**	−.12**	−.15**	.03	−.05†	−.04	−.02	4.01	.58	1–5
2. Rel Sat W2	.27**	.38**	.10**	.46**	−.04	−.10**	.03	−.09**	−.04	−.01	8.10	1.95	1–10
3. Autonomy	.13**	.12**	.11**	.13**	.03	.03	−.08*	−.03	−.04	.04	3.90	.76	1–5
4. Rel Sat W1	.36**	.48**	.11**	.28**	−.05^†	−.08*	.05^†	−.10**	−.06^†	.01	8.43	1.74	1–10
5. Rel duration	−.13**	.03	−.02	.01	–	.52**	.02	.17**	.05	.00	9.27	5.61	0–20+
6. Children	−.15**	−.04	−.02	−.02	.52**	–	.10**	.23**	.07†	−.01	1.18	1.10	0–6
7. Trad GRA	−.01	.05†	−.07*	.02	.09**	.13**	.45**	.11**	.12**	−.17**	2.60	1.16	1–5
8. Housework	.09**	.00	.01	.02	−.18**	−.26**	−.22**	−.56**	.46**	−.10**	3.83	.83	1–5
9. Child care	.10*	.04	.01	.05	−.06†	−.09*	−.11**	.36**	−.42**	−.17**	3.70	.73	1–5
10. East Germany	.01	−.01	.09**	−.01	.00	−.01	−.20**	.10**	.11**	–	.20	.40	0–1
M	3.79	8.18	4.02	8.56	9.27	1.18	2.73	2.16	2.34	.20
SD	.62	1.79	.72	1.65	5.61	1.10	1.09	.81	.70	.40
Range	1–5	1–10	1–5	1–10	0–20+	0–6	1–5	1–5	1–5	0–1

Note. Correlations between male and female partners’ scores are along the diagonal. Rel, relationship; Sat, satisfaction; W1, Wave 1; W2, Wave 2; Trad GRA, traditional gender role attitude; East Germany, living in Eastern Germany at baseline.

**p < .001. *p < .01. †p < .05 (two-tailed).

Actor–partner interdependence model

Next, we tested the measurement of the latent constructs (each partner’s emotion work and autonomy) with confirmatory factor analyses. A conceptually permissible covariance was added between dyadic coping and constructive conflict behavior for female partners to improve model fit. These CFA models fit the data well (emotion work: χ ² (5) = 4.025, RMSEA = .000 [90% CI = .000, .028], CFI = 1.000, TLI = 1.002, SRMR = .008; autonomy: χ ² (5) = .899, RMSEA = .000 [90% CI = .000, .000], CFI = 1.000, TLI = 1.008, SRMR = .004). We then proceeded to compute the APIM to test associations among male and female partners’ emotion work and relationship satisfaction (Hypotheses 1 and 2). Equality constraints were applied to corresponding pathways (one at a time) and a χ² difference test was computed to test whether path coefficients substantively differed within and between male and female partners. If constraining paths to equality reduced model fit (i.e., a statistically significant chi-square difference), then paths substantively differed and the constraint was dropped (Kenny et al., 2006). All within- (e.g., actor) and between-gender (e.g., partner) pathways significantly differed.

The control variables were then added to the model. Baseline relationship satisfaction was regressed only on future relationship satisfaction (with covariances specified among baseline male and female relationship satisfaction and male and female emotion work), and the other control variables were regressed on all variables in the model. Nonsubstantive paths were pruned one by one to produce the most parsimonious final model, with female partners’ relative child care involvement, male and female relative housework involvement, and male and female traditional gender role attitudes being omitted.

The final model is presented in Figure 1, and model fit indices reveal good fit to the data. In terms of actor effects and in line with Hypothesis 1, each partner’s own emotion work was linked to increases in their own relationship satisfaction 1 year later. The female emotion work-relationship satisfaction pathway was, however, significantly stronger than the male pathway (χ ² _diff (1) = 19.86, p < .000). In terms of partner effects, female partner emotion work was associated with higher relationship satisfaction for men, but male partner emotion work was not linked to female relationship satisfaction (β = .05, p = .119). These findings contribute more nuanced support for our first hypothesis. The link between female emotion work and her own relationship satisfaction was significantly stronger than the link from her emotion work to the male partner’s relationship satisfaction (χ ² _diff (1) = 29.51, p < .000). Similarly, the association between male emotion work and his own relationship satisfaction was significantly stronger than the nonsignificant partner effect from his emotion work and the female partner’s relationship satisfaction. Overall, the strongest predictor of Wave 2 relationship satisfaction for both partners was female partners’ prior emotion work provision, supporting Hypothesis 2.

Figure 1.

Final actor–partner interdependence model for male and female partners’ emotion work and relationship satisfaction (n = 1,932). Standardized estimates; control variables regressed on all variables in model. Dotted line represents nonsignificant association. Model fit indices: χ ² (62) = 102.542; RMSEA = .024 (90% CI = .015, .032); CFI = .987; TLI = .976; SRMR = .020. ***p < .001 (two-tailed). RMSEA, root mean square error of approximation; CI, confidence interval; CFI, comparative fit index; TLI, Tucker–Lewis index; SRMR, standardized root mean square residual.

Turning to the control variables, Wave 1 relationship satisfaction foretold future satisfaction, as anticipated. Having more children was linked to lower emotion work provision (females: β = −.21; males: β = −.17, ps < .001). Men in longer-term relationships provided less emotion work (β = −.08, p = .018), and those in longer-term relationships reported higher relationship satisfaction (females: β = .05, p = .028; males: β = .08, p = .001). Men’s higher relative child care involvement was linked to their own increased emotion work performance (β = .12, p < .001). Finally, living in Eastern Germany was linked to higher emotion work performed by female (β = .07, p = .032) and male (β = .18, p < .001) partners and men’s increased relationship satisfaction (β = .05, p = .024).

Latent variable interaction models

To test whether autonomy moderated associations between emotion work and relationship satisfaction (Hypothesis 3), we computed latent moderated structural equation models with Maslowsky et al. s’ (2015) two-step procedure. First, we computed a baseline model that examined direct effects of male and female partners’ emotion work provision and autonomy on relationship satisfaction without interaction terms. The key results from the baseline model are presented in Table 2. Notably, the only significant (albeit small) direct association that emerged was the positive link between male partner autonomy and his own relationship satisfaction.

Table 2.

Baseline and latent variable interaction model results for male and female partners’ autonomy (n = 1,932).

Baseline model (no interactions)
Pathways		ß		SE
Female Autonomy → Female Relationship Satisfaction		−.03		.03
Female Autonomy → Male Relationship Satisfaction		−.02		.03
Male Autonomy → Male Relationship Satisfaction		.06*		.03
Male Autonomy → Female Relationship Satisfaction		.02		.03
Latent variable interaction models
Outcome variable		ß	SE	H_o	D
Female relationship satisfaction
Interaction	Female Emotion Work × Female Autonomy	−.11	.06	−40700.30	10.12**
	Female Emotion Work × Male Autonomy	.01	.06	−40705.34	.04
	Male Emotion Work × Male Autonomy	.06*	.03	−40702.95	4.81*
Male relationship satisfaction
Interaction	Male Emotion Work × Male Autonomy	−.02	.04	−40705.11	.50
	Male Emotion Work × Female Autonomy	.04	.04	−40704.43	1.85
	Female Emotion Work × Female Autonomy	.02	.06	−40705.11	.50

Note. H_o , log-likelihood value; D, log-likelihood ratio test statistic. Baseline model fit indices: χ² (75) = 142.798; RMSEA = .022 (90% CI = .016, .027); CFI = .986; TLI = .978; SRMR = .018. Baseline model H_o = −40705.36. RMSEA, root mean square error of approximation; CI, confidence interval; CFI, comparative fit index; TLI, Tucker–Lewis index; SRMR, standardized root mean square residual. One degree of freedom used for all model comparisons. The relationship satisfaction of the opposite sex partner was included as a covariate in each latent interaction model.

**p < .01. *p < .05 (two-tailed).

Next, latent variable interaction models were computed one at a time because of the computational complexity of estimating these models (six in total; Muthén, 2011). These interaction models were identical to the baseline model with the addition of the relevant interaction term and represented all combinations of the focal variables for both partners (female relationship satisfaction was regressed on Female Emotion Work × Female Autonomy, Female Emotion Work × Male Aautonomy, and Male Emotion Work × Male Autonomy; male relationship satisfaction was regressed on Male Emotion Work × Male Autonomy, Male Emotion Work × Female Autonomy, and Female Emotion Work × Female Autonomy). Importantly, the relationship satisfaction of the opposite sex partner was included as a covariate in each interaction model. To determine if the latent variable interactions were significant, the relative fit of the baseline model with each separate interaction model was compared (one at a time) using a log-likelihood ratio test. If the ratio test is significant (i.e., a statistically significant chi-square value is attained), then retaining the baseline model would result in a significant loss in model fit relative to the interaction model and the researcher should retain the interaction model for that particular effect (Maslowsky, Jager, & Hemken, 2015).

The main results from the latent interaction models are also presented in Table 2. The only significant latent variable interaction was between male emotion work and male autonomy with higher female relationship satisfaction at Wave 2 (β = .06, p = .042). This finding was further supported by the significant log-likelihood ratio test comparing the male emotion work x male autonomy → female relationship satisfaction interaction model to the baseline model (χ² (1) = 4.81, p = .028). The moderating effect of male autonomy on the association between male emotion work and female relationship satisfaction is presented in Figure 2. Simple slopes tests were computed in Mplus to determine whether each interaction slope significantly differed from zero. The low male autonomy slope (one standard deviation below the male autonomy mean) was not different from zero (unstandardized results: β = .10, SE = .07, p = .188), while the high male autonomy slope (one standard deviation above mean) was significantly different from zero (unstandardized results: β = .39, SE = .09, p < .000). Frequent emotion work from male partners was only associated with higher female partner relationship satisfaction in the future when male partners also reported high autonomy, supporting our third hypothesis.

Figure 2.

Interaction plot illustrating the moderating effects of male partner autonomy on the association between male partner emotion work and female partner relationship satisfaction.

Turning to the other log-likelihood ratio results, the significant ratio test comparing the baseline model with the Female Emotion Work × Female Autonomy → female relationship satisfaction model suggests female autonomy may be an important moderator, yet the interaction term was not significant (p = .087) and counter to our prediction. The baseline model was retained, but further research is needed on the role of female autonomy in the emotion work-relationship satisfaction link. The remaining four log-likelihood ratio tests were nonsignificant. Taken together, these findings partially support Hypothesis 3.

Discussion

Guided by RDS and gender relations perspectives, this study examined longitudinal links between male and female partners’ emotion work provision and relationship satisfaction, potential gender differences in these associations, and whether autonomy moderated associations among the focal variables. The first notable finding was that providing emotion work for a partner was linked to one’s own higher relationship satisfaction in the future, controlling for prior satisfaction levels. This finding supports our first hypothesis and provides an important contribution to the literature by corroborating and extending results from several cross-sectional studies (e.g., Erickson, 1993; Holm et al., 2001; Minnotte et al., 2010), demonstrating emotion work can improve partners’ perceptions of their relationship quality 1 year later.

Individuals who perform emotion work may be engaging in the so-called “labor of love” because, although boosting a partner’s positive emotions may be time-consuming, psychologically demanding, or contrary to one’s personal mood (Erickson, 2011), it seems to enhance—not undermine—the provider’s relationship satisfaction. Perhaps feelings of burnout that may accompany emotion work efforts become mitigated by the rewarding end goal of building a positive relationship climate in which both partners benefit. Related research on sacrifice in intimate relationships (i.e., putting aside self-interest to benefit a partner) and emotion regulation processes supports this goal-oriented reasoning. Individuals who made sacrifices to achieve positive relationship outcomes (e.g., to make a partner happy) reported higher relationship satisfaction and less relationship conflict (Impett, Gable, & Peplau, 2005) and achieving desired goals in a social interaction buffered the negative impacts of strong emotion regulatory efforts on participants’ well-being (Wong, Tschan, & Semmer, 2017). Exploring how motivations and intentions behind emotion work processes contribute to personal and relational well-being presents an exciting opportunity for future research.

In addition to the overall positive link between emotion work and relationship satisfaction, the findings also reveal notable gender differences when considering the strength of associations between these variables: women’s emotion work provision was a stronger predictor of both male and female partners’ relationship satisfaction compared to men’s emotion work (supporting our second hypothesis). In other words, women’s emotion-enhancing efforts were especially salient in producing a positive relationship climate. These findings support arguments on how sociocultural norms underpinning the traditional male breadwinner and female homemaker family model continue to structure gender relations within the private context of an intimate union, despite the proliferation of alternative work and relational arrangements in contemporary society (e.g., Ferree, 2010; Risman, 2004). This traditional model casts women as competent emotional supporters who possess an ‘innate ability’ (and desire) to maintain family intimacy due to their biological childbearing capacity and traditional role as caregivers (Duncombe & Marsden, 1993). If gendered norms prescribe emotional care as an indicator of a female’s appropriate performance as a partner and woman, then male partners’ relationship satisfaction may be dependent on this support and female partners’ satisfaction on the performance of this work.

But why are the relational benefits of emotion work efforts strongest under the purview of women? One explanation may be that the actual quality of emotion work provided by female versus male partners is notably different, with the former providing more effective emotional support than the latter. Several studies found that although men and women engage in similar supportive behaviors in observational settings, they differ in how they tailor that support to their partners’ needs and perceive overall support exchanges in their relationships (e.g., Neff & Karney, 2005). Although women provided positive support to male partners encountering stress, several studies showed men engaged in more negative behaviors (Neff & Karney, 2005) or were unable to conceal their own negative emotions (Thomeer et al., 2015) when their partners were experiencing stress. Aligning with general trends in support exchanges, the emotion-specific support provided by female partners may be more helpful than what men provide, leading men to experience a stronger boost in relationship satisfaction while ascribing less importance to how their own emotive performance shapes their relationship cognitions.

Finally, the findings revealed one intriguing interaction effect that partially supported our third hypothesis: men’s autonomy moderated the association between male partner emotion work and female partner relationship satisfaction (variables that were otherwise unassociated). Specifically, the association between male emotion work and female relationship satisfaction was strengthened when male partners were more autonomous. If male partners were less autonomous, however, their emotion work efforts had no influence on their partners’ relationship satisfaction. Autonomous men—that is, men who can maintain their sense of self and personal goals amidst influence from others—may perform emotion work with sincere interests, even in the face of opposition from conventional standards of masculinity (e.g., independence, stoicism; Holmes, 2015). While rigid norms of masculinity are sustained in many ways, one prominent way is through men restricting their own expressiveness while policing each other’s behaviors to ensure they align with these standards (Connell & Messerschmidt, 2005; Duncombe & Marsden, 1993). As such, autonomous men may be better equipped to stay true to their personal convictions about being emotionally attentive partners and reject judgment that may challenge their ‘manliness’ for engaging in these actions. In turn, women may report more satisfying ties if they interpret their partners’ emotion work as genuinely motivated, especially if females expect less (but still desire) affective efforts from their male partners. These findings suggest men’s emotion-enhancing efforts can positively impact their partners’ relationship appraisals, so long as these efforts are accompanied by a strong sense of autonomy.

Limitations

Some limitations must be considered when interpreting these findings. First, although the items comprising the emotion work scale are similar to items in other studies (e.g., Erickson, 2005; Minnotte et al., 2010), they were derived from condensed versions of pre-existing scales. Pairfam is an omnibus study that collects data on many constructs from multiple informants to elucidate the breadth of familial experiences, which necessitates the use of shorter measures. As such, some of the present study’s constructs were assessed with 1 item (e.g., relationship satisfaction) or condensed scales (e.g., gender role ideology). The gender role ideology scale in particular—which focused exclusively on women’s career and family roles—precluded our ability to directly test proposed explanations for the salience of men’s autonomy on women’s relationship satisfaction, necessitating a broader scale that taps male and female partners’ gender roles in emotionally relevant contexts and perceptions of others’ gender ideologies. These measurement disadvantages are, however, offset by the opportunity to understand relational dynamics from a national sample comprised of both partners. Relatedly, given emotion work’s relatively recent empirical introduction into couples research (e.g., Erickson, 1993), the construct has been inconsistently measured across studies and contains additional dimensions not assessed in this study. For example, in addition to emotion work’s intrapersonal versus interpersonal components (Hochschild, 1979), it is also plausible that it contains a positive (e.g., boost one’s own or a partner’s positive emotions) versus negative (e.g., reduce one’s own or a partner’s negative emotions) dimension. Future work could investigate the underlying dimensions of emotion work processes to further refine the conceptualization and measurement of this multi-faceted construct. Finally, same-sex couples were excluded from the analysis due to their small sample size (n = 34), yet some scholars suggest individuals enact emotion work differently in same-sex or different-sex partnerships (Umberson et al., 2015). Considering how emotion work processes contribute to relationship appraisals based on intersections of gender and sexual orientation represents an important area for future research.

Conclusion

The present study examined the interrelations of emotion work, relationship satisfaction, and autonomy, as well as gender differences in these associations. The findings revealed emotion work was linked to heightened relationship satisfaction 1 year later for male and female partners, but women’s emotion work provision was particularly impactful for predicting both partners’ relationship satisfaction. In addition, male partner’s emotion work was only linked to women’s enhanced relationship satisfaction if he also reported a high level of autonomy. Taken together, these results reveal longer-term benefits that accrue when partners perform a “labor of love” by engaging in behaviors that boost each other’s positive emotions. Nevertheless, there is diversity in how these processes unfold. While men benefit from emotion work irrespective of their partners’ underlying motives, effort alone is not enough for women—rather, their partners’ emotion work must be accompanied by autonomous motivations to strengthen their relationship satisfaction. When managing the emotional climate of an intimate union, it seems as though women’s emotion work yields important relationship dividends, but men who “phone it in” with their affective efforts will not be remunerated.

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by Insight Development Grant 430-2014-00682 from the Social Sciences and Humanities Research Council of Canada (SSHRC).

References

Baxter

(2002). Patterns of change and stability in the gender division of household labour in Australia, 1986-1997. Journal of Sociology, 38, 399–424. doi:10.1177/144078302128756750

Blais

M. R.

Sabourin

Boucher

Vallerand

(1990). Toward a motivational model of couple happiness. Journal of Personality and Social Psychology, 59, 1021–1031. doi:10.1037/0022-3514.59.5.1021

Bodenmann

(2000). FDCT-N: Fragebogen zur erfassung des dyadischen copings als generelle tendenz [Questionnaire for the assessment of dyadic coping as a general tendency]. Fribourg, Switzerland: Institute for Family Studies and Consulting.

Brüderl

Hank

Huinink

Nauck

Neyer

F. J.

Walper

… Wilhelm

(2015). The German family panel (pairfam) (ZA5678 data file version 6.0.0). Cologne, Germany: GESIS Data Archive. doi:10.4232/pairfam.5678.6.0.0

Carothers

B. J.

Reis

H. T.

(2013). Men and women are from earth: Examining the latent structure of gender. Journal of Personality and Social Psychology, 2, 385–407. doi:10.1037/a0030437

Carver

Joyner

Udry

J. R.

(2003). National estimates of adolescent romantic relationships. In Florsheim

(Ed.), Adolescent romantic relations and sexual behaviour: Theory, research, and practical implications (pp. 23–56). Mahwah, NJ: Erlbaum.

Cheung

Lucas

R. E.

(2014). Assessing the validity of single-item life satisfaction measures: Results from three large samples. Quality of Life Research, 23, 2809–2818. doi:10.1007/s11136-014-0726-4

Connell

R. W.

Messerschmidt

J. W.

(2005). Hegemonic masculinity: Rethinking the concept. Gender & Society, 19, 829–859. doi:10.1177/0891243205278639

Cooke

L. P.

(2007). Persistent policy effects on the division of domestic tasks in reunified Germany. Journal of Marriage and Family, 69, 930–950. doi:10.1111/j.1741-3737.2007.00422.x

10.

Curran

M. A.

McDaniel

B. T.

Pollitt

A. M.

Totenhagen

C. J.

(2015). Gender, emotion work, and relationship quality: A daily diary study. Sex Roles, 73, 157–173. doi:10.1007/s11199-015-0495-8

11.

Duncombe

Marsden

(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. doi:10.1177/0038038593027002003

12.

Erickson

R. J.

(1993). Reconceptualizing family work: The effect of emotion work on perceptions of marital quality. Journal of Marriage and the Family, 55, 888–900. doi:10.2307/352770

13.

Erickson

R. J.

(2005). Why emotion work matters: Sex, gender, and the division of household labour. Journal of Marriage and Family, 67, 337–351. doi:10.1111/j.0022-2445.2005.00120.x

14.

Erickson

R. J.

(2011). Emotional carework, gender, and the division of household labour. In Garey

A. I.

Hansen

K. V.

Ehrenreich

(Eds.), At the heart of work and family: Engaging the ideas of Arlie Hochschild (pp. 61–73). New Brunswick, NJ: Rutgers University Press.

15.

Ferree

M. M.

(2010). Filling the glass: Gender perspectives on families. Journal of Marriage and Family, 72, 420–439. doi:10.1111/j.1741-3737.2010.00711.x

16.

Finkel

E. J.

Simpson

J. A.

Eastwick

P. W.

(2017). The psychology of close relationships: Fourteen core principles. Annual Review of Psychology, 68, 383–411. doi:10.1146/annurev-psych-010416-044038

17.

Funk

J. L.

Rogge

R. D.

(2007). Testing the ruler with item response theory: Increasing precision of measurement for relationship satisfaction with the couples satisfaction index. Journal of Family Psychology, 21, 572–583. doi:10.1037/0893-3200.21.4.572

18.

Furman

Buhrmester

(1985). Children’s perceptions of the personal relationships in their social networks. Developmental Psychology, 21, 1016–1024. doi:10.1037/0012-1649.21.6.1016

19.

Helgeson

V. S.

Fritz

H. L

. (1998). A theory of unmitigated communion. Personality and Social Psychology Review, 2, 173–183. doi:10.1207/s15327957pspr0203_2

20.

Hendrick

S. S.

Dicke

Hendrick

(1998). The relationship assessment scale. Journal of Social and Personal Relationships, 15, 137–142. doi:10.1177/0265407598151009

21.

Hill

P. B.

Arránz Becker

. (2004). Stabilität von partnerschaft und ehen [Stability of partnerships and marriages]. Retrieved from http://www.soziologie.rwth-aachen.de/aw/cms/website/zielgruppen/mes/forschung_und_projekte/∼tlc/stab/lang/de

22.

Hochschild

A. R.

(1979). Emotion work, feeling rules, and social structure. American Journal of Sociology, 85, 551–575. doi:10.1086/227049

23.

Holm

K. E.

Werner-Wilson

R. J.

Cook

A. S.

Berger

P. S.

(2001). The association between emotion work balance and relationship satisfaction of couples seeking therapy. The American Journal of Family Therapy, 29, 193–205. doi:10.1080/019261801750424316

24.

Holmes

(2015). Men’s emotions: Heteromasculinity, emotional reflexivity, and intimate relationships. Men and Masculinities, 18, 176–192. doi:10.1177/1097184X14557494

25.

Horne

R. M.

Johnson

M. D.

Galambos

N. L.

Krahn

H. J

. (2017). Time, money, or gender? Predictors of the division of household labour across life stages. Sex Roles. Advance online publication. doi:10.1007/s11199-017-0832 -1

26.

Huinink

Brüderl

Nauck

Walper

Castiglioni

Feldhaus

(2011). Panel analysis of intimate relationships and family dynamics (pairfam): Conceptual framework and design. Journal of Family Research, 23, 77–101.

27.

Impett

E. A.

Gable

S. L.

Peplau

(2005). Giving up and giving in: The costs and benefits of daily sacrifice in intimate relationships. Journal of Personality and Social Psychology, 89, 327–344. doi:10.1037/0022-3514.89.3.327

28.

Johnson

M. D.

Horne

R. M.

Galovan

A. M.

(2016). The developmental course of supportive dyadic coping in couples. Developmental Psychology, 52, 2031–2043. doi:10.1037/dev0000216

29.

Kenny

D. A.

Kashy

D. A.

Cook

W. L.

(2006). Dyadic data analysis. New York, NY: Guilford Press.

30.

Knee

C. R.

Patrick

Vietor

N. A.

Nanayakkara

Neighbors

(2002). Self- determination as a growth motivation in romantic relationships. Personality and Social Psychology Bulletin, 28, 609–619. doi:10.1177/0146167202288005

31.

Lerner

R. M.

Johnson

S. K.

Buckingham

M. H.

(2015). Relational developmental systems based theories and the study of children and families: Lerner and Spanier (1978) revisited. Journal of Family Theory & Review, 7, 83–104. doi:10.1111/jftr.12067

32.

Little

T. D.

(2013). Longitudinal structural equation modeling. New York, NY: The Guilford Press.

33.

Lucas

R. E.

Diener

Suh

(1996). Discriminat validity of well-being measures. Journal of Personality and Social Psychology, 71, 616–628. doi:10.1037/0022-3514.71.3.616

34.

Maslowsky

Jager

Hemken

(2015). Estimating and interpreting latent variable interactions: A tutorial for applying the latent moderated structural equations method. International Journal of Behavioral Development, 39, 87–96. doi:10.1177/0165025414552301

35.

Minnotte

K. L.

Pederson

Mannon

S. E.

(2010). The emotional terrain of parenting and marriage: Emotion work and marital satisfaction. The Social Science Journal, 47, 747–761. doi:10.1016/j.soscij.2010.07.011

36.

Muthén

L. K.

(2011). Latent variable interaction [Blog post]. Retrieved February 2, 2017, from http://www.statmodel.com/discussion/messages/11/4627.html?1300840234

37.

Muthén

L. K.

Muthén

B. O.

(1998–2012). Mplus user’s guide (7th ed.). Los Angeles, CA: Muthén & Muthén.

38.

Neff

L. A.

Karney

B. R.

(2005). Gender differences in social support: A question of skill or responsiveness? Journal of Personality and Social Psychology, 88, 79–90. doi:10.1037/0022-3514.88.1.79

39.

Noom

M. J.

Dekovic

Meeus

W. H. J.

(2001). Conceptual analysis and measurement of adolescence autonomy. Journal of Youth and Adolescence, 30, 577–595. doi:10.1023/A:1010400721676

40.

Overton

W. F.

(2013). A new paradigm for developmental science: Relationism and relational-developmental systems. Applied Developmental Science, 17, 94–107. doi:10.1080/10888691.2013.778717

41.

Pederson

D. E.

Minnotte

K. L.

Mannon

S. E.

Kiger

(2011). Exploring the relationship between types of family work and marital well-being. Sociological Spectrum, 31, 288–315. doi:10.1080/02732173.2011.557038

42.

Risman

B. J.

(2004). Gender as social structure: Theory wrestling with activism. Gender & Society, 18, 429–450. doi:10.1177/0891243204265349

43.

Ryan

R. M.

Deci

E. L.

(2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55, 68–78. doi:10.1037//0003-006X.55.1.68

44.

Strazdins

Broom

D. H.

(2004). Acts of love (and work): Gender imbalance in emotional work and women’s psychological distress. Journal of Family Issues, 25, 356–378. doi:10.1177/0192513X03257413

45.

Thomeer

M. B.

Reczek

Umberson

(2015). Gendered emotion work around physical health problems in mid- and later-life marriages. Journal of Aging Studies, 32, 12–22. doi:10.1016/j.jaging.2014.12.001

46.

Umberson

Thomeer

M. B.

Lodge

A. C.

(2015). Intimacy and emotion work in lesbian, gay, and heterosexual relationships. Journal of Marriage and Family, 77, 542–556. doi:10.1111/jomf.12178

47.

VanLaningham

Johnson

D. R.

Amato

(2001). Marital happiness, marital duration, and the U-shaped curve: Evidence from a five-wave panel study. Social Forces, 79, 1313–1341. doi:10.1353/sof.2001.0055

48.

Wilcox

W. B.

Nock

S. L.

(2006). What’s love got to do with it? Equality, equity, commitment, and women’s marital quality. Social Forces, 84, 1321–1345. doi:10.1353/sof.2006.0076

49.

Wong

Tschan

Semmer

N. K.

(2017). Effort in emotion work and well-being: The role of goal attainment. Emotion, 17, 67–77. doi:10.1037/emo0000196