Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Dec 1.
Published in final edited form as: Fam Relat. 2024 May 20;73(5):3073–3088. doi: 10.1111/fare.13043

Understanding marital stability through work–family experiences in proximal and distal contexts: Comparing United States and Japan

Chengfei Jiao 1, Joseph G Grzywacz 2
PMCID: PMC11756913  NIHMSID: NIHMS1996683  PMID: 39867815

Abstract

Objective:

Guided by Gottman’s framework of marital stability and the ecological theories, the present study aims to understand the relationships between work–family spillover and marital stability within two levels of context—the relational and social cultural contexts.

Background:

The relational context of marriage is manifested by spousal relationships—spousal support and strain, which would moderate the relationship between work–family spillover and marital stability. Identified relationships also unfold within sociocultural contexts.

Method:

This study uses data from the Midlife in the United States 2 (MIDUS 2) and Midlife in Japan (MIDJA) projects to explore these dynamics. The current study involved 500 Japanese and 1,800 American participants who were married and employed at the time of data collection.

Results:

Results from multigroup path models revealed cultural differences. The relational context—spousal support and strain—played moderating roles in the associations between work–family spillover and marital stability in both countries. However, distinct sociocultural patterns emerged as spousal strain showed a stronger association among Japanese participants, whereas spousal support was more prominent among Americans in relation to marital stability.

Conclusion:

The results support Gottman’s contention that positive and negative features of marriage are related to marital stability directly and indirectly by shaping the impacts of work–family spillover and sociocultural expectations of work and family.

Implications:

Practitioners and organizations should recognize the role of spouses in addressing the negative effects of work–family spillover in marriage, as well as incorporating clients’ or employees’ cultural backgrounds when addressing marital concerns.

Keywords: work–family spillover, spousal support and strain, marital stability, culture, contextual factors

High or increasing divorce rates are a global concern. Although divorce rates in the United States have declined by 38% since peaking at 22.6 divorces per 1,000 married women in 1979, 14 out of every 1,000 married women in the United States sought divorce in 2021 (Loo, 2023). Countries like the United Kingdom, Norway, and South Korea have experienced a tripling of divorce rates since the 1970s (Ortiz-Ospina & Roser, 2020). In Japan, divorce increased by 60% between 1980 to 2012, causing concern, although it is low by international standards (Margolis, 2023). Countries such as Mexico and Turkey also continue to observe rising divorce rates. Given the negative impact of divorce on the mental health of adults, such as depression, loneliness, and diminished self-esteem, as well as on children’s psychosocial development (American Psychological Association, n.d.), it is essential to understand the factors that might influence marital stability.

A well-fitted work and family arrangement is closely related to marital outcomes; however, the concepts of “work–family balance” or “work–family fit” remain poorly conceived (Grzywacz & Carlson, 2007). Previous literature has predominantly used “work–family conflict” as the primary operational form of a poor-fitted work–family arrangement wherein the responsibilities of one role interfere or conflict with the responsibilities of another (Greenhaus & Beutell, 1985). Results from a meta-analysis (Fellows et al., 2016) indicate a consistent negative association between work–family conflict and indicators of marital quality. However, there is also a body of research that focuses on positive experiences at the work–family interface, sometimes referred to as “facilitation,” “enrichment,” and “positive spillover” (e.g., Grzywacz & Butler, 2005). Previous literature has suggested that these indicators of well-fitted work-family arrangement have also been linked with marital outcomes, such as satisfaction and quality (e.g., Heller & Watson, 2005; Liu et al., 2016). No previous research on marital outcomes, however, has captured both positive and negative aspects of a well-fitted work and family arrangement despite evidence of their distinctness (e.g., Grzywacz & Marks, 2000).

Furthermore, a meta-analysis by Fellows et al. (2016), using 33 papers published between 1986 to 2014 and including 49 samples from Europe, North America, and Asia, highlighted the potential moderating role of contextual factors in shaping the associations of positive and negative experiences at the work–family interface with marital outcomes. Importantly, myriad contextual factors range from proximal features of the marriage itself to more distal features of community, economy, and sociocultural norms across geographical regions. To gain a deeper understanding of the mechanisms underlying the associations between indicators of work–family fit and marital experiences, it is crucial to consider the impacts of these contextual factors.

Therefore, this study aims to determine how positive and negative indicators of work–family fit shape marital stability within the proximal context of the marriage itself and the distal sociocultural context. Following Gottman’s (1993) framework of marital stability, we incorporate the relational context of marriage and posit that the levels of positivity and negativity within the marriage (i.e., support and strain) are directly related to marital stability and will modify the impact of work–family experiences, namely negative and positive spillover from work to family. Second, based on the ecological theories (e.g., Ogbu, 1981), we aim to acknowledge the sociocultural context, recognizing that cultural contexts can create substantial variation in the meaning of “work” and “family,” as well as expectations of “good marriages” from one country to the next (Berry, 2022; Hong et al., 2021).

LITERATURE REVIEW

Relational context

Gottman (1993) posited that two features of the marital relationship could be used to determine if a marriage is developing toward marital stability or dissolution: the relative positivity and negativity in relational interactions. Gottman (1993) further suggested that the experience of instability in marriage could be exaggerated by situational realities, such as conflicts and demands and the attributions assigned to these behaviors by the partner. Support and strain, two features of every social relationship (Due et al., 1999; House et al., 1988), could characterize the levels of positivity and negativity in a marriage. Support refers to the availability and potential receipt of useful resources from a social partner (e.g., a spouse) for meeting basic or higher level needs (House et al., 1988). Within the context of marriage, spousal support manifests in instrumental assistance (e.g., helping with chores) or emotional connection. Strain is not the opposite of support, but rather the emotional experience that may accompany fulfilling the demands of the spouse, a partner’s critique, or the inevitability of disagreements among individuals in a long-term relationship (Antonucci & Jackson, 1987; Hung et al., 2019).

Following Gottman’s theorizing, spousal support and strain can be considered as elements of the relational contexts because each can wax and wane across the course of a marriage. They can amplify or buffer the effects of events occurring in the marital union. Specifically, the harmful effects of negative work–family spillover for marriages can be neutralized by spouse support (i.e., positivity), while spouse support (i.e., positivity) could exaggerate the beneficial effects of positive work–family spillover for marriage (Leatham & Duck, 1990; Repetti, 1989; Walen & Lachman, 2000). Conversely, the harmful potential consequences of negative work–family spillover for marriage can be exaggerated under conditions of high spousal strain (i.e., negativity), and spousal strain (i.e., negativity) would attenuate the benefits of positive work–family spillover for marriage (Repetti, 1989; van Steenbergen et al., 2014).

Sociocultural context

Eco-cultural models of human behavior (Mistry & Wu, 2010; Ogbu, 1981) ask researchers to acknowledge and accommodate the ever-present yet often overlooked sociocultural context of individual and social behavior. In the marital context, the understanding of “good marriages” and “appropriate behavior” by spouses varies across Western and Eastern cultures (Taniguchi & Kaufman, 2014). In Western societies, relational features like romance, emotional support, and attentiveness to partners’ personal achievement are often seen as fundamental aspects of marriage (Kamo, 1993). Conversely, in Eastern Asian countries, emotive aspects of romance and support are considered less prioritized in marriage than instrumental support for family survival (Taniguchi & Kaufman, 2014). Moreover, previous literature suggests that Eastern cultures tend to exhibit lower levels of emotional involvement and open communication in romantic relationships compared to Western cultures while emphasizing the importance of relational harmony (Kamo, 1993; Li et al., 2016; Yen & Yang, 2011). Given these differences, the immediate marital context appears to vary across Eastern and Western individuals, thereby creating the possibility that support and strain in the marriage may have distinct implications for marital stability and correspondingly distinct implications for modifying the relation of work–family experiences with marital stability.

From a work–family perspective, cultural variations are anticipated in the experiences of work–family spillover between Western and Eastern cultures. Work is perceived as a means of achieving personal accomplishment distinct from one’s familial needs in Western societies, such as the United States (Lu & Gilmour, 2004). In contrast, work is regarded as a resource for supporting family welfare in East Asian cultures, leading individuals to experience less distress prioritizing work responsibilities in a family context (Lu & Gilmour, 2004; Lu et al., 2008). Consequently, individuals in Western societies might encounter greater tension between work and family demands due to their tendency to view work and family as separate aspects of life (Greenhaus & Beutell, 1985). However, most work–family studies have focused on Western societies, such as the United States. Limited research has been conducted in other regions, including East Asia, despite the considerable distress experienced by Asian employees due to inflexible working hours (Le et al., 2020; Ling & Poweli, 2001; Lu et al., 2008). Therefore, it is crucial to investigate the work–family spillover and marital outcomes from a cross-cultural perspective.

CURRENT STUDY

Based on theory and previous empirical studies, it is crucial to study contextual factors when investigating the putative influence of a well-fitted work and family arrangement on marital stability. The first aim of this study was to understand the association between work–family spillover and marital stability within the relational context of spousal support and strain; therefore, we hypothesized the following:

H1.

Spousal support strengthens the relationship of greater positive work–family spillover with better marital stability while buffering the relationship of greater negative work–family spillover with poorer marital stability.

H2.

Spousal strain weakens the relationship of greater positive work–family spillover with better marital stability but intensifies the relationship of greater negative work–family spillover and poorer marital stability.

Furthermore, because our paper focuses on examining the moderating roles of relational and sociocultural contexts on the relationship between work–family spillover and marital stability, we need to test the direct relationships between the independent and dependent variables:

H3.

Greater positive and less negative work–family spillover are correlated with greater marital stability.

H4.

Greater spousal support and less spousal strain are correlated with marital stability.

This study’s second aim was to investigate the influence of sociocultural context on the above hypothesized associations by comparing the United States and Japan. These two countries were selected due to their representation of distinctive expectations of marriage. The United States emphasizes emotional support and self-fulfillment within marriage (Kamo, 1993), whereas Japan values cooperation and strives to avoid conflicts in intimate relationships (Taniguchi & Kaufman, 2014). As a result, we anticipated sociocultural variations in the significance of spousal support and strain in shaping work–family and marital stability. However, given the limited empirical research on sociocultural variations in the direct relationship between work–family spillover and marital stability, as well as the moderating roles of spousal support and strain, any potential differences between these two countries are considered exploratory in nature.

METHOD

Sample and data

This study employed data from the National Survey of Midlife in the United States 2 (MIDUS 2), collected in 2004, and Midlife in Japan (MIDJA), collected in 2008. MIDUS 2 and MIDJA are part of the longitudinal follow-up of MIDUS 1, a national survey conducted by the Mac-Arthur Midlife Research Network from 1995 to 1996. The purpose of MIDUS 1 was to investigate how behavioral, social, and psychological factors explain the age differences in physical and mental health. Participants were invited to answer self-administered surveys and telephone interviews, and over 7,000 U.S. adults from 48 states (aged 25–74 years) participated. MIDUS 2 employed the same protocols, and participants responded to questions like those in MIDUS 1. MIDJA occurred during the first wave of longitudinal follow-up of MIDUS 1, involving over 1,000 adults from Tokyo, Japan (aged 30–79 years). Both projects utilized probability samples, including wide age and socioeconomic diversity. MIDUS data collection was reviewed and approved by the Education and Social/Behavioral Sciences and the Health Sciences Institutional Review Boards at the University of Wisconsin–Madison. For the present study, we utilized data obtained from self-administered surveys in MIDUS 2 and MIDJA, thereby focusing on cross-sectional data collected at the individual level.

The core sample of MIDUS 2 and MIDJA consisted of 3,487 and 1,027 participants, respectively. For the current study, only participants who were married and employed at the time of data collection were selected, resulting in a final sample of 500 Japanese and 1,800 American respondents. The average years of marriage were 24.16 years (SD = 9.67 years) and 13.78 years (SD = 9.67 years) for Japanese and American participants, respectively. The majority of the participants were male (60.6% for Japanese, 53.1% for American) and reported having just or more than enough income to meet daily needs (59.5% for Japanese, 84.8% for American). Most American participants reported having at least or above college qualifications (70.3%), whereas most Japanese participants did not (57.4%). Descriptive information is presented in Table 1.

TABLE 1.

Descriptive information and correlation among key variables.

Variable PWFS NWFS Spousal support Spousal strain Gender Years of marriage Education Finance Marital stability
PWFS — −.01 .07** .01 .03 .01 .13** .10** .03
NWFS .13** — −.14** .24** −.02 −.04 .03 −.20** −.17**
Spousal support .28** −.06 — —.65** −.14** −.01 —.02 .12** .70**
Spousal strain −.03 .22** −.46** — .06** .08 .04 −.14** −.62**
Gender .08 .08 −.11* .05 — −.08 .00 −.05* −.06
Years of marriage −.08 −.35** −.09* −.11* .05 — .04 .04 .11*
Education .03 .08 .12** .01 −.38** −.21** — .11* —.01
Finance .06 −.05 .17** −.09** .01 −.04 .09 — .15**
Marital stability .09 −.23** .48** −.57** −.12** .17** .01 .09 —
Japan (N = 500)
Mean (SD) or % 10.44 (3.29) 9.12 (3.09) 2.88 (.72) 2.23 (.57) 39.4% 24.16 (9.67) 42.6% 59.5% .06 (1.70)
United States (N = 1,800)
Mean (SD) or % 11.79 (2.74) 10.24 (2.68) 3.62 (.54) 2.17 (.60) 46.9% 13.78 (9.67) 70.3% 84.8% .01 (1.80)
Open in a new tab

Note: NWFS = negative work–family spillover; PWFS = positive work–family spillover. Gender: 0 = male, 1 = female; Education: 0 = below college qualification, 1 = at least or above college qualification; Finance: 0 = having just enough income to meet daily needs, 1 = just or more than enough income to meet daily needs. Correlations above the diagonal refer to the American sample, whereas below the diagonal refer to the Japanese sample. Marital stability was operationalized as the sum of standardized item scores.

*

p < .05.

**

p < .01.

Measures

The same instruments were used in MIDUS 2 and MIDJA.

Positive and negative work–family spillover

Each aspect of work–family spillover, positive and negative, was assessed using four items (e.g., Grzywacz & Marks, 2000). Sample items included “The things you do at work make you a more interesting person at home” (i.e., positive work–family spillover; α = .81 for Japan; α = .68 for the United States) and “Your job reduces the effort you can give to activities at home” (i.e., negative work–family spillover; α = .81 for the Japanese; α = .80 for the United States). All items were rated on a Likert scale from 1 (all of the time) to 5 (never). Previous studies have reported acceptable reliabilities for the positive and negative work–family spillover scales (α > .70; e.g., Grzywacz & Marks, 2000). Items were first reverse-coded and summed so that high scores reflect a higher level on the respective scale.

Spousal support and strain

Each aspect of the relational context of marriage—support and strain received from a spouse— was measured using six items (e.g., Grzywacz & Marks, 2000). A sample item for spousal support was “How much does your spouse or partner really care about you?” (α = .92 for Japan; α = .90 for the United States). A sample item for spousal strain was “How often does your spouse or partner make too many demands on you?” (α = .87 for the Japanese; α = .87 for the United States). Participants responded to all items on a Likert scale from 1 (often) to 4 (never). Previous studies have reported high reliabilities for the spousal support and strain scales (α > .75, e.g., Schuster et al., 1990; Walen & Lachman, 2000). Items were reverse-coded and then averaged so that high scores reflect a higher level on the respective scale.

Marital stability

Two items developed by Booth et al. (1983) were employed to assess participants’ perceptions of marital stability. The first item asked how often the participants have thought that their relationship was in trouble during the past year (1 = never to 5 = all the time). The second item asked the participants to rate the likelihood of eventually separating from their partners (1 = very likely to 4 = not likely at all). These items were reverse-coded to indicate lower levels of marital risk and higher levels of marital stability. These two items had high item-to-total correlations in the original study, ranging from .63 to .75 (Booth et al., 1983). Given the different scales used for each item, the scores were standardized and then summed, with higher scores indicating greater marital stability.

Covariates

Covariates include participants’ gender (1 = male, 2 = female), education (1 = below college qualifications, 2 = at least or above college qualifications), and current financial situation (0 = not enough to meet daily needs, 1 = just enough or more than enough to meet daily needs). Additionally, length of marriage (in years) was considered a covariate given that marital stability is the outcome of interest.

Statistical analysis

Measurement equivalence

As cross-national data were utilized in this study, it is essential to assess measurement equivalence to ensure that key variables were understood similarly across countries (Ariely & Davidov, 2012). Measurement equivalent tests were conducted by using Mplus (Version 8) for measures of positive and negative work–family spillover and spousal support and strain. Item-level descriptive and confirmatory factor analysis was conducted in each sample for the mentioned measures. Configural, metric, and scalar variances were assessed, with at least achieving configural invariance required to proceed with the main analyses. To evaluate variances on different levels, we applied the cut-off points proposed by Chen (2007), whereby measurements show invariance across countries when the decrease in the comparative fit index does not exceed .01. However, equivalence testing for marital stability was not possible because the measure consisted of only two items, resulting in an unidentified model for the metric equivalence test.

Main analyses

To test H1 and H2, multisample path models were conducted using Mplus (Version 8) to examine the moderation roles of spousal support and strain. Two path models were tested to test H1 and H2, with spousal support and spousal strain as distinct moderators on the association of positive and negative work–family spillover with marital stability, while adjusting for the effects of covariates. To avoid potential collinearity, each model’s independent variables and moderator were centered (Smith & Sasaki, 1979). Interaction terms were created by multiplying the centered variables (Bedeian & Mossholder, 1994). Thus, two interaction terms were created for each model: “positive work–family spillover * spousal support” and “negative work–family spillover * spousal support” to test H1 in Model 1, and “positive work–family * spousal strain” and “negative work–family spillover * spousal strain” to test H2 in Model 2 because prior studies indicated the potential correlation between spousal support and strain (Abbey et al., 1985), Model 1 treated spousal strain as a covariate when spousal support was considered the moderator and vice versa in Model 2.

To achieve the study’s second aim, we conducted multigroup path models using Mplus (Version 8) to test group differences between the Japanese and American respondents. The freely estimated model was assessed as the baseline model, followed by the imposition of equality constraints on all paths across groups (Aiken et al., 1991). A significantly poorer fit suggests the presence of overall group differences. We revised the constrained model to identify specific paths that differ significantly by removing one equality constraint at a time based on the modification indices. The revised model was subjected to a nested chi-square test in Mplus, comparing it with the previous model, where a significant chi-square change (Δχ2) indicated a cultural difference in the specific path. This process continued until no further equality constraints could be removed without a significant Δχ2. Furthermore, the distributions were not symmetric by checking the histograms of the independent, dependent, and moderating variables. Thus, MLR was used as the Mplus estimator because it adjusts the mean of chi-square for nonnormality with missing data (Asparouhov & Muthén, 2006). Descriptive and correlational analyses by groups were conducted using IBM SPSS (Version 27.0) before the main analyses.

RESULTS

Measurement equivalence testing

Before conducting equivalence tests, the global fit for key variables was tested in the two countries separately by correlating the error variances of specific items. Item-level descriptive statistics are summarized (see Table 1 in the supplemental materials). The results of confirmatory factor analysis showed that all measures had excellent to reasonable model fits in the two samples (see Table 2 in the supplemental materials). The results of measurement equivalence tests indicated that all key measures achieved at least configural invariance with acceptable comparative fit index/Tucker–Lewis index and satisfactory root-mean-square error of approximation (see Supplemental Table 3), indicating that the patterns of factor loadings remained the same across American and Japanese participants. Additionally, measures for negative work–family spillover and spousal strain achieved metric invariance, meaning the item loadings of each item on the corresponding measure were equivalent across the two groups (Chen, 2007; see Supplemental Table 3). Therefore, we proceeded with the main analyses.

Descriptive statistics

Table 1 summarizes descriptive statistics and correlations among the key variables. We conducted independent sample t tests in SPSS to compare the means for positive work–family spillover, negative work–family spillover, spousal support, spousal strain, and marital stability between American and Japanese participants. There was no observed difference in marital stability, but significant differences in positive work–family spillover, t = − 8.42 (704.50), p < .01, d = .45; negative work–family spillover, t = − 7.31 (723.71), p < .01, d = .39; spousal support, t = −21.31 (658.03), p < .01, d = .45; and spousal strain, t = 1.86 (2283), p < .03, d = .27, across the American and Japanese samples. The American sample reported greater positive work–family spillover, negative work–family spillover, and spousal support but less spousal strain than the Japanese sample.

Greater negative work–family spillover was associated with lower marital stability in both samples (r = −.23, p < .01 for the Japanese; r = −.17, p < .01 for the Americans). However, no associations between positive work–family spillover and marital stability were significant. Furthermore, in both samples, greater spousal support was associated with higher marital stability (r = .48, p < .01 for the Japanese; r = .70, p < .01 for the Americans), whereas greater spousal strain was associated with lower marital stability (r = −.57, p < .01 for the Japanese; r = −.62, p < .01 for the Americans).

Multiple-group path analysis

Model 1

We first tested H1—whether spousal support strengthens the relationship of greater positive work–family spillover with better marital stability while buffering the relationship of greater negative work–family spillover with poorer marital stability. The results of path models are presented in Figure 1, indicating that the interaction term of spousal support and negative work–family spillover had a small but significant association with marital stability (β = .06, SE = .03, p = .048 for the Japanese; β = .03, SE = .02, p = .045 for the Americans). The positive work–family spillover and spousal support interaction term was nonsignificant. Thus, H1 was partially supported.

FIGURE 1.

FIGURE 1

Open in a new tab

The moderating effect of spousal support on the association between work–family spillover and marital risk.

Note. N = 500 for Japan; N = 1,800 for the United States. Gender: 0 = male, 1 = female. Standardized coefficients before “/” are for the Japanese group, and those after “/” are for the American group. The effects of gender, age, years of marriage, education, and financial situation were controlled, and only significant covariates are shown in the diagram.

*p < .05. **p < .01.

The observed interaction between spousal support and negative work–family spillover was explored by dividing participants into two groups—low spousal support (below the mean) and high spousal support (above the mean). Among Japanese participants, the relationship between negative work–family spillover and marital stability was stronger for those in the low spousal support group (β = − .14, p = .038) compared to the high spousal support group (β = − .05, p = .362). A marginal difference was also observed among American participants reporting low spousal support (β = −.09, p = .015) and those reporting high support (β = −.08, p = .011).

These results suggest that spousal support buffered the association between negative work–family spillover and marital stability among both the Japanese and Americans.

Regarding the direct relationship between work–family spillover and marital stability, the results showed that negative work–family spillover had a low and marginally significant association with marital stability for both the Japanese (β = − .04, SE = .02, p = .050) and the Americans (β = −.03, SE = .02, p = .050), whereas positive work–family spillover was not. Therefore, H3 was partially supported. As for covariates, the length of marriage was positively related to marital stability for both groups (β = .16, SE = .03, p = .000 for the Japanese; β = .11, SE = .02, p = .000 for the Americans), with no group differences found. Education and financial situation were not significant predictors of marital stability for either the Japanese or the Americans.

The second aim was to investigate the influence of sociocultural context on the hypothesized associations by comparing the United States and Japan. The freely estimated model was fully saturated, thus perfectly fitting the data. After constraining the structural parameters across the Japanese and American groups, the overall model fit was significantly worse than the fully unconstrained model (Δχ2 = 62.133, Δdf = 10; p = .000), suggesting that at least one path differed between the two groups. The modification indices suggested releasing the equality constraints from spousal support to marital stability, gender to marital quality, and then spousal strain to marital stability one at a time. Each revised model had a significant chi-square change, indicating cultural differences in these paths. Table 4 in the supplemental materials shows the results of the chi-square change by removing each constraint.

Cultural difference was found regarding the direct path from spousal support to marital stability between Japanese and American participants (see Table 4 in the supplemental materials, Revised Model a). The association was stronger for Americans (β = .51, SE = .03, p = .000) than for Japanese participants (β = .33, SE = .04, p = .000). In contrast, the association between spousal strain and marital stability was stronger for Japanese participants (β = − .39, SE = .04, p =. 000) than for Americans (β = −.29, SE = .02, p = .000, see Table 4 in the supplemental materials, Revised Model c). Thus, H4 was supported. In addition, gender was significantly associated with marital stability for both groups (β = .08, SE = .04, p = .035 for the Japanese; β = .04, SE = .02, p = .018 for the Americans) with a significant group difference, indicating that women tended to report higher marital stability in America but low marital stability in Japan (see Table 4 in the supplemental materials, Revised Model b).

Model 2

Next we tested H2—whether spousal strain weakens the relationship of greater positive work–family spillover with better marital stability but intensifies the relationship of greater negative work–family spillover and poorer marital stability. The path model (see Figure 2) showed that the interaction term of negative work–family spillover and spousal strain was negatively related to marital stability (β = −.04, SE = .02, p = .021 for both the Japanese and Americans). The positive work–family spillover and spousal strain interaction term was nonsignificant. Therefore, H2 was partially supported.

FIGURE 2.

FIGURE 2

Open in a new tab

The moderating effect of spousal support on the association between work–family spillover and marital risk.

Note: N = 500 for Japan; N = 1,800 for the United States. Gender: 0 = male, 1 = female. Standardized coefficients before “/” are for the Japanese group, and those after “/” are for the American group. The effects of gender, age, years of marriage, education, and financial situation were controlled, and only significant covariates are shown in the diagram.

*p < .05. **p < .01.

To examine the interaction between spousal strain and negative work–family spillover, participants were divided into two groups—low spousal strain (below the mean) and high spousal strain (above the mean). Among Japanese participants, the relationship between negative work–family spillover and marital stability was only significant for those in the high spousal strain group (β = −.19, p = .004) but not for the low spousal strain group (β = −.06, p = .323). Similar findings were observed among American participants, although the difference in associations were marginal (β = −.05, p = .114 for the high strain group; β = −.01, p = .638 for the low strain group). These results suggested that spousal strain modified the association between negative spillover and marital stability in both groups.

Moreover, results showed that negative work–family spillover was marginally associated with marital stability for the Japanese (β = −.04, SE = .02, p = .050) and the Americans (β = −.03, SE = .02, p = .050), but positive work–family spillover was not. Therefore, H3 was partially supported. In addition, among the covariates, the length of marriage was correlated with martial stability (β = .17, SE = .03, p = .000 for the Japanese; β = .12, SE = .02, p = .001 for the Americans), but education and financial situation were not.

Potential group differences were also tested. The fully unconstrained model was fully saturated, thus perfectly fit the data. After constraining the structural parameters to be equal across the Japanese and American groups, the overall model fit was significantly worse than the fully unconstrained model (Δχ2 = 80.287, Δdf = 10; p = .000). This suggests that the paths (as a whole) had significant differences across the two groups. Similar to Model 1, the modification indices suggested releasing equality constraints from spousal support to marital stability, gender to marital stability, and spousal strain to marital stability one at a time. Each revised model had a significant chi-square change indicating cultural differences in these paths. Table 5 in the supplemental materials shows the results of chi-square change by removing each constraint.

The same group-difference patterns are found as in Model 1. For the direct relationship between spousal support and marital stability (see Table 5 in the supplemental materials, Revised Model a), the association was stronger among Americans (β = .51, SE = .03, p = .000) compared with Japanese participants (β = .31, SE = .04, p = .000). Conversely, the link between spousal strain and marital stability was stronger for the Japanese (β = − .41, SE = .04, p = .000) than for the Americans (β = −.28, SE = .02, p = .000; see Table 5 in the supplemental materials, Revised Model c). Thus, H4 was supported. Additionally, group differences were found regarding the role of gender (β = −.08, SE = .04, p = .028 for Japanese; β = .04, SE = .02, p = .023 for Americans; see Table 5 in the supplemental materials, Revised Model b).

DISCUSSION

Drawing on Gottman’s (1993) theory and ecocultural theories (Mistry & Wu, 2010; Ogbu, 1981), we examined how proximal and distal contexts shape association between work–family experiences and marital stability. Results from the analysis of cross-national data make several contributions to the literature. Specifically, we hypothesized that within the relational context, spousal support and strain could function as indicators of positivity and negativity in marriage, moderating the relationship between work–family spillover and marital stability. To explore cultural differences, we compared the United States and Japan. The United States represents a marital belief emphasizing emotional support and personal achievement, whereas Japan represents a traditional East Asian family ideology that promotes cooperation and conflict avoidance in intimate relationships.

The main contribution of this research is the evidence that the immediate context of the marriage may modify how experiences in daily life contribute to marital success. Specifically, we found that spousal support, as the relative positivity in marital interactions, buffered the association between negative work–family spillover and marital stability for Japanese and American participants. In contrast, spousal strain, as the relative negativity in marital interactions, exacerbated the negative association between negative work–family spillover and marital stability in both groups. The results are consistent with Gottman’s (1993) thinking that relational positivity and negativity serve as an important context for marital stability. The results also extend Fellows et al.’ (2016) findings by indicating that one source of heterogeneity in the correlation of work–family conflict (or a manifestation of a poorly fitted work and family arrangement) with marital outcomes is attributed to the marital context itself.

The greater sensitivity of negative work–family spillover (vis-à-vis positive work–family spillover) to modifying effects of the relational context is noteworthy. It is plausible that individuals tend to process negative life events or information more than positive ones, including negative emotions, parenting, feedback, and so forth (Baumeister et al., 2001). Thus, when negative work–family spillover (as a negative event) occurs in their lives, individuals are more likely to be sensitive to the effects of their relational context. For instance, they might perceive support from their spouse as more beneficial in their marriage when experiencing negative work–family spillover. Likewise, spousal strain intensifies the effects of negative work–family spillover. These results highlight the possibility that it is not the absolute amount of positivity or negativity in the marital relationship that is important but the positivity-to-negativity ratio that is meaningful (Gottman, 1993).

Furthermore, as suggested by the ecocultural theories (Mistry & Wu, 2010; Ogbu, 1981), notable differences were observed between Japanese and American sociocultural contexts. Specifically, spousal strain had a stronger relationship with marital stability than spousal support among the Japanese, whereas spousal support was a stronger factor concerning marital stability than spousal strain among Americans. This finding is consistent with previous research indicating that harmony and cooperation are prioritized in Japanese marriages (e.g., Taniguchi & Kaufman, 2014) and that American couples prioritize emotional connection (e.g., Kamo, 1993). Our finding further emphasizes that spousal roles are perceived differently across cultures. One interpretation of this result is that in Japan, where cooperation and harmony are valued in marriage, individuals might feel obligated to meet the demands of their spouse and consider the spouse’s problems as their own (e.g., Li et al., 2016; Taniguchi & Kaufman, 2014), thus prioritizing “strain” over support from their spouse. In contrast, in the United States, where romance and self-fulfillment are valued, marriage entails more self-serving purposes, such as seeking emotional support and affection from one’s spouse (Kamo, 1993). In this regard, the current study further confirmed these cultural beliefs, indicating that spousal strain is more relevant to the Japanese culture, whereas spousal support is more relevant to the American culture.

Nevertheless, there are also many similarities in the observed associations in the Japanese and American data. Our results indicated similarities in the associations between positive and negative work–family spillover and marital stability in both countries, particularly when considering the relational context. This finding contradicts previous research (e.g., Hassan et al., 2010) arguing that individuals in Western cultures may experience more difficulties in balancing work and family roles than those in East Asian societies (e.g., Japanese). Furthermore, our findings suggest that the moderation roles of spousal support and strain on the associations between work–family spillover and marital stability were equally significant for both American and Japanese participants. This contrasts with the belief that emotional support and spousal interactions are less valued in marriage in East Asian cultures compared to Western cultures (Kamo, 1993; Yen & Yang, 2011). This result may be attributed to the cultural emphasis on relational harmony in Japanese culture, despite relatively lower emphasis on personal achievement and happiness (Kamo, 1993). Consequently, couples in Japan may also be influenced by the support and strain they receive from their spouses. Thus, the present study reinforces the notion that spousal relationships are valuable aspects of marriage in both cultures, and that we should consider work–family spillover in relation to marital outcomes as a global matter rather than through a cultural lens. More nuanced cultural research is needed to enhance our understanding of the interaction between work–family experiences and spousal relationships in marriage.

Although not the primary focus of our study, cultural differences were observed in the association between gender and marital stability among the participants. Female Japanese participants reported lower levels of marital stability, whereas female American participants reported higher levels. This could be attributed to a stronger adherence to traditional gender roles in Japan, limiting the power of wives in marriages and increasing the vulnerabilities. This, in turn, might lead to a sense that their marriage is at risk, especially when they are employed and face the pressure of fulfilling domestic responsibilities alongside work (e.g., Ono, 2009). These findings highlight the need for future research on the role of gender in work–family spillover, spousal relationships, and marriages, particularly within the Japanese context.

Study limitations

The contributions of this study must be interpreted considering its limitations. The MIDUS 2 and MIDJA data sets only contain cross-sectional data, which means no causality can be inferred from the current results. Previous research has also suggested the possibility of a crossover effect between spouses, where the work–family spillover experienced by one spouse influences the spillover experienced by the other spouse, as well as the support and strain received, ultimately impacting their overall marital experiences (Lavner & Clark, 2017; van Steenbergen et al., 2014). Thus, future research should incorporate dyadic data to examine the relationship between work–family spillover and marital outcomes within the spousal support and strain framework. Additionally, the data were collected in the 2000s, and there may have been marital and work values shift in both countries over the past decade. Lastly, our study could not account for the influence of social constraints within each society on marital stability. For instance, the impact of gender stereotypes on the experiences of Japanese women in marriage may differ from that of American women (Nemoto, 2008; Ono, 2009). These limitations suggest that the current findings should be interpreted with caution.

IMPLICATIONS

Despite the above limitations, this study made several valuable contributions to previous literature. First, we demonstrated that the immediate relational context—spousal support and strain levels—requires attention when studying the relationship between work–family spillover and marital experiences. Second, we demonstrated meaningful sociocultural differences and similarities in how work–family experiences are associated with marital stability across Japan and the United States. These results are valuable to practice-oriented professionals looking for strategies to help working families better “balance” work and family.

Specifically, our results suggest that corporate or workplace initiatives that minimize negative work-to-family spillover, such as family-supportive supervisory practices, may have universal value in Eastern and Western cultural contexts. However, such practices may have an even greater impact in distinct cultural contexts if they aligned with cultural expectations in marriage; that is, emphasize opportunities for nurturing marital harmony and minimizing opportunities for marital strain in Japan, as opposed to promoting spousal support in the United States. These results, therefore, offer a way to culturally tailor common strategies for promoting work–life balance, especially in global organizations. However, the results also offer some potential insight into family life educators working independently of organizations by suggesting culturally distinctive strategies for minimizing the threat of negative work-to-family spillover for marriages.

CONCLUSION

In conclusion, our study found that when considering the relational context, the association between work–family spillover and marital stability diminishes among American and Japanese participants. Using population-level data adds to the generalizability of our findings across the American and Japanese populations. We identified both sociocultural differences and similarities in our models of marital stability. Specifically, the role of spousal strain was accentuated in the Japanese context, which values cooperation and avoids conflicts in intimate relationships. By contrast, among Americans, who value emotional support and self-fulfillment, the salience of spousal support was accentuated. Despite these differences, which we attribute to possible sociocultural shaping of marital expectations and meaning-making around everyday work and family responsibilities, our findings also suggested that spousal relationships are valuable assets in marriage across American and Japanese contexts. Overall, our study findings emphasize both proximal relational and distal sociocultural contexts shape marital stability.

Supplementary Material

Supplement, Table 1

Footnotes

SUPPORTING INFORMATION

Additional supporting information can be found online in the Supporting Information section at the end of this article.

REFERENCES

  1. Abbey A, Abramis DJ, & Caplan RD (1985). Effects of different sources of social support and social conflict on emotional well-being. Basic and Applied Social Psychology, 6(2), 111–129. 10.1207/s15324834basp0602_2 [DOI] [Google Scholar]
  2. Aiken LS, West SG, & Reno RR (1991). Multiple regression: Testing and interpreting interactions. Sage. [Google Scholar]
  3. American Psychological Association. (n.d.). Divorce and child custody. https://www.apa.org/topics/divorce-child-custody
  4. Antonucci TC, & Jackson JS (1987). Social support, interpersonal efficacy, and health: A life course perspective. In Carstensen LL & Edlstein BA(Eds.), Handbook of clinical gerontology (pp. 291–311). Pergamon Press. [Google Scholar]
  5. Ariely G, & Davidov E. (2012). Assessment of measurement equivalence with cross-national and longitudinal surveys in political science. European Political Science, 11, 363–377. 10.1057/eps.2011.11 [DOI] [Google Scholar]
  6. Asparouhov T, & Muthén B. (2006). Robust chi square difference testing with mean and variance adjusted test statistics (Mplus Web Notes No. 10). https://www.statmodel.com/download/webnotes/webnote10.pdf
  7. Baumeister RF, Bratslavsky E, Finkenauer C, & Vohs KD (2001). Bad is stronger than good. Review of General Psychology, 5(4), 323–370. 10.1037/1089-2680.5.4.323 [DOI] [Google Scholar]
  8. Bedeian AG, & Mossholder KM (1994). Simple question, not so simple answer: Interpreting interaction terms in moderated multiple regression. Journal of Management, 20(1), 159–165. 10.1177/014920639402000108 [DOI] [Google Scholar]
  9. Berry JW (2022). The forgotten field: Contexts for cross-cultural psychology. Journal of Cross-Cultural Psychology, 53(7–8), 993–1009. 10.1177/00220221221093810 [DOI] [Google Scholar]
  10. Booth A, Johnson D, & Edwards JN (1983). Measuring marital instability. Journal of Marriage and Family, 45(2), 387–394. 10.2307/351516 [DOI] [Google Scholar]
  11. Chen FF (2007). Sensitivity of goodness of fit indexes to lack of measurement invariance. Structural Equation Modeling, 14(3), 464–504. 10.1080/10705510701301834 [DOI] [Google Scholar]
  12. Due P, Holstein B, Lund R, Modvig J, & Avlund K. (1999). Social relations: Network, support and relational strain. Social Science & Medicine, 48(5), 661–673. 10.1016/S0277-9536(98)00381-5 [DOI] [PubMed] [Google Scholar]
  13. Fellows KJ, Chiu H-Y, Hill EJ, & Hawkins AJ (2016). Work–family conflict and couple relationship quality: A meta-analytic study. Journal of Family and Economic Issues, 37(4), 509–518. 10.1007/s10834-015-9450-7 [DOI] [Google Scholar]
  14. Gottman JM (1993). A theory of marital dissolution and stability. Journal of Family Psychology, 7(1), 57–75. 10.1037/0893-3200.7.1.57 [DOI] [Google Scholar]
  15. Greenhaus JH, & Beutell NJ (1985). Sources of conflict between work and family roles. Academy of Management Review, 10(1), 76–88. 10.2307/258214 [DOI] [Google Scholar]
  16. Grzywacz JG, & Butler AB (2005). The impact of job characteristics on work-to-family facilitation: Testing a theory and distinguishing a construct. Journal of Occupational Health Psychology, 10(2), 97–109. 10.1037/1076-8998.10.2.97 [DOI] [PubMed] [Google Scholar]
  17. Grzywacz JG, & Carlson DS (2007). Conceptualizing work-family balance: Implications for practice and research. Advances in Developing Human Resources, 9(4), 455–471. 10.1177/1523422307305487 [DOI] [Google Scholar]
  18. Grzywacz JG, & Marks NF (2000). Reconceptualizing the work–family interface: An ecological perspective on the correlates of positive and negative spillover between work and family. Journal of Occupational Health Psychology, 5(1), 111–126. 10.1037/1076-8998.5.1.111 [DOI] [PubMed] [Google Scholar]
  19. Hassan Z, Dollard MF, & Winefield AH (2010). Work-family conflict in East vs Western countries. Cross Cultural Management, 17(1), 30–49. 10.1108/13527601011016899 [DOI] [Google Scholar]
  20. Heller D, & Watson D. (2005). The dynamic spillover of satisfaction between work and marriage: The role of time and mood. Journal of Applied Psychology, 90(6), 1273–1279. 10.1037/0021-9010.90.6.1273 [DOI] [PubMed] [Google Scholar]
  21. Hong P, Abdelmoneium AO, Badahdah AM, & Grzywacz JG (2021). A conceptual model of work-family balance among Qatari adults. Journal of Cross-Cultural Psychology, 52(4), 406–423. 10.1177/00220221211011251 [DOI] [Google Scholar]
  22. House JS, Umberson D, & Landis KR (1988). Structures and processes of social support. Annual Review of Sociology, 14, 293–318. 10.1146/annurev.so.14.080188.001453 [DOI] [Google Scholar]
  23. Hung M, Voss MW, Bounsanga J, Graff T, & Birmingham WC (2019). Assessing spousal support and health in an aging population: Support and strain amidst changing social dynamics. Social Work in Health Care, 58(4), 345–367. 10.1080/00981389.2019.1569577 [DOI] [PubMed] [Google Scholar]
  24. Kamo Y. (1993). Determinants of marital satisfaction: A comparison of the United States and Japan. Journal of Social and Personal Relationships, 10(4), 551–568. 10.1177/0265407593104005 [DOI] [Google Scholar]
  25. Lavner JA, & Clark MA (2017). Workload and marital satisfaction over time: Testing lagged spillover and crossover effects during the newlywed years. Journal of Vocational Behavior, 101, 67–76. 10.1016/j.jvb.2017.05.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Le H, Newman A, Menzies J, Zheng C, & Fermelis J. (2020). Work-life balance in Asia: A systematic review. Human Resource Management Review, 30(4), Article 100766. 10.1016/j.hrmr.2020.100766 [DOI] [Google Scholar]
  27. Leatham G, & Duck S. (1990). Conversations with friends and the dynamics of social support. In Duck S. & Silver RC(Eds.), Personal relationships and social support (pp. 1–29). Sage. [Google Scholar]
  28. Li A, Robustelli BL, & Whisman MA (2016). Marital adjustment and psychological distress in Japan. Journal of Social and Personal Relationships, 33(7), 855–866. 10.1177/0265407515599678 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Ling Y, & Poweli GN (2001). Work-family conflict in contemporary China. Beyond an American-based model. International Journal of Cross Cultural Management, 1(3), 357–373. [Google Scholar]
  30. Liu H, Ngo HY, & Cheung FM (2016). Work–family enrichment and marital satisfaction among Chinese couples: A crossover-spillover perspective. International Journal of Stress Management, 23(2), 209–231. 10.1037/a0039753 [DOI] [Google Scholar]
  31. Loo J. (2023). Divorce rate in the United States: Geographic variation, 2022 (Family Profiles FP-23–24). National Center for Family and Marriage Research. 10.25035/ncfmr/fp-23-24 [DOI] [Google Scholar]
  32. Lu L, & Gilmour R. (2004). Culture, self and ways to achieve SWB: A cross-cultural analysis. Journal of Psychology in Chinese Societies, 5(1), 51–79. [Google Scholar]
  33. Lu L, Kao S-F, Chang T-T, Wu H-P, & Cooper CL (2008). Work/family demands, work flexibility, work/family conflict, and their consequences at work: A national probability sample in Taiwan. International Journal of Stress Management, 15(1), 1–21. 10.1037/1072-5245.15.1.1 [DOI] [Google Scholar]
  34. Margolis E. (2023, August 21). Navigating the notorious gauntlet of divorce in Japan. The Japan Times. https://www.japantimes.co.jp/community/2023/08/21/how-tos/navigating-the-notorious-gauntlet-of-divorce-in-japan/ [Google Scholar]
  35. Mistry J, & Wu J. (2010). Navigating cultural worlds and negotiating identities: A conceptual model. Human Development, 53(1), 5–25. 10.1159/000268136 [DOI] [Google Scholar]
  36. Nemoto K. (2008). Postponed marriage: Exploring women’s views of matrimony and work in Japan. Gender & Society, 22(2), 219–237. 10.1177/0891243208315868 [DOI] [Google Scholar]
  37. Ogbu JU (1981). Origins of human competence: A cultural-ecological perspective. Child Development, 52(2), 413–429. 10.2307/1129158 [DOI] [Google Scholar]
  38. Ono H. (2009). Husbands’ and wives’ education and divorce in the United States and Japan, 1946–2000. Journal of Family History, 34(3), 292–322. 10.1177/0363199009337996 [DOI] [Google Scholar]
  39. Ortiz-Ospina E, & Roser M. (2020). Marriages and divorces. Our World in Data. https://ourworldindata.org/marriages-and-divorces [Google Scholar]
  40. Repetti RL (1989). Effects of daily workload on subsequent behavior during marital interaction: The roles of social withdrawal and spouse support. Journal of Personality and Social Psychology, 57(4), 651–659. 10.1037/0022-3514.57.4.651 [DOI] [PubMed] [Google Scholar]
  41. Schuster TL, Kessler RC, & Aseltine RH Jr. (1990). Supportive interactions, negative interactions, and depressed mood. American Journal of Community Psychology, 18(3), 423–438. 10.1007/BF00938116 [DOI] [PubMed] [Google Scholar]
  42. Smith KW, & Sasaki MS (1979). Decreasing multicollinearity: A method for models with multiplicative functions. Sociological Methods & Research, 8(1), 35–56. 10.1177/004912417900800102 [DOI] [Google Scholar]
  43. Taniguchi H, & Kaufman G. (2014). Gender role attitudes, troubles talk, and marital satisfaction in Japan. Journal of Social and Personal Relationships, 31(7), 975–994. 10.1177/0265407513516559 [DOI] [Google Scholar]
  44. van Steenbergen EF, Kluwer ES, & Karney BR (2014). Work–family enrichment, work–family conflict, and marital satisfaction: A dyadic analysis. Journal of Occupational Health Psychology, 19(2), 182–194. 10.1037/a0036011 [DOI] [PubMed] [Google Scholar]
  45. Walen HR, & Lachman ME (2000). Social support and strain from partner, family, and friends: Costs and benefits for men and women in adulthood. Journal of Social and Personal Relationships, 17(1), 5–30. 10.1177/0265407500171001 [DOI] [Google Scholar]
  46. Yen P-C, & Yang W-S (2011). A comparative study of marital dissolution in East Asian societies: Gender attitudes and social expectations towards marriage in Taiwan, Korea and Japan. Asian Journal of Social Science, 39(6), 751–775. 10.1163/156853111X619210 [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement, Table 1
NIHMS1996683-supplement-Supplement__Table_1.docx (24.7KB, docx)

RESOURCES