Husband–Wife Religious Discordance,Marital Satisfaction,and Risk of Marital Dissolution in Two Generations

Abstract

We examined whether three types of husband–wife religious discordance (denomination, intensity, and attendance) were associated with the risk of marital dissolution over three decades. Using multigroup path analysis, we tested the mediating and moderating roles of husbands’ and wives’ marital satisfaction in the above associations in two generations. Using the Longitudinal Study of Generations data, we selected 193 second-generation married couples and 173 third-generation married couples. Results showed that denomination discordance was positively associated with marital dissolution in both generations, but discordance in religious attendance and intensity had no such association. Moderation was found with respect to wives’ marital satisfaction which reduced the risk of marital dissolution more in couples of different denominations compared with couples of the same denomination. Mediating effects were not found with respect to marital satisfaction. Implications are discussed in terms of the continued importance of religious identity to the long-term stability of marriages.

Keywords

religion marital dissolution marital satisfaction generations Longitudinal Study of Generations

Research has consistently shown a positive relationship between religiosity in married couples and better marital outcomes. Stronger religious beliefs and more frequent religious service attendance in married couples have been shown to be associated with greater marital satisfaction and marital stability (Jenkins, 1991; Mahoney, Pargament, Tarakeshwar, & Swank, 2001). Independent of religious traits themselves, religious homogamy—agreement between husbands and wives in their religious beliefs, behaviors, and identities; religious concordance is frequently used interchangeably—predict greater marital happiness and satisfaction in couples, and lower risk of divorce (Schafer & Kwon, 2017; Schramm, Marshall, Harris, & Lee, 2012; Sherkat & Ellison, 1999).

The present analysis extends the literature on the consequences of religious homogamy in marriages by prospectively examining the relationship between three types of marital religious discordance and risk of divorce, and the role that marital satisfaction plays in mediating and moderating this relationship. Furthermore, we examine marital satisfaction of both husbands and wives and compare mediating and moderating processes in two generations, one which came of marriageable age in the 1950s and another in the 1970s. Specifically, we use longitudinal data spanning several decades to examine whether discordance in religious denomination, religious intensity, and religious attendance within married couples, predict the long-term survival of their marriages. In addition, we examine the role that marital satisfaction plays in mediating and moderating the above associations, focusing on differences by gender and generational position.

Religion and Marital Outcomes

Research has consistently found that religiosity and religious concordance in married couples have a beneficial impact on the quality and longevity of their relationships. These studies have shown associations between the degree of religiosity of husbands and wives and marital quality indicators, such as greater marital satisfaction (Ellison, Burdette, & Bradford Wilcox, 2010; Jenkins, 1991; Schramm et al., 2012; Shehan, Bock, & Lee, 1990) and lower rates of divorce (Kalmijn, de Graaf, & Janssen, 2005; Wittberg, 1999). Married couples who exhibit greater religious homogamy enjoy better marital quality (Dudley & Kosinski, 1990; Ortega, Whitt, & William, 1988), with stronger associations for couples who are more religious (Curtis & Ellison, 2002; Heaton & Pratt, 1990; Jenkins, 1991; Myers, 2006).

Explanations for why shared religiousness benefit married couples tend to focus on institutional forces. Most religions promote beliefs and behaviors that are conducive to marital success, such as fidelity, honesty, and forgiveness which not only promote harmony but also promote reconciliation in the event of conflict (Lambert & Dollahite, 2006). Religious couples are often members of a congregation and/or a tight-knit religious community that encourage behaviors and reinforce social norms of the faith that encourage marital stability (Sherkat & Ellison, 1999; Wolfinger & Wilcox, 2008). Religious networks employ peer approval and acceptance, as well as fear of rejection from the group, as social tools to induce normative behavior among their members (Amato & Rogers, 1997).

Other reasons for marital benefits exhibited by religious couples are rooted in interpersonal and social explanations. Married couples who report higher religiosity tend to report more frequent joint attendance at religious services (Sherkat & Ellison, 1999). Since joint attendance is an observable act on the part of the couple, it represents a powerful signal of religious concordance (Bahr & Chadwick, 1985), provides an additional forum for marital interaction and bonding, and creates communal bonds with other like-minded couples (Call & Heaton, 1997). Through religious participation, the couple gains acceptance and approval of the religious community, which itself becomes a bulwark against marital instability.

Denominational similarity may also function to make marriages more harmonious through a common religious identity, shared religious values, and greater religious continuity, particularly when children are concerned. Put another way, denominational differences may cause strain in marriages due to the difficulties of reconciling differing expectations and lack of family acceptance, as has been shown among Catholic–Protestant mixed marriages in Northern Ireland (Wright, Rosato, & O’Reilly, 2017). Similarity in the religious denomination of husbands and wives may confer marital benefits by virtue of eliminating a potential source of conflict. In addition to these practical benefits, the common religious identity may help promote the sanctification of the marriage. Studies of family relations show sanctification of marriage by religious couples has a positive effect on the marital relationship (Ellison, Henderson, Glenn, & Harkrider, 2011; Mahoney, 2010; Mahoney et al., 1999; Mahoney, Pargament, Murray-Swank, & Murray-Swank, 2003).

Even if spouses share a denomination, they may still differ in the degree to which they religiously identify themselves. Married couples in which one partner identifies as being more religious than the other tend to suffer more from conflict than married couples in which couples share the same religious intensity (Sherkat & Ellison, 1999). Alternatively, there is evidence to suggest that couples in religiously heterogamous marriages may simply be less religious than couples in religiously homogenous marriages (Shehan et al., 1990).

Hypothesis 1: Marital religious discordance, measured as denomination, intensity, and attendance, will be associated with marital dissolution.

Marital Religious Discordance and Dissolution: The Role of Marital Quality

Research on religion and marriage treat marital quality and marital stability independently but rarely together. We suggest that marital satisfaction likely mediates the relationship between religious discordance and risk of marital dissolution, as has been shown with respect to other risk factors of divorce such as marital conflict, young age at marriage, and low income (Amato, 2010; Conger et al., 1990; Gottman, 1993; Karney & Bradbury, 1995). Marital satisfaction may also serve as a moderator of this relationship if more satisfied marriages are better able than less satisfied marriages to withstand strains deriving from religious discordance between spouses—thereby reducing the strength with which religious discordance is associated with marital dissolution.

Marital satisfaction may also condition the impact of religious discordance on risk of marital dissolution since satisfied marriages could withstand the stresses of religious disagreements in those couples, or be a sign that those stresses did not occur in religiously divergent couples. Viewed from another perspective, religious disagreements in the presence of low satisfaction may lead to elevated risk of marital dissolution because of the tenuous ties signified by both conditions.

Past research into the mechanisms through which religious factors might affect marital stability have used qualitative methods to examine questions such as how religiosity affects marital conflict (Lambert & Dollahite, 2006). In this study, using the concepts of mediation and moderation (Baron & Kenny, 1986; Hayes, 2013), we examine whether marital satisfaction acts as a conditional factor between religious factors and marital stability. Religious factors may have greater or lesser impact on marital dissolution depending on the level of marital satisfaction that already exists in the marriage. For example, a marriage with low satisfaction may be more likely to result in dissolution due to exogenous factors like religious homogamy than one that has high levels of marital satisfaction. There are two possible mechanisms through which this might occur (see Figure 1). Religious heterogamy may reduce the level of marital satisfaction, which in turn increases the risk of marital dissolution (mediation), or marital satisfaction may reduce the impact of religious heterogamy on the risk of marital dissolution (moderation). Proposed mediation and moderation models are depicted graphically in Figure 1 and form the basis for the following hypotheses:

Hypothesis 2: Marital satisfaction will mediate the association between husband–wife religious discordance and marital dissolution.

Hypothesis 3: Marital satisfaction will moderate the association between husband–wife religious discordance and marital dissolution.

Figure 1.

Mediation model (top) and moderation model (bottom).

Gender and Generational Differences

Studies have consistently demonstrated that women tend to be more religious than men (Miller & Hoffmann, 1995). In addition, research appears to suggest that religion is more important for women, especially in the case of religiously conservative wives. Wives’ religiosity has been found to have greater effect on marital satisfaction than husbands’ religiosity or couples’ coreligiosity (Jenkins, 1991) and marital disputes have been observed to be more frequent when there were differences in core religious beliefs and the wife was more religiously conservative (Curtis & Ellison, 2002). The risk of divorce may be lower when the husband is more religiously conservative, but it is higher when the wife’s religious beliefs are more conservative (Vaaler, Ellison, & Powers, 2009).

Gender differences have been shown in the association between religious indicators and marital outcomes with religious orientations of wives having greater consequences for the quality of the union than the religious orientations of husbands (Call & Heaton, 1997; Curtis & Ellison, 2002; Vaaler et al., 2009). Given the greater sensitivity to religious considerations observed among women, we advance the following hypothesis:

Hypothesis 4: The relationships among religious discordance, marital satisfaction, and marital dissolution will be stronger for women than for men.

Evidence suggests that the impact of religious discordance on marital outcomes has weakened over time such that religious differences have become better tolerated in more contemporary marriages (Myers, 2006). The changing religious landscape has accelerated the rate of interdenominational marriages in the United States. In 2014, 19% of people married in the 1950s had a spouse of a different religion to their own, and this increased to 30% for people married in the 1980s (Pew Research Center, 2015; see also Chaves & Anderson, 2012). Other evidence has shown similar historical increases for specific religious combinations such as marriages between Catholics and Protestants and between Jews and Gentiles between the 1970s and 1990s (Kalmijn, 1998). The decline in religious homogamy has been explained in terms of growth in the importance of alternative institutions for mate selection, such as the workplace, the softening of religious ideology as a basis for gender role allocation in a marriage (Mahoney, 2010; Myers, 2006), as well as the weakening authority of parents in controlling the marital choices of their young adult children (Rosenfeld, 2007).

It is also important to note that there has been a steady decline in religiousness in the United States over the past few decades (Chaves & Anderson, 2012; Lugo et al., 2008). Many baby boomers have turned away from the social conventions and institutions of their parents, including their religious beliefs and the religious institutions with which they were associated. The decline in religious authority of religion and the growing impact of secular culture in daily life implies that the relative significance of religion on marriage has also declined. Thus, we anticipate that these historical changes will result in generational differences in the association between religious discordance and marital outcomes, and propose the following hypothesis:

Hypothesis 5: The relationships among religious discordance, marital satisfaction, and marital dissolution will be weaker in the more recent generation which will be more tolerant of religious differentiation in their marriages.

Method

Sample

To test our hypotheses, we used nine waves (Ws) of data from the Longitudinal Study of Generations. The Longitudinal Study of Generations began in 1971 (W1) as a survey of intergenerational relations among 2,044 members of 358 three-generation families in which the oldest generation lived in Southern California (for more details, see Bengtson, 2013). In 1985 (W2), the study became longitudinal with repeated data collection in 1988 (W3), 1991 (W4), 1994 (W5), 1997 (W6), 2000 (W7), 2005 (W8), and 2016 (W9). Data were collected mostly by mail-back questionnaires, until W8 when an online survey option was provided and used by about half of respondents, and in W9 used by about 80%. Critical to the current investigation is that spouses were recruited into the study and continually added as the study progressed through time.

This present study focused on married couples in the second generation (G2) and third generation (G3), corresponding to marriages mostly taking place 1940-1955 and 1970-1985, respectively. Among the initial sample of 701 G2s at W1, we selected 193 married dyads who reported their religious denomination and marital satisfaction (n = 386 G2s; 55.1%), and then traced their marital outcomes for 34 years through W8 (G2s were not surveyed at W9). We chose the W3 survey as the baseline for G3 because the most marriages were found in this wave. Among the initial sample of 738 G3s at W3, we selected 173 married couple dyads who reported religious denominations and marital satisfaction (n = 346 G3s; 46.9%) and then traced their marital outcomes for 29 years through W9.

Measures

Husband–Wife Religious Discordance

We operationalized husband–wife religious discordance along three dimensions corresponding to denomination, intensity, and attendance (see Hwang, Silverstein, & Brown, 2018, for a similar approach). Measures of religiosity were taken for G2 dyads at W1 and for G3 dyads at W3. Religious denomination was measured by an open-ended question (“What is your current religious affiliation?”), and responses were sorted into 79 denominational categories based on an established classification scheme (Steensland et al., 2000). We combined these 79 categories into 7 exclusive categories: Mainline Protestant, Evangelical Protestant, Catholic, Jewish, Church of Latter Day Saints (Mormon), others, and none (for more details, see Hwang et al., 2018). Based on dyadic disagreement/agreement with respect to the above categories, religious denomination discordance between husbands and wives was assessed with a binary variable: discordant (1) and concordant (0).

Religious intensity was measured by a single item (“Do you consider yourself to be religious?”), and response options ranged on a 4-point scale from not at all religious (1) to very religious (4). We created an intensity discordance variable, subtracting the lower partner’s score from the higher partner’s score, reflecting an absolute value of the difference in religious intensity between husbands and wives (0-3).

Religious participation was measured by a single item (“How often do you attend religious service?”) and respondents answered on a 6-point scale ranging from never (1) to more than once a week (6). We created an attendance discordance variable by subtracting the lower partner’s score from the higher partner’s score, reflecting the absolute differences in religious attendance (0-5).

Marital Satisfaction

Marital satisfaction was measured for husbands and wives by 10 items from the Gilford–Bengtson Martial Satisfaction Scale (Gilford & Bengtson, 1979). The items asked how often the couple engaged in five positive and five negative couple activities. Respondents answered on a 5-point scale ranging from hardly ever (1) to almost always (5), and five negatively worded items. An example of a positive item is “You have a good time together” and an example of a negative item is “One of you becomes critical and belittling.” An average score was calculated from the 10 items, reverse coding the negative items, with higher scores reflecting greater marital satisfaction. Reliability α was .84 for husbands and .87 for wives in the G2 generation measured at W1, and .84 for husbands and .86 for wives in the G3 generation measured W3.

Marital Dissolution

We constructed a measure of marital dissolution that was measured by participants’ reports of their marital status from W2 to W8 (1985-2005) for G2 dyads from W4 to W9 (1991-2016) for G3 dyads. Marital dissolution included both separation and divorce, thus responses across all waves were combined and recorded as a binary variable: divorced or separated (1) and married (0).

Control Variables

All control variables were derived from W1 for G2s and W3 for G3s. Participants’ socioeconomic characteristics included age (in years), education (1 = grade school and 8 = postgraduate), race (0 = non-White and/or Hispanic, 1 = White non-Hispanic), number of children, experience of divorce, and marital duration. Household annual income was represented with a measure based on fifty ordered categories (from 1 = less than $1,000 to 50 = $50,000 or more) as reported by G2s at W1 and 12 ordered categories (from 1 = less than $10,000 to 12 = $110,000 or more) as reported by G3s at W3.

Analytic Strategy

After presenting descriptive information about study variables, we present the results of multigroup path analyses that examine the mediating and moderating roles of marital satisfaction in the association between husband–wife religious discordance and marital dissolution. Before estimating those models, we tested for structural invariance to assess whether path estimates were equal across the two generations using a chi-square difference test. To assess overall model fit we used the comparative fit index where a value >.90 indicates a good fit, and the root mean square error of approximation where a value <.06 indicates a good fit (Hu & Bentler, 1999).

In both mediation and moderation models, religious discordance variables and marital satisfaction were centered at their means to minimize multicollinearity (Aiken & West, 1991). A Sobel test (Preacher & Leonardelli, 2001) was conducted in evaluating the mediating role of marital satisfaction. The proportion of missing data in study variables was less than 1% in both G2 and G3 generations. To address missing values, full information maximum likelihood estimation was performed, such that all cases and missing data patterns were considered in the analysis (Acock, 2005).

It is important to note several methodological decisions that were made in estimating the models. First, we did not incorporate the timing of marital dissolution—which would be required for event history analysis—because dates of divorce and separation were not consistently provided across surveys. Therefore, we predicted only whether the event occurred in the time intervals studied. Second, we used a linear probability, rather than a logistic, specification to predict the dichotomous outcome of marital dissolution. Methodologists find little difference between the two models when event probabilities are not extreme, and recommend using the former under this condition because of the straightforward interpretation of its probabilistic coefficients, particularly when used in path analysis (Hellevik, 2009).

Results

Descriptive Characteristics

Baseline religious characteristics are shown for husbands and wives by generation in Table 1. In general, wives reported stronger religiosity than husbands across the measures considered. Generational differences were also found with G3s less religious than G2s. In terms differences within couples, husband–wife religious denomination discordance was significantly higher in G3 dyads (38.2%) at Wave 3, compared with G2 dyads (23.3%) at Wave 1 (see Table 1). However, religious intensity discordance and attendance discordance were not significantly different between the generations.

Table 1.

Religious Characteristics of Husbands and Wives at Baseline by Generation.

Variables	G2 Husbands (n = 193)	G2 Wives (n = 193)	G3 Husbands (n = 173)	G3 Wives (n = 173)	G2 Dyads (n = 193)	G3 Dyads (n = 173)
Variables	n (%)	n (%)	n (%)	n (%)	n (%)	n (%)
Religious denomination
Mainland Protestant	81 (42.0)	73 (37.8)	24 (13.9)	39 (22.5)
Evangelical Protestant	22 (11.4)	25 (13.0)	26 (15.0)	25 (14.5)
Catholic	38 (19.7)	41 (21.2)	27 (15.6)	33 (19.1)
Jewish	24 (12.4)	21 (10.9)	14 (8.1)	12 (6.9)
Mormon	7 (3.6)	9 (4.7)	10 (5.8)	12 (6.9)
Others	6 (3.1)	8 (4.1)	3 (1.7)	5 (2.9)
None	15 (7.8)	16 (8.3)	69 (39.9)	47 (27.2)
Religious intensity
Not at all religious	22 (11.4)	12 (6.2)	36 (20.8)	15 (8.7)
Somewhat religious	43 (22.3)	37 (19.2)	45 (26.0)	62 (35.8)
Moderately religious	97 (50.3)	101 (52.3)	63 (36.4)	62 (35.8)
Very religious	30 (15.5)	43 (22.3)	27 (15.6)	33 (19.1)
Religious attendance
Never	57 (29.5)	60 (31.1)	57 (32.9)	44 (25.4)
Once a year or so	54 (28.0)	52 (26.9)	36 (20.8)	36 (20.8)
Several times a year	13 (6.7)	9 (4.7)	24 (13.9)	24 (13.9)
About once a month	9 (4.7)	6 (3.1)	12 (6.9)	11 (6.4)
Once a week or so	38 (19.7)	46 (23.8)	28 (16.2)	49 (28.3)
More than once a week	18 (9.3)	16 (8.3)	13 (7.5)	8 (4.6)
Husband–wife religious denomination
Concordance					148 (76.7)	107 (61.8)
Discordance^a					45 (23.3)	66 (38.2)
Husband–wife religious intensity
Intensity concordance					89 (46.1)	73 (42.2)
Intensity discordance^b					103 (53.5)	98 (56.0)
Husband–wife religious attendance
Attendance concordance					104 (53.9)	96 (55.5)
Attendance discordance^b					81 (42.0)	72 (41.6)

Note. G2 = baby boomer’s parent generation; G3 = baby boomer generation; W = wave. G2 data come from W1 (1971) and G3 data come from W3 (1988).

Significant mean difference between G2 and G3 (p < .001). ^bNonsignificant mean difference between G2 and G3 (p > .05).

Table 2 provides sociodemographic statistics of study variables at baseline for each generation. The average of G2 husbands and wives were mid-forty at W1, whereas G3s were mid-30s at W3. Although G3s are younger than G2s, G3s had more experiences of divorce before their current marital relationships compared with G2s. Regarding three types of husband–wife religious discordance, G3 dyads reported more discordance in denomination compared with G2 dyads (t = −3.17, p < .001). In addition, G3 dyads reported more incidents of marital dissolution over the course of the study compared with G2 dyads (t = −2.87, p < .01). Marital satisfaction was not significantly different between G2s and G3s in all waves.

Table 2.

Husbands’ and Wives’ Characteristics at Baseline by Generation.

	G2		G3		t Test between
	M (SD)		M (SD)		G2 and G3, Husbands	G2 and G3, Wives	G2 and G3, Dyads
	Husbands	Wives	Husbands	Wives	G2 and G3, Husbands	G2 and G3, Wives	G2 and G3, Dyads
Age	45.95 (5.19)	42.89 (4.82)	36.99 (4.14)	35.16 (4.02)	18.34***	16.71***
Education	5.25 (1.67)	4.58 (1.13)	5.37 (1.80)	4.77 (1.54)	−0.65	−1.33
White	0.89 (0.31)	0.90 (0.29)	0.92 (0.26)	0.95 (0.19)	0.99	1.92
Experience of divorce	0.11 (0.31)	0.13 (0.33)	0.26 (0.49)	0.25 (0.49)	−3.45***	−2.71**
Household income	20.59^a (9.38)		6.12^b (2.55)
Number of children	3.41 (1.62)		1.76 (1.19)				11.17***
Marital duration	20.82 (5.96)		10.70 (5.69)				16.61***
Religious discordance
Denomination	0.23 (0.42)		0.38 (0.48)				−3.17***
Attendance	0.69 (1.03)		0.66 (0.95)				0.29
Intensity	0.64 (0.69)		0.71 (0.72)				−0.94
Marital dissolution	0.15 (0.35)		0.27 (0.44)				−2.87**
Marital satisfaction	3.90 (0.57)	3.91 (0.68)	3.80 (0.61)	3.91 (0.63)	1.61	0.31

Note. G2 = baby boomer’s parent generation; G3 = baby boomer generation; W = wave. G2 data come from W1 (1971) and G3 data come from W3 (1988).

20 Is between $19,000 and $20,999 in W1. ^b6 Is between $50,000 and $59,999 in W3.

p < .05. **p < .01. ***p < .001.

Mediation Model

To test the mediation model, multigroup path analysis was conducted, with goodness-of-fit statistics found in Table 3. First, we tested the mediation model for G2s and G3s separately. finding a good model fit in each generation. Next, we conducted a multigroup analysis, comparing a constrained model in which all parameters were held equal across groups, to an unconstrained model in which all parameters were free to vary across groups, the chi-square difference between the models was not significant (Δχ² = 23.874, Δdf [degrees of freedom] = 22, p > .05), suggesting that the more parsimonious constrained model, in which estimates were held constant across G2s and G3s, is preferred over the unconstrained model.

Table 3.

Model Fit Information of G2-G3 Multigroup Path Models.

Model	χ²	df	χ²_diff	Δdf	CFI	RMSEA
Mediation model
Single group analysis
G2 (n = 193)	55.859	55			.998	.009
G3 (n = 173)	57.130	55			.996	.015
Multigroup analysis
Unconstrained model	112.991	110			.997	.009
Constrained model	136.865	132	23.874	22	.995	.010
Moderation model
Single group analysis
G2 (n = 193)	106.517	121			1.000	.000
G3 (n = 173)	134.225	121			.983	.025
Multigroup analysis
Unconstrained model	240.755	242			1.000	.000
Constrained model	260.978	264	20.223	22	1.000	.000

Note. df = degrees of freedom; CFI = comparative fit index; RMSEA = root mean square error of approximation; G2 = baby boomer’s parent generation; G3 = baby boomer generation.

Coefficients for the constrained mediation model are presented in Table 4. With respect to direct effects between religious discordance and marital outcomes, husband–wife denomination discordance was positively associated with marital dissolution (b = .152, p < .001). That is, couples with different religious denominations experienced greater risk of divorce or separation compared with couples with the same religions denomination. Neither attendance discordance nor intensity discordance was associated with marital dissolution. Wives’, but not husband’s, marital satisfaction was negatively associated with marital dissolution (b = −.075, p = .027). However, the three types of religious discordance were not associated with either husbands’ or wives’ marital satisfaction. These results do not support our second hypothesis concerning the mediating role of marital satisfaction.

Table 4.

Path Coefficients in the G2-G3 Constrained Multigroup Model.

Path	Mediation model		Moderation model
Path	Estimate	SE	Estimate	SE
Predicting marital dissolution
Denomination discordance → Marital dissolution	.152***	.044	.150***	.044
Attendance discordance → Marital dissolution	.034	.020	.035	.021
Intensity discordance → Marital dissolution	−.029	.028	−.027	.028
Husbands’ marital satisfaction → Marital dissolution	−.027	.039	−.027	.038
Wives’ marital satisfaction → Marital dissolution	−.075*	.034	−.075*	.034
Predicting husbands’ marital satisfaction
Denomination discordance → Husbands’ marital satisfaction	−.013	.070
Attendance discordance → Husbands’ marital satisfaction	.024	.033
Intensity discordance → Husbands’ marital satisfaction	−.037	.045
Predicting wives’ marital satisfaction
Denomination discordance → Wives’ marital satisfaction	−.080	.077
Attendance discordance → Wives’ marital satisfaction	−.011	.037
Intensity discordance → Wives’ marital satisfaction	.013	.050
Interaction terms
Denomination discordance × Husbands’ marital satisfaction → Marital dissolution			−.070	.089
Attendance discordance × Husbands’ marital satisfaction → Marital dissolution			−.028	.047
Intensity discordance × Husbands’ marital satisfaction → Marital dissolution			−.003	.053
Denomination discordance × Wives’ marital satisfaction → Marital dissolution			−.171*	.078
Attendance discordance × Wives’ marital satisfaction → Marital dissolution			.013	.039
Intensity discordance × Wives’ marital satisfaction → Marital dissolution			.096	.050

Note. G2 = baby boomer’s parent generation; G3 = baby boomer generation; SE = standard error.

p < .05. ***p < .001.

Moderation Model

In testing for the moderating role of marital satisfaction as specified in our third hypotheses, we conducted multigroup path analysis by generation, finding again that the constrained model was preferred over the unconstrained model (Δχ² = 20.223, Δdf = 22, p > .05). The coefficients for this model are found in Table 4. Results indicate that wives’ marital satisfaction moderated the association between denomination discordance and marital dissolution in both generations (b = −.171, p = .027). Figure 2 presents predicted values based on estimates from this model. Couples in which wives had weaker marital satisfaction experienced greater relative risk of marital dissolution related to denominational discordance compared with couples in which wives had stronger marital satisfaction. Conversely, one can interpret the results in terms of high martial satisfaction among wives buffering the negative impact of denominational discordance on marital survival. In contrast, husbands’ marital satisfaction did not moderate the association between any of the three types of religious discordance and marital dissolution.

Figure 2.

Predicted probability of marital dissolution by denominational discordance and wives’ marital satisfaction.

Discussion

In this analysis, we examined whether three types of husband–wife religious discordance were associated with risk of marital dissolution over the course of several decades. Using multigroup path analysis, we focused on the mediating and moderating roles of marital satisfaction in the above associations across two adjacent generations. Our goal was to clarify mechanisms through which husband–wife religious differences influence marital outcomes.

Our first hypothesis, husband–wife religious denomination discordance would predict a risk of marital dissolution, was confirmed. These results support previous research indicating that differences in religious denomination between husbands and wives are related to marital instability (Mullins, Brackett, Bogie, & Pruett, 2004). The negative effect of denomination discordance on marital dissolution was similar across two generations, a surprising result given that other studies have found smaller effects in younger generations (Mahoney, 2010; Myers, 2006). We speculate that this may be related to the different ages at which each generation entered the study with the older generation having longer surviving marriages at the start of the observation period.

In contrast to denomination differences, discrepancies in husband–wife religious intensity and religious attendance were not associated with marital dissolution. Marriages appear to better tolerate religious difference when that difference is one of magnitude than of kind. Research suggests that denominational subcultures are marked out by differences in values that are seen as central to denominational core beliefs and self-identity, but that there may be large variations in attitudes toward other, noncore values within denominations (Gay, Ellison, & Powers, 1996). Greater tolerance of religious differences in magnitude versus kind among religiously discordant couples may reflect the differing types and levels of religious differences the denominations themselves are able to tolerate before they experience a schism. This would be consistent with findings that partners who share core religious beliefs and values report greater marital satisfaction (Ellison et al., 2010). Husbands and wives with different denominations may also have fewer opportunities to spend time together engaging in joint religious activities, and be more likely to have nonoverlapping social circles which may increase social distance between them (Mullins et al., 2004). Finally, denominational concordance may allow couples to sanctify their marriage, even when differences in the magnitude of their beliefs exist, in ways not possible when there is denominational discordance. The sharing of similar beliefs and engaging in joint religious activities between couples have been found to be more important for sanctification of marital relations than the couples’ specific beliefs (Lichter & Carmalt, 2009). Similar results were found in other research with respect to denominational discordance and emotional distance in parent–adult children relations, demonstrating that the risk imposed by categorical religious differences extends to multiple family relationships (Hwang et al., 2018).

Our second hypothesis, husbands’ and wives’ marital satisfaction will mediate the association between husband–wife denomination discordance and marital dissolution, was not supported in both generations. The direct importance of marital discordance in religious denomination for marital survival suggests that denomination as a fundamental characteristic exerts a structural and potentially irreconcilable impact on marriages that is not explained by satisfaction with the marriage itself.

Our third hypothesis regarding marital satisfaction as a moderator of the relationship between husband–wife denomination discordance and marital dissolution was supported. Moderation was found only for marital satisfaction expressed by wives, demonstrating greater marital protection afforded by women’s tolerance of religious differences with their spouses. In part, this result affirms other research showing that women’s perspective on religion and family life is more consequential than it is for men. In addition, this study supports Bernard’s (1972) research indicating that husbands and wives have different subjective experiences of marriage. Given that wives are less satisfied with marriage than their husbands (Fowers, 1991), the finding of this study suggests that wives’ marital satisfaction is a more crucial factor enhancing or alleviating the impact of religious denomination discordance on their marital relations, compared with husbands’ marital satisfaction.

Contrary to the expectation described by our fourth hypothesis, we found no generational difference in either mediation or moderation models. Considering that the proportion of religiously unaffiliated people in the United States has increased most dramatically in post-baby boom generations (Pew Research Center, 2015), future research should examine religious processes in marriages within more contemporary generations. Furthermore, the growth of cohabitation as a living arrangement, particularly for couples in younger generations, behooves researchers to examine how religion plays a role in nonmarital partnerships.

Although this study provides insights into the marital implications of husband–wife religious discordance, we interpret our findings cautiously for several reasons. First, we employed a simple binary variable to represent denomination discordance of all types across seven denominational categories. Sample size restrictions prevented examination of specific types of discordance which may be wider than others. For example, one might expect that denominational differences will be less salient when both spouses have Christian affiliations, than when they are from traditions with wider theological differences (such as one partner having a Christian and another a non-Christian denominational identity, or no denominational identity). Second, the sample consisted of mostly southern California families, tended to be lower middle class and higher, and underrepresented minority families. Therefore, our findings may not be generalizable to other populations. Third, we did not account for the timing of marital dissolution, nor were we able to match the generations based on age, calling for more detailed information on marital timing and stronger intercohort research designs. Fourth, we examined the interaction terms with respect to various combinations of religious homogamy and heterogamy between husbands and wives but all results were not significant. However, this issue should be addressed in future research because it is possible that same-faith couples who both attend frequently have the greatest marital longevity.

Although it is understood that different external sources of stress affect relationship well-being in same-sex couples (Cao et al., 2017), the current study did not address questions of marital satisfaction or longevity in same-sex couples for several reasons. First, same-sex marriages were not legally recognized in most of the United States during data collection for Waves 7 (2000) and 8 (2005), which means that individuals who identified as lesbian, gay, or bisexual (LGB) may not have identified as married if they were in a long-term relationship with someone of the same sex. Second, less than 4% (n = 22) of G3 survey respondents identified as LGB, and of those only 64% indicated that they were living with their partner, resulting in too small a sample to support any meaningful analysis of outcomes. And finally, sexual identity does not always predict sexual behavior (Institute of Medicine, 2011) which means that individuals identifying as LGB may choose to marry someone of the opposite sex, and because we did not ask respondents in previous waves of the survey to specify the sex of their partner, we cannot assume they were in a same-sex relationship.

Because same-sex marriage has been legal throughout the United States since the landmark 2015 Supreme Court case Obergefell v Hodges, and because same-sex marriages experience sources of stress that do not affect opposite-sex marriages (Cao et al., 2017), future studies of marital satisfaction and longevity should distinguish between same-sex and opposite-sex marital partnerships. To enable these distinctions, future surveys should collect data not only on respondent sexual identity but also on the sex of the respondent’s marital partner.

In conclusion, our results indicate that discordance in religious denomination between husbands and wives is a risk factor predicting marital dissolution in two different generations, and that wives’ marital satisfaction can influence this risk. We hope that these results clarify our understanding of the link between religious differences and marital outcomes, and provide researchers, as well as practitioners, further knowledge about marital stress that might be imposed on religiously heterogeneous couples and gender imbalances in how this may be expressed in terms of their consequences for marital durability.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The project described was supported by Award Numbers 56407 by John Templeton Foundation and R21AG050429 by National Institute on Aging.

ORCID iDs

Woosang Hwang

Joonsik Yoon

References

Acock

A. C.

(2005). Working with missing values. Journal of Marriage and Family, 67, 1012-1028. doi:10.1111/j.1741-3737.2005.00191.x

Aiken

L. S.

West

S. G.

(1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage.

Amato

P. R.

(2010). Research on divorce: Continuing trends and new developments. Journal of Marriage and Family, 72, 650-666. doi:10.1111/j.1741-3737.2010.00723.x

Amato

P. R.

Rogers

S. J.

(1997). A longitudinal study of marital problems and subsequent divorce. Journal of Marriage and the Family, 59, 612-624. doi:10.2307/353949

Bahr

H. M.

Chadwick

B. A.

(1985). Religion and family in Middletown, USA. Journal of Marriage and the Family, 47, 407-414. doi:10.2307/352140

Baron

R. M.

Kenny

D. A.

(1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182. doi:10.1037/0022-3514.51.6.1173

Bengtson

V. L.

(2013). Families and faith: How religion is passed down across generations. New York, NY: Oxford University Press.

Bernard

(1972). The future of marriage. New York, NY: World Publishing.

Call

V. R. A.

Heaton

T. B.

(1997). Religious influence on marital stability. Journal for the Scientific Study of Religion, 36, 382-392. doi:10.2307/1387856

10.

Cao

Zhou

Fine

Liang

Mills-Koonce

W. R.

(2017). Sexual minority stress and same-sex relationship well-being: A meta-analysis of research prior to the U.S. nationwide legalization of same-sex marriage. Journal of Marriage and Family, 79, 1258-1277. doi:10.1111/jomf.12415

11.

Chaves

Anderson

(2012). Continuity and change in American religion, 1972–2008. In Marsden

P. V.

(Ed.), Social trends in American life: Findings from the General Social Survey since 1972 (pp. 212-239). Princeton, NJ: Princeton University Press.

12.

Conger

R. D.

Elder

G. H.

Lorenz

F. O.

Conger

K. J.

Simons

R. L.

Whitbeck

L. B.

. . . Melby

J. N.

(1990). Linking economic hardship to marital quality and instability. Journal of Marriage and the Family, 52, 643-656. doi:10.2307/352931

13.

Curtis

K. T.

Ellison

C. G.

(2002). Religious heterogamy and marital conflict findings from the National Survey of Families and Households. Journal of Family Issues, 23, 551-576. doi:10.1177/0192513X02023004005

14.

Dudley

M. G.

Kosinski

F. A.

(1990). Religiosity and marital satisfaction: A research note. Review of Religious Research, 32, 78-86. doi:10.2307/3511329

15.

Ellison

C. G.

Burdette

A. M.

Bradford Wilcox

(2010). The couple that prays together: Race and ethnicity, religion, and relationship quality among working-age adults. Journal of Marriage and Family, 72, 963-975. doi:10.1111/j.1741-3737.2010.00742.x

16.

Ellison

C. G.

Henderson

A. K.

Glenn

N. D.

Harkrider

K. E.

(2011). Sanctification, stress, and marital quality. Family Relations, 60, 404-420. doi:10/dd67hq

17.

Fowers

B. J.

(1991). His and her marriage: A multivariate study of gender and marital satisfaction. Sex Roles, 24, 209-221.

18.

Gay

D. A.

Ellison

C. G.

Powers

D. A.

(1996). In search of denominational subcultures: Religious affiliation and “pro-family” issues revisited. Review of Religious Research, 38, 3-17. doi:10.2307/3512537

19.

Gilford

Bengtson

V. L.

(1979). Marital satisfaction in three generations: Positive and negative dimensions. Journal of Marriage and the Family, 41, 387-398. doi:10.2307/351705

20.

Gottman

J. M.

(1993). A theory of marital dissolution and stability. Journal of Family Psychology, 7, 57-75. doi:10.1037/0893-3200.7.1.57

21.

Hayes

A. F.

(2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York, NY: Guilford Press.

22.

Heaton

T. B.

Pratt

E. L.

(1990). The effects of religious homogamy on marital satisfaction and stability. Journal of Family Issues, 11, 191-207. doi:10.1177/019251390011002005

23.

Hellevik

(2009). Linear versus logistic regression when the dependent variable is a dichotomy. Quality & Quantity, 43, 59-74.

24.

L. T.

Bentler

P. M.

(1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6, 1-55. doi:10.1080/10705519909540118

25.

Hwang

Silverstein

Brown

(2018). Parent–adult child religious discordance: Consequences for intergenerational solidarity across several decades. Journal of Family Issues, 39, 1545-1572.

26.

Institute of Medicine. (2011). The health of lesbian, gay, bisexual, and transgender (LGBT) people: Building a foundation for better understanding. Washington, DC: National Academies Press.

27.

Jenkins

K. W.

(1991). Religion and families. In Bahr

S. J.

(Ed.), Family research: A sixty-year review, 1930-1990 (pp. 235-288). Lexington, MA: Lexington Books.

28.

Kalmijn

(1998). Intermarriage and homogamy: Causes, patterns, trends. Annual Review of Sociology, 24, 395-421. doi:10/dtpvgf

29.

Kalmijn

de Graaf

P. M.

Janssen

J. P. G.

(2005). Intermarriage and the risk of divorce in the Netherlands: The effects of differences in religion and in nationality, 1974–94. Population Studies, 59, 71-85. doi:10.1080/0032472052000332719

30.

Karney

B. R.

Bradbury

T. N.

(1995). The longitudinal course of marital quality and stability: A review of theory, methods, and research. Psychological Bulletin, 118, 3-34. doi:10.1037/0033-2909.118.1.3

31.

Lambert

N. M.

Dollahite

D. C.

(2006). How religiosity helps couples prevent, resolve, and overcome marital conflict. Family Relations, 55, 439-449. doi:10.1111/j.1741-3729.2006.00413.x

32.

Lichter

D. T.

Carmalt

J. H.

(2009). Religion and marital quality among low-income couples. Social Science Research, 38, 168-187. doi:10/d93f8w

33.

Lugo

Stencel

Green

Smith

Cox

Pond

. . . Keeter

(2008). U.S. religious landscape survey: Religious affiliation. Retrieved from http://www.pewforum.org/2008/02/01/u-s-religious-landscape-survey-religious-affiliation/

34.

Mahoney

(2010). Religion in families, 1999–2009: A relational spirituality framework. Journal of Marriage and Family, 72, 805-827. doi:10.1111/j.1741-3737.2010.00732.x

35.

Mahoney

Pargament

K. I.

Jewell

Swank

A. B.

Scott

Emery

Rye

(1999). Marriage and the spiritual realm: The role of proximal and distal religious constructs in marital functioning. Journal of Family Psychology, 13, 321-338. doi:10.1037/0893-3200.13.3.321

36.

Mahoney

Pargament

K. I.

Murray-Swank

(2003). Religion and the sanctification of family relationships. Review of Religious Research, 44, 220-236. doi:10/bngwsx

37.

Mahoney

Pargament

K. I.

Tarakeshwar

Swank

A. B.

(2001). Religion in the home in the 1980s and 1990s: A meta-analytic review and conceptual analysis of links between religion, marriage, and parenting. Journal of Family Psychology, 15, 559-596. doi:10.1037/0893-3200.15.4.559

38.

Miller

A. S.

Hoffmann

J. P.

(1995). Risk and religion: An explanation of gender differences in religiosity. Journal for the Scientific Study of Religion, 34, 63-75. doi:10.2307/1386523

39.

Mullins

L. C.

Brackett

K. P.

Bogie

D. W.

Pruett

(2004). The impact of religious homogeneity on the rate of divorce in the United States. Sociological Inquiry, 74, 338-354. doi:10.1111/j.1475-682X.2004.00095.x

40.

Myers

S. M.

(2006). Religious homogamy and marital quality: Historical and generational patterns, 1980–1997. Journal of Marriage and Family, 68, 292-304. doi:10.1111/j.1741-3737.2006.00253.x

41.

Ortega

S. T.

Whitt

H. P.

William

J. A.

(1988). Religious homogamy and marital happiness. Journal of Family Issues, 9, 224-239. doi:10.1177/019251388009002005

42.

Pew Research Center. (2015). America’s changing religious landscape. Retrieved from http://www.pewforum.org/2015/05/12/americas-changing-religious-landscape/

43.

Preacher

K. J.

Leonardelli

G. J.

(2001). Calculation for the Sobel test. Retrieved from http://quantpsy.org/sobel/sobel.htm

44.

Rosenfeld

M. J.

(2007). The age of independence: Interracial unions, same-sex unions, and the changing American family. Cambridge, MA: Harvard University Press.

45.

Schafer

M. H.

Kwon

(2017). Religious heterogamy and partnership quality in later life. Journals of Gerontology: Series B, gbx072. doi:10.1093/geronb/gbx072

46.

Schramm

D. G.

Marshall

J. P.

Harris

V. W.

Lee

T. R.

(2012). Religiosity, homogamy, and marital adjustment: An examination of newlyweds in first marriages and remarriages. Journal of Family Issues, 33, 246-268. doi:10.1177/0192513X11420370

47.

Shehan

C. L.

Bock

E. W.

Lee

G. R.

(1990). Religious heterogamy, religiosity, and marital happiness: The case of Catholics. Journal of Marriage and the Family, 52, 73-79. doi:10.2307/352839

48.

Sherkat

D. E.

Ellison

C. G.

(1999). Recent developments and current controversies in the sociology of religion. Annual Review of Sociology, 25, 363-394. doi:10.1146/annurev.soc.25.1.363

49.

Steensland

Robinson

L. D.

Wilcox

W. B.

Park

J. Z.

Regnerus

M. D.

Woodberry

R. D.

(2000). The measure of American religion: Toward improving the state of the art. Social Forces, 79, 291-318.

50.

Vaaler

M. L.

Ellison

C. G.

Powers

D. A.

(2009). Religious influences on the risk of marital dissolution. Journal of Marriage and Family, 71, 917-934. doi:10.1111/j.1741-3737.2009.00644.x

51.

Wittberg

(1999). Families and religions. In Sussman

M. B.

Steinmetz

S. K.

Peterson

G. W.

(Eds.), Handbook of marriage and the family (2nd ed., pp. 503-523). New York, NY: Plenum Press.

52.

Wolfinger

N. H.

Wilcox

W. B.

(2008). Happily ever after? Religion, marital status, gender and relationship quality in urban families. Social Forces, 86, 1311-1337.

53.

Wright

D. M.

Rosato

O’Reilly

(2017). Influence of heterogamy by religion on risk of marital dissolution: A cohort study of 20,000 couples. European Journal of Population, 33, 87-107.