اثرات اختلافات زناشویی والدین و طلاق بر روی زندگی معنوی و مذهبی جوانان

Abstract A growing literature reveals that parental divorce and marital discord can have undesirable effects on the mental health and social well-being of children, some of which extend well into adulthood. Our study augments this body of work by focusing on the interplay of divorce and discord in shaping the religious and spiritual lives of young adults. Several discrete subgroups of young adults are identified in terms of parental marital status and degree of parents’ marital conflict, and multiple religious and spiritual outcomes are considered. Data are taken from the National Survey on the Moral and Spiritual Lives of Young Adults from Divorced and Intact Families, a nationwide US telephone survey of approximately 1500 young adults ages 18–35 conducted in 2001. Findings confirm that persons raised by parents in intact, happy, low-conflict marriages tend to score higher on most religious and spiritual outcomes. However, offspring from divorced families and those from intact high-conflict families differ on some outcomes, but not others. Indicators of traditional institutional religious practices and beliefs appear more vulnerable to the effects of parental divorce and discord than personal spiritual beliefs and practices. Overall, findings reveal a rich but complex set of relationships between family background and religious and spiritual lives among young adults.

1. Introduction Divorce rates in the US rose sharply beginning in the 1960s, before leveling off in the early-mid 1980s at historically high levels (Heaton, 2002 and Cherlin, 2004). These patterns have raised concerns among scholars, policymakers, and the public about the implications of parental divorce for offspring. Although much of the research on this topic has centered on outcomes during childhood and adolescence, several studies also demonstrate that the effects of parental divorce can reach well into adulthood, and there is mounting evidence that children raised in intact, two-parent families tend to fare better on a host of psychological, social, and behavioral outcomes than those raised in other types of households (McLanahan and Sandefur, 1995, Amato and Booth, 1997 and Amato, 2001). More recent studies have attempted to disentangle the effects of parental divorce per se from those of parental marital discord, in an effort to determine (a) whether the observed influence of divorce may be due partly to exposure to parents’ marital conflict, and also (b) whether it is better for children if quarreling parents separate as opposed to remaining married (Amato et al., 1995, Booth and Amato, 2001 and Yu et al., 2010). Scholars have studied the effects of parental divorce and/or marital conflict on a wide array of child and young adult outcomes, ranging from psychological well-being, to behavior problems, to the quality of parent–child relationships (Amato and Booth, 2001, Amato, 2001 and Strohschein, 2005). To date, however, few of these studies have focused on the implications of divorce and marital discord for the religious and spiritual lives of young adults. This oversight is surprising for several reasons. Parents’ religious identities, practices, and beliefs are by far the strongest predictors of those of their offspring (Hoge et al., 1982, Glass et al., 1986 and Sherkat, 1991). The intergenerational transmission of faith commitments has long been a central topic of research interest among sociologists of religion, and it also has important implications for religious organizations, their leaders, and their activities (Sherkat, 1991, Sherkat and Wilson, 1995 and Lawton and Bures, 2001). There is a modest body of evidence showing that adolescents and young adults raised in intact, two-parent households are more likely than others to be religious, and specifically adopt the affiliations and practices of their parents (Regnerus and Uecker, 2006 and Uecker et al., 2007). This is particularly the case when couples’ marital happiness is high (Myers, 1996), and among offspring who report close relationships with their parents (Sherkat and Wilson, 1995). However, this literature is characterized by at least two notable limitations: (a) studies in this tradition have not explored the interplay of parental divorce and marital conflict on the religious and spiritual lives of offspring; and (b) only a handful of outcomes in this domain (e.g., affiliation and denominational “switching,” religious attendance) have been examined, yielding an incomplete picture of the reach of parental marital relations into the religious and spiritual lives of young adults. Our study contributes to this emerging body of work in several specific ways. Drawing lessons from the broader literature on the effects of parental divorce and marital quality on adjustment outcomes among offspring, we use data from an unique sample of approximately 1500 young adults (ages 18–35), evenly divided between children from intact families consisting of two biological parents, and children who experienced parental divorce prior to age 15 (Marquardt, 2005). For the purposes of this analysis, we distinguish among seven groups of young adults: (a) those raised by parents in intact, happy marriages; (b) those whose parents were in unhappy but low-conflict marriages; (c) those whose parents were in high-conflict marriages; (d) offspring of divorce whose parents experienced high levels of conflict during both pre- and post-divorce periods; (e) offspring of divorce whose parents experienced consistently low levels of conflict, both before and after the split; (f) offspring of divorce whose parents experienced high pre-divorce conflict, followed by reductions in conflict after the divorce; and (g) offspring of divorce whose parents exhibited minimal conflict preceding the divorced, but experienced elevated levels of conflict during the post-divorce period. We then estimate a series of multivariate models gauging the net differences among these divorce/marital discord categories on an array of religious and spiritual outcomes in young adulthood, including: (a) frequency of religious attendance and prayer; (b) (dis)interest in religion as a source of truth and meaning; (c) skepticism toward the religiousness and religious sincerity of parents; (d) images of God; (e) positive and negative experiences of God; and (f) religious and spiritual identities. These findings cast fresh light on the complex implications of parental marital discord and divorce on young adults’ religiousness and spirituality in an era of relatively high rates of divorce, family change, and religious ferment.

4. Results Descriptive statistics on all variables used in these analyses are presented in Table 1. Roughly 21% of NSMS respondents were raised in what we have termed intact, happy low-conflict families, while another 14% grew up in intact, unhappy low-conflict families, and 13% came from intact, high-conflict families. Offspring from highly conflictual divorces – i.e., those with high-conflict during both pre- and post-divorce periods – make up approximately 18% of the NSMS sample, while another 12% of respondents experienced parental divorces in which high levels of conflict prior to the divorce were reduced following the split. For the remaining 22% of NSMS respondents, parental divorces were not preceded by significant conflict among parents – at least, not that the offspring could recall. Roughly two-thirds of these young adults (15% of NSMS sample) recalled little or no conflict in the post-divorce phase, while a relatively small proportion of the NSMS respondents (7.5%) reported that levels of perceived parental conflict were greater in the aftermath of divorce than during the pre-divorce period. Table 1. Descriptive statistics. Variable N Mean St.dev. Range Intact happy, low-conflict marriage 1462 .21 – 0–1 Intact unhappy, low-conflict marriage 1462 .14 – 0–1 Intact unhappy, high-conflict marriage 1462 .13 – 0–1 Divorce, high-conflict before and after 1462 .18 – 0–1 Divorce, high-conflict before only 1462 .12 – 0–1 Divorce, low-conflict before and after 1462 .15 – 0–1 Divorce, low-conflict before only 1462 .07 – 0–1 Religious attendance 1508 2.51 1.17 1–4 Frequency of prayer 1491 3.17 1.03 1–4 Religious disinterest 1480 4.22 2.02 2–8 More religious than mother 1490 1.80 1.04 1–4 More religious than father 1481 2.20 1.19 1–4 Doubt mother’s religious sincerity 1479 1.49 .86 1–4 Doubt father’s religious sincerity 1461 1.69 .98 1–4 Positive experiences of God 1494 3.29 .85 1–4 Absence of doubts about God 1493 3.49 .79 1–4 Positive God images 1497 3.59 .65 1–4 Image of God as angry 1467 1.56 .87 1–4 Religious and spiritual 1432 .57 – 0–1 Religious but not spiritual 1432 .05 – 0–1 Spiritual but not religious 1432 .17 – 0–1 Neither religious nor spiritual 1432 .21 – 0–1 Raised no religion 1469 .12 – 0–1 Raised conservative Protestant 1469 .30 – 0–1 Raised Catholic 1469 .29 – 0–1 Raised Mormon 1469 .03 – 0–1 Raised mainline Protestant 1469 .16 – 0–1 Raised other religion 1469 .09 – 0–1 Peak attendance in childhood 1504 3.14 1.13 1–4 Peak prayer in childhood 1510 2.98 1.16 1–4 Female 1510 .58 – 0–1 Nonwhite 1510 .17 – 0–1 Age 1504 28.77 4.61 18–35 Married 1504 .57 – 0–1 Cohabiting 1504 .05 – 0–1 Divorced or separated 1504 .06 – 0–1 Education 1502 5.02 1.42 1–7 Household income 1434 2.87 1.20 1–5 Table options Turning to the young adults’ religious and spiritual variables that will be treated as outcomes in this study, we find that the average respondent attends religious services roughly once a month (2.51 on a 1–4 metric) and for many respondents, prayer is a regular part of their lives (mean of 3.17 on a 1–4 metric). In general, religious disinterest seems relatively low in the overall sample of young adults (4.22 on a 2–8 metric), suggesting that many respondents believe that religion is a source of existential meaning and ultimate truth. Levels of skepticism regarding parents’ religiousness are relatively low. Respondents are disinclined to believe that they are “more religious than [their] mother ever was” (mean of 1.80 on a 1–4 metric) or to doubt the sincerity of their mother’s religion (mean of 1.49 on a 1–4 metric). Skepticism toward paternal religiousness tends to be somewhat higher; on (mean scores of 2.20 and 1.69, respectively). The average respondent tends to report high levels of positive experience with God (3.29 on a 1–4 metric) and minimal negative experiences of God (3.48 on a 1–4 metric. Positive images of God (e.g., loving, forgiving, and all-powerful) are widespread in the NSMS sample (mean score of 3.59 on a 1–4 metric), while images of God as “angry” are much less common (1.56 on a 1–4 metric). Finally, a majority of respondents (57%) report being both religious and spiritual; only 5% are religious but not spiritual, while 17% identify as spiritual but not religious, and the remaining 21% are neither religious nor spiritual. With regard to childhood religious upbringing, the largest segment of NSMS respondents were raised in conservative Protestant faiths (30%), followed by Catholic (29%), mainline Protestant (16%), no religion (12%), other faiths (9%), and Mormon (3%). Levels of childhood religious attendance (3.14 on a 1–4 metric) and prayer (2.98 on a 1–4 metric) were relatively high. The average age of NSMS respondents is 28.77; majorities are female (58%) and non-Hispanic white (83%). Most respondents (60%) were married, while only small percentages were cohabiting (5%) or divorced or separated (6%) at the time of the NSMS survey. Nearly three in five respondents (57%) have children. According to the figures in Table 1, the average respondent has completed some college but lacks a four-year (BA or BS) degree, and has an annual household income of roughly $40K. Table 2, Table 3, Table 4 and Table 5 present the results of multivariate ordered logistic regression models estimating the net effects of parental divorce and marital quality on various facets of young adults’ religiousness. In the models displayed in the table, five of the family background (divorce/discord) groups – each of which is identified with a dummy variable – are compared with a reference category. Two sets of analyses are performed for each religious/spiritual outcome variable. In one set of analyses, the reference category consists of young adults from intact, high-conflict families, which are presumably the most problematic of the intact family groupings. In the second set of analyses, the comparison group consists of young adults from divorced families in which levels of conflict among parents were consistently low, both before and after the split. To conserve space, the parameter estimates for covariates are not tabled, and their estimated net effects are not discussed in the text, but further information is available from the authors. Table 2. Estimated net effects of parental divorce and marital conflict on young adults attendance, prayer, and religious disinterest: ordered logistic regression. ATTEND PRAYER NOEXIS Intact, happy low-conflict 1.58⁎⁎,aaa 1.25 .54⁎⁎⁎,aa Intact, unhappy, low-conflict 1.27a 1.33 .83 Divorce, high-conflict before and after .91 1.37 .87 Divorce, high-conflict before only .68+ 1.10 1.25a+ Divorce, low-conflict before and after .81 1.37 .89 Divorce, low-conflict before only 1.18 1.45 1.06 Intact, unhappy, high-conflict [Ref.] 1.00 1.00 1.00 N 1341 1332 1319 Notes: Cell entries are odds ratios. ATTEND = frequency of current attendance; PRAYER = frequency of current prayer; and NOEXIS = disinterest in religion. All models control for: female, age, nonwhite, raised no religion, raised conservative Protestant, raised Catholic, raised LDS, raised other religion, peak attendance in childhood, peak prayer frequency in childhood, married, cohabiting, divorced or separated, children, education, and household income. ⁎⁎⁎ p < .001, difference from unhappy high conflict marriages. ⁎⁎ p < .01, difference from unhappy high conflict marriages. + p < .05, difference from unhappy high conflict marriages. aaa p < .001, difference from divorce, low-conflict before and after. aa p < .01, difference from divorce, low-conflict before and after. a p < .05, difference from divorce, low-conflict before and after. a+ p < .10, difference from divorce, low-conflict before and after. Table options Table 3. Estimated net effects of parental divorce and marital conflict on young adults perceptions of parental religious sincerity and relative religiousness: ordered logistic regression. MOREMA MOREPA MADBT PADBT Intact, happy, low-conflict 0.48⁎⁎⁎,aaa 0.35⁎⁎⁎,aaa 0.25⁎⁎⁎,aaa 0.21⁎⁎⁎,aaa Intact, unhappy, low-conflict 1.00 0.86 0.65+ 0.61⁎ Divorce, high-conflict before and after 2.09⁎⁎⁎,aaa 1.61⁎,aa 1.45+,aaa 1.62⁎,aaa Divorce, high-conflict before only 1.31 1.62⁎,a 1.05aa 1.27a Divorce, low-conflict before and after 1.01 1.01 0.58⁎ 0.79 Divorce, low-conflict before only 1.53+ 1.18 1.19aa 1.15 Intact, unhappy high-conflict [Ref.] 1.00 1.00 1.00a 1.00 N 1325 1318 1324 1318 Notes: Cell entries are odds ratios. MOREMA = R more religious than mother ever was; MOREPA = R more religious than father ever was; MADBT = R doubts mother’s religious sincerity; and PADBT = R doubts father’s religious sincerity. All models control for: female, age, nonwhite, raised no religion, raised conservative Protestant, raised Catholic, raised LDS, raised other religion, peak attendance in childhood, peak prayer frequency in childhood, married, cohabiting, divorced or separated, children, education, and household income. ⁎⁎⁎ p < .001, difference from intact, unhappy, high-conflict. ⁎ p < .05, difference from intact, unhappy, high-conflict. + p < .10, difference from intact, unhappy, high-conflict. aaa p < .001, difference from divorce, low-conflict before and after. aa p < .01, difference from divorce, low-conflict before and after. a p < .05, difference from divorce, low-conflict before and after. Table options Table 4. Estimated net effects of parental divorce and marital conflict on God images, God experiences, and doubts about God. POSGODc ANGRYb POSGODEXc NODOUBTc Intact, happy, low-conflict .058+ .63+ .057+ .129⁎⁎⁎ Intact, unhappy, low-conflict .079⁎ .72 .061⁎ .058+ Divorce, high-conflict before and after .031 1.04 .053 .030+ Divorce, high-conflict before only .052+ .91 .040 .010a Divorce, low-conflict before and after .043 .65+ .059+ .083⁎ Divorce, low-conflict before only .042 1.03 .046 .069⁎ Intact, unhappy, high-conflict [Ref.] – 1.00a+ –a+ –a N 1326 1308 1329 1329 Notes: “b” = cell entries are odds ratios from ordered logistic regression models, and “c” = cell entries are standardized Betas from OLS regression models. POSGOD = image of God as positive; ANGRY = image of God as angry; POSGODEX = positive experiences of God; and NODOUBT = absence of doubts about God. All models control for: female, age, nonwhite, raised no religion, raised conservative Protestant, raised Catholic, raised LDS, raised other religion, peak attendance in childhood, peak prayer frequency in childhood, married, cohabiting, divorced or separated, children, education, and household income. ⁎⁎⁎ p < .001, difference from intact, unhappy, high-conflict. ⁎ p < .05, difference from intact, unhappy, high-conflict. + p < .10, difference from intact, unhappy, high-conflict. a p < .05, difference from divorce, low-conflict before and after. a+ p < .10, difference from divorce, low-conflict before and after. Table options Table 5. Estimated net effects of parental divorce and marital quality on religious and spiritual identities of young adults: multinomial logistic regression (N = 1273). RNS SNR NRNS Intact, happy, low-conflict .73 .31⁎⁎⁎,aaa .57+ Intact, unhappy, low-conflict .20+,a+ .52⁎,a .76 Divorce, high-conflict before and after .80 .72 .74 Divorce, high-conflict before only .70 1.46 1.19 Divorce, low-conflict before and after .63 .98 .69 Divorce, low-conflict before only .66 .97 .94 Intact, unhappy, high-conflict [Ref.] 1.00 1.00 1.00 Notes: Cell entries are odds ratios. RNS = religious, not spiritual; SNR = spiritual, not religious; NRNS = neither religious nor spiritual; vs. religious and spiritual. All models control for: female, age, nonwhite, raised no religion, raised conservative Protestant, raised Catholic, raised LDS, raised other religion, peak childhood attendance, peak childhood prayer frequency, married, cohabiting, divorced or separated, children, education, household income. ⁎⁎⁎ p < .001, difference from intact, unhappy, high-conflict marriage. ⁎ p < .05, difference from intact, unhappy, high-conflict. + p < .10, difference from intact, unhappy, high-conflict. aaa p < .001, difference from divorce, low-conflict before and after. a p < .05, difference from divorce, low-conflict before and after. a+ p < .10, difference from divorce, low-conflict before and after. Table options Table 2 focuses on young adults’ frequency of religious attendance and personal prayer, as well as their levels of religious disinterest. Several findings are particularly noteworthy. First, young adults from intact happy families attend religious services much more often than their counterparts from intact high-conflict families; indeed, the cumulative odds of religious attendance are more than 50% greater (OR = 1.58, p < .01) for the former group. However, few other significant differences involving offspring from intact high-conflict families emerge here. On average, young adults from intact happy families and those from intact unhappy families attend services more often than offspring from low-conflict divorces (p < .001 and p < .01, respectively). Second, levels of religious disinterest are dramatically lower among young adults from intact happy families as compared with those from intact high-conflict families (OR = .54, p < .001); a similar pattern emerges when this group is contrasted with offspring from low-conflict divorces (p < .01). Third, there is also an interesting pattern involving children from divorces in which high levels of pre-divorce parental conflict were reduced following the split: (a) Attendance levels are marginally lower for this group than for offspring from intact high-conflict marriages (OR = .68, p < .10); and (b) levels of religious disinterest are higher among this group than among children from consistently low-conflict divorces (p < .10). Fourth, these analyses reveal no effects of family background on the frequency of young adults’ current prayer activity. In Table 3, turning to young adults’ skepticism regarding their parents’ religiousness, several intriguing patterns surface. First, offspring from intact happy families are dramatically less inclined than the offspring from intact high-conflict families to: (a) doubt the sincerity of mother’s religion (OR = .25, p < .001) or father’s religion (OR = .21, p < .001); or (b) to believe that they are more religious than their mothers (OR = .48, p < .001) or their fathers (OR = .35, p < .001) ever were. Young adults from intact happy families are also much less skeptical of their parents’ religion – at least, measured with these items – than their counterparts from families with low-conflict divorces (all differences p < .01 or greater). Second, young adults who experienced high-conflict divorces growing up – i.e., those with high levels of discord during both pre- and post-divorce periods – hold much more negative views of parental religion than either of the comparison groups in our analyses. Specifically, they are more doubtful of their mother’s religious sincerity (OR = 1.45, p < .10) and their father’s religious sincerity (OR = 1.62, p < .05), and they tend to believe that they are more religious than their mother ever was (OR = 2.09, p < .001) and more religious than their father ever was (OR = 1.61, p < .05). These young adults from high-conflict divorces are also much more skeptical about parents’ religiousness than their counterparts from consistently non-conflictual divorces (all differences p < .01 or greater). Third, offspring from families in which high pre-divorce conflict was diminished following the split are only distinctive in a few regards: (a) They are inclined to believe that they are more religious than their fathers ever were, compared with offspring from intact high-conflict families (OR = 1.62, p < .05), and with offspring from consistently low-conflict divorces (p < .05). (b) They are more doubtful about the sincerity of mother’s religion (p < .01) and father’s religion (p < .05) than offspring of consistently low-conflict divorces. Fourth, compared with their counterparts from intact high-conflict families, the offspring from divorces in which parental conflict increased from low to high levels following the split are marginally more likely (OR = 1.53, p < .10) to believe that they are more religious than their mothers ever were. They are also more doubtful of the sincerity of their mother’s religion than young adults who experienced consistently low-conflict divorces while growing up (p < .01). Table 4 displays the results of multivariate ordered logit and OLS regression models gauging the net effects of parental divorce and marital quality on young adults’ images of God and positive and negative experiences vis-a-vis God. Findings for these outcomes appear much weaker than those presented in Table 2 and Table 3. First, compared to persons raised in intact high-conflict families, those who grew up in intact happy families express somewhat more positive God images (β = .058, p < .10) and more positive experiences of God (β = .057, p < .10); however, both patterns are only marginally statistically significant. They are marginally less prone to view God as angry (OR = .63, p < .10), and they are much less prone to report doubts about God (β = .129, p < .001). None of these differences surface when young adults from intact happy families are compared with those who experienced consistently low-conflict parental divorces. Second, respondents from intact unhappy families report more positive experiences of God (β = .061, p < .05) than those from intact high-conflict families. In a marginally significant finding, they also report somewhat fewer doubts about God (β = .058, p < .10). Third, compared with respondents from intact, high-conflict families, those in which there were high levels of pre-divorce conflict declined following the split: (a) report fewer doubts about God (β = .083, p < .05); and (b) are marginally less prone to experience God as angry (OR = .65, p < .10). Fourth, offspring from consistently low-conflict divorces report fewer doubts about God (β = .083, p < .05), and marginally more positive experiences of God (β = .059, p < .10), than those from intact high-conflict families, although these groups do not differ on the other outcomes in Table 4. Other interesting differences surface among the offspring of divorce themselves. For example, the offspring from consistently high-conflict divorces are more inclined to view God as angry than those from consistently low-conflict divorces (p < .05), and also marginally more prone to report more doubts about God (p < .10). In addition, compared to those young adults who experienced consistently low-conflict divorces, their counterparts whose parents exhibited low levels of conflict prior to the split but elevated conflict in the post-divorce period are marginally more prone to experience God as angry (p < .10), while those from divorces with high pre-divorce conflict and reduced post-divorce conflict tended to report more doubts about God (p < .05). Finally, in Table 5 we present the results of a multinomial logistic regression model, estimating the net effects of parental divorce and marital quality on the relative odds of identifying as (a) religious but not spiritual, (b) spiritual but not religious, or (c) neither religious nor spiritual, as opposed to both religious and spiritual, which is the omitted category in this model. These results are quite straightforward. First, the relative odds of identifying as “spiritual but not religious” are an estimated 69% lower (OR = .31, p < .001) for young adults from intact happy families as compared to those from intact high-conflict families. A similar, albeit more muted, pattern is observed for offspring from intact but unhappy families (OR = .52, p < .05). These two groups are also significantly different (p < .001 and p < .05, respectively) from young adults who experienced consistently low-conflict divorces while growing up. Second, in a pattern that is marginally significant, the relative odds of identifying as “neither religious nor spiritual” are 43% lower (OR = .57, p < .10) for respondents from intact happy families, as compared with those from intact high-conflict families. Third, respondents raised in intact but unhappy families are especially disinclined to identify as “religious but not spiritual” (OR = .20, p < .05). Finally, compared to offspring from consistently low-conflict divorce, those who experienced high pre-divorce parental conflict followed by reduced post-divorce conflict are marginally more inclined to opt for “neither religious nor spiritual” identity (p < .10). These are the only significant family background effects that surface for this outcome.