Are second and third marriages more likely less likely or equally likely to end in divorce?

J Marriage Fam. Author manuscript; available in PMC 2014 Apr 1.

Published in final edited form as:

J Marriage Fam. 2013 Apr 1; 75(2): 276–287.

Published online 2013 Mar 14. doi: 10.1111/jomf.12008

PMCID: PMC3636559

NIHMSID: NIHMS429454

Abstract

A random multistate sample of married individuals (N = 1,931) was used to explore whether more positive attitudes toward divorce and weaker commitment to marriage may contribute to the greater instability of remarriages than first marriages. Remarried adults, whether or not they brought children from a previous union into the remarriage, reported marital quality (happiness and conflict) equal to those in first marriages. They also reported more positive attitudes toward divorce, which were associated with higher divorce proneness (i.e., thinking about and taking actions toward divorce). Marriage type interacted with marital quality to predict divorce proneness, such that the association between low marital quality and divorce proneness was stronger for remarried individuals than for those in first marriages. This suggests that remarried adults may be more likely than adults in first marriages to take steps toward divorce when experiencing marital distress, possibly reflecting a weaker commitment to marriage.

Keywords: divorce, marital commitment, marital conflict, marital stability, relationship commitment, remarriage

In the past 25 years, remarriage has become increasingly common. In 2000 – 2001, remarriages represented around 29% of existing marriages (Amato, Johnson, Booth, & Rogers, 2003) and 38% of new marriages (Kreider, 2005). Many adults who remarry have children from previous unions; in 2000, 11% of married couples had a stepchild, defined as a child from a previous union of either spouse, living with them (Amato, Booth, Johnson, & Rogers, 2007). Unfortunately, couples who remarry are at relatively higher risk for divorce than those in first marriages (e.g., Bulanda & Brown, 2007). In contrast to 33% of first marriages, 39% of remarriages end in the first 10 years (Bramlett & Mosher, 2001). Although most studies do not differentiate between remarrying couples who bring children to the marriage and those who do not, there is some evidence that children from previous unions may further increase risk for divorce, especially for women (Bramlett & Mosher, 2002; Teachman, 2008; White & Booth, 1985).

It is somewhat surprising, then, that remarriages are not generally characterized by poorer relationship quality than first marriages. Research indicates that remarried individuals and those in first marriages do not differ on important indictors of marital quality, such as happiness, interaction, disagreements, or problems (Amato et al., 2007; Bulanda & Brown, 2007), or on long-term trajectories of marital happiness (Kamp Dush, Taylor, & Kroeger, 2008). Although the presence of children from previous unions may predict lower remarital satisfaction (Bir-Akturk & Fisiloglu, 2009; White & Booth, 1985), and more disagreements (Stewart, 2005), these challenges do not appear to result in lower average relationship quality in remarriages than first marriages. Thus, existing research suggests that the elevated risk of divorce in remarriages is not a consequence of relative marital unhappiness or high levels of marital problems.

Instead, it is possible that the increased rates of divorce in remarriage may be explained by differences between first marriages and remarriages in attitudes toward divorce, defined as perceived acceptability of divorce as a solution to marital distress, and commitment to marriage, defined as the tendency to stay in a marriage even when encountering marital difficulties (cf. Amato & DeBoer, 2001). According to social psychological theories, the experience of divorce from a previous spouse is likely to influence attitudes toward divorce, shifting them toward greater acceptance. Cognitive dissonance theory (Festinger, 1957) states that dissonance arises if an individual behaves in a way that is discrepant with his or her attitudes. Because dissonance is inherently uncomfortable, the individual is motivated to reduce it, which can be accomplished by changing the attitude to be more consistent with behavior. In the case of divorce, a person who does not view divorce as acceptable but divorces his or her spouse may reduce the resulting dissonance by coming to view divorce more favorably. Self-perception theory (Bem, 1972) similarly suggests that people often determine what their attitudes are by examining their own behaviors. According to this theory, individuals who divorce may evaluate their divorce attitudes on the basis of their own behavior of divorcing and conclude that they are generally in favor of it. Consistent with these theories, individuals who have remarried following divorce reported more positive attitudes toward divorce (i.e., greater endorsement of divorce as an acceptable solution to marital distress) than those in first marriages (Booth & Edwards, 1992; White & Booth, 1985). Furthermore, longitudinal research has shown that the experience of divorce predicts increases in prodivorce attitudes within individuals (Amato & Booth, 1991; Thornton, 1985).

Similarly, we theorized that having divorced a previous spouse may decrease the extent to which remarried individuals are committed to their new marriage. Because they have experienced firsthand that marriage did not last forever, that they were able to make it through a divorce, and that they were able to form a new relationship after ending their first marriage, remarried people may be more willing and more likely to choose divorce as a way to handle problems in their remarriage. In other words, they may be less committed to their marriages. According to commitment theory, individuals with a strong commitment to marriage view couple problems as solvable, believe that they can and should work to solve them, and generally behave in ways that promote marital health and longevity (e.g., Stanley, Rhoades, & Whitton, 2010). In contrast, those with low marital commitment may be less optimistic that marital issues can be fixed and therefore less likely to remain in a marriage that is not going well (Amato & DeBoer, 2001). It is possible, then, that remarried individuals may be more likely to divorce than those in first marriages because, if they encounter marital conflict or unhappiness, they are more likely to take steps toward ending the marriage.

Findings regarding the intergenerational transmission of divorce provide some support for this hypothesis. Compared to adults with nondivorced parents, those with divorced parents report not only more positive general attitudes toward divorce (Segrin, Taylor, & Altman, 2005) but also less commitment to their own marriages (Whitton, Rhoades, Stanley, & Markman, 2008), and they are more likely to consider divorce themselves if their marital satisfaction is low (e.g., Amato & DeBoer, 2001). These findings suggest that personal observation of divorce as a response to marital problems — by one’s parents — may lower commitment to staying married in the face of marital distress. It is possible that, similarly, experiencing one’s own divorce may be linked with less commitment to staying in a marriage when problems arise. Indeed, previous research has found that remarried individuals report more “willingness to leave” a marriage, as indicated by less endorsement of the belief that divorce is morally wrong and a greater perceived ability to financially and emotionally cope with a future divorce, than do individuals in first marriages (White & Booth, 1985). However, we are aware of no studies that have tested whether remarried adults are more likely to have taken steps toward divorce if they are experiencing marital dissatisfaction, which would be a strong indicator of lower commitment to their current marriage. In addition, previous research has generally not differentiated between remarried adults with versus without children from a previous relationship. Yet, because conflict over nonshared children is common in remarriages (Stanley, Markman, & Whitton, 2002), and because some remarried parents may feel a stronger allegiance to their children than to their new spouse (e.g., Anderson & Greene, 2011), it is possible that adults who brought children from previous relationships to their remarriage may be less committed to their relationships than those who did not.

In sum, we propose that individuals in remarriages may have higher divorce rates not because their relationships are of lower quality but partly because they have more positive attitudes toward divorce and lower commitment to marriage; that is, remarried couples, especially those who have brought children to the remarriage, may tolerate less marital unhappiness or conflict before taking steps toward divorce. We put forth several specific hypotheses. The first three, which aimed to replicate previous findings, concerned main effects of remarriage on marital quality, attitudes about divorce, and divorce proneness (i.e., the extent to which individuals have thought about or taken steps towards divorcing their spouse):

Hypothesis 1: Individuals in remarriages, with or without children from a previous relationship, will not report lower marital quality (in terms of happiness or marital conflict) than those in first marriages.

Hypothesis 2: Individuals in remarriages, with or without children from a previous relationship, will report more positive attitudes toward divorce (i.e., greater acceptance of divorce as an acceptable solution to marital distress) and higher divorce proneness than those in first marriages.

Hypothesis 3: Positive attitudes toward divorce will be associated with more divorce proneness, partially accounting for associations between remarriage and divorce proneness.

Extending previous research, we also predicted that the influence of marriage type and attitudes toward divorce on divorce proneness would differ by level of marital quality:

Hypothesis 4: Marital quality will moderate the association between marriage type and divorce proneness, such that being remarried (with or without children from a previous relationship) versus in a first marriage will be linked with higher levels of divorce proneness more strongly among individuals who report poorer marital quality (i.e., less marital happiness or higher marital conflict). This would indicate a lower commitment to marriage among remarried than first-married adults.

Hypothesis 5: Similarly, holding positive attitudes toward divorce will predict current divorce proneness more strongly among couples who report lower marital quality.

Finally, we hypothesized that positive attitudes toward divorce and marriage type would interact in the prediction of divorce proneness:

Hypothesis 6: Among individuals who have previously divorced, we did not expect to see greater divorce proneness when faced with marital unhappiness if they do not have the risk factor of more positive attitudes toward divorce. Therefore, we hypothesized that being remarried would be associated with more divorce proneness among less happily married individuals in the presence of more positive attitudes toward divorce, but not in the presence of more negative attitudes toward divorce.

We tested these hypotheses in a sample of individuals living in the southeastern and central United States, randomly sampled for telephone interviews in 2001 (Johnson et al., 2002). The sample was sufficiently large to provide adequate power for tests of moderation and allowed for statistical control of several selection variables that might be associated with both marriage type and the relationship variables of marital quality and divorce proneness (i.e., gender, age, education, race, marriage length, number of children in the home, and religiosity).

Method

Participants

Participants were recruited as part of the Oklahoma Marriage Initiative Statewide Baseline Survey (Johnson et al., 2002). In 2001, interviewers used random-digit dialing methods to contact households in Oklahoma and three surrounding states. In addition, to ensure adequate inclusion of individuals with low income levels, interviewers called households on a list provided by the Oklahoma Department of Human Services of people receiving Medicaid due to financial need. Among the 5,766 adults contacted, 58% agreed to complete the telephone survey, yielding a sample of 3,344 individuals (3,023 recruited by random-digit dialing and 303 via the Medicaid list; 2,323 living in Oklahoma and 1,021 living in Arkansas, Kansas, and Texas). Only one adult (≥18 years old) was interviewed per household. Women were more likely than men to answer the telephone, likely because calls were made during the day. Consequently, the overall sample included more women (70%) than men (30%). For the present analyses, we selected participants who were currently married (n = 1,977; 59%). Because our hypotheses centered around the experience of a previous divorce, we excluded the small number of individuals in a second marriage who did not report a past divorce (and presumably were widowed; n = 46; 2%), leaving a final sample of 1,931. Of these, 1,379 (71%) were in their first marriage, and 552 (29%) were in remarriages (i.e., their second or higher order marriage following a divorce). Of the 552 remarried participants, 425 (77%) remarried with at least one child from a previous relationship. Demographic characteristics are presented in Table 1, by marriage type.

Table 1

Means and Standard Deviations of All Variables by Marriage Type (N = 1,931)

Measures

Attitudes toward divorce were measured with three items tapping the belief that divorce is an acceptable solution for marital distress. Participants rated their level of agreement with three statements: (a) “When married people realize that they no longer love each other, they should get a divorce even if they have children”; (b) “Sure, divorce is bad, but a lousy marriage is even worse”; and (c) “When there are children in the family, parents should stay married even if they do not get along” on a scale that ranged from 1 (strongly agree) to 5 (strongly disagree). Items, (re)coded so higher scores indicated greater acceptance of divorce, were averaged. Internal consistency was good (α = .70).

Divorce proneness was assessed with items from the Marital Instability Index (Booth, Johnson, & Edwards, 1983), which was designed to assess instability among currently intact couples. Participants responded to five questions: (a) “Have you ever thought your marriage might be in trouble?”, (b) “Has the thought of getting a separation or divorce crossed your mind?”, (c) “Have you discussed divorce or separation from your spouse with a close friend?”, (d) “Have you or your spouse ever seriously suggested the idea of divorce?”, and (e) “Have you and your spouse talked about consulting an attorney regarding a possible divorce or separation?” on a 6-point Likert-type scale (1 = “never”; 2 = “yes, but not within the last 3 years”; 3 = “yes, within the last 3 years”; 4 = “yes, within the last year”; 5 = “yes, within the last 6 months”; 6 = “yes, within the last 3 months”). Higher scores represent greater instability and are highly predictive of marital dissolution (Booth, Johnson, White, & Edwards, 1985). In this sample, internal consistency was excellent (α = .85).

Marital quality variables

We assessed two important indices of marital quality: (a) marital happiness and (b) marital conflict. Marital happiness was based on one item: “Taking things altogether, how would you describe your marriage?” (1 = “not too happy,” 2 = “pretty happy,” 3 = “very happy”). Consistent with most previous research using this scale, we converted it into a dichotomous variable (1 = very happy, 0 = pretty happy or not too happy). Because reports of marital happiness tend to be very positive, with the majority of people choosing the “very happy” option, people who choose “pretty happy” are admitting that their marriages are less than ideal and are thought to have more in common with the “not very happy” group, which tends to be smaller than 3% of the sample (Glenn, 1998; 3.1% in the current sample). This dichotomous measure is highly discriminating of other outcomes, including global well-being (Glenn, 1998). As a check of this measure’s validity, we also collected another measure, of marital satisfaction, based on two items (α = .78) drawn from the General Social Survey: (a) “All in all, how satisfied are you with your marriage?” (1 = completely satisfied, 5 = not at all satisfied; reversed) and (b) “Taking things altogether, how would you describe your marriage”? (1 = not too happy, 2 = pretty happy, 3 = very happy). The two measures were highly correlated (r = .83, p < .001) and all analyses, when re-run using the continuous satisfaction measure, yielded identical findings. Therefore, only results using the dichotomous marital happiness measure are presented, for consistency with previous research and for ease of interpretation.

Marital conflict was assessed with a four-item measure that has previously demonstrated evidence of reliability and validity (e.g., Stanley et al., 2002). Participants rated the frequency of four destructive conflict patterns on a 3-point scale (1 = never or almost never, 2 = once in awhile, 3 = frequently): (a) “Little arguments escalate into ugly fights with accusations, criticisms, name-calling, or bringing up past hurts”; (b) “My partner criticizes or belittles my opinions, feelings, or desires”; (c) “My partner seems to view my words or actions more negatively than I mean them to be”; (d) and “When we argue, one of us withdraws … that is, doesn’t want to talk about it anymore; or leaves the scene” (α = .76). Higher scores indicate greater negative conflict. In this sample, marital happiness and conflict showed a moderate negative association (r = −.44, p < .001), supporting their validity and the utility of examining them separately.

Marriage type

All participants were currently married. On the basis of participant responses to two questions — “Have you ever been divorced?” and “Do you have a child from a prior marriage or prior relationship?” — we created two dummy-coded variables to represent marriage type: (a) remarried without previous children (i.e., previously divorced and has no children from prior relationships) and (b) remarried with previous children (i.e., previously divorced and has at least one child from a prior relationship). Never-divorced served as the omitted comparison group.

Control variables

We controlled for the standard variables of gender (0 = female, 1 = male), age at interview, education (1 = less than high school, 2 = high school graduate, 3 = some college, 4 = college graduate, 5 = postgraduate degree), race (coded as two dummy variables representing Black and Hispanic, with White serving as the omitted comparison group), number of children currently living in the home, religiosity (“All things considered, how religious would you say that you are?”: 1 = not at all religious, 2= slightly religious, 3 = moderately religious, 4 = very religious), and marriage length (in years). Sample descriptives on all control variables are displayed by marriage type in the top half of Table 1.

Results

Very few (2.0%) of cases were missing data on one or more of the variables included in analyses. Twenty-four participants (1.2%) had missing data for religiosity; all other items were missing less than 1%. Because so few cases were missing any data, we elected to use listwise deletion (see Allison, 2002).

To test Hypotheses 1 and 2, we compared individuals on marital happiness, marital conflict, divorce acceptance, and divorce proneness by marriage type (first marriages vs. remarriages without children from prior relationships vs. remarriages with children from prior unions), first using a one-way analysis of variance and then using an analysis of covariance including all control variables as covariates. The bottom half of Table 1 displays uncontrolled means and F values from the analyses of variance in plain text and adjusted means and F values from analyses of covariance in italics. For F tests, dfnumerator = 2, and dfdenominator varied from 1,904 to 1,928, depending on missing data on individual variables. Planned comparisons were used to test for statistically reliable differences between groups. Cohen’s d effect sizes, which describe the difference between groups in standard deviation units, were used to interpret the magnitude of differences. In survey data, effect sizes <0.2 are interpreted as weak, those from 0.20 through 0.39 as moderate, and ≥0.4 as strong (Amato et al., 2007).

The results displayed in Table 1 show that, consistent with the hypothesis that remarried individuals would not report poorer marital quality than those in first marriages, marital happiness and marital conflict did not differ by marriage type, whether or not the control variables were included as covariates in the model. Also as expected, individuals in first marriages reported less positive attitudes toward divorce than did those in remarriages without children from a previous union, t(2,014) = 3.29, p < .01 (d = 0.30; moderate effect) and those in remarriages with children from previous relationships, t(2,014) = 8.26, p < .001 (d = 0.46; strong effect). Inclusion of controls did not alter the statistical significance or effect sizes of differences in divorce attitudes between participants in first marriages and remarriages of both types (ps < .01; ds = 0.24 for those without previous children and .42 for those with previous children). In uncontrolled analyses, divorce proneness differed by marriage type; first-married participants reported less divorce proneness than participants in remarriages without prior children, t(1,928) = 2.28, p < .05 (d = 0.24; moderate effect) but did not differ in divorce proneness from those who remarried after having children in a previous relationship, t(1,928) = 1.80, p = .07. The differences in divorce proneness by marriage type were not significant in models that included statistical controls. (This finding regarding the main effect of marriage type on divorce proneness should be interpreted in the context of the interaction between marriage type and marital quality, presented below.)

Next, we used hierarchical linear regression analyses to test the remaining hypotheses. We conducted separate sets of regressions using the two different indicators of marital quality (happiness and conflict) as predictors. In all models, we included all statistical controls and the two dummy-coded variables representing remarriage without previous children and remarriage with previous children, with first marriages representing the comparison group. For each set of regressions, Model 1 included only main effects of marriage type, the marital quality variable, and attitudes toward divorce. In this model, a significant coefficient for divorce attitudes would support Hypothesis 3 (that divorce attitudes would predict divorce proneness, controlling for selection factors and marital quality). In Model 2, we added all two-way interaction terms involving marriage type, divorce attitudes, and the marital quality variable (i.e., either happiness or conflict). Significant interactions between the dummy-coded variables for marriage type and the marital quality variables would support Hypotheses 4 and 5. Specifically, negative coefficients for the interactions between the remarriage variables and happiness, and positive coefficients for the interactions between remarriage and conflict, would indicate that poorer marital quality is associated more strongly with divorce proneness in remarried than in first-married individuals. Similarly, a significant interaction between divorce attitudes and the marital quality variables would support the hypothesis that positive attitudes toward divorce are associated more strongly with divorce proneness when marital quality is lower, versus higher. Model 2 also included two three-way interactions (between remarriage without previous children, attitudes toward divorce, and marital quality and between remarriage with previous children, attitudes toward divorce, and marital quality), to test whether being in a remarriage of either type predicted greater divorce proneness when faced with poorer marital quality more strongly if attitudes toward divorce are more positive (Hypothesis 6).

The results are presented in Table 2. As expected, in Model 1 more positive attitudes toward divorce were associated with greater divorce proneness, controlling for demographic variables, marriage length, marriage type, and the marital quality variable. However, including attitudes toward divorce in the equation did not appreciably reduce the association between marriage type and divorce proneness as would be expected if it served as a mediator. Consistent with Hypothesis 4, that remarried individuals may have relatively weak commitment to marriage, there were significant two-way interactions between each marital quality variable (happiness and conflict) and both remarriage types (with or without previous children). These findings indicate that being in a remarriage was associated with greater divorce proneness more strongly among individuals who reported lower, versus higher, marital quality. In other words, among participants who reported being less than very happy with their marriage and who reported higher marital conflict, those in remarriages had higher divorce proneness than those in first marriages.

Table 2

Summary of Hierarchical Regression Models Predicting Divorce Proneness From Marital Quality (Marital Happiness or Marital Conflict) by Marriage Type

The findings regarding associations between attitudes toward divorce and divorce proneness were somewhat complex. The significant three-way interactions among remarriage with previous children, divorce attitudes, and each of the marital quality variables (happiness and conflict) indicate that being in a remarriage to which one brought children from a previous relationship was associated with greater divorce proneness when faced with low marital quality more strongly in the presence of more, versus less, positive attitudes toward divorce. In contrast, the nonsignificant coefficients for the two-way interactions between divorce attitudes and the marital quality variables indicate that holding more positive attitudes did not strengthen the association between marital quality and divorce proneness among first-married and remarried participants with no previous children. (This finding applies only to these two groups because of the coding of variables, in which first marriage was the comparison group, and the nonsignificant interactions between the remarriage without children variable and the marital quality variables).

To help elucidate these complex findings, in Table 3 we report mean standardized divorce proneness scores by marriage type, positive versus negative attitudes toward divorce (i.e., those above and below the mean for divorce attitudes, respectively), and marital quality, adjusted for all control variables. Estimated group means from analyses in which happiness was used as the index of marital quality are presented in the top panel; those in which marital conflict was used are presented in the bottom panel. Because the divorce-proneness scores are standardized, they can be interpreted as the number of standard deviations that the average score for the given group is above or below the full sample mean. Values in the top panel demonstrate a clear main effect of marital happiness, with individuals in very happy marriages showing lower divorce proneness (displayed in the top row) than those in less than very happy marriages (displayed in the bottom row). Among participants who were happy in their marriages, the level of divorce proneness was fairly constant and low (0.20 – 0.37 SD below the mean) across the three types of marriage and across groups with positive and negative attitudes toward divorce. In contrast, among participants who reported being less than very happy, being remarried — with or without children from previous unions — was associated with higher divorce proneness (Ms = 0.65 – 1.10) than was first marriage (Ms = 0.31 – 0.51). Values in the bottom panel of Table 3 show that, in a parallel fashion, there was a clear main effect of marital conflict; participants in marriages with lower than average conflict consistently reported lower divorce proneness (0.22 – 0.48 SD below the mean) than those in high-conflict marriages. Among participants who reported higher than average marital conflict, however, remarriage was associated with higher divorce proneness (Ms = 0.41 – 0.83) than was first marriage (Ms = 0.24 – 0.39). This pattern of findings, reflected in the significant two-way interactions between each type of remarriage and the marital quality variables, supports the hypothesis that remarried individuals are more likely to consider, and take steps toward, divorce when experiencing some marital problems than are those in first marriages.

Table 3

Standardized Divorce Proneness Scores by Marriage Type, Attitudes Toward Divorce, and Either Marital Happiness (Top Panel) or Marital Conflict (Bottom Panel)

Contrary to what we predicted, once all statistical controls were in place, attitudes toward divorce did not predict divorce proneness among individuals in first marriages and remarriages that did not involve children from previous relationships. Within each of these two groups, being less than very happy and reporting higher than average marital conflict were each associated with similarly high levels of divorce proneness for those with positive and those with negative attitudes toward divorce. In contrast, the divorce proneness of less than very happily married individuals who brought children to their remarriage differed according to their attitudes toward divorce (reflecting the three-way interaction among being remarried with children from a previous relationship, divorce attitudes, and marital happiness); those who had more positive divorce attitudes showed more divorce proneness (M =1 .10) than those with more negative divorce attitudes (M = 0.65). This same pattern was observed when marital conflict was examined as the index of marital quality: Among participants in high-conflict remarriages with children from previous unions, those who had more positive divorce attitudes showed more divorce proneness (M = 0.67) than those with more negative divorce attitudes (M = 0.41).

Next, we conducted follow-up analyses to help elucidate the unexpected finding that attitudes toward divorce predicted divorce proneness among relatively unhappily remarried individuals only if they had children from a previous relationship. We re-ran the regression analyses twice: first using only the two attitudes-toward-divorce items that specifically asked about the acceptability of divorce when children are involved and then using only the one item that did not mention children. We had two hypotheses: that (a) the child-related items might be more strongly predictive of divorce proneness among unhappily remarried adults with children who had already experienced a family transition than those who did not and (b) the non-child item might interact with the marital quality of remarried individuals who did not have children from a previous union, whereas the full divorce attitudes measure did not. However, the pattern of findings using these two alternate divorce-acceptance measures did not differ in any way from analyses that used the full measure.

Finally, for exploratory purposes, we checked for gender differences in the associations of marriage type and divorce attitudes with divorce proneness. There were no significant two-way interactions between gender and any of the predictor variables, and no three-way interactions between gender and the two-way interaction terms (e.g., remarriage × marital quality). This suggests that the observed associations, including the moderating effects of marital quality level, did not differ for men and women.

Conclusion

Overall, the results of this study suggest that the higher rates of divorce seen in remarriages than in first marriages may be partially explained by differences in attitudes toward divorce and marital commitment, reflected by the tendency to take steps toward divorce when experiencing lower marital quality. Replicating previous findings in other data sets (e.g., Amato et al., 2003, Bulanda & Brown, 2007), we found that remarriages were no less happy and no more conflicted than were first marriages. Extending that work, we also found that having a child from a previous union was not associated with lower happiness or more conflict in remarriage. Together, these findings suggest that poorer marital quality does not explain the heightened risk for divorce in remarriages versus first marriages (e.g., Bramlett & Mosher, 2001), or that in remarriages with versus without children from previous unions (e.g., Teachman, 2008).

Individuals in remarriages did, however, report more positive attitudes toward divorce than those in first marriages, replicating findings from data collected in the 1980s (Booth & Edwards, 1992; White & Booth, 1985). In line with cognitive dissonance and self-perception theories, which posit that individuals’ behaviors influence their attitudes (Bem, 1972; Festinger, 1957), these findings suggest that the experience of divorcing a previous spouse may lead remarried individuals to hold more accepting attitudes toward divorce as a solution to marital distress. This is important because, consistent with previous studies (Amato, 1996; Booth & Edwards, 1992), we found that greater endorsement of divorce as a solution to marital distress was associated with higher divorce proneness, even when controlling for demographic factors that might account for a selection effect and for marital quality, a robust predictor of divorce proneness. Therefore, the current findings support hypotheses that the relatively positive attitudes toward divorce among adults in remarriages may raise their risk for marital disruption.

Novel to the current study is the finding that remarried individuals reported greater divorce proneness than individuals in first marriages, but only when in lower quality marriages. Specifically, remarried individuals who were very happy with their marriage, and those who reported lower than average marital conflict, did not show more divorce proneness than those in high-quality first marriages. In contrast, among participants in lower quality marriages (i.e., those who were less than very happy with their marriages and who reported higher than average marital conflict), remarried adults did report greater divorce proneness than those in first marriages. This pattern of findings is consistent with the perspective that remarriages may be characterized by relatively weak commitment to marriage; compared to those in first marriages, remarried adults appear to be more likely to take steps toward divorce when they encounter marital problems. Given that scores on this divorce-proneness measure are highly predictive of marital dissolution (Booth et al., 1985), remarried individuals may also be more likely to actually divorce their partners in response to low marital quality.

An interesting and somewhat unexpected finding was that having more negative attitudes toward divorce (i.e., viewing divorce as a less acceptable response to marital discord) buffered the effects of low marital quality on divorce proneness, but only among people who brought children from a previous relationship into a remarriage. Holding negative views about divorce may inhibit divorce proneness in these couples because parents worry about the potential effects of divorce, and potentially a second or third divorce, on their children. Conversely, remarried parents who hold relatively positive attitudes about divorce, which includes viewing divorce as not particularly harmful to children, may be more likely to take steps to dissolve low-quality marriages because they want to protect their children from exposure to marital conflict. Given the significant complexities and challenges involved in forming and maintaining stepfamilies (e.g., Bray & Kelly, 1998), which this data set was not able to address, future research is needed to explore more deeply how nonshared children (from one’s own and one’s partner’s previous relationships) are associated with remarital quality and stability.

Several limitations of the study should be noted. First, the data were cross-sectional, prohibiting conclusions about the direction of effects. We focused on the conditions under which marital quality may predict divorce proneness; however, it is also likely that thinking about and taking steps toward divorce may negatively affect marital quality (Amato & Rogers, 1999). Second, although we controlled for many potential preexisting differences between individuals in first marriages and remarriages (e.g., religiosity, education), it is still possible that selection effects contributed to findings. Because remarried individuals appear to more quickly divorce when encountering marital disharmony than first-married individuals, the group of people who remain in remarriages may be selected for those in higher quality relationships, reducing group differences in marital quality by marriage type. In addition, this sample of married adults, which did not include any unpartnered or cohabiting individuals, may hold more positive attitudes about marriage than the general population. In particular, adults who choose to remarry after divorce, rather than cohabit with their new partner, may be a select group that highly values the institution of marriage. To the extent that this is true, it is particularly noteworthy that we still observed more positive attitudes toward divorce and weaker marital commitment among remarried than first-married participants. Third, risk for divorce was operationalized by self-reported divorce proneness rather than actual divorce rates. Future longitudinal research is needed to directly test whether remarried adults are more likely than first-married adults to dissolve their unions when martially distressed. Fourth, only 25% to 34% of the variance in divorce proneness was accounted for by the models, suggesting that other, unmeasured factors are related to adults’ decisions to take steps toward divorce and contribute to the differences in divorce proneness between marriage types. Furthermore, although there were moderate effect sizes between first marriages and remarriages in attitudes toward divorce, the effects of marital commitment would best be considered small (accounting for only 2% additional variance beyond main effects). Finally, the sample was drawn from one region of the United States; our findings may not generalize to other regions.

Despite its limitations, this study makes an important contribution to the literature. The findings suggest that more positive attitudes toward divorce and a weaker commitment to marriage may represent mechanisms that partially explain the heightened risk for divorce in remarriages. Extending previous evidence that going through a divorce may weaken people’s commitment to the institution of marriage as a lifelong institution (i.e., general attitudes about marriage and the acceptability of divorce; Amato & Booth, 1991), the present findings suggest that it may also weaken their personal commitment to their own current marriages, as evidenced by the greater divorce proneness shown by remarried compared to first-married adults when encountering relatively low marital quality. For individuals who brought children to their remarriage, holding more positive attitudes toward divorce placed them at even greater risk for marital instability when in a marriage that was less than very happy or above average in conflict.

These findings, which together suggest that couples and families formed by remarriage are vulnerable to disruption in the face of marital dissatisfaction, have important clinical implications. Although there are certainly times when quickly exiting a bad marriage is likely best for all involved (e.g., if domestic violence is present), commitment to staying married even through rough times is generally associated with improvements in relationship quality over time (e.g., Rusbult & Buunk, 1993). For remarried individuals with children from past unions, such commitment may be particularly important to making it past the turbulent family relationships that are common during the first years of remarriage (Bray & Kelly, 1998). In this context, the present findings support targeted intervention efforts to reduce risk for divorce in couples who are remarrying, and they suggest that such efforts include program content focused on the value of commitment to staying married through periods of unhappiness or conflict.

Acknowledgments

Scott M. Stanley’s and Howard J. Markman’s work on this study was supported by grants from the Eunice Kennedy Shriver National Institute of Child Health & Human Development (R01HD047564 and R01 HD053314, respectively). The development of this data set was supported by the Oklahoma Department of Human Services and the Oklahoma Marriage Initiative. We gratefully acknowledge Paul Amato, who provided valuable advice on statistical analyses and helpful comments on a draft of this article.

References

• Allison PD. Missing data. Thousand Oaks, CA: Sage; 2002. [Google Scholar]
• Amato PR. Explaining the intergenerational transmission of divorce. Journal of Marriage and the Family. 1996;58:628– 640. [Google Scholar]
• Amato PR, Booth A. The consequences of divorce for attitudes toward divorce and gender roles. Journal of Family Issues. 1991;12:306–322. doi: 10.1177/019251391012003004. [CrossRef] [Google Scholar]
• Amato PR, Booth A, Johnson D, Rogers SJ. Alone together: How marriage in America is changing. Cambridge, MA: Harvard University Press; 2007. [Google Scholar]
• Amato PR, DeBoer DD. The transmission of marital instability across generations: Relationship skills or commitment to marriage? Journal of Marriage and Family. 2001;63:1038–1051. doi: 10.1111/j.1741-3737.2001.01038.x. [CrossRef] [Google Scholar]
• Amato PR, Johnson DR, Booth A, Rogers SJ. Continuity and change in marital quality between 1980 and 2000. Journal of Marriage and Family. 2003;65:1–22. doi: 10.1111/j.1741-3737.2003.00001.x. [CrossRef] [Google Scholar]
• Amato PR, Rogers SJ. Do attitudes toward divorce affect marital quality? Journal of Family Issues. 1999;20:69–86. doi: 10.1177/019251399020001004. [CrossRef] [Google Scholar]
• Anderson ER, Greene SM. “My child and I are a package deal”: Balancing adult and child concerns in repartnering after divorce. Journal of Family Psychology. 2011;25:741–750. doi: 10.1037/a0024620. [PubMed] [CrossRef] [Google Scholar]
• Bem DJ. Self-perception theory. In: Berkowitz L, editor. Advances in experimental social psychology. Vol. 6. San Diego, CA: Academic Press; 1972. pp. 1–62. [Google Scholar]
• Bir-Akturk E, Fisiloglu H. Marital satisfaction in Turkish remarried families: Marital status, stepchildren, and contributing factors. Journal of Divorce & Remarriage. 2009;50:119–147. doi: 10.1080/10502550802365755. [CrossRef] [Google Scholar]
• Booth A, Edwards JN. Starting over: Why remarriages are more unstable. Journal of Family Issues. 1992;13:179–194. doi: 10.1177/019251392013002004. [PubMed] [CrossRef] [Google Scholar]
• Booth A, Johnson D, Edwards JN. Measuring marital instability. Journal of Marriage and the Family. 1983;45:387–394. doi: 10.2307/351516. [CrossRef] [Google Scholar]
• Booth A, Johnson D, White LK, Edwards JN. Predicting divorce and permanent separation. Journal of Family Issues. 1985;6:331–346. doi: 10.1177/019251385006003005. [PubMed] [CrossRef] [Google Scholar]
• Bramlett MD, Mosher WD. First marriage dissolution, divorce, and remarriage: United States. Advance data from vital and health statistics; no. 323. National Center for Health Statistics; 2001. [Google Scholar]
• Bramlett MD, Mosher WD. Cohabitation, marriage, divorce, and remarriage in the United States. Washington, DC: U.S. Government Printing Office; 2002. [PubMed] [Google Scholar]
• Bray JH, Kelly J. Stepfamilies: Love, marriage and parenting in the first decade. New York: Broadway Books; 1998. [Google Scholar]
• Bulanda JR, Brown SL. Race – ethnic differences in marital quality and divorce. Social Science Research. 2007;36:945–967. doi: 10.1016/j.ssresearch.2006.04.001. [CrossRef] [Google Scholar]
• Festinger L. A theory of cognitive dissonance. Evanston, IL: Row and Peterson; 1957. [Google Scholar]
• Glenn ND. The course of marital success and failure in five American 10-year marriage cohorts. Journal of Marriage and the Family. 1998;60:569– 576. [Google Scholar]
• Johnson CA, Stanley SM, Glenn ND, Amato PR, Nock SL, Markman HJ, Dion MR. Marriage in Oklahoma: 2001 baseline statewide survey on marriage and divorce (OKDHS Report No. S02096) Oklahoma City: Oklahoma Department of Human Services; 2002. [Google Scholar]
• Kamp Dush CM, Taylor MG, Kroeger RA. Marital happiness and psychological well-being across the life course. Family Relations. 2008;57:211– 226. [PMC free article] [PubMed] [Google Scholar]
• Kreider RM. Current Population Reports. Washington, DC: U.S. Census Bureau; 2005. Number, timing, and duration of marriages and divorces: 2001; pp. 70–97. Retrieved from http://www.census.gov/prod/2005pubs/p70-97.pdf. [Google Scholar]
• Rusbult CE, Buunk BP. Commitment processes in close relationships: An interdependence analysis. Journal of Social and Personal Relationships. 1993;10:175–204. doi: 10.1177/026540759301000202. [CrossRef] [Google Scholar]
• Segrin C, Taylor ME, Altman J. Social cognitive mediators and relational outcomes associated with parental divorce. Journal of Social and Personal Relationships. 2005;22:361–377. doi: 10.1177/0265407505052441. [CrossRef] [Google Scholar]
• Stanley SM, Markman HJ, Whitton SW. Communication, conflict and commitment: Insights on the foundations of relationship success from a national survey. Family Process. 2002;41:659– 675. [PubMed] [Google Scholar]
• Stanley SM, Rhoades GK, Whitton SW. Commitment and the securing of romantic attachment. Journal of Family Theory & Review. 2010;2:243–257. doi: 10.1111/j.1756-2589.2010.00060. [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Stewart SD. Boundary ambiguity in stepfamilies. Journal of Family Issues. 2005;26:1002–1029. doi: 10.1177/0192513X04273591. [CrossRef] [Google Scholar]
• Teachman J. Complex life course patterns and the risk of divorce in second marriages. Journal of Marriage and Family. 2008;70:294–305. doi: 10.1111/j.1741-3737.2008.00482.x. [CrossRef] [Google Scholar]
• Thornton A. Changing attitudes toward separation and divorce: Causes and consequences. American Journal of Sociology. 1985;90:856– 872. [Google Scholar]
• White LK, Booth A. The quality and stability of remarriages: The role of stepchildren. American Sociological Review. 1985;50:689– 698. [Google Scholar]
• Whitton SW, Rhoades GK, Stanley SM, Markman HJ. Effects of parental divorce on marital commitment and confidence. Journal of Family Psychology. 2008;22:789–793. doi: 10.1037/a0012800. [PMC free article] [PubMed] [CrossRef] [Google Scholar]