“It’s complicated”

Abstract

Given previous research on the risks associated with cycling in young adult dating relationships, the present study examines the frequency with which cyclical dating relationships (relationships that end and renew) persist into cohabitation and marriage, the characteristics of these relationships, and the constraints associated with cycling during these stages using a nationally representative sample of cohabiting (n = 323) and married (n = 752) couples. Using retrospective accounts, results suggest that over one-third of cohabiters and one-fifth of spouses have experienced a breakup and renewal in their current relationship. Additionally, partners who have experienced cycling are at greater risk for further cycling and experiencing greater constraints to permanently ending the relationship, greater uncertainty in their relationship’s future, and lower satisfaction.

Keywords

Churning cohabitation marital satisfaction relationship cycling structural constraints

The study of on-again–off-again relationships is relatively new. This process of ending and renewing a romantic relationship has been referred to as relationship cycling (Dailey, Pfiester, Jin, Beck, & Clark, 2009) or churning (Halpern-Meekin, Manning, Giordano, & Longmore, 2013b) and falls conceptually in between the more traditional relationship categories of together and broken up (see Dailey, Middleton, & Green, 2012). Recent research suggests that about 30–50% of young adult dating partners have experienced at least one breakup and reconciliation with their current partner (e.g., Dailey, Pfiester, et al., 2009; Halpern-Meekin et al., 2013b; Vennum, 2011). Unfortunately, partners in renewed relationships have been found to be at greater risk for relationship distress. Although researchers have yet to determine causal ordering, compared to stably together relationships (relationships that have been continually maintained), relationship cycling is associated with lower commitment and satisfaction, poorer communication, greater uncertainty, and higher levels of verbal abuse and physical violence (e.g., Dailey et al., 2012; Dailey, Pfiester et al., 2009; Halpern-Meekin, Manning, Giordano, & Longmore, 2013a; Vennum, 2011).

The majority of studies on correlates of this form of instability have been conducted with young adult populations, with the expectation that relationship instability may be more common during this developmental period. Although relationship exploration is theorized as an aspect of emerging adulthood (Arnett, 2005), Halpern-Meekin et al. (2013a) argue that this specific type of instability may be a sign of relationship distress that spans developmental periods. Stanley and Markman (1992) suggest that when dating relationships transition into cohabitation and marriage, constraints that encourage the continuance of the relationship regardless of possible relationship problems or mutual commitment to the future of the relationship increase. The accrual of constraints in the presence of the lower dedication and relationship quality reported by cyclical partners may put these relationships at risk for further declines in stability and relationship quality by encouraging the continuation of a relationship that would have otherwise ended had the constraints not been present (Stanley & Markman, 1992). The purpose of this study is to add to the current understanding of relationship cycling by exploring the extent to which cyclical dating relationships persist into cohabitation and marriage, the characteristics of the cyclical relationships that do persist, and how constraints relate to cycling across these transitions.

Background

Recently, researchers have found that about 60% of young adults have experienced the ending and renewing of a dating relationship with the same partner (Dailey, Pfiester et al., 2009) and about 30–50% of young adult dating partners have experienced at least one breakup and reconciliation with their current partner (Dailey, Pfiester, et al., 2009; Halpern-Meekin et al., 2013b; Vennum, 2011). Unfortunately, partners in these cyclical relationships report lower explicit decision making, commitment, and satisfaction, along with greater conflict and uncertainty than noncyclical partners (Dailey, Pfiester, et al., 2009; Halpern-Meekin et al., 2013b; Vennum, 2011). Finding that cyclical couples also report greater verbal abuse and physical violence than stably together or separated partners, Halpern-Meekin et al. (2013a) hypothesize that the lower relationship quality reported by cyclical partners may be due, in part, to greater challenges with conflict management.

It is concerning that those partners who experience the most conflict are those that are choosing to renew their relationship instead of permanently ending it (Halpern-Meekin et al., 2013a). Halpern-Meekin et al. (2013a) suggest that the greater intimate self-disclosure and relationship length (indicating greater investment) reported by cyclical dating partners (Halpern-Meekin et al., 2013b) may contribute to partners seeking to renew the relationship in spite of the risks. As dating partners transition into cohabitation and marriage, cyclical partners may encounter additional forces that encourage the continuation of the relationship.

We conceptualize cycling during marriage as trial separations rather than divorce and remarriage to the same individual. Compared to partners who do not renew, young adults in cyclical dating relationships more often stay in contact after breaking up and report an implicit understanding that the relationship has not ended but has been redefined (Dailey, Rosetto, Pfiester, & Surra, 2009). Applying this framework to marital relationships, we equate ending the relationship permanently with divorce whereas trial separations during marriage would continue the pattern of redefining the relationship through time apart.

Although research on renewing cohabiting and marital relationships is sparse, research findings with young adult partners suggest that relationship renewal during these stages is less common than is found in dating relationships. Unlike young adult dating relationships, reconciliation of cohabiting and marital relationships appears rare, with only 10% of young adult cohabiting partners reconciling within 4 years of the separation (Binstock & Thornton, 2003). This reunion is often short-lived with one-third of those who reconciled separating again within a year (Binstock & Thornton, 2003). Further, ending and renewing a cohabiting relationship decreases the chances of partners proceeding to marriage (Binstock & Thornton, 2003). Interestingly, 32–35% of young adults who separate from their spouse reconcile at least once (Binstock & Thornton, 2003; Wineberg, 1994), although 50% of those that reconcile separate again within 3 years (Binstock & Thornton, 2003). Since previous research on the prevalence of cycling has been conducted with young adult samples, we do not propose a specific hypothesis on the prevalence of cycling during cohabitation and marriage and simply wonder to what extent cyclical dating relationships transition into cohabitation and marriage in a nationally representative sample. Given previous findings, though, we do expect that married partners will be less likely to report having experienced a breakup and renewal while dating or cohabiting than current cohabiting partners, and partners who experienced cycling during a previous relationship stage will be more likely to experience cycling during subsequent relationship stages (i.e., cycling while dating will be related to cycling during cohabitation).

Risky transitions in romantic relationships

We use Stanley and Markman’s (1992) commitment model to conceptualize the risks associated with cyclical relationships transitioning into cohabitation and marriage. Grounded in social exchange principals, this model makes an important distinction between the motivation to stay and the cost to leaving a relationship. These forces are conceptualized as two meta-constructs: dedication and constraint (Stanley & Markman, 1992). Dedication includes a personal desire to build, maintain, and invest in the quality of the relationship for the benefit of both partners, whereas constraints encourage the continuance of the relationship by making termination of the relationship more financially, socially, or psychologically costly (Stanley & Markman, 1992).

Dedication to the relationship promotes actions (i.e., sacrifice) that serve in the best interests of the couple, and having confidence in the future of the relationship increases the likelihood that partners will further invest in it (Stanley, Rhoades, & Whitton, 2010). Constraints, on the other hand, serve to make a relationship harder to end regardless of the quality of the relationship (Stanley & Markman, 1992). Constraints are not inherently bad unless they are accrued in the presence of low dedication and relationship quality (Stanley, Rhoades, & Markman, 2006; Stanley, Rhoades, Amato, Markman, & Johnson, 2010). When partners accrue additional constraints before clarifying their dedication to the relationship (such as moving in together prior to making a commitment to marriage), they increase the forces impelling the relationship forward (Stanley et al., 2006). Stanley et al. (2006) call this inertia. Inertia increases the risk of relationships continuing into marriage that otherwise would have ended had the constraints not been present (Stanley et al., 2006). This is similar to couples making transitions that are event-driven (such as one’s lease expiring; Surra & Hughes, 1997), rather than based on careful deliberation about the future of the relationship. Stanley et al. (2006) suggest that event-driven transitions are more likely to occur when partners do not thoroughly evaluate the consequences of moving through relationship transitions. Without a conscious intent to make the relationship work (sliding versus deciding), partners run the risk of moving through transitions (such as cohabitation and marriage) that accrue constraints to ending the relationship without engaging in the pro-relationship behaviors needed to make the relationship function long term (Stanley et al., 2006), leading to further distress and instability.

For example, the process of moving in together is not clearly defined for many cohabiters. Lindsay (2000) reported that most couples say cohabitation “just happened.” Rhoades (2005) found that this ambiguity about the future of the relationship can be reflected in partners’ reasons for cohabiting. Specifically, men were more likely than their female partners to endorse moving in together because they could not see a future together but did not want to break up (Rhoades, 2005). Accordingly, researchers have found that spouses who move in together without first making explicit their dedication to the future of their relationship (committing to marry or becoming engaged before cohabiting) report more negative interactions, uncertainty, and proneness for divorce compared with partners who moved in after engagement (Kline et. al., 2004; Stanley et al., 2010).

Should the characteristics of cyclical dating relationships (i.e., greater sliding and uncertainty, Vennum, 2011 ; and lower dedication, Dailey, Pfiester et al., 2009; Halpern-Meekin et al., 2013b, than noncyclical dating partners) persist through the transition to cohabitation or marriage, these partners may be at greater risk for later distress than noncyclical couples. In the presence of uncertainty and lower dedication to the relationship, cyclical partners may be less prone to actions that are in the best interest of the couple than noncyclical partners, thereby reducing relationship quality and increasing the chances of further dissolution (Stanley & Markman, 1992). This is particularly dangerous because unlike dating partners, cohabiting and married partners are more likely to have tangible resources (structural constraints) that the couple shares (Rhoades, Stanley, & Markman, 2010), possibly providing additional pressure to maintain the relationship.

Given the greater uncertainty and sliding reported by cyclical versus noncyclical partners (Vennum, 2011), we expect that cyclical partners who transition to cohabitation will be less likely to make an explicit commitment to marriage prior to cohabiting. Further, due to ambiguous transitions, we expect that cohabiters and spouses with a history of cycling will report greater uncertainty in the future of their relationship and lower satisfaction than partners without a history of cycling.

Accrued constraints

Several constraints may be particularly relevant for cyclical couples progressing into cohabitation and marriage. Pulling from Rusbult’s (1983) investment model, Stanley and Markman (1992) include the amount of investment in the relationship within the realm of constraints. Accordingly, relationship length has been found to be a stabilizing feature in cohabiting (e.g., Manning, 2004) and marital (e.g., White & Booth, 1991) relationships as joint investments have been found to increase over time (Rhoades, 2005). Given the greater relationship length reported by young adults in cyclical dating relationships compared to their noncyclical counterparts (e.g., Dailey, Pfiester et al., 2009; Halpern-Meekin et al., 2013b), we expect that cyclical cohabiting and married couples will report longer courtships than noncyclical couples.

Stanley, Rhoades, et al. (2006) also suggest that children may impact relationship stability. Although the presence of children themselves do not appear to have an effect on the stability of cohabiting relationships (e.g., Rhoades, Stanley, & Markman, 2010), factors related to caring for children are often cited as reasons to not permanently end a relationship (e.g., Knoester & Booth, 2000). We expect that partners with a history of cycling will be more likely to have children under the age of 18 and to report child care as an important factor in the decision to cohabit than noncyclical partners.

Economic factors may also influence stability. For example, financial strain has been found to be positively associated with relationship instability (Cutrona, Russell, Burzette, Wesner, & Bryant, 2011) but negatively associated with permanently ending a marriage (Knoester & Booth, 2000). Joint investments, such as owning a home, further encourage the continuation of relationships (Rhoades, 2005). Since financial resources can serve either as a stressor that ends relationships or a constraint that keeps people together, we do not propose a specific hypothesis regarding the direction of the relationship of cycling with income and home ownership, but do expect that they will be related.

Current investigation

Given the greater risks for distress in cyclical relationships, we expand on extant literature by asking three important questions: (1) to what extent do cyclical dating relationships transition into cohabitation and marriage, (2) how do the characteristics of cyclical and noncyclical relationships differ during cohabitation and marriage, and (3) how are the constraints associated with cohabitation and marriage related to cycling during these stages?

In order to answer these questions, we used a nationally representative sample to examine whether cycling is present throughout the life course and across demographic groups. Further, using dyadic data allowed us to control for the effects of gender in our analyses. We first examined the prevalence of cycling in cohabiting and married couples across relationships stages. We next examined several key characteristics of relationships with a history of cycling versus those without. Finally, we examined the presence of constraints for cyclical and noncyclical cohabiting and married couples. The answers to these questions provide us with greater perspective on whether cycling is a young adult phenomenon or a pattern that has implications for adult committed relationships.

Methods

Sample

The secondary data used in this study were drawn from a larger study on married and cohabiting heterosexual couples conducted in the US by the National Center for Family and Marriage Research in collaboration with Knowledge Networks (see Knowledge Networks, 2010). For clarity throughout the article, cohabiting partners will be referred to as men and women and married partners will be referred to as husbands and wives. Knowledge Networks maintains a national panel of adults (age 18 and older) selected through random digit dialing and address-based sampling methodologies. Members complete a demographic profile that determines their eligibility for inclusion in specific studies. Members who were randomly selected for participation in this study received an e-mail letting them know a survey was available. Participants received a laptop if they did not have one, and other incentives, such as points redeemable for cash, were provided to those who already had computer access.

As males are less likely to respond than females, the survey was originally given to 266 cohabiting and 1500 married males to complete and give to their partners. Additionally, a supplementary “opt-in” panel of 184 heterosexual cohabiting couples was recruited through online advertisements. This resulted in a total sample of 323 cohabiting couples. Of the 1,060 husbands that completed the survey, only 752 of their wives also completed the survey, resulting in a total married sample of 752 couples in which both spouses completed the survey. Because we chose to include the opt-in panel of cohabiting couples to increase our analytic power, we used weights provided by Knowledge Networks to adjust the sample to the distributions provided by the Current Population Survey. The resulting sample is nationally representative of U.S. married and cohabitating heterosexual adults 18–64 years old. The descriptive characteristics of the currently cohabiting and married samples are displayed in Table 1.

Table 1.

Descriptive statistics for cohabiting and married females and males.

	Cohabiting (n = 323)		Married (n = 752)
	Female	Male	Female	Male
Variables	M (SD) or %	M (SD) or %	M (SD) or %	M (SD) or %
Age (years)	34.78 (12.73)	37.29 (12.45)	44.16 (11.61)	45.48 (11.23)
Race
White, non-Hispanic	55.60	56.60	76.00	73.70
Black, non-Hispanic	16.60	21.10	4.30	5.30
Hispanic	18.50	14.70	13.30	14.40
Other	6.70	4.40	5.30	4.40
Multiracial	2.50	3.10	1.10	2.00
Education
≥Some college	53.60	58.80	61.80	63.10
≤High school diploma	46.40	41.20	38.20	36.90
Previously married	34.10	36.30	n/a	n/a

Note. Number of previous marriages was only asked of cohabiting partners in the secondary data set.

Measures

Cycling

To determine whether cohabiting couples had ever experienced a breakup and renewal, they responded no (0) or yes (1) to, “Did you and your current partner ever separate and get back together?” Partners answered this question in regard to current and previous relationship stages (while dating, cohabiting, and/or while married). Since the perception of cycling by one partner is likely to impact the relationship, couples in which at least one partner indicated they had broken up and gotten back together at least once were labeled cyclical (4.4% of cohabiting couples disagreed and 1.6% of spouses disagreed).

Commitment to marriage prior to cohabiting

To assess ambiguity around the transition to cohabitation and marriage, currently cohabiting partners were asked whether they and their partner had already decided to get married before living together. Married partners who had cohabited were asked if they had decided to marry before cohabiting. Response options for both questions were no (0) or yes (1). Both currently cohabiting partners and spouses who had lived together prior to marriage were asked whether being ready to commit to marriage yet was a factor in their decision to cohabit (1) or not (0).

Relationship uncertainty

One item was used to measure uncertainty in the future of the relationship: “What are the chances you and your spouse/partner will break up in the future?” Participants indicated that there was no chance (1), a little chance (2), a 50–50 chance (3), a pretty good chance (4), or an almost certain chance (5).

Relationship satisfaction

Three questions assessed relationship satisfaction. Two of these questions asked participants to indicate if they were very dissatisfied (1), somewhat dissatisfied (2), neither satisfied nor dissatisfied (3), somewhat satisfied (4), or very satisfied (5): “Taking all things considered, how satisfied are you with your relationship with your spouse or partner?” and “how satisfied are you with how well your spouse or partner listens to you?” The third question asked participants to rate their relationship with their current partner on a scale from completely unhappy (1) to completely happy (10). Coefficient αs ranged from .76 to .82 for men and women in the cohabiting and married samples.

Constraints

Length of courtship was assessed by asking participants to report the year and month they began dating. Spouses were also asked the year and month they were married. Months were converted to decimals and added to years to represent the overall length of courtship. Participants were also asked to indicate how many children they had under the age of 18 in the household. This was recoded into no children (0) or children present (1). Participants who had cohabited were asked whether shar[ing] in caring for a child/children was a factor influencing their decision to cohabit (1) or not (0). Each item was dummy coded as to whether the participant indicated the factor affected their decision (1) or not (0). Participants further indicated their household income (less than $24,999, $25,000–$49,999, $50,000–$74,999, $75,000–$99,999, and greater than $100,000), and whether they currently owned their home (ownership status of living quarters: rented for cash or occupied without payment of cash rent (0) or owned or being bought by you or someone in your household (1)).

Analysis plan

The sample had <1% missing data, so list-wise deletion was used in Statistical Package for Social Sciences. As recommended by Kenny, Kashy, and Cook (2006), we first assessed the degree of nonindependence in our data by conducting partial correlations for our variables in which we controlled for the effects of the between-dyads variable (cycling) for each of our outcome variables. Partners’ scores were highly correlated, indicating a large degree of nonindependence. For example, spouses’ and currently cohabiting partners’ reports of satisfaction in the relationship and relationship uncertainty were correlated above .5.

Because our data were nonindependent and our primary interest was the differences between cyclical and noncyclical partners rather than between the members of each couple, we analyzed males’ and females’ responses separately in order to control for gender when conducting analysis with noncontinuous outcomes. For continuous outcomes, we used a regression procedure recommended by Kenny et al. (2006) for use with nonindependent data and included several controls. In the currently cohabiting sample, we controlled for having experienced previous cohabiting and marital relationships (see Rhoades, 2005). Given previous findings on the impact of premarital cohabitation on marital quality (see Stanley et al., 2010), we controlled for whether or not spouses had lived together prior to marriage. For the χ2 analysis, φ is reported as a measure of effect size. For the multiple regressions, model R2 is reported along with Pearson and semi-partial correlations to indicate the unique contribution of each independent variable.

Results

For simplicity, the majority of the statistics are displayed in corresponding tables.

The prevalence of relationship renewals

Prevalence across relationship stages

We first examined how common a history of cycling was for currently cohabiting and married partners and whether a history of cycling was less prevalent in more committed relationship stages. Overall, married partners were less likely to report that their relationship had been cyclical prior to marriage (23% of spouses) than current cohabiting partners (37% current cohabiters; χ2(1, N = 2,151) = 87.30, p < .01, φ = .21). About 25% (n = 80) of cohabiting couples reported a breakup and renewal while dating and about 22% (n = 70) reported a breakup and renewal while cohabiting (see Table 2 for frequencies of cycling across relationship stages). Similarly, just over 23% (n = 177) of married couples reported that they had broken up and gotten back at least once prior to marriage, with the majority (87%, n = 153) of those breakups occurring while the couple was dating. Although about half (54.5%) of the married couples in our sample reported that they had cohabited before marriage, very few (6%, n = 45) of those who had cohabited before marriage broke up and got back together while cohabiting. Similarly, a small number of spouses (just over 6%, n = 49) indicated that they had experienced a trial separation during the course of their marriage.

Table 2.

Descriptive statistics for cycling in cohabiting and married couples.

	Cohabiting couples (n = 323)		Married couples
	Cohabiting couples (n = 323)		No premarital cohabitation (n = 353)		Premarital cohabitation (n = 394)
% Cyclical while	N	%	N	%	N	%
Dating	80	25	60	8	93	12
Cohabiting	70	22	–	–	45	6
Married (trial separation)	–	–	20	3	29	4

Note. Couples could report cycling while dating and cohabiting, thus the percentages are not mutually exclusive.

Patterns of cycling

We also hypothesized that previous cycling would increase the risk for further cycling. As expected, cohabiting partners who had broken up and renewed their dating relationship were more likely to break up and renew while cohabiting than partners who had not experienced a breakup and renewal while dating, χ2(1, N = 323) = 8.85, p < .01, φ = .17. Similarly, spouses who had cohabited and were cyclical while dating were more likely to break up and renew while cohabiting than cohabiting spouses who were not cyclical while dating, χ2(1, N = 381) = 14.66, p < .001, φ = .20. This amounted to 48% of spouses who broke up and got back together while living together prior to marriage having already experienced a breakup and reconciliation while dating. Although not a significant difference, 30% of spouses who had experienced a trial separation had broken up and gotten back together prior to marriage compared with 24% of spouses who had not experienced a trial separation, χ2(1, N = 744) = 1.09, p = .30, φ = .04.

Cyclical relationship characteristics

We expected that cohabiting and married partners who had experienced a previous breakup and renewal would be less likely than noncyclical partners to have made an explicit commitment to marriage prior to cohabiting and be more likely to report greater uncertainty in the future of their relationship and lower relationship satisfaction.

Lack of commitment to marriage

Contrary to expectations, currently cohabiting partners with a history of cycling were no less likely to have decided to marry prior to moving in together than cohabiting partners without a history of cycling (see Table 3 for full percentages and χ2 values). Interestingly, the findings were not the same for currently married couples who had cohabited prior to marriage. Husbands, but not wives, who had experienced premarital cycling were less likely to have made the decision to marry their partner prior to moving in together (34%) compared to husbands who did not experience cycling prior to marriage (48%). Similarly, husbands, but not wives, who had experienced premarital cycling (28%) were more likely to endorse that they cohabited because they were not ready for marriage compared to noncyclical husbands (19%). Currently cohabiting cyclical and noncyclical partners did not differ in their endorsement of this reason for cohabiting.

Table 3.

Factors influencing the decision to cohabit for cyclical and noncyclical cohabiting and married partners.

	Females								Males
	Cyclical		Noncyclical						Cyclical		Noncyclical
Variable	N	%	N	%	χ2	df	p	φ	N	%	N	%	χ2	df	p	φ
Currently cohabiting
Decision to marry prior to cohabitation	41	33	67	34	.04	1	.87	.01	48	38	76	38	.003	1	.96	.00
Reason cohabited—not ready to marry	41	33	78	40	1.64	1	.20	.07	52	42	69	35	1.49	1	.22	.07
Reason cohabited—child care	38	30	31	16	9.52	1	.01	.17	44	35	28	14	19.24	1	.001	.24
Currently married
Decision to marry prior to cohabitation	54	46	146	53	1.41	1	.23	.06	39	34	133	48	7.05	1	.01	.13
Reason cohabited—not ready to marry	34	30	67	24	1.02	1	.31	.05	33	28	53	19	4.11	1	.05	.10
Reason cohabited—child care	22	19	18	7	13.66	1	.001	.19	12	10	22	8	.59	1	.44	.04

Follow-up analysis revealed gender differences for husbands and wives who were cyclical while dating on whether lack of readiness to commit to marriage was a reason they decided to cohabit (χ2(1, N = 94) = 10.38, p < .01, φ = .33) and their perception of if they had decided to marry each other prior to cohabitation (χ2(1, N = 92) = 11.66, p < .01, φ = .36). Specifically, of those spouses who were cyclical prior to marriage, about 28% of husbands and 23% of wives endorsed not being ready for marriage as a reason to live together unmarried and about 67% of cyclical husbands reported that the couple had not decided to marry prior to cohabiting versus about 55% of cyclical wives.

Uncertainty and satisfaction

To account for the nonindependence of our data, we used the regression procedure suggested by Kenny et al. (2006) in which two separate regressions are run for each dependent variable. The first regression, which uses the difference between partners’ scores as the outcome variable, tells us the main effect of gender as well as the interaction of gender with cycling (whether the difference between females and males on the outcome variable differs depending on whether or not they have a history of cycling). The second regression uses the sum of the partners’ scores to examine the main effect of cycling on the outcome variable, controlling for a history of cohabitation and marriage (Kenny et al., 2006).

In the first set of regressions, cohabiting and married males reported greater uncertainty in the future of the relationship as well as greater satisfaction compared to their female counterparts, although the differences were small (refer to Table 4 for descriptive statistics by gender). The difference between males’ and females’ scores on the variables of interest did not differ by whether or not the couple had experienced a breakup and renewal, nor by whether partners had been previously married or had previously cohabited (details are available from the authors).

Table 4.

Relationship characteristics of cyclical and noncyclical cohabiting and married partners.

	Females				Males
	Cyclical		Noncyclical		Cyclical		Noncyclical
Variable	M	SD	M	SD	M	SD	M	SD
Currently cohabiting
Relationship length	9.54	8.62	6.68	6.64	9.88	8.79	6.29	5.96
Uncertainty about future	2.26	1.01	1.83	.77	2.01	.91	1.81	.76
Satisfaction	14.98	4.21	16.72	2.99	16.26	2.92	17.37	2.60
Currently married
Length of courtship	4.17	3.42	2.63	2.31	4.31	3.86	2.56	2.24
Uncertainty about future	1.57	.88	1.39	.65	1.61	.86	1.44	.65
Satisfaction	16.20	4.04	17.12	3.10	16.67	3.36	17.70	2.82

Note. Relationship and courtship lengths are in years.

Of greater interest, the results of the second set of regressions (see Table 5 for summed regression coefficients) revealed that cohabiting and married participants with a history of cycling reported greater uncertainty in the future of the relationship and lower relationship satisfaction than noncyclical couples, controlling for previous committed relationship experience in the currently cohabiting sample and premarital cohabitation in the currently married sample. Several controls were also significantly related to uncertainty and satisfaction. Specifically, currently cohabiting men who had cohabited in previous relationships reported greater uncertainty and lower satisfaction in the current relationship. Currently cohabiting women who had been previously married reported less uncertainty about the future of their current cohabiting relationship. In the married sample, if the couple cohabited prior to marriage, they were more uncertain in the future of their relationship, controlling for premarital cycling.

Table 5.

Summary of summed dyadic regression analyses for cohabiting and married partners.

Variable	Courtship length				Relationship uncertainty				Relationship satisfaction
Variable	B	SE B	β	r, r _sp	B	SE B	β	r, r _sp	B	SE B	β	r, r _sp
Currently cohabiting
Constant	8.86***	1.35			3.42***	.15			34.96***	.56
Cycling	8.40***	1.54	.28	.22, .28	.56**	.17	.18	.20, .17	−2.65***	.64	−.22	−.24, −.22
M previously cohabit	1.67	1.63	.06	.13, .05	.60**	.18	.20	.19, .17	−2.83***	.67	−.24	−.22, −.22
M previously married	6.34***	1.91	.21	.28, .17	.15	.21	.05	.02, .04	.25	.79	.02	.02, .02
F previously cohabit	−4.95**	1.61	−.17	−.04, −.16	.25	.18	.08	.12, .07	−.16	.67	−.01	−.06, −.01
F previously married	7.32***	1.85	.24	.28, .20	−.60**	.21	−.19	−.13, −.16	1.32	.77	.11	.10, .09
R2	.20				.11				.12
F	15.56***				7.33***				8.39***
Married
Constant	4.39***	.29			2.73***	.07			35.23***	.32
Premarital cycling	3.06***	.44	.24	.26, .24	.32**	.11	.12	.12, .11	−1.84***	.49	−.14	−.15, −.14
Premarital cohabit	1.56***	.38	.15	.18, .14	.22*	.09	.09	.10, .09	−.79	.42	−.07	−.09, −.07
R2	.09				.02				.03
F	36.53***				7.93***				9.97**

M: male; F: female.

Note. For currently cohabiting participants, M and F scores were summed to create the dyadic scale score. For married participants, husbands’ and wives’ scores were summed to create the dyadic scale score. Cycling is dummy coded such that cycling = 1. Similarly, the control variables indicating previous cohabiting relationships, previous marriages, or premarital cohabitation are dummy coded so that 1 = the presence of that experience.

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

Accrued constraints

Length of courtship

We expected that cyclical couples would be more likely to report constraints (longer courtships, children, and lack of financial resources) in their relationships than noncyclical couples. Using the dyadic regressions described above to analyze differences in relationship length for cohabiting and married couples, we found that spouses differed in their reports of courtship length (likely due to the retrospective nature of the data), but current cohabiting partners did not. For both married and currently cohabiting partners, couples with a history of cycling reported longer courtships than couples who had not cycled (see Tables 4 and 5). Specifically, cyclical cohabiting partners reported being together an average of 3 years longer than noncyclical partners, and spouses who had experienced cycling prior to marriage reported courting just over a year and a half longer, on average, than spouses who had not experienced premarital cycling. Several controls were significant: current cohabiters who had been previously married reported longer courtships, currently cohabiting women who had cohabited prior to the current relationship reported shorter courtships, and spouses who cohabited reported longer courtships.

Children

We further expected that cyclical partners who had or were currently living together would be more likely to report the presence of children under 18 in the house and that desiring to share in child care were important factors influencing their decision to cohabit (see Table 3). As expected, currently cohabiting cyclical women (30%) and men (35%) were more likely to report that sharing in child care was an important factor when they were deciding whether or not to live together without being married compared with 16% of noncyclical women and 14% of noncyclical men. Interestingly, this pattern only held true for wives in the currently married sample, with 19% of wives who had cohabited and experienced a breakup and renewal prior to marriage endorsing child care as an important reason for cohabiting compared with 7% of wives who did not experience a breakup and renewal prior to marriage. Consistent with expectations, cyclical currently cohabiting partners (51%) were more likely to report children less than 18 years of age in the house than noncyclical partners (37%, see Table 6). Contrary to expectations, there was not a significant difference in reports of the current presence of children in the home between spouses who had experienced premarital cycling and spouses who were not cyclical prior to marriage.

Table 6.

Constraints for cyclical and noncyclical cohabiting and married partners.

Variable	Cyclical		Noncyclical
Variable	N	%	N	%	χ2	df	p	φ
Currently cohabiting
Children under age 18	114	51	148	37	10.72	1	.001	.13
Current household income					2.67	4	.615	.07
$0–$24,999	71	32	104	27
$25,000–$49,999	75	33	125	33
$50,000–$74,999	34	15	63	16
$75,000–$99,999	27	12	61	16
$100,000+	18	8	32	8
Home ownership	75	40	159	49	4.34	1	.04	.09
Currently married
Children under age 18	102	53	291	47	2.56	1	.11	.06
Current household income					17.23	4	.002	.15
$0–$24,999	45	24	83	13
$25,000–$49,999	47	25	132	21
$50,000–$74,999	37	20	141	23
$75,000–$99,999	22	12	123	20
$100,000+	39	21	143	23
Home ownership	123	65	477	77	10.78	1	.001	.12

Financial resources

Several economic constraints were also expected to impact cyclical partners. Although cyclical cohabiting couples were not significantly more likely to report a lower household income than noncyclical partners, spouses who had been cyclical prior to marriage were more likely to report a lower current household income than spouses who had not broken up and renewed prior to marriage with 49% of cyclical spouses reporting a household income less than $50,000 annually versus only 34% of noncyclical spouses (see Table 6 for distribution of income by quartiles). Both currently cohabiting and married partners with a history of cycling were less likely to own their home compared to noncyclical partners. Specifically, 40% of currently cohabiting cyclical partners reported owning their home versus almost half of noncyclical current cohabiting partners and 65% of spouses who had experienced premarital cycling owned their home compared to 77% of spouses who had stable courtships.

Discussion

The current study used the Commitment Model by Stanley and Markman (1992) to conceptualize the prevalence and correlates of relationship cycling during cohabitation and marriage in a nationally representative sample. Given the risks associated with cycling in young adult dating relationships, we sought the answer to three important questions: (1) to what extent do cyclical dating relationships transition into cohabitation and marriage, (2) do the characteristics of cyclical relationships persist across these transitions, and (3) how are the constraints associated with cohabitation and marriage related to cycling during these stages?

The prevalence of relationship renewals

In answer to our first question, we found that a history of cycling was not uncommon in cohabiting and marital relationships. Of current cohabiting couples, 37% reported ever having experienced a breakup and renewal (25% while dating and 22% while cohabiting). This frequency of reconciliations during cohabitation is substantially higher than that found in previous literature (Binstock & Thornton, 2003). Several factors may account for this higher rate. First, previous research used young adult cohabiting couples, whereas we used a nationally representative cohabiting sample with an average age of mid-30s. Given the older age of our sample, participants may be more likely to have accrued constraints, encouraging renewal. Second, the sample used by Binstock and Thornton (2003) was collected about two decades earlier than the current sample. It may be that rates of cycling have increased as social and economic shifts have contributed multiple pathways to relationship formation (e.g., Amato, 2011).

Married partners were less likely to report cycling during courtship than currently cohabiting couples, with only 23% of spouses reporting their courtship was cyclical. Consistent with previous research (Binstock & Thornton, 2003), very few spouses experienced cycling while cohabiting (6%). Given that about half of spouses indicated they had cohabited prior to marriage, this percentage is in stark contrast to the percentage of cohabiters who reported cycling while living together. This suggests that perhaps some cyclical courtships do not progress onto marriage due to low relationship quality and dedication, although we do not have the data to test this hypothesis. More research is needed exploring cyclical cohabiters’ decision making around the transition to marriage.

Also surprising given the high rates of renewal found in young adult marriages (Binstock & Thornton, 2003), quite a small number of spouses in our nationally representative sample had experienced a trial separation during their marriage (6%), suggesting that marriage is perhaps a more stable relationship state across the life course than during young adulthood. Supporting this hypothesis, although cycling while dating was associated with cycling during cohabitation, premarital cycling did not predict cycling during marriage. Again, it may be that the transition to marriage accrues greater constraints than the transition to cohabitation, thereby increasing the stability (although not necessarily the quality) of the marriage.

Constraints and relationship quality

Given that 37% of currently cohabiting partners and 23% of spouses had experienced cycling, we next assessed the characteristics of these relationships. As suggested by Stanley and Markman (1992), two forces, dedication and constraints, contribute to the stabilizing of relationships. Thus, it may be that those cyclical dating relationships that persist into cohabitation or marriage are of higher dedication and quality (e.g., partners have improved their relationship or addressed the issues that precipitated the original breakup) than those that do not transition into cohabitation or marriage or that there are constraints that are encouraging the continuation of the relationship regardless of lower levels of dedication and relationship quality. Our results, on average, support the latter hypothesis.

Relationship quality

Although we did not have a direct measure of dedication to the relationship in the current data set, we were able to assess whether partners in cyclical relationships were more likely to enter into cohabitation without making an explicit commitment to marriage than partners in cyclical relationships, potentially putting them at greater risk for later distress (Stanley & Markman, 2006). Contrary to expectations, no difference was detected between currently cohabiting cyclical and noncyclical partners in how likely they were to have decided to marry prior to moving in together.

Differences did emerge, though, for spouses who cohabited prior to marriage. Husbands who had experienced premarital cycling were less likely to report that the couple had made the decision to marry prior to cohabiting compared to husbands who had not experienced premarital cycling. No difference was found between cyclical and noncyclical wives. This same gender pattern was reflected in the endorsement of cohabiting due to not being ready for marriage: husbands who were cyclical prior to marriage were significantly more likely to support this statement than husbands who were noncyclical. Again, no differences were found for cyclical versus noncyclical wives or current cohabiters.

To test whether a difference existed between cyclical partners’ perceptions, we examined whether spouses who were cyclical while dating differed in their readiness for marriage prior to cohabiting. We found that husbands were significantly more likely to report that the couple had not decided to marry prior to cohabiting and that they were deciding to cohabit because they were not ready for marriage than their wives. It may be that husbands and wives who had experienced premarital cycling were not on the same page regarding the status of their relationship during the transition to cohabitation. This would be consistent with the greater uncertainty, poorer communication, and greater sliding (Vennum, 2011) reported by cyclical dating partners. From the inertia perspective (Stanley et al., 2006), the lack of clearly formed commitment prior to cohabitation would place these cyclical couples at greater risk of continuing on to marriage without a clearly formed commitment due to accrued constraints, potentially leading to fewer pro-relationship behaviors, and hence, lower marital quality.

Consistent with this premise, both cohabiting and married couples with a history of cycling reported greater uncertainty in the future of their relationship and lower satisfaction than noncyclical couples, although these effects were small (Cohen, 1988). That premarital cycling may have a lingering effect on marital quality has important implications. Although there is the possibility that cyclical couples may renew due to actual improvements in the relationship (Dailey et al., 2011), these findings suggest that on average, couples who experience cyclical courtships and proceed to marriage do not have relationship quality equivalent to that of couples without a history of cycling. Given the variety of trajectories of marital satisfaction found in previous research (e.g. Anderson, Van Ryzin, & Doherty, 2010), though, we expect that not all cyclical couples follow the same trajectory and suggest that further research explore what distinguishes cyclical couples who maintain higher quality marriages versus those that do not. More research is needed to understand heterogeneity in cyclical relationships.

Interestingly, having experienced a previous committed relationship had different effects depending on whether it was cohabitation or marriage. Cohabiting men who had cohabited previously were more uncertain and less satisfied in the current cohabiting relationship whereas currently cohabiting women who had been previously married reported less uncertainty in the future of the current relationship. Having experienced the ending of a marriage may prompt growth and conscious evaluation of what a person desires in relationships (Schneller & Arditti, 2004), leading to increases in conscious decision making around transitions in the current relationship, thus reducing uncertainty. The ending of previous cohabiting relationships may not produce the same result if they were entered into with less commitment to the future of the relationship. Similarly, spouses who had cohabited prior to marriage reported greater uncertainty in the future of their marriage, possibly due to a greater ratio of constraints to dedication operating to encourage their transition to marriage.

The presence of constraints

Given the presence of lower dedication and relationship quality (Dailey, Pfiester et al., 2009) and greater sliding (Vennum, 2011) reported by young adults in cyclical versus noncyclical dating relationships, we suspected that cyclical couples who move into cohabitation and marriage may also report greater constraints to permanently ending the relationship than noncyclical couples. Our hypothesis was supported, with cyclical couples being more likely to report substantially longer courtships, the presence of children under the age of 18, that help with child care was a factor influencing their decision to cohabit, and fewer financial resources than noncyclical couples. In line with our expectations, the longer courtships reported by cyclical couples suggest that investments in the relationship accrue over time. Courtship length was also impacted by previous relationship history, lengthening courtship for some and shortening it for others. More research is needed on the influence of previous cohabitation and marriage on decision making, dedication, and uncertainty in current relationships.

The presence of children and lack of financial resources varied slightly by relationship status. Specifically, cohabiting cyclical partners were more likely to report having children under the age of 18 in the home and that help with child care was an important factor in their decision to cohabit than noncyclical cohabiting partners, although this did not hold true in the currently married sample. Recent research findings suggest that it is common for unmarried birth partners to move into cohabitation as a response to the birth of a child (Reed, 2006), suggesting children may encourage the transition of cyclical dating couples into cohabitation. Since the marriages in our sample averaged over 20 years in length and spouses are reporting on their current status, we cannot know whether children under the age of 18 were present during their transition into cohabitation or marriage.

Although both cyclical cohabiting and married couples were less likely to own their home than noncyclical couples, only spouses who had been cyclical prior to marriage reported a lower income than their noncyclical counterparts. Home ownership may be a sign of stability that is harder to achieve for cyclical couples across relationship stages. Other indicators of financial resources not present in the current study may be more relevant to stability in cohabiting unions, such as earning potential and full-time employment (e.g., Gibson-Davis, Edin, & McLanahan, 2005; Manning & Smock, 1995). Further, it may be that financial strain or worry, rather than simply income must be considered (taking into account the impact of debt, the presence of children, cost of living, etc.) when assessing how finances may prevent cyclical partners from stabilizing their relationships (e.g., Tach & Edin, 2011).

Limitations

Although the data set was nationally representative and dyadic, the data were cross-sectional rather than longitudinal and therefore left several questions having to do with timing unanswered. For instance, we do not have data on the quality of the relationship when spouses transitioned into marriage or what constraints were operating at that time. Given the retrospective nature of some of the data, there may also be some inaccuracy in participants’ reports of the past aspects of their relationships. Further, we did not have access to people whose relationships permanently ended to compare constraints and relationship quality. We also do not have information on the nature (how long, how they were interpreted, did partners change residences, etc.) of the separations and reconciliations cyclical partners experienced during cohabitation or marriage. Without this information, it is possible that partners’ interpretations of what constitutes a separation may differ. Although we classified couples in which one partner indicated the relationship was cyclical as cyclical, it may be that they could have more closely represented the noncyclical couples. It is also likely that the number of times partners have ended and renewed their relationship would be related to the level of accrued constraints and relationship quality, although we were unable to assess this hypothesis.

Implications for research and practice

Research up to this point on cycling in romantic relationships identifies couples post breakup and renewal. Longitudinal studies are needed to determine the precipitating factors of the initial breakup and renewal, how relationship dynamics (e.g., uncertainty, trust, attachment, dedication, satisfaction, etc.) change throughout these periods of breakup and renewal and through transitions to progressive stages of relationship development (e.g., cohabitation, childbearing, and marriage), and whether certain individuals are at greater risk for cyclical relationships throughout their life. Additionally, we are not yet able to distinguish between cyclical relationships that stabilize versus those that are at risk for a subsequent dissolution or separation, which would be important for intervention with these couples.

Our findings also suggest that more research is needed on the function of dedication and the accrual of constraints in cyclical relationships. Cyclical partners may temporarily end the relationship due to lower quality, but renew due in part to accrued investments and constraints making permanently ending the relationship more challenging. Additionally, although we found support for the increased role of structural/material constraints in cyclical relationships, we were not able to explore the role of perceived or felt constraints in cyclical relationships, which may further impact relationship stability (Rhoades et al., 2010). Further understanding of the forces operating in cyclical relationships would help inform interventions with premarital couples.

Conclusion

Our findings suggest that the ending and renewing of relationships is not simply a developmentally normal characteristic of young adult relationships and may be a sign of relationship distress that persists across transitions to cohabitation and marriage. The present findings add to our understanding of how ambiguity, dedication, and accrued constraints interact to encourage the continuation of patterns of cycling and the associated relationship characteristics. Further research is needed following cyclical relationships over time and through the transition to marriage to better understand the risk and resiliency factors present in these relationships.

Footnotes

Funding

This research was supported in part by a grant from United States Department of Health and Human Services.

References

Amato

P. R.

(2011) Relationship sequences and trajectories: Women’s family formation pathways in emerging adulthood. In Fincham

F. D.

Cui

(Eds.), Romantic relationships in emerging adulthood (pp. 27–43). New York, NY: Cambridge University Press.

Anderson

J. R.

Van Ryzin

M. J.

Doherty

W. J.

(2010). Developmental trajectories of marital happiness in continuously married individuals: A group-based modeling approach. Journal of Family Psychology, 24, 587–596.

Arnett, J. J. (2005). Emerging adulthood: Understanding a new way of coming of age. In J. J. Arnett & J. L. Tanner (Eds.), Emerging adults in America: Coming of age in the 21st century (pp. 3–19). Washington, DC: American Psychological Association.

Binstock

Thornton

(2003). Separations, reconciliations, and living apart in cohabiting and marital unions. Journal of Marriage and Family, 65, 432–443. doi:10.1111/j.1741-3737.2003.00432.x

Cohen

(1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.

Cutrona

C. E.

Russell

D. W.

Burzette

R. G.

Wesner

K. A.

Bryant

C. M.

(2011). Predicting relationship stability among midlife African American couples. Journal of Consulting and Clinical Psychology, 79, 814–825. doi:10.1037/a0025874

Dailey

R. M.

Jin

Pfiester

Beck

(2011). On-again/off-again dating relationships: What keeps partners coming back? Journal of Social Psychology, 151, 417–440. doi:10.1080/00224545.2010.503249

Dailey

R. M.

Middleton

A. V.

Green

E. W.

(2012). Perceived relational stability in on-again/off-again relationships. Journal of Social and Personal Relationships, 29, 52–76. doi:10.1177/0265407511420192

Dailey

R. M.

Pfiester

Jin

Beck

Clark

(2009). On-again/off-again dating relationships: How are they different from other dating relationships? Personal Relationships, 16, 23–47. doi:10.1111/j.1475-6811.2009.01208.x

10.

Dailey

R. M.

Rossetto

K. R.

Pfiester

Surra

C. A.

(2009). A qualitative analysis of on-again/off-again romantic relationships: “It’s up and down, all around.” Journal of Social and Personal Relationships, 26, 443–466. doi: 10.1177/0265407509351035

11.

Gibson-Davis

C. M.

Edin

McLanahan

(2005). High hopes but even higher expectations: The retreat from marriage among low-income couples. Journal of Marriage and Family, 67, 1301–1312. doi:10.1111/j.1741-3737.2005.00218.x

12.

Halpern-Meekin

Manning

W. D.

Giordano

P. C.

Longmore

M. A.

(2013a). Relationship churning, physical violence, and verbal abuse in young adult relationships. Journal of Marriage and Family, 75, 2–12. doi:10.1111/j.1741-3737.2012.01029.x

13.

Halpern-Meekin

Manning

W. D.

Giordano

Longmore

(2013b). Relationship churning in emerging adulthood: On/off relationships and sex with an ex. Journal of Adolescent Research, 28, 166–188. doi:10.1177/0743558412464524

14.

Kenny

D. A.

Kashy

D. A.

Cook

W. L.

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

15.

Kline

G. H.

Stanley

S. M.

Markman

H. J.

Olmos-Gallo

P. A.

St Peters

Whitton

S. M.

Prado

L. M

. (2004). Timing is everything: Pre-engagement cohabitation has increased risk for poor marital outcomes. Journal of Family Psychology, 18, 311–318. doi:10.1037/ 0893-3200.18.2.311

16.

Knoester

Booth

(2000). Barriers to divorce: When are they effective? When are they not? Journal of Family Issues, 21, 78–99. doi:10.1177/019251300021001004

17.

Knowledge Networks. (2010). Knowledge panel design summary. Retrieved from http://www.knowledgenetworks.com/knpanel/docs/KnowledgePanel(R)-Design-Summary-Description.pdf

18.

Lindsay

J. M.

(2000). An ambiguous commitment: Moving in to a cohabiting relationship. Journal of Family Studies, 6, 120–134.

19.

Manning

W. D.

(2004). Children and the stability of cohabiting couples. Journal of Marriage and Family, 66, 674–689. doi:10.1111/j.0022-2445.2004.00046.x

20.

Manning

W. D.

Smock

P. J.

(1995). Why Marry? Race and the transition to marriage among cohabiters. Demography, 32, 509–520. doi:10.2307/2061671

21.

Reed

J. M.

(2006). No crossing the “extra line”: How cohabiters with children view their unions. Journal of Marriage and Family, 68: 1117–1131. doi:10.1111/j.1741-3737.2006.00318.x

22.

Rhoades

G. K

. (2005) A longitudinal study of cohabiting couples’ reasons for cohabitation, relationship quality, and psychological well-being. Retrieved from ProQuest Digital Dissertations. (AAT 3278541)

23.

Rhoades

G. K.

Stanley

S. M.

Markman

H. K.

(2010). Should I stay or should I go? Predicting relationship stability from four aspects of commitment. Journal of Family Psychology, 24, 543–550. doi:10.1037/a0021008.

24.

Rusbult

C. E.

(1983). A longitudinal test of the investment model: The development (and deterioration) of satisfaction and commitment in heterosexual involvements. Journal of Personality and Social Psychology, 45, 101–117. doi:10.1037/0022-3514.45.1.101

25.

Schneller

D. P.

Arditti

J. A

. (2004). After the breakup: Interpreting divorce and rethinking intimacy. Journal of Divorce and Remarriage, 42, 1–37. doi:10.1300/J087v42n01_01

26.

Stanley

S. M.

Markman

H. J.

(1992). Assessing commitment in personal relationships. Journal of Marriage and Family, 54, 595–608. doi: 10.2307/353245

27.

Stanley

S. M.

Rhoades

G. K.

Amato

P. R.

Markman

H. J.

Johnson

C. A.

(2010). The timing of cohabitation and engagement: Impact on first and second marriages. Journal of Marriage and Family, 72, 906–918. doi:10.1037/0893-3200.20.1.117

28.

Stanley

S. M.

Rhoades

G. K.

Markman

H. J.

(2006). Sliding versus deciding: Inertia and the premarital cohabitation effect. Family Relations, 55, 499–509. doi: 10.1111/j.1741-3729.2006.00418.x

29.

Stanley

S. M.

Rhoades

G. K.

Whitton

S. W.

(2010). Commitment: Functions, formation, and the securing of romantic attachment. Journal of Family Theory and Review, 2, 243–257. doi: 10.1111/j.1756-2589.2010.00060.x

30.

Surra

C. A.

Hughes

D. K.

, (1997). Commitment processes in accounts of the development of premarital relationships. Journal of Marriage and Family, 59, 5–21.

31.

Tach

Edin

(2011). The relationship contexts of young disadvantaged men. Annals of the American Academy of Political and Social Science, 635, 76–94. doi:10.1177/0002716210393680

32.

Vennum

A. V.

(2011). Understanding young adult cyclical relationships. (Doctoral dissertation) Retrieved from http://etd.lib.fsu.edu/.

33.

White

L. K.

Booth

(1991). Divorce over the life course: The role of marital happiness. Journal of Family Issues, 12, 5–21. doi: 10.1177/019251391012001002.

34.

Wineberg

(1994). Marital reconciliation in the United States: Which couples are successful? Journal of Marriage and Family, 56, 80–88.