Premarital Parenthood and Newlyweds’ Marital Trajectories

Abstract

An increasing number of couples in the United States are entering their first marriage having already had a child together, raising important questions about whether and how these couples’ marriages differ from newlywed couples who enter marriage without children. The current study used 5 waves of data collected over the first 4.5 years of marriage from a sample of ethnically diverse, first-married newlywed couples living with low incomes to examine the effects of premarital parenthood on couples’ self-reported satisfaction, observed communication, and marital dissolution over time. Among couples who entered marriage with a shared biological child (premarital parents), satisfaction levels were lower, communication was less effective, less positive, and more negative than couples entering marriage without children. Rates of change in marital functioning did not differ between groups, but the rate of marital dissolution was twice as high among premarital parents (19.1%) relative to couples who were not parents at the start of marriage (9.5%). These between-group differences remained robust after controlling for several demographic differences (race, age, education, household income, work status, relationship length, premarital cohabitation), and initial differences in communication did not account for between-group differences in marital satisfaction or dissolution. Together, these findings indicate that newlywed marriage unfolds in similar ways for low-income couples with and without children at the start of marriage, but couples raising children at the time of marriage have greater vulnerability to marital distress and dissolution. Further research is needed to characterize this risk and how interventions can offset it.

One of the most significant shifts in family formation in the United States over the last several decades has been increases in the percentage of children born to unmarried parents. Overall, the percentage of all births to unmarried mothers has increased steadily, from 18% in 1980 to 28% in 1990 to 33% in 2000 to 40% in 2014, with elevated rates of nonmarital childbearing among Black women (70%) and Hispanic women (53%) relative to non-Hispanic White (29%) and Asian or Pacific Islander women (16%; Hamilton et al., 2015). Because a significant number of these women do eventually go on to marry either their child’s father or another partner (Martinez, Daniels, & Chandra, 2012), there are more adults – particularly ethnic minority, economically disadvantaged adults – who already have a child when they marry for the first time. For example, recent data indicate that 57% of Black women in first marriages and 20% of White women in first marriages had a premarital birth (Hayford, Stykes, & Guzzo, 2014); among men, 54% of Black fathers, 37% of Hispanic fathers, and 17% of White fathers entered their first marriage with one or more children (Payne, 2012).

This departure from traditional pathways of family formation, in which adults typically became parents only after getting married, raises new questions about whether and how premarital parenthood affects newlyweds’ marital trajectories. These questions also have important practical implications. Recent federal policy initiatives have sought to promote marriage in the hope that successfully encouraging unmarried parents to get married would increase couples’ relationship stability and ultimately confer benefits to their children (e.g., Administration for Children and Families Archives, n.d.; Heath, 2012; Johnson, 2012). Understanding whether the marriages of couples who became parents before marrying do indeed resemble those of newlywed couples who enter marriage without children—or whether premarital parenthood is associated with increased risk for poor marital outcomes—can inform these efforts.

The current study addressed these gaps using longitudinal data from a sample of ethnically diverse first-time newlywed couples living with low incomes. Specifically, we examine and compare trajectories of marital satisfaction and observed communication, as well as marital dissolution rates, among couples entering marriage with a shared biological child together (n = 115) and couples entering marriage without children (n = 263).¹ Doing so allowed us to examine whether newlywed marriage changes in similar ways for premarital parents and nonparents, and whether premarital parenthood affects couples’ risk for poor marital outcomes.

Understanding the Experiences of Premarital Parents

To date, there has been limited attention to the experiences of newlywed couples who enter marriage with children or how they compare to newlywed couples who enter marriage without children. The couple and family literature has largely focused on the experiences of first-time newlywed couples without children (e.g., Karney & Bradbury, 1997; Kurdek, 1998; Lavner, Bradbury, & Karney, 2012), the transition to parenthood among married couples (e.g., Belsky & Rovine, 1990; Cox, Paley, Burchinal, & Payne, 1999; Lawrence, Rothman, Cobb, Rothman, & Bradbury, 2008), and stepfamilies (e.g., Larson & Allgood, 1987; Peek, Bell, Waldren, & Sorell, 1988), none of which can speak directly to the experiences of this emerging demographic. Understanding these families’ experiences during the newlywed years is particularly important given that the newlywed years have long been recognized as a time of significant risk and change for many couples – they are a period during which marital satisfaction declines on average (e.g., Kurdek, 1998; Lavner & Bradbury, 2010) and when the risk of marital dissolution is greatest (Bramlett & Mosher, 2001).

There are competing hypotheses regarding how premarital parenthood may be associated with first-time newlyweds’ marital trajectories. Meta-analytic findings indicate that having children places unique demands on couples and is associated with lower marital quality (e.g., Mitnick, Heyman, & Smith Slep, 2009; Twenge, Campbell, & Foster, 2003), suggesting that couples entering marriage with children will experience more negative marital outcomes (e.g., lower marital quality, higher marital dissolution) over the early years of marriage relative to couples entering marriage without children. Couples entering marriage with children could also be at greater risk for poor marital outcomes because they may represent less established relationships that only proceeded to marriage because of the child, not because of the couples’ relationship per se. However, the opposite hypothesis is also plausible: Premarital parents may have marital outcomes that are similar to or more positive than newlywed couples who enter marriage without children. Because couples who enter marriage with children have more experience coping with stress, they may better weather the transition to marriage and other stressors during the newlywed years, consistent with previous findings on stress inoculation in marriage (Neff & Broady, 2011). Moreover, given that few unmarried couples who have children together go on to marry (Carlson, McLanahan, & England, 2004), those who do may represent a select group with the strongest relationships, suggesting they may fare at least as well as newlywed couples who enter marriage without children. The presence of children could also serve as a barrier to marital dissolution (e.g., Knoester & Booth, 2000; Levinger, 1965, 1976), suggesting that divorce rates among premarital parents may be similar to or lower than those among couples who enter marriage without children, regardless of any differences in marital quality.

There have been few studies comparing marital outcomes among couples with and without children when they entered marriage. All must be interpreted cautiously given that they are based on data from over two decades ago (when premarital parenthood was less common) and because none were limited to couples with a shared biological child, raising concerns that the findings may also represent difficulties from navigating stepfamily-like relationships. Data from the National Survey of Family Growth, Cycle 5 (conducted in 1995) indicate that women who had at least one child at the start of marriage had higher divorce rates over the first 5 years of their first marriage than women without children at the start of marriage (Bramlett & Mosher, 2002). Another study of 297 adults sampled in 1997 – who had been married an average of 7 years and only 6% of whom had children prior to marriage – found that having a child prior to marriage was associated with higher levels of marital discord and lower levels of marital harmony (Amato & Booth, 2001). Lastly, findings from a 16-year study of Black and White first-time newlywed couples that began in 1986 indicate that couples who entered marriage with a child were more likely to have a relatively dissatisfied marital trajectory than a highly satisfied marital trajectory (Birditt, Hope, Brown, & Orbuch, 2012). Together, these studies suggest that premarital parenthood is likely to be associated with poorer marital outcomes, though whether these patterns hold for couples with a shared biological child and if they are true in more recent years when premarital childbearing is more common is unclear.

The Current Study

This study addressed these gaps by examining marital trajectories over the first 4.5 years of marriage among a sample of ethnically diverse, first-married newlywed couples living with low incomes.² We used multiple waves of self-report data on satisfaction and observational data of couples’ communication to examine differences in couples’ marital trajectories over time between newlyweds who entered marriage with a shared biological child and newlyweds who entered marriage without children, in addition to examining between-group differences in rates of marital dissolution. This design has several strengths. First, we examine experiences of premarital parenthood in the population in which premarital parenthood is most common (e.g., Hayford et al., 2014; Payne, 2012). Second, our prospective, longitudinal data beginning soon after couples married allow us to separate differences in couples’ initial level of marital functioning (i.e., intercepts) from differences in changes in functioning over time (i.e., slopes), making them ideally suited for addressing questions about whether premarital parent couples begin their marriages with lower levels of marital quality, whether their marriages become increasingly distressed over time, or some combination of these patterns. Third, the use of observed communication data along with self-reported marital satisfaction provides a robust assessment of couples’ marital quality over time and is important given the central role of couples’ communication in theoretical models of relationship deterioration (e.g., Jacobson & Margolin, 1979; Reis & Patrick, 1996; Woodin, 2011) and in recent federal interventions aimed at strengthening unmarried parents’ relationships (e.g., Wood, Moore, Clarkwest, & Killwald, 2014). Fourth, we considered whether a range of demographic characteristics (race, age, education, household income, work status, relationship length, premarital cohabitation, children during marriage) differed between couples who entered marriage with children or without children and whether these factors accounted for any between-group differences in marital quality or marital dissolution, improving upon prior work that did not consider these factors.

Method

The current data collection was approved by the Human Subjects Protection Committee at the RAND Corporation (Protocol number: k0256–07-02; Title of study: Development and Maintenance of Low-income Newlywed Marriages).

Sampling

The sampling procedure was designed to yield first-married newlywed couples in which both partners were of the same ethnicity (Hispanic, African American, or Caucasian), living in neighborhoods with a high proportion of low-income residents in Los Angeles County. Recently married couples were identified through names and addresses on marriage license applications. Addresses were matched with Census data to identify applicants living in low-income communities, defined as Census block groups wherein the median household income was no more than 160% of the 1999 federal poverty level for a 4-person family. Next, names on the licenses were weighted using data from a Bayesian Census Surname Combination, which integrates Census and surname information to produce a multinomial probability of membership in each of four racial/ethnic categories (Hispanic, African American, Asian, and Caucasian/other). Couples were chosen using probabilities proportionate to the ratio of target prevalences to the population prevalences, weighted by the couple’s average estimated probability of being Hispanic, African American, or Caucasian, which are the three largest racial/ethnic groups among people living in poverty in Los Angeles County (U.S. Census Bureau, 2002; see also Elliott et al., 2013). A total of 3,793 couples were contacted through the addresses they listed on their marriage licenses and offered the opportunity to participate in a longitudinal study of newlywed development. Of the 3,793 couples contacted, 2,049 could not be reached and 1,522 responded to the mailing and agreed to be screened for eligibility. Of those, 824 couples were screened as eligible, and 658 of them agreed to participate in the study, with 431 couples actually completing the study.

Participants

The total sample comprised 431 couples (862 spouses) identified with the above procedures. At baseline, marriages averaged 4.8 months in duration (SD = 2.5). Mean ages were 26.3 (SD = 5.0) for women and 27.9 (SD = 5.8) for men. Median household income was $45,000. For husbands, 21% had less than a high school (HS) degree, 27% had a HS degree, 33% had some college, and 20% had graduated college. For wives, 15% had less than a HS degree, 25% had a HS degree, 32% had some college, and 28% had graduated college. Twelve percent of couples were African American, 12% were Caucasian, and 76% were Hispanic.

Procedure

For the first four assessments, couples were visited in their homes by two trained interviewers who described the IRB-approved study and obtained written informed consent from each participant. The marital satisfaction measure was administered orally to participants by an interviewer who entered their responses immediately via computer. After completing this and other self-report measures individually, partners were reunited for three 8-minute videotaped discussions. For the first interaction, which was designed to assess problem-solving behaviors, partners were asked to identify a topic of disagreement in their relationship and then to devote 8 minutes to working toward a mutually satisfying resolution of that topic. For the second interaction, which was designed to assess social support behaviors, one randomly chosen spouse was asked to “talk about something you would like to change about yourself” while the partner was instructed to “be involved in the discussion and respond in whatever way you wish.” Spouses were instructed to avoid selecting or discussing any topics that were sources of tension or difficulty within the relationship. After a short break, a third discussion was held that was identical to the second discussion, with the roles reversed. Couples were debriefed and paid $75 for participating.

Interviewers returned at 9 months (T2 = 14 months into marriage), 18 months (T3 = 23 months into marriage), and 27 months after baseline (T4 = 32 months into marriage) and administered the same interview and observational protocol. Couples who reported that they had divorced or separated did not complete the interview. Couples who could not be reached at any time point, or who reported that they had temporarily separated, were recontacted at the next time point. Couples were debriefed following each interview. Couples were paid $75 for T1, $100 for T2, $125 for T3, and $150 for T4. Data collection for T1 through T4 took place between 2009 and 2013.

The T5 assessment was designed as a more limited follow-up and thus procedures for this assessment differed from the previous four assessments in two respects. First, this assessment did not include a home visit. Instead, all couples were contacted by telephone and administered self-report questionnaires. Accordingly, observational communication data were not collected at this time point. Second, all data collection occurred within a two-month calendar window (February and March 2014), resulting in unequal spacing of visits between the Time 4 and Time 5 visits across the sample (whereas the previous visits occurred at similar intervals for all couples). On average, Time 5 occurred 23 months after T4 (SD = 5.0), which was 50 months after baseline/55 months into marriage. Each individual was compensated $25 for the T5 interview.

Over the 5 waves of the study, 93 couples (21.6%) were lost to attrition and 55 couples (12.8%) divorced or legally separated. Attrition did not differ by premarital parenthood status (χ²(1) = 2.69, p = .101, φ = .08).³ The mean number of assessments completed was 4.08 and the median number completed was 5.⁴

Behavioral Observation

Videotaped discussions from T1-T4 were scored by 16 trained coders using the Iowa Family Interaction Rating Scales (IFIRS; Melby et al., 1998). Coders – five of whom were native Spanish speakers – coded only in their native language. At Wave 1, principal axis factor analysis was applied to the IFIRS codes, which were formed by averaging each individual’s scores for each code across the three discussion tasks, to investigate their latent structure. The scree plot suggested three factors – positivity, negativity, and effectiveness – for husbands and for wives (Cattell, 1966), which explained 35.7% of the total variance for husbands and 34.7% of the total variance for wives. Adding a fourth factor accounted for only an additional 3.6% of the variance for husbands and 5.1% for wives and was not indicated by the scree plot (for details, see Williamson et al., 2013).

Measures

Premarital parenthood.

At baseline, participants were asked “Who lives in your current household (besides the two of you)?” followed by 15 categories, one of which was “Your or your spouse’s children, including biological, adopted, step, and foster children.” They were then asked, “How many of your children who are living with you are your biological child with your spouse/your biological child from another partner/your adopted child with your spouse/your stepchild/your foster child?” Couples who endorsed the presence of a biological child(ren) with their spouse and no other categories of children were coded as 1 = premarital parent (n = 115, 27%). Of these, 73% had 1 child, 19% had 2 children, 4% had 3 children, and 4% had 4 children. Couples with no children in any of the categories were coded as 0 = no child at the time of marriage (n = 263, 61%). Couples who had a step-child (n = 24, 5%) or a step-child and a biological child (n = 29, 7%) were excluded from analyses given the small cell sizes, resulting in 378 couples for all analyses. No adopted or foster children were present in the sample.

Marital satisfaction.

Marital satisfaction was conceptualized as spouses’ global sentiment toward the relationship and was assessed by summing responses on an eight-item questionnaire. Five items asked how satisfied the respondent was with certain areas of their relationship (e.g., “satisfaction with the amount of time spent together”) and were scored on a 5-point scale (1 = very dissatisfied, 2 = somewhat dissatisfied, 3 = neutral, 4 = somewhat satisfied, 5 = very satisfied). Three items asked to what degree the participant agreed with a statement about their relationship (e.g., “how much do you trust your partner”) and were scored on a 4-point scale (1 = not at all, 2 = not that much, 3 = somewhat, 4 = completely). Total scores could thus range from 8 to 37. Cronbach’s α exceeded .70 for husbands and wives across all time points (Table 1).

Table 1.

Alphas and Intraclass Correlation Coefficients (ICCs) for Communication and Marital Satisfaction over Time

		Husband		Wife
Time	Domain	Alpha	ICC	Alpha	ICC
T1	Positivity	.74	.83	.74	.81
	Negativity	.80	.72	.78	.77
	Effectiveness	.78	.74	.78	.80
	Marital satisfaction	.70	--	.70	--
T2	Positivity	.67	.81	.72	.86
	Negativity	.79	.85	.81	.78
	Effectiveness	.76	.77	.77	.75
	Marital satisfaction	.78	--	.75	--
T3	Positivity	.65	.83	.67	.82
	Negativity	.82	.89	.81	.88
	Effectiveness	.65	.79	.71	.77
	Marital satisfaction	.76	--	.77	--
T4	Positivity	.68	.77	.79	.79
	Negativity	.76	.74	.78	.78
	Effectiveness	.72	.84	.85	.85
	Marital satisfaction	.83	--	.79	--
T5	Marital satisfaction	.78	--	.81	--

Positivity.

A composite positivity behavioral scale was created by averaging an individual’s scores on the group enjoyment, positive mood, warmth/support, physical affection, humor/laugh, endearment, and listener responsiveness codes. At each time point, a positivity score was calculated for each of the three discussion tasks, and these three scores were averaged to form the final positivity score for each individual. Cronbach’s α exceeded .65 and intra-class correlation coefficients (ICCs) exceeded .75 for husbands and wives across all time points (Table 1).

Negativity.

A composite negativity behavioral scale was created by averaging an individual’s scores on the angry coercion, contempt, denial, disruptive process, dominance, hostility, interrogation, and verbal attack codes. At each time point, a negativity score was calculated for each of the three discussion tasks, and these three scores were averaged to form the final negativity score for each individual. Cronbach’s α exceeded .75 and ICCs exceeded .70 for husbands and wives across all time points (Table 1).

Effectiveness.

A composite effectiveness, or problem-solving skill, behavioral scale was created by averaging an individual’s scores on the assertiveness, communication, effective process, solution quality, and solution quantity. At each time point, an effectiveness score was calculated for each of the three discussion tasks, and these three scores were averaged to form the final effectiveness score for each individual. Cronbach’s α exceeded .65 and ICCs exceeded .70 for husbands and wives across all time points (Table 1).

Divorce.

Couples were coded as divorced (0 = no, 1 = yes) if they reported at any time point that they had divorced or permanently separated.

Demographics.

We considered several demographic characteristics at baseline, including (1) Race, defined as Caucasian, Hispanic, or African American, (2) Age, calculated by subtracting the participant’s birthdate from the date of the interview, (3) Education, the highest level of education obtained (1 = less than a high school degree, 2 = high school degree, 3 = some college, 4 = college degree or higher), (4) Household income, calculated by averaging husband and wife reports of their household income, (5) Work status, coded with 1 = working full-time, 2 = working part-time, 3 = not working by choice, 4 = unemployed and looking for work, (6) Relationship length, assessed in months using participants’ responses to the question, “Thinking about your relationship since you first got together, how many years have you been together with [spouse’s name]?”, and (7) Premarital cohabitation, assessed in months using participants’ response to the question, “How long did you live together before getting married?”; participants who did not live together prior to marriage received a 0. Additionally, we examined whether participants had children during marriage based on responses at T2-T5 to the question “Have you and [SPOUSE NAME] had a child or children since we last spoke with you?”

Correlations between all study variables at Time 1 are shown in Table 2.

Table 2.

Correlations Between Main Study Variables at Time 1 for Husbands and Wives (N = 378 Couples)

	1	2	5	4	5	6	7	8	9	10	11	12	13	14
1. Premarital parenthood	--	.01	.17**	−.21***	−.35***	−.30***	−.14**	−.32***	.21***	.40***	−.16**	−.20***	.24***	−.12*
2. African American	.01	--	--	--	.02	−.08	−.07	−.02	−.14**	−.05	−.01	.01	−.01	−.17**
3. Hispanic	.17**	--	--	--	−.28***	−.33***	−.22***	−.07	.09	−.05	−.03	−.14**	.24***	.13*
4.Caucasian	−.21***	--	--	--	.32***	.46***	.32***	.10	.01	.10*	.05	.16**	−.29***	−.02
5. Education level	−.31***	−.01	−.33***	.41***	--	.51***	.28***	.37***	−.01	− 23***	.09	.12*	−.17**	.24***
6. Household income	−.30***	−.08	−.33***	.46***	.52***	--	.50***	.36***	.10	−.01	.06	.11*	− 23***	.17**
7. Age	−.15**	−.04	−.21***	.29***	.21***	.42***	--	.25***	.29***	.30***	.08	.09	−.26***	.08
8. Work status	−.07	−.25***	.08	.11*	.14*	.19***	.07	--	−.01	−.12*	.08	.10	−.06	.18**
9. Premarital cohabitation	.21***	−.14**	.09	.01	−.04	.10	.23***	.05	--	.64***	.03	−.17**	.03	.05
10. Relationship length	.40***	−.05	−.05	.10*	−.18**	−.01	.27***	.01	.64***	-	.02	−.10	−.07	−.04
11. Marital satisfaction	−.16**	−.01	−.05	.07	.03	.07	.01	−.03	−.02	−.06	--	.26***	−.16**	.16**
12. Positivity	−.21***	.02	−.17**	.18***	.19***	.13*	.01	.01	−.23***	−.19***	.16**	--	−.13*	.27***
13. Negativity	.26***	−.03	.25***	−.27***	−.12*	−.16**	19***	−.01	.04	−.05	−.13*	−.21***	--	−.15**
14. Effectiveness	−.15**	−.19***	.11*	.02	.21***	.15**	.06	.01	−.07	−.12*	.06	.35***	−.14**	--

Notes. Husbands’ correlations are below the diagonal and wives’ correlations are above the diagonal. Correlations between binary and continuous variables are point-biserial correlations, and correlations between two binary variables are phi.

^*p < .05.

^**p < .01.

^***p < .001.

Results

Preliminary Analyses

Demographic differences by child status.

We began by examining baseline differences between couples entering marriage with children (premarital parents) and couples without children at the time of marriage (nonparent newlyweds) in demographic characteristics (Table 3). Relative to couples who started marriage without children, couples who started marriage with children were more likely to be Hispanic, less likely to be Caucasian, were younger, had lower levels of education, had lower household incomes, had been in their relationships longer at the time of marriage, and had lived together longer prior to marriage. Further, wives were less likely to be working than wives in couples who did not start marriage with a child. In contrast, couples with and without children at the time of marriage did not significantly differ in husbands’ work status or their likelihood of having a child over the course of the study.

Table 3.

Demographics by Premarital Parenthood Status

	Descriptive Statistics		Difference Between Groups
Demographic Characteristics	Nonparent Newlyweds	Premarital Parents	Test statistic	Effect size
Caucasian	17.9%	2.6%	16.24***	−.21
African American	8.4%	8.7%	0.01	.01
Hispanic	73.8%	88.7%	10.50**	.17
Husband age	27.9 (5.3)	26.2 (5.2)	2 94**	.33
Wife age	26.1 (4.6)	24.7 (4.9)	2.69**	.30
Husband education	2.76 (1.02)	2.07 (0.92)	6.24***	.71
Wife education	3.01 (0.97)	2.23 (0.97)	7.14***	.80
Household income	$65,102 ($47,469)	$36,528 ($24,958)	6.02***	.75
Husband work status	85.6%	80.0%	1.82	−.07
Wife work status	76.0%	43.5%	37.89***	−.32
Premarital relationship length (years)	4.1 (2.8)	5.6 (3.9)	−4.19***	.44
Premarital cohabitation (months)	9.4 (19.9)	38.2 (45.9)	−8 49***	.81
Marital childbearing	58%	51%	1.52	.07

Notes. Descriptive statistic is Mean (SD) for continuous variables and percentage for categorical variables. Test statistic is t for continuous variables and χ² for categorical variables. Effect size is Cohen’s d for continuous variables and φ for categorical variables.

^*p < .05.

^**p < .01.

^***p < .001.

Satisfaction and communication trajectories.

We examined couples’ marital satisfaction, positivity, negativity, and effectiveness trajectories over time using dyadic growth curve modeling in a multilevel modeling framework in Stata v14 (StataCorp, 2015). Growth curve analytic techniques allow for a two-level data analysis. Level 1 estimates within-subject trajectories of change (growth curves) for a variable, described by two parameters: an intercept (initial level of the variable) and a slope (rate of change over time). Level 2 examines between-subject differences in these parameters using individual-level predictors. Missing data was accommodated through the use of Maximum Likelihood estimation.

Husbands’ and wives’ data were estimated simultaneously within the same equations using the dual intercept and slope model outlined by Raudenbush, Brennan, and Barnett (1995), which allows for separate intercepts, slopes, and random effects for husbands and wives. Time was uncentered so that the intercept terms (B_w00 and B_h00) could be interpreted as the value at baseline (Time 1), and each follow-up assessment was set as the number of months elapsed since baseline. Values for Time 2 (9), Time 3 (18), and Time 4 (27) were the same for all couples, and values for Time 5 varied by couple. We used the following Level 1 equation:

$${\text{Y}}_{\text{ti}}(\text{Outcome})=(wife)\text{it}[{\pi }_{\text{w}0\text{i}}+{\pi }_{\text{w}1\text{i}}{(Time)}_{\text{it}}]+(husband)\text{it}[{\pi }_{\text{h}0\text{i}}+{\pi }_{\text{h}1\text{i}}{(Time)}_{\text{it}}]+{\text{e}}_{\text{it}}$$

We conducted four separate models, one for marital satisfaction, one for positivity, one for negativity, and one for effectiveness, each of which are represented in the equation by “Outcome.” In the equation, i = couple and t = time; (wife)_it is an indicator for wives and (husband)_it is an indicator for husbands; π_w0i and π_h0i are the expected level of the outcome variable for wives and husbands, respectively, within couple i at baseline (T1); π_w1i and π_h1i are the average rates of change per month in the outcome variable for the wives and husbands, respectively, within couple i; and e_it is a random error of measurement.

Results indicated that husbands’ and wives’ marital satisfaction declined over time [husbands: intercept = 33.85 (SE = 0.16), slope = −.03 (SE = 0.01), p < .001, effect size r = 0.22; wives: intercept = 33.15 (SE = 0.18), slope = −.04 (SE = 0.01), p < .001, effect size r = 0.28]. For observed positivity, husbands and wives had stable trajectories over time [husbands: intercept = 2.34 (SE = 0.03), p < .001, slope = −.002 (SE = 0.01), p = .15, effect size r = 0.07; wives: intercept = 2.30 (SE = 0.03), p < .001, slope = −.001 (SE = 0.01), p = .71, effect size r = 0.02]. For observed negativity, husbands declined over time and wives had stable trajectories over time [husbands: intercept = 1.86 (SE = 0.03), p < .001, slope = −.003 (SE = 0.01), p = .017, effect size r = 0.12; wives: intercept = 1.91 (SE = 0.03), p < .001, slope = .001 (SE = 0.01), p = .99, effect size r = 0.01). For observed effectiveness, husbands and wives declined over time (husbands: intercept = 4.10 (SE = 0.04), p < .001, slope = −.012 (SE = 0.01), p < .001, effect size r = 0.29; wives: intercept = 4.23 (SE = 0.04), p < .001, slope = −.012 (SE = 0.01), p < .001, effect size r = 0.31).

Premarital Parenthood and Couples’ Marital Trajectories

Next, we examined whether there were differences in marital trajectories between couples who started marriage with children and couples who started marriage without children. To do so, we added premarital parenthood status as a Level 2 predictor for intercepts and slopes in the equations described above and again ran separate models for marital satisfaction, positivity, negativity, and effectiveness. Results, shown in Table 4, indicated initial differences in all domains of functioning, such that couples entering marriage with children had lower initial marital satisfaction, observed positivity, and observed effectiveness, and higher initial observed negativity than couples without children at the time of marriage. However, premarital parenthood status did not predict slopes, indicating that satisfaction and observed communication changed in similar ways for both groups of couples (see Figure 1).

Table 4.

Differences in Marital Satisfaction and Communication Trajectories between Couples Entering Marriage With and Without a Child

	Intercept				Linear Slope
Model	Nonparent Newlywed β (SE)	Premarital Parent β (SE)	χ2	Effect size φ	Nonparent Newlywed β (SE)	Effect size r	Premarital Parent β (SE)	Effect size r	χ2	Effect size φ
No Controls
Husbands’ satisfaction	34.17 (0.19)	33.12 (0.29)	8.86**	.15	−0.033 (0.007)	.22	−0.013 (0.012)	.06	1.76	.07
Wives’ satisfaction	33.58 (0.21)	32.14 (0.32)	14.45***	.20	−0.038 (0.009)	.23	−0.047 (0.014)	.18	0.31	.03
Husbands’ positivity	2.44 (0.04)	2.12 (0.06)	23.71***	.25	−0.002 (0.002)	.05	−0.004 (0.003)	.07	0.50	.04
Wives’ positivity	2.40 (0.03)	2.06 (0.05)	27.33***	.27	0.001 (0.002)	.01	0.001 (0.003)	.02	0.05	.01
Husbands’ negativity	1.77 (0.03)	2.07 (0.05)	24.86***	.26	−0.002 (0.002)	.08	−0.005 (0.002)	.10	0.81	.05
Wives’ negativity	1.82 (0.03)	2.12 (0.03)	22.47***	.24	0.001 (0.002)	.04	−0.002 (0.003)	.05	1.40	.06
Husbands’ effectiveness	4.18 (0.05)	3.89 (0.08)	8.98**	.15	−0.010 (0.002)	.20	−0.018 (0.004)	.23	3.06	.09
Wives’ effectiveness	4.30 (0.05)	4.04 (0.08)	7.32**	.14	−0.010 (0.002)	.22	−0.018 (0.004)	.24	2.97	.09
Models with Demographic Controls
Husbands’ satisfaction	34.62 (0.94)	33.40 (0.94)	9.96**	.16	−0.033 (0.008)	.22	-0.017(0.012)	.07	1.24	.06
Wives’ satisfaction	34.04 (0.93)	32.43 (0.96)	15.08***	.20	−0.037 (0.009)	.22	−0.046(0.014)	.17	0.29	.03
Husbands’ positivity	2.54 (0.14)	2.33 (0.15)	9.19**	.16	−0.001 (0.002)	.05	−0.004 (0.003)	.08	0.76	.04
Wives’ positivity	2.44 (0.09)	2.23 (0.10)	10.86**	.17	0.001 (0.002)	.01	−0.001 (0.003)	.01	0.02	.01
Husbands’ negativity	1.83 (0.14)	2.12 (0.14)	19.30***	.23	−0.002 (0.002)	.08	−0.005 (0.003)	.10	0.67	.04
Wives’ negativity	1.87 (0.14)	2.15 (0.15)	18.43***	.22	0.001 (0.002)	.04	−0.002 (0.003)	.05	1.29	.06
Husbands’ effectiveness	3.82 (0.21)	3.62 (0.21)	3.82	.10	−0.010 (0.002)	.21	−0.018 (0.004)	.23	2.68	.08
Wives’ effectiveness	3.89 (0.21)	3.74 (0.22)	2.39	.08	−0.010 (0.002)	.21	−0.018 (0.004)	.23	2.48	.08
Models with Demographic Controls and Communication Controls
Husbands’ satisfaction	33.17 (1.16)	32.31 (1.17)	4.74*	.11	−0.033 (0.008)	−.22	−0.020 (0.013)	−.08	0.80	.04
Wives’ satisfaction	32.65 (1.16)	31.24 (1.19)	11.51***	.17	−0.039 (0.009)	−.23	−0.047 (0.014)	−.17	0.25	.02

Notes. One unit of time equals one month. Demographic controls included race, age, education, household income, work status, relationship length, and premarital cohabitation (entered simultaneously). Communication controls included positivity, negativity, and effectiveness (entered simultaneously with demographic controls). Significant differences between couples with children at the start of marriage (premarital parents) and couples without children at the start of marriage (nonparent newlyweds) based on the χ² are shown in bold. Effect size φ is the effect size for the difference between premarital parents and nonparent newlyweds and is equal to √(χ²/n). Effect size r is the effect size for the slope terms and is equal to Z/√N.

^*p < .05.

^**p < .01.

^***p < .001.

Figure 1.

Differences in Husbands’ and Wives’ Self-Reported Marital Satisfaction and Observed Communication Trajectories by Premarital Parenthood Status with 95% Confidence Intervals.

We examined whether these results remained robust controlling for between-group demographic differences. To do so, we entered race, age, education, household income, work status, relationship length, and premarital cohabitation simultaneously with premarital parenthood at Level 2 in the models for marital satisfaction, positivity, negativity, and effectiveness. These results, also shown in Table 4, indicated that the pattern of results stayed the same for marital satisfaction, positivity, and negativity after controlling for these demographic characteristics, such that premarital parenthood status predicted intercepts but not slopes. These findings indicate that the differences in satisfaction, positivity, and negativity between couples who started their marriages with children and couples who started their marriages without children were not accounted for by between-group differences in demographic characteristics. However, when the demographic variables were added to the model for effectiveness, couples who started marriage with and without children no longer differed significantly in their intercepts; slopes were still not significantly different.

Premarital Parenthood and Divorce

We examined differences in divorce rates over the five waves of the study between couples entering marriage with and without children using chi-square tests and logistic regression. Of the 378 couples, 47 (12%) divorced by Time 5. Couples who entered marriage with a child were significantly more likely to divorce than couples who entered marriage without a child (19.1% vs. 9.5%, χ²(1) = 6.81, p = .009, φ = .13; also see Table 5, Model 1).

Table 5.

Logistic Regressions Predicting Divorce

	Model 1			Model 2			Model 3
	B	p	OR	B	p	OR	B	p	OR
Constant	−2.25	<.001	.11	.79	.634	2.19	.53	.807	1.70
Premarital parenthood	.81	.010	2.25	.87	.024	2.38	.87	.038	2.38
Caucasian				−.44	.560	0.65	−.55	.479	0.58
Hispanic				−.96	.062	0.38	−.87	.116	0.42
Husband age				−.03	.577	.97	−.04	.448	.96
Wife age				−.07	.241	.93	−.05	.398	.95
Husband education				−.42	.034	.65	−.40	.060	.67
Wife education				.25	.247	1.28	.30	.194	1.34
Household income				.01	.406	1.00	.01	.362	1.00
Husband work status				−.11	.542	.90	−.09	.637	.92
Wife work status				.03	.882	1.03	.04	.839	1.04
Relationship length				−.01	.902	.99	−.01	.919	.99
Premarital cohabitation				−.01	.858	1.00	.01	.938	1.01
Husbands’ positivity							.01	.982	1.01
Wives’ positivity							.11	.784	1.12
Husbands’ negativity							−.27	.435	.76
Wives’ negativity							.21	.528	1.24
Husbands’ effectiveness							−.02	.931	.98
Wives’ effectiveness							−.11	.664	.90

Note. Significant effects (p < .05) are shown in bold. Variables were block entered in each model, such that Model 1 included premarital parenthood, Model 2 included premarital parenthood and all demographic variables entered simultaneously, and Model 3 included premarital parenthood, all demographic variables, and all communication variables entered simultaneously.

As with satisfaction and communication, we analyzed whether these results remained robust after controlling for demographic differences (race, age, education, household income, work status, relationship length, and premarital cohabitation). In a logistic regression model in which premarital parenthood status and the demographic variables were entered simultaneously as predictors of divorce, premarital parenthood remained significant (B = 0.87, p = .024, OR = 2.38; see Table 5, Model 2). These findings indicate that the difference in divorce rates between newlywed couples who entered marriage with a child and newlywed couples who entered marriage without a child was not accounted for by between-group differences in initial demographic characteristics.

Premarital Parenthood, Initial Communication, Marital Satisfaction, and Divorce

Lastly, given between-group differences in initial communication, we examined whether couples’ initial communication accounted for between-group differences in marital satisfaction and divorce. For marital satisfaction, we included husbands’ and wives’ baseline positivity, negativity, and effectiveness along with the demographic predictors and premarital parenthood as Level 2 predictors of satisfaction intercepts and slopes, finding that the results remained the same: intercepts significantly differed between the two groups of couples and slopes did not (see Table 4). For divorce, we included husbands’ and wives’ baseline positivity, negativity, and effectiveness in the logistic regression model along with the demographic variables and premarital parenthood, again finding that the pattern of results remained the same: premarital parenthood significantly predicted divorce (B = .87, p = .038, OR = 1.70; see Table 5, Model 3). Together, these results indicate that differences in marital satisfaction trajectories and in 4-year divorce rates between couples who entered marriage with children and couples who entered marriage without children were not accounted for by initial differences in their communication.

Discussion

Our results indicated significant differences in self-reported satisfaction, observed communication, and marital dissolution over the first 4.5 years of marriage between newlywed couples who entered marriage with a shared biological child and couples who entered marriage without children. Specifically, in premarital parent couples, husbands and wives started their marriage with lower levels of self-reported satisfaction, observed positivity, and observed effectiveness, and with higher levels of observed negativity, though groups did not differ in their rates of change in marital quality over time. Furthermore, divorce rates over the 4.5 years were twice as high (19.1%) among couples who entered marriage with children relative to couples who entered marriage without children (9.5%). Accordingly, first-married newlywed couples who start their marriage with children have significantly worse marital outcomes over time relative to first-married newlywed couples who start their marriage without children, consistent with earlier findings (Amato & Booth, 2001; Birditt et al., 2012; Bramlett & Mosher, 2002).

Several other findings provide important context for these results. First, newlywed couples who entered marriage with children had different demographic profiles than couples who entered marriage without children. Specifically, these couples were more likely to be Hispanic, less likely to be Caucasian, were younger, had lower levels of education, had lower household incomes, were less likely to be employed (wives only), had been in their relationships longer, and had lived together longer prior to marriage. Couples did not differ in their likelihood of having children over the course of marriage, however. These differences are notable in light of our sampling strategy, in which all couples were recruited from the same low-income communities where the median household income was no more than 160% of the federal poverty level. As such, demographic differences between groups cannot simply be attributed to differential sampling techniques, but rather suggest that there are meaningful demographic differences underlying differences in premarital parenthood status even among couples within the same communities. More generally, these patterns are consistent with other work showing that premarital parenthood is especially common among racial and ethnic minorities and among individuals living with lower incomes (e.g., Hayford et al., 2014; Martinez et al., 2012; Payne, 2012), and speak to diverging pathways of family formation among disadvantaged couples and their relatively more advantaged peers.

Importantly, although these demographic differences between premarital parents and nonparent newlyweds were robust, they generally did not account for between-group disparities in the marital outcomes described above. Specifically, differences in couples’ self-reported marital satisfaction, observed positivity, observed negativity, and marital dissolution between couples who entered marriage with a child and couples who entered marriage without a child remained significant after controlling for demographic factors, indicating that between-group differences in couples’ marital outcomes based on premarital parenthood status were not attributable to demographic differences between groups. Similarly, couples’ initial differences in observed communication did not account for differences in marital satisfaction or dissolution based on premarital parenthood status, indicating that although couples who entered marriage with a child had worse communication than did couples who entered marriage without a child, communication deficits did not explain between-group differences in marital satisfaction or divorce.

Before discussing the implications of these results, we first outline several caveats. First, this study focused only on comparing newlywed couples who entered marriage with biological children to newlywed couples who entered marriage without children. As such, our findings cannot speak to how these couples’ trajectories may differ from the relationship trajectories of unmarried parents who decide not to marry. Lacking longitudinal data on this group of couples, we can only speculate that because the couples in our sample chose to marry, they may be more committed to their relationships than unmarried parent couples and may have less risky profiles in other respects as well, which would suggest that relationship outcomes would be worse among a comparable group of unmarried parent couples. Additional research comparing unmarried parent couples who remain unmarried with unmarried parent couples who eventually marry is needed to examine these issues. We also note that the premarital parent subsample consisted only of couples who had a biological child together. We were unable to examine couples in which the child was the biological child of only one partner (i.e., stepfamilies) because of the small cell size, but these families should be considered in future research. It will also be important for research to extend beyond the early years of marriage to consider how these patterns continue to evolve over time, and to use more standard measures of marital satisfaction to facilitate comparisons across samples. Lastly, although we included a robust set of demographic controls (i.e., race, age, education, household income, work status, relationship length, and length of premarital cohabitation), couples who enter marriage with children may differ from couples who enter marriage without children in other domains that affect their marital trajectories (e.g., religiosity, social attitudes); these differences should be considered in future research.

Notwithstanding these limitations, these findings have important implications for our understanding of how premarital parenthood predicts the developmental course of newlywed marriage. Overall, couples entering marriage with children had poorer marital trajectories on average relative to couples who entered marriage without children (i.e., lower levels of satisfaction, positivity, and effectiveness, higher levels of negativity and divorce). Critically, the differences in marital quality were found at the start of marriage (intercept effects), rather than emerging gradually over time (slope effects). On the one hand, this pattern could be interpreted as a minimizing/stabilizing of risk: these couples start their marriages worse off, but how their marriages change is no different from couples who entered marriage without children. On the other hand, these strong initial differences in marital quality did not dissipate as time passed (i.e., there were not more positive slope effects), indicating that these between-group differences remain constant over time. Whether this relatively lower level of marital quality poses other problems for these couples and their children over longer-term follow-up is an important question for future research. This may well be the case given that there was already an indication that divorce rates were significantly elevated among this group, suggesting that this early, continued risk may have meaningful consequences.

Our findings also raise new questions about the psychological processes underlying differences in marital outcomes between couples with and without children at the start of marriage. Importantly, our results cast doubt on communication deficits as a key contributor to between-group differences in relationship quality and dissolution: although groups differed in communication such that premarital parents had less positive, less effective, and more negative communication at the start of marriage, these differences did not explain between-group differences in satisfaction and divorce. These patterns suggest that other explanations are needed for the observed differences in these outcomes. One simple explanation is that by definition, premarital parents had an additional stressor at the start of marriage (children) that nonparent newlyweds did not have. A robust literature shows that stressors like child-rearing are negatively associated with romantic relationship functioning (e.g., Mitnick et al., 2009; Randall & Bodenmann, 2009), and these stressors may be particularly salient and meaningful among disadvantaged couples like those studied here (e.g., Jackson et al., 2016; Waller, 2008). In addition to the stress associated with raising children, the presence of children could amplify the effects of other stressors, further explaining between-group effects. For example, lower household income may be more problematic when it has to support more people, and less predictable work hours could pose more difficulties when childcare needs to be managed. Future research should investigate these possibilities to better understand why newlywed couples who enter marriage with children fare worse on average relative to couples who enter marriage without children. Better understanding the factors that explain poorer marital outcomes among couples entering marriage with children relative to couples who enter marriage without children is particularly important in order to inform the design of federal policies and programs designed to support their relationships, as these findings suggest that these programs’ current focus on communication (e.g., Dion, Avellar, & Clary, 2010) does not target a key mechanism explaining the higher risk for marital distress among these couples.

Finally, at a broader level, this study sheds new light on change in marriage over the newlywed years. Husbands’ and wives’ marital satisfaction declined on average, consistent with several decades of findings (e.g., Karney & Bradbury, 1997; Kurdek, 1998). More novel are our findings regarding longitudinal patterns of observed communication, which provide a more complete picture of how marriage changes. Optimistically, husbands’ and wives’ positivity remained stable, husbands’ negativity declined, and wives’ negativity remained stable, indicating that couples’ positive and negative communication does not deteriorate over the early years of marriage like satisfaction does. In contrast, communication effectiveness declined for husbands and wives. These findings indicate that couples’ ability to problem-solve and provide support effectively, including their ability to generate high quality solutions, is the one domain of communication where newlywed couples have increasing difficulty over the early years of marriage, suggesting that greater attention to understanding the role of effectiveness in marital functioning may be warranted. More generally, the similarity in patterns of stability and change between couples entering marriage with and without children is striking given that these couples differed in many ways: premarital parents were much more established in their relationships by virtue of having a biological child together, having been in the relationship longer, and having lived together longer prior to marriage. Viewed in this light, the fact that their marriages unfolded in a similar manner to that of couples who entered marriage without children is surprising and suggests that the degree to which the newlywed years represent a time of transition and change for couples is similar regardless of whether couples begin them solely as a couple or as parents as well.

In sum, these results indicate that marriage is not a panacea for unmarried parents; newlywed couples who enter marriage with children have worse marital trajectories and are at greater risk for marital dissolution over the early years of marriage relative to newlywed couples who enter marriage without children, though their marriages change in similar ways. Given that these differences are not accounted for by demographic factors or by communication deficits, further research examining the factors that contribute to these disparities will be important for understanding risk and resilience in this growing segment of the population and for informing programs and policies aimed at improving their relationships.