Continuity and Change in Stepfather–Stepchild Closeness Between Adolescence and Early Adulthood

Abstract

This study draws on nationally representative data from Waves I and III of the National Longitudinal Study of Adolescent to Adult Health (Add Health) to prospectively examine the factors associated with different patterns of closeness between stepchildren and their stepfathers over the transition to adulthood in stably married stepfamilies (N = 881). Results indicate much variability in how these relationships unfold over time, although a sizable minority of youth remained close to their stepfathers. Consistent with family systems theory, the quality of other family relationships is particularly important to understanding children’s relationships with stepfathers as they develop over time.

The rise in divorce during the 1960s and 1970s in the United States led to a corresponding increase in remarriage and stepfamily formation. Although the divorce rate declined during the 1980s, the percentage of children born to never-married mothers has continued to increase, and many of these mothers and their children eventually form stepfamilies. Although it is difficult to obtain precise counts of their prevalence (Pryor, 2014), estimates suggest that more than 10% of all two-parent families in the United States are married or cohabiting stepfamilies (Kreider & Ellis, 2011) and that approximately 30% of all U.S. children will spend some time in a stepfamily (Bumpass, Raley, & Sweet, 1995). Given that children usually live primarily with their mothers when biological parents separate, most stepfamilies involve residential stepfathers (Stewart, 2007)—the focus of the current study.

Earlier research has raised concerns about the implications of stepfamily formation for children’s well-being, indicating that children in stepfamilies generally have lower well-being than children in households with two biological parents and tend to show little or no advantage over children in single-parent households (Amato, 2010; Sweeney, 2010). More recent studies, however, have indicated great variability in stepfamily functioning and the quality of parent–child relationships (Ganong, Coleman, & Jamison, 2011; King, Thorsen, & Amato, 2014), with positive stepfamily functioning and parent–child relationships associated with greater child well-being (King, 2006).

Although recent research has elucidated factors associated with positive relationships between stepfathers and stepchildren (e.g., Ganong et al., 2011; Jensen & Shafer, 2013; King et al., 2014), we still know little about how these relationships unfold over time, especially as stepchildren enter young adulthood. Most research on stepfamilies has focused on minor children who are still living at home, and the few studies that examine adult stepchild–stepparent relationships usually capture these relationships at one point in time and/or rely on retrospective reports of earlier relationship characteristics (Ganong et al., 2011; Kalmijn, 2013; Kinniburgh-White, Cartwright, & Seymour, 2010).

The current study draws on Waves I and III of the National Longitudinal Study of Adolescent to Adult Health (Add Health) to prospectively examine patterns of stepfather–stepchild closeness between adolescence and young adulthood in a sample of married mother–stepfather families. We focus on married mother–stepfather for several reasons. Add Health did not collect information on adolescents’ relationships with nonresident stepparents (who are usually stepmothers) or with parents’ cohabiting partners. Family processes and children’s relationships with stepparents tend to differ depending on whether the stepparent is resident or nonresident, a stepmother or a stepfather, and married or cohabiting with the biological parent. For example, stepchildren tend to have stronger bonds with resident stepparents than with nonresident stepparents, and being a stepparent is often more difficult for stepmothers than for stepfathers (Ganong et al., 2011). Children who live with fathers after parental separation are a more selective group (Herrerias, 1995), and children sometimes live in these arrangements as a result of difficulties experienced by biological mothers (e.g., financial, employment, or emotional difficulties; difficulty in handling the children). Most cohabiting relationships either transition to marriages or dissolve within a few years of forming, with different implications for children’s well-being (Cherlin, 2014). Our sample of married stepfather families includes those formed directly through marriage and those that began as cohabiting partnerships but transitioned into marriage before the Wave I interview. Compared with other stepfamily forms, married stepfather families may be better able to offer support to stepchildren during the tenuous transition to adulthood (Berger & McLanahan, 2015).

The first aim of our study is to document how common different patterns of stepfather–stepchild relationship stability and change are between adolescence and young adulthood. That is, how many relationships remain stable (either close or not close), and how many change (become close or less close)? We conduct a parallel analysis of patterns of closeness to biological fathers for adolescents living in two-biological-parent families to gain further insight into how the prevalence of different relationship trajectories in stepfather families may be similar to, or different from, relationship trajectories in two-biological-parent families. The second aim of our study is to examine the factors associated with these different patterns of closeness to stepfathers over time. In particular, we focus on factors that (a) distinguish relationships that remain close from those that remain not close, (b) are associated with relationships becoming close over time, and (c) are associated with relationships becoming less close over time. These patterns may be consequential for stepchildren’s well-being, and they may have implications for the quality of stepfather–stepchild ties in later life.

Background

How Does Closeness to Stepfathers Change Over Time, and Why?

Prior research is mixed on how the stepfather–stepchild relationship changes over time, with some studies suggesting improvement, others suggesting stability, and still others suggesting that the relationship becomes less close (Ganong & Coleman, 1994; Stewart, 2005, 2007). It is likely that multiple trajectories exist that are hidden by averaging across all families. Several qualitative studies suggest that stepparent–stepchild relationships vary greatly in quality, from being very close and loving to being problematic and destructive to being disengaged (Ganong et al., 2011). It is unclear, however, which patterns are most common in the general population, or how much change these relationships undergo over time as stepchildren enter young adulthood. In one of the few studies to explore how stepparent–stepchild relationships develop over the transition to adulthood, Ganong et al. (2011) found six distinctive patterns among the retrospective reports of 49 adult stepchildren, some of which reflected stability in relationships and others of which reflected change (we return to this study in the Discussion section, where we note similarities and differences in our findings).

These different relationship trajectories likely depend on the larger context in which these relationships form and develop. We know of no research that has prospectively examined patterns of closeness to stepfathers and the factors that are associated with changes in these relationships from adolescence into young adulthood. Findings from studies that have examined the correlates of positive stepfather–stepchild relationships at a single point in time or retrospectively suggest a number of factors that may be relevant.

Consistent with family systems theory (Minuchin, 1974) and existing research on stepfamily dynamics (e.g., King et al., 2014), we assume that other family relationships are particularly important in understanding children’s relationships with stepfathers, and we further propose that these other relationships will continue to be associated with stepfather–stepchild relationships as they develop over time. Family systems theory holds that all parts of a family system are interconnected, with each subsystem (e.g., mother–child dyad) affecting all other subsystems (e.g., stepfather–stepchild dyad) (Broderick, 1993). In particular, we view the quality of the mother–stepfather relationship, the mother–child relationship, the nonresident father–child relationship, and the adolescents’ perceptions of family belonging as potential predictors of stepfather–stepchild closeness patterns.

Many studies have reported a positive link between marital quality and the parent–child relationship (Erel & Burman, 1995), including studies of stepfamilies (Jensen & Howard, 2015). Parents in supportive marriages may be more emotionally available to respond to children’s needs and thereby develop closer relationships with them. A good marriage might encourage stepfathers to develop and maintain affinity with stepchildren (Ganong, Coleman, Fine, & Martin, 1999) and encourage mothers to support a close relationship between the stepfather and her children (Marsiglio, 2004). Children may also be more likely to accept and respond positively to stepfathers whom they perceive as making positive contributions to the family, including when they see their mothers in happy marriages (Ganong et al., 2011).

Scholars have also noted mothers’ pivotal role in enhancing, or detracting from, family functioning in stepfamilies (Pryor, 2014), and a number of studies suggest that a positive mother–child relationship can help facilitate the development of positive ties between adolescents and their stepfathers (King, Amato, & Lindstrom, 2015; Weaver & Coleman, 2010). Given the strong association between the quality of the mother–child relationship and the quality of the stepfather–child relationship, it is likely that the closeness of the mother–child bond, and any changes to it, will be associated with corresponding patterns of stepfather–stepchild closeness. Indeed, when children leave the household and no longer live with stepfathers, mothers may be especially important in keeping children connected to their stepfathers.

Research is less clear about whether children’s relationships with nonresident biological fathers have implications for stepfather–stepchild relationships. Children may be more likely to develop close ties to their stepfathers if they do not have close ties to, or much contact with, their nonresident biological fathers (Dunn, Cheng, O’Conner, & Bridges, 2004; Ganong & Coleman, 2004). A few studies have suggested, however, that an adolescent’s relationship with each father is largely independent (King, 2006; King, Amato, & Lindstrom, 2015; White & Gilbreth, 2001). The number of studies that have examined this topic is small, and none has considered the possible long-term implications for the stepfather–stepchild relationship.

Another indicator of the quality of family relationships that goes beyond dyadic relationships is the adolescent’s perception of family belonging. A family-level construct, family belonging encompasses feelings of inclusion within one’s family, including feelings of being paid attention to, of being understood, and of having fun together (Goodenow, 1992; Leake, 2007). Although the quality of an individual’s relationships with each family member undoubtedly influences perceptions of family belonging, feelings of belonging are conceptually distinct from these dyadic relationships. Empirical evidence supports this distinction (King, Boyd, & Thorsen, 2015).

We considered several background characteristics that may be associated with patterns of closeness to stepfathers (see also Jensen & Howard, 2015, for a review). With respect to gender, several studies suggest that boys tend to report better relationships with stepfathers than do girls (Jensen & Shafer, 2013; Pasley & Moorefield, 2004) and that adolescent girls in stepfamilies are more likely than boys to disengage from their families (Hetherington, Bridges, & Insabella, 1998). One study of adult stepchildren, however, found that girls were more likely than boys to report that relationships with stepparents improved over time (Ganong et al., 2011). With respect to age, older adolescents tend to report weaker ties with stepfathers than do younger adolescents (King et al., 2014). Studies examining racial-ethnic differences in stepfamily relationships have yielded mixed findings regarding whether stepfather–stepchild relationships are closer in White or in minority families (Hofferth & Anderson, 2003; King et al., 2014; Stewart, 2007). Parental education and family income are generally associated with parental involvement (Amato, 1998; Cooksey & Fondell, 1996), although not all studies report significant differences in adolescent reports of closeness to stepfathers by socioeconomic status (e.g., King, Boyd, & Thorsen, 2015).

The length of time a stepfamily has been together has been found to be associated with closer stepfather–stepchild ties in both adolescence (King et al., 2014) and adulthood (Arranz Becker, Salzburger, Lois, & Nauck, 2013; Kalmijn, 2013). Longer durations afford greater opportunities for stepfathers to invest in their stepchildren and promote relationship building and bonding (Ganong et al., 1999; van der Pas, van Tilburg, & Silverstein, 2013). Children who acquire residential stepfathers early in life are more likely to accept stepfathers as parent figures, and they may be especially likely to form attachments to stepfathers that endure over time (Ganong et al., 2011). We also controlled for whether the stepchild was coresiding with the mother and stepfather at Wave III, as coresiding with parents in adulthood may be associated with both more positive (e.g., more shared activities and support) and more negative (e.g., conflict) aspects of parent–child ties (Aquilino, 1997; White & Rogers, 1997).

Stepfamilies and the Transition to Adulthood

The current study captures youth during critical points in the life course. Youth have many developmental tasks to accomplish during the transition to adulthood, and parents often continue to serve as crucial sources of support to them during this period (Lareau, 2015; Stewart, 2007). The prevalence of several types of risk behavior also peaks during emerging adulthood, not adolescence (Arnett, 2000). Close, supportive parent–child relationships are thought to be especially important for adolescents in stepfamilies because these adolescents are more vulnerable to peer influence and at greater risk for poor outcomes (Hetherington & Kelly, 2002).

Studies suggest that stepfathers do not decline in importance in supporting their stepchildren after those children grow up and leave the household (Klaus, Nauck, & Steinbach, 2012), although the form of this support may change. Ganong et al. (2011, 408) reported a number of ways that stepparents continue to provide important support to stepchildren as they develop. For example, in childhood and adolescence, stepparents helped stepchildren with homework, coached their athletic teams, and taught them how to drive. During emerging adulthood, stepparents drove stepchildren to explore colleges, fixed their cars, and helped them move. Thus, stepchildren who maintain close relationships with stepfathers, or whose relationships improve over time, are likely advantaged during the transition to adulthood. Stepchildren who are not close to stepfathers, or who are no longer close to stepfathers in young adulthood, are likely less able to draw on the support and resources of stepparents.

Emerging adulthood involves a number of transitions and new role demands, including the completion of education, moving out of the parental home, forming intimate relationships, and becoming economically self-sufficient. Young adulthood is also a time when individuals often redefine or renegotiate their relationships with parents. Young adults also have more developed cognitive and empathic abilities than do adolescents, which may lead them to think more deeply or reflectively about these relationships—perhaps to better appreciate what parents have done for them, or alternatively, to question how they were raised. The way parent–child relationships are redefined at this point is likely to determine the future trajectory of these relationships. For perhaps the first time, young adults have the option of not having relationships with stepfathers. When stepchildren no longer have to live with stepfathers or depend on them financially, will stepchildren continue to nurture these ties or let them diminish?

It is therefore important to better understand how stepfather–stepchild relationships develop over time and to identify the factors that are associated with their development (Ganong et al., 2011). In addition to the potential benefits for stepchildren (White, 1994a, 1994b), stepfathers may also benefit from being able to maintain close relationships with stepchildren, especially as they confront the challenges of aging, as adult children often serve as crucial sources of support for elderly parents (van der Pas et al., 2013). By focusing on the period between adolescence and young adulthood, the current study addresses a gap in research on stepchild–stepfather relationships. Drawing on a large, nationally representative sample, this study prospectively examines how closeness to stepfathers changes (or remains stable) during the transition to adulthood and considers a wide variety of intra- and interpersonal factors that may be associated with these relationship trajectories.

Method

Data and Sample

We used data from Waves I and III of Add Health (Harris et al., 2009). The Wave I in-home interview includes 20,745 adolescents in Grades 7–12 during the 1994–1995 school year and is nationally representative when appropriate sample weights are used. Parent data (n = 17,670) were collected from one parent, usually the biological mother. In 1996, 14,738 of the adolescents were reinterviewed (the Wave II sample design purposefully excluded adolescents who were in Grade 12 at Wave I). Wave III data were collected from 15,197 of the original adolescent respondents in 2001–2002 when these youth were in their early adult years, age 18–26 years (including some of the youth who were not interviewed at Wave II). The primary analytic sample for the current study was restricted to adolescents with valid sample weights who reported living with their biological mothers and married stepfathers in Wave I, and who were reinterviewed in Wave III and reported that their mothers and stepfathers were still living together (n = 881). This sample included both stepfamilies that began as cohabiting partnerships and transitioned into marriage before Wave I and those formed directly through marriage. The comparison sample consisted of adolescents with valid sample weights who reported living with their biological mothers and fathers in Wave I, and who were reinterviewed in Wave III and reported that their biological parents were still living together (n = 6378).

Measures

The dependent variable, stepfather–child closeness, was a categorical variable measuring adolescents’ reported closeness with their stepfathers between Wave I and Wave III. This variable was created using a one-item measure of stepfather–child closeness at Wave I and an identical item at Wave III, which asked respondents to indicate how close they felt to their stepfathers (1 = not at all, 2 = very little, 3 = somewhat, 4 = quite a bit, 5 = very much). Values of 1 to 3 were combined and considered “not close” and values of 4 or 5 were combined and considered “close.” The final variable was then created by running a cross-tabulation of the respective dichotomous indicators of closeness at Wave I and Wave III and creating a four-category variable (1 = remain close, 2 = remain not close, 3 = become close, 4 = become less close between Wave I and Wave III). This coding scheme has been used in previous research using the Add Health data (e.g., Scott, Booth, King, & Johnson, 2007). There were 131 cases in which adolescents did not answer the question about their stepfathers at Wave I because they referred to them as their “mother’s husband.” For these cases, the Wave II value for stepfather–child closeness was used as a substitute for closeness at Wave I. A similar measure of biological father–child closeness was created for the comparative analysis.

The adolescent’s age at Wave I was measured in years ($\stackrel{‒}{x}$ = 15.57, SE = .06). Gender was coded as a binary variable (0 = male; 1 = female, 54%). Race was coded as a set of dummy variables, with White as the reference category (65% White, 14% Black, 14% Hispanic, 8% other race). The mother’s level of education was coded as a set of dummy variables with “less than high school education” as the reference category (14% less than high school education, 33% high school degree or equivalent, 35% some college, 19% college degree). The number of years in the stepfamily at Wave I was measured as a continuous variable ($\stackrel{‒}{x}$ = 8.09, SE = .16). Whether the adolescent lived with parents at Wave III was a binary variable(0 = did not live with their mothers and stepfathers at Wave III; 1 = lived with their mothers and stepfathers at Wave III, 27%).

The mother–stepfather relationship included three items from the Wave I mother interview that were used to create a scale that was standardized to have a mean of 0 and a standard deviation of 1: how happy her current relationship with the stepfather is (1 = completely unhappy to 10 = completely happy; $\stackrel{‒}{x}$ = 8.75, SE = .07), whether they talked about separating in the past year (0 = yes; 1= no, 87%), and how much they fight (1 = a lot to 4 = not at all; $\stackrel{‒}{x}$ = 2.92, SE = .03)

Mother–child closeness was measured with one item from Wave I in which adolescents were asked how close they felt to their mothers (1 = not at all, 2 = very little, 3 = somewhat, 4 = quite a bit, 5 = very much; $\stackrel{‒}{x}$ = 4.57, SE = .03). Change in mother–child closeness is a change score measuring the difference between the one-item measure of mother–child closeness at Wave III and the one-item measure of mother–child closeness at Wave I.

Nonresident father–child relationship quality included three items from Wave I that were used to create a scale (α = .88) that was standardized to have a mean of 0 and a standard deviation of 1: how close adolescents felt to their nonresident father (1= not close at all to 5 = extremely close; $\stackrel{‒}{x}$ = 2.66, SE = .08), how often in the previous 12 months they had stayed overnight with him (0 = not at all to 5 = more than once a week; $\stackrel{‒}{x}$ = 1.13, SE = .09), and how often in the previous 12 months they had talked to him in person or on the telephone, or received a letter from him (0 = not at all to 5 = more than once a week; $\stackrel{‒}{x}$ = 2.05, SE = .10). We also tested alternative specifications of the nonresident father–child relationship that paralleled the measures for mothers (i.e., a single item of how close adolescents felt to their nonresident father in Wave I and a change score measuring the difference between this item at Wave III and Wave I). The results for Wave I closeness were similar to those reported for nonresident father–child relationship quality, and changes in closeness were not significantly associated with closeness patterns to stepfathers.

Family belonging was measured with the mean of four items from Wave I that asked adolescents to rate (1 = not at all to 5 = very much) how much they feel their family understands them, how much they feel like they have fun in their family, how much they want to leave home (reverse coded), and how much they feel their family pays attention to them ($\stackrel{‒}{x}$ = 3.68, SE = .03; α = .77).

Analysis Plan

We began by examining the prevalence of the four patterns of stepfather–stepchild closeness between adolescence and young adulthood (remain close, remain not close, become close, become less close). For comparative purposes, we also conducted a parallel analysis of patterns of closeness to biological fathers for adolescents living in two-biological-parent families.

We then examined the factors associated with these different patterns of closeness to stepfathers. First we examined mean differences for all the predictor variables (adolescent age, gender, race, family income, mother’s education, years in a stepfamily, living with parents, mother–stepfather relationship quality, mother–child closeness, changes in mother–child closeness, nonresident father–child relationship quality, and feelings of family belonging) by group membership in one of the four patterns of closeness to stepfathers. Then we examined group differences in a multivariate framework with multinomial logistic regression. This modeling procedure is appropriate when the dependent variable (four closeness groups) has more than two nominal categories. With four closeness groups, six group comparisons can be made; however, only three bear directly on our research questions (remain close vs. remain not close, become close vs. remain not close, and become less close vs. remain close). To highlight these three comparisons, we reported significant bivariate group differences in a separate column from the other group differences in Table 1. For the multinomial logistic regression model, we report results from the three main contrasts of interest in Table 2, with results from the remaining contrasts reported in Appendix A.

Table 1.

Means (and Standard Errors) for Predictor Variables by Closeness to Stepfathers Group (N = 881)

Variable	Remain close (1) n = 408	Remain not close (2) n = 197	Become close (3) n = 141	Become not close (4) n = 135	Differences of interest p < .05	Other differences p < .05
Age	15.57 (.09)	15.55 (.11)	15.89 (.13)	15.23 (.16)	1 > 4	3 > 4
Female	.51 (.02)	.59 (.04)	.67 (.04)	.45 (.04)	None	2, 3 > 4 1 > 3
Child’s racea
Black	.15 (.02)	.12 (.02)	.12 (.03)	.14 (.03)	None	None
Hispanic	.14 (.02)	.13 (.03)	.15 (.03)	.15 (.03)	None	None
Other race	.08 (.01)	.07 (.02)	.04 (.02)	.12 (.03)	None	None
Log family income	3.67 (.04)	3.62 (.06)	3.74 (.06)	3.56 (.07)	None	None
Mother’s educationb
High school	.35 (.02)	.32 (.03)	.31 (.04)	.31 (.04)	None	None
Some college	.35 (.02)	.34 (.03)	.40 (.04)	.30 (.04)	None	None
College degree	.17 (.02)	.21 (.03)	.16 (.03)	.22 (.04)	None	None
Years in stepfamily	8.95 (.24)	7.22 (.32)	7.85 (.39)	7.05 (.36)	1 > 2, 4	1 > 3
Live with parents at Wave III	.28 (.02)	.22 (.03)	.25 (.04)	.30 (.04)	None	None
Mother–stepfather relationship	.11 (.04)	−.17 (.07)	.06 (.06)	−.10 (.08)	1 > 2, 4 3 > 2	None
Mother–child closeness	4.79 (.03)	4.12 (.07)	4.30 (.07)	4.82 (.04)	1 > 2	1, 4 > 3 4 > 2
Change in mother– child closeness	−.06 (.03)	−.08 (.09)	.26 (.07)	−.68 (.10)	1 > 4 3 > 2	2 , 3 > 4 3 > 1
Nonresident father– child relationship	−.09 (.04)	.17 (.08)	.06 (.08)	−.06 (.08)	2 > 1	2 > 4
Family belonging	3.94 (.04)	3.19 (.06)	3.30 (.06)	3.96 (.05)	1 > 2	1 > 3 4 > 2, 3

^aOmitted reference group is White.

^bOmitted reference group is less than a high school education.

Table 2.

Multinomial Logistic Regression Results for Closeness to Stepfathers (for Group Comparisons of Interest) (N = 881)

	Remain close (1) vs. Remain not close (2)		Become close (3) vs. Remain not close (2)		Become not close (4) vs. Remain close (1)
	b	SE	b	SE	b	SE
Age	.02	(.07)	.13	(.08)	−.03	(.07)
Female	−.26	(.21)	.32	(.25)	−.15	(.22)
Child’s racea
Black	.25	(.33)	−.11	(.38)	.14	(.33)
Hispanic	−.43	(.32	−.19	(.35)	.31	(.33)
Other race	−.13	(.40)	−.62	(.53)	.74*	(.37)
Log family income	.07	(.16)	.26	(.18)	−.17	(.16)
Mother’s educationb
High school	−.54	(.36)	−.36	(.41)	−.14	(.36)
Some college	−.45	(.37)	−.13	(.42)	−.23	(.36)
College degree	−.64	(.42)	−.61	(.47)	.13	(.41)
Years in stepfamily	.09***	(.02)	.02	(.03)	−.10***	(.03)
Live with parents at Wave III	.20	(.25)	.29	(.28)	.18	(.24)
Mother–stepfather relationship	.40***	(.15)	.35*	(.17)	−.41***	(.15)
Mother–child closeness	1.34***	(.20)	.65**	(.19)	−.46	(.27)
Change in mother–child closeness	.82***	(.14)	.69***	(.16)	−1.04***	(.14)
Nonresident father–child relationship	−.42***	(.13)	−.15	(.15)	.02	(.13)
Family belonging	1.13***	(.16)	.09	(.18)	.03	(.18)

Note. b = unstandardized coefficient. SE = standard error.

^aOmitted reference group is White.

^bOmitted reference group is less than a high school education. F(48, 120.5) = 6.40, p < .000.

^*p < .05,

^**p < .01,

^***p < .001.

Results are based on weighted data, with standard errors adjusted for clustering and stratification in the Add Health sample design using STATA Version 13. Missing data were handled using the multiple imputation procedure ICE.

Results

Patterns of Closeness

How common are different patterns of stepfather–stepchild closeness between adolescence and young adulthood?

The most common pattern (46%) was one in which stepchildren reported feeling close to stepfathers in adolescence and in young adulthood. Stability also characterized the next-largest group (23%), but at the opposite extreme, with stepchildren not feeling close to stepfathers in adolescence or in young adulthood. Almost as many adolescents reported becoming less close to stepfathers over time (close in adolescence but not close in young adulthood, 15%) as reported improving relationships (not close in adolescence) to the point of feeling close to stepfathers in young adulthood (16%).

How do closeness patterns in stepfather families compare to patterns in two-biological-parent families?

Although stability in father–child closeness was also more common than change for adolescents living with two biological parents, closeness patterns diverged markedly from those of adolescents in stepfather families in their prevalence. The majority of adolescents, 74%, reported feeling close to their biological fathers in adolescence and in young adulthood. Only 6% reported not feeling close to their fathers at both time points. Somewhat fewer adolescents reported changes in closeness to fathers in two-biological-parent families than did adolescents in stepfather families, although the percentage who reported becoming close to their biological fathers (9%) was fairly similar to the percentage who reported becoming less close to their biological fathers over time (11%).

Predicting Patterns of Closeness

Table 1 reports bivariate results, with the means for all of the predictor variables displayed by group membership in one of the four patterns of closeness to stepfathers. Table 2 reports multivariate results from the multinomial logistic regression models for the group comparisons of interest. Because the bivariate and multivariate results are very similar with regard to significant differences, we focus on the results in Table 2 in the following discussion of the findings, noting the one instance in which a significant bivariate difference evident in Table 1 became nonsignificant in the multivariate model.

What factors differentiated stepchildren who reported feeling close to their stepfathers in both adolescence and young adulthood from those who did not feel close to their stepfathers at either time?

Stepchildren who consistently reported feeling close to their stepfathers had spent more years of their childhood living with stepfathers (see column 1 in Table 2). These youth also reported feeling closer to their mothers in adolescence and were more likely to report that they became closer to their mothers in young adulthood than were youth who were not close to their stepfathers. In addition, these youth reported stronger feelings of family belonging in adolescence, and their mothers reported having higher-quality relationships with stepfathers. Stepchildren who reported having lower-quality relationships with nonresident fathers during adolescence were also more likely to remain close to their stepfathers.

Which factors differentiated stepchildren whose relationships with stepfathers improved from adolescence to young adulthood from those who continued to report not feeling close to their stepfathers?

Stepchildren who reported becoming closer to their stepfathers in young adulthood were significantly more likely to report that they were closer to their mothers in adolescence, and more likely to report becoming closer to their mothers in young adulthood, than were stepchildren who continued to report not feeling close to their stepfathers. The mothers of children who reported becoming close to their stepfathers in young adulthood also reported having higher-quality relationships with stepfathers when the children were adolescents.

Which factors differentiated stepchildren whose relationships with stepfathers became less close from adolescence to young adulthood from those who continued to report feeling close to their stepfathers?

Stepchildren who reported no longer feeling close to their stepfathers in young adulthood were significantly less close to their mothers in young adulthood as well than were stepchildren who continued to report feeling close to their stepfathers. These stepchildren also spent fewer years living with stepfathers before the Wave I adolescent interview, and their mothers were significantly more likely to report having lower-quality relationships with the stepfathers. The bivariate results suggested that adolescents who no longer felt close to stepfathers in young adulthood were younger at the time of the Wave I interview than those who remained close, but this difference was no longer significant in the multivariate model.

Which factors were unrelated to patterns of stepfather–stepchild closeness from adolescence to young adulthood?

Whether the stepfather–stepchild relationship remained close, remained not close, increased to become close, or declined to no longer be close did not appear to depend on the child’s age, gender, or race, family income, the mother’s education level, or whether the child was living with the mother and stepfather at the time of the Wave III interview. Consistent with a family systems perspective, the most significant predictors of the closeness of the stepfather–stepchild relationship were largely those that captured the quality of other family relationships.

Several other factors were tested in preliminary analyses. None was significantly associated with closeness patterns, and they were not included in the final model in the interest of parsimony. Wave I variables included the presence of full, step, and half siblings; whether the adolescent was U.S. born; the stepfather’s education level; the mother’s number of prior relationships; whether the adolescent was born in marriage; whether the adolescent called the stepfather “mother’s husband” at Wave I; and adolescent religiosity. We also tested several variables from Wave III that captured the adolescent’s own family formation experiences, including whether they had children of their own and whether they were currently married or cohabiting, as well as whether they were currently in school at Wave III. Despite some suggestion in the literature that parent–child relationships become more positive as children move into adult roles such as marriage and parenthood (Aquilino, 1999; Fischer, 1981; Umberson, 1992), we found no association between family formation experiences and patterns of closeness to stepfathers.

Discussion

The current study examined stepchildren’s reports of closeness to their stepfathers in adolescence and young adulthood in married mother–stepfather families, with an eye toward identifying factors that predict patterns of closeness over time. Results revealed much variability in how these relationships unfolded over the transition to adulthood, but the most common pattern (46%) was one in which stepchildren remained close to their stepfathers. It is this group of youth in stepfather families who may fare best in the transition to adulthood by being able to draw on the social capital inherent in positive parent–child relationships, and they may attain positive outcomes more similar to those of children growing up with two biological parents.

The next-largest group (23%) reported not feeling close to their stepfathers, which persisted over time. Although stability in stepfather–stepchild relationships over time was most common, some stepchildren (16%) reported that they became closer to their stepfathers in young adulthood, and almost as many (15%) reported that they became less close to their stepfathers. It should be kept in mind that these results are based on mother–stepfather families that remained married over time, which likely overestimates the percentage of positive relationships that would be found in the larger population of U.S. stepfamilies. Future research should examine how stepfather–stepchild relationships unfold in cohabiting stepfamilies and in stepfamilies in which unions dissolve.

With one notable exception, these patterns were consistent with the six patterns of stepchild–stepparent relationship development that emerged in Ganong et al.’s (2011) study of adult stepchildren. Two of their typologies (accepting as a parent, liking from the start) correspond most closely with our “remain close” group. Three of their typologies (rejecting the stepparent, coexisting, and accepting with ambivalence) likely have much in common with our “remain not close” group. The final typology in their study (changing trajectories), however, was largely composed of relationships that moved from disliking the stepparent to an improved relationship, similar to our “become close” group. Yet we also found a group that experienced becoming less close in adulthood, and in our national sample it was as prevalent as the group that became close. Thus, although stability in relationship closeness was more common, it appears that a fair number of stepchildren eventually become closer to their stepfathers and that a number of stepchildren become less close. Future research would benefit from more attention to the factors that are associated with both of these types of changes.

Ganong et al. (2011) noted that improvement in stepparent–stepchild relationships did not appear to be due to a single event but rather was something that unfolded over time, as stepchildren started to see how stepparents were benefiting them or their parents (or both). A similar process may have occurred for many of the stepchildren in our “become close” group, although we could not explore this with the Add Health data. It is unclear whether stepchildren who became less close did so gradually or whether certain events triggered this change. Future research would also benefit from an examination of how these trajectories are associated with stepchild (and stepparent) well-being. Although we might expect that stepchildren who remain close to their stepfathers will be advantaged relative to stepchildren who remain not close, it is less clear whether relationships that become close are linked with better outcomes for youth making the transition to adulthood or whether there are any consequences for youth who become less close to their stepfathers in adulthood.

The analysis of closeness patterns for adolescents living in two-biological-parent families was also instructive, illustrating that children in stepfamilies are at greater risk than those in two-biological-parent families for experiencing relationships that become less close over time, or of never having had a close relationship in the first place. Although we know from prior research that father–child relationships in two-biological-parent families are closer on average than stepfather–stepchild relationships (Lansford, Ceballo, Abbey, & Stewart, 2001), results from the current study highlight the greater variability in stepfamily relationships, as well as the greater likelihood of experiencing a decline in closeness to stepfathers in adulthood despite their parents remaining in stable marriages. The importance of parent–child ties for positive child outcomes in both adolescence and young adulthood, coupled with the fact that the vast majority of children in two-biological-parent families feel close to their biological fathers in adolescence and only a relatively small percentage experience a decline in closeness by young adulthood, suggests one possible mechanism underlying the relatively advantaged position and better outcomes that youth in two-biological-parent families evidence in both adolescence and during the transition to adulthood.

Whether stepfather–stepchild relationships remained close, remained not close, became close, or became less close, two factors stood out in distinguishing between these four patterns of closeness from adolescence to young adulthood: the closeness of the mother–child relationship and the quality of the relationship between the mother and the stepfather. Prior research had suggested that mother–child relationships are key for understanding children’s ties to stepfathers, with several studies reporting cross-sectional associations between the quality of the mother–child and stepfather–child relationships (e.g., Dunn et al., 2004; King et al., 2014). A few studies also have reported that closeness to mothers before a stepfather’s entry is associated with adolescents reporting closer ties to stepfathers within the first year of stepfather entry (e.g., King, Amato, & Lindstrom, 2015). The current study expands this line of evidence by prospectively considering patterns of children’s relationships with mothers and stepfathers between adolescence and young adulthood.

Our findings suggest that mother–child closeness in adolescence is positively associated with either remaining close to stepfathers or becoming closer to stepfathers during the transition to adulthood. Furthermore, changes in mother–child closeness appear to go hand in hand with changes in stepfather–child closeness. Adolescents who became closer to their mothers over time were significantly more likely to either remain close to their stepfathers or to become closer to them in young adulthood (compared to remaining not close to them). Likewise, adolescents who reported becoming closer to their mothers over time were significantly less likely to decline in closeness to their stepfathers during young adulthood. The current study cannot determine the direction of causal influence, but both prior research (e.g., King, 2009) and theory (e.g., Doherty, Kouneski, & Erickson, 1998) suggest that a large part of the association runs from the mother–child relationship to the stepfather–child relationship rather than the other way around. As we are not aware of any other studies that have examined how changes in mother–child relationships are linked to changes in stepfather–stepchild relationship during the transition to young adulthood, more attention to this issue is warranted.

A high-quality marriage between the mother and the stepfather was also a significant predictor of patterns of stepfather–stepchild closeness, and the current study benefited from having reports of marital quality from the mothers themselves (ruling out the possibility that this association was a result of children who felt close to their stepfathers perceiving that their parents had better marriages). Mothers who reported higher-quality relationships with stepfathers in the Wave I interview were more likely to have children who either remained close to stepfathers over time or became close to them in young adulthood, and they were less likely to have children who reported that stepfather–stepchild relationships were no longer close in young adulthood.

Previous findings on the association between adolescents’ relationships with nonresident biological fathers and resident stepfathers have been mixed, and a number of studies have suggested that adolescents’ relationships with each father are largely independent (e.g., King, 2009). In the current study, however, we found that the quality of the nonresident father–adolescent relationship differentiated children who reported feeling close to their stepfathers in both adolescence and young adulthood from those who did not feel close to their stepfathers at either time. This suggests that ties with nonresident biological fathers may have long-term effects that are masked in the short term or in cross-sectional studies.

Even after controlling for the closeness of the mother–child relationship and the quality of the mother–stepfather relationship, the adolescent’s feelings of family belonging distinguished between those who remained close to their stepfathers and those who remained not close. Researchers have suggested that although family belonging may be more challenging to attain in stepfamilies than in two-biological-parent families, it may be particularly beneficial for positive stepfamily functioning and stepchild well-being (King, Boyd, & Thorsen, 2015). Future research should direct more attention to this relatively understudied construct.

Taken together, the significant associations we found between other family relationships and the development of stepfather–stepchild closeness are consistent with family systems theory and the tenet that all parts of the family system are interconnected and influence one another. The only nonrelationship factor found to be significantly associated with closeness to stepfathers was the length of time the stepfamilies had lived together, a finding consistent with prior research. Our study is unable to illuminate the key mechanisms at play, but they likely involve a combination of factors that are fostered by longer durations such as more time spent together, affinity-building behaviors, acceptance of the stepparent as a parent figure, and coming to appreciate what the stepfather brings to the family (Ganong et al., 2011).

A number of limitations of the current study are worth noting. Our study examines only one aspect of the stepfather–stepchild relationship (closeness), which is assessed at only two time points and relies on stepchild reports. Future research would benefit from a richer set of measures that could examine multiple aspects of this relationship at multiple time points. Future research would also benefit from including the perspective of stepfathers. Although beyond the scope of the current study, future research would also benefit from considering other stepfamily forms and by following stepfamilies that dissolve over time. Our results speak to stepchildren living in stable married mother–stepfather families and may not apply to other stepfamily types.

The current study contributes to the growing literature on factors associated with the development of close stepfather–stepchild ties. We extend prior research in this area by using nationally representative data to prospectively examine stepchildren’s closeness to their stepfathers at two critical junctures in the life course: adolescence and young adulthood. Our findings revealed much variability in how these relationships unfolded over time, although a sizable minority of youth remained close to their stepfathers over the transition to adulthood. Consistent with family systems theory, the quality of other family relationships are particularly important in understanding children’s relationships with stepfathers as they develop over time.

Appendix 1. Multinomial Logistic Regression Results for Closeness to Stepfathers (for Other Group Comparisons) (N = 881)

	Become not close (4) vs. Remain not close (2)		Become close (3) vs. Remain close (1)		Become not close (4) vs. Become close (3)
	b	SE	b	SE	b	SE
Age	−.01	(.18)	.11	(.07)	−.13	(.09)
Female	−.41	(.26)	.58*	(.23)	−.73**	(.28)
Child’s racea
Black	.39	(.41)	−.36	(.35)	.50	(.43)
Hispanic	−.12	(.37)	.24	(.34)	.07	(.41)
Other race	.60	(.44)	−.49	(.49)	1.22*	(.54)
Log family income	−.04	(.19)	.19	(.17)	−.30	(.21)
Mother’s educationb
High school	−.68	(.43)	.18	(.37)	−.32	(.46)
Some college	−.68	(.44)	.31	(.38)	−.55	(.46)
College degree	−.51	(.49)	.04	(.44)	.09	(.52)
Years in stepfamily	−.01	(.03)	−.06*	(.03)	−.04	(.03)
Live with parents at Wave III	.38	(.30)	.09	(.26)	.09	(.31)
Mother–stepfather relationship	−.02	(.17)	−.04	(.17)	−.37	(.19)
Mother–child closeness	.84***	(.26)	−.68**	(.22)	.19	(.28)
Change in mother–child closeness	−.22	(.13)	−.13	(.17)	−.91***	(.18)
Nonresident father–child relationship	−.40*	(.16)	.27*	(.14)	−.25	(.17)
Family belonging	1.16***	(.21)	−1.04***	(.17)	1.07***	(.22)

Note. b = unstandardized coefficient. SE = standard error.

^aOmitted reference group is White.

^bOmitted reference group is less than a high school education. F(48, 120.5) = 6.40, p < .000.

^*p < .05,

^**p < .01,

^***p < .001.