Kay Pasley & Brad S Moorefield. Handbook of Contemporary Families. Editor: Marilyn Coleman & Lawrence H Ganong. Sage Publication. 2004.
Given the changing social, legal, and economic contexts (Arnup, 1999), it is not surprising that the heterogeneity of the American family has become more visible. This diversity is evident in increases in the numbers of people that have experienced divorce, remarriage, and cohabitation since the 1970s (Teachman, 2000), and academic discourse about this diversity has followed these trends.
One outcome of the increased diversity is greater attention given to the study of stepfamilies. Historically, stepfamilies were most often formed following the death of a parent; now they usually follow parental divorce. In addition, stepfamilies are more broadly defined than in the past, reflecting their greater diversity. Stepfamilies are now recognized as consisting of any biological or adoptive parent (heterosexual, gay, or lesbian) with a child from a prior relationship who elects to marry or to cohabit (the latter have been called nonlegal stepfamilies; Bumpass, Raley, & Sweet, 1995).
In this chapter, we briefly summarize the literature, noting insights gleaned especially from research since the new millennium. We identify conceptual, methodological, and empirical topics that need to be examined more completely, and we offer suggestions for addressing them.
Remarriage and stepfamilies were the focus of few research efforts before 1970. In fact, beyond two articles addressing the incidence of these marital transitions (Glick & Norton, 1971; Troll, 1971), few studies were published on stepparents in these early years. The body of knowledge regarding stepfamilies has grown immensely since then. By 1980, stepfamilies had received some attention by researchers, as evident in decade reviews on nontraditional family forms (Macklin, 1980) in the Journal of Marriage and the Family (JMF) that included stepfamilies. In another JMF review, stepfamilies were classified as noninstitutions, and terms like reconstituted, blended, and binuclear families were used to describe them (Price-Bonham & Balswick, 1980). By 1990, there were well over 200 scholarly publications from which to draw in a JMF decade review, justifying an entire article addressing research on remarriage and stepfamilies (Coleman & Ganong, 1990). However, the 1990s were seen as “a period of enormous productivity in the study of remarriage and stepfamilies” (Coleman, Ganong, & Fine, 2000, p. 1288), and the resulting review was derived from the examination of over 850 publications. Thus, an increasing number of studies are contributing to the knowledge base on stepfamilies. (For a comprehensive bibliography of the empirical literature on stepfamilies, visit the Stepfamily Association of America Web site at www.saafamilies.org.)
The Population Profile of the United States: 2000 (U.S. Bureau of the Census, 2002), weighting the 1996 Survey of Income and Program Participation (SIPP) data by population controls based on the decennial census, reported that 17% of all children in the United States under the age of 18 lived in a remarried stepfamily household and another 5% lived with a parent and the parent’s cohabiting partner. Because these figures exclude nonresident stepfamilies, our best guess is that legal and nonlegal resident and nonresident stepfamilies compose about 30% of the households containing children under 18 years of age. These are estimates because the National Center for Health Statistics suspended collection of detailed information on marriage, divorce, and remarriage in 1996, making it is impossible to identify accurately the number and characteristics of remarriages and stepfamilies. Further, recent estimates, based on national surveys completed before the 2000 decennial census, contain inconsistencies because data were gathered retrospectively or because different comparison groups were used in various reports (e.g., data from both men and women or from women of certain age ranges). In addition, the number of children residing with a biological parent and stepparent cannot be determined from many of the reports from the U.S. Bureau of the Census. Data typically are reported on whether children are living with two adults, but the relationship of the adults to the children in the household is not known.
Kreider and Fields (2002), reporting findings from a nationally representative sample of men and women ages 15 years and older in 1996 from the SIPP data, estimated that about 50% of all first marriages ended in divorce (50% for men regardless of age group, and 44-52% for women depending on the referenced age group). Further, 54% of men and 60% of women had been married once, about 13% of both men and women had been married twice, and about 3% of both genders had been married three or more times. The median duration of first marriages that ended in divorce was 7.8 years, time from first separation to divorce was 0.8 years, median time between divorce and remarriage was 3 years, and duration of remarriages that ended in divorce was 7.3 years for men and 6.8 years for women. These results are similar to data from earlier periods (e.g., Norton & Moorman, 1987), with two notable exceptions: In the more recent data, individuals were spending about 1 year longer between divorce and remarriage, and remarried couples that had divorced remained together longer than the median of 4.5 years reported earlier (Norton & Miller, 1993).
Recent reports from the National Survey of Family Growth showed that black women experienced the highest rate of relationship disruption and were the least likely to remarry; they also were more likely to end a second marriage (Bramlett & Mosher, 2002). Social class indicatorslower median family income, male unemployment, poverty, and receipt of welfare also are associated with instability in remarriage for women, as is the presence of children (Bumpass, Sweet, & Martin, 1990; Furstenberg & Cherlin, 1991; Wilson & Clarke, 1992).
Research on Remarriage and Stepfamily Adjustment
Before the 1970s, mentions of the quality of remarriages or stepfamily adjustment were rare. By 1980, there were several studies with conflicting results regarding whether individuals in remarriages were more or less happy than those in first marriages (e.g., Glenn & Weaver, 1977; Renne, 1971; White, 1979). During the 1980s, the few researchers that focused on marital processes or stepfamily functioning (Coleman & Ganong, 1990) generally found few differences in marital satisfaction (e.g., Vemer, Coleman, Ganong, & Cooper, 1989) and well-being (e.g., Nock, 1981; Weingarten, 1980) between individuals in remarriages and first marriages.
Stepfamilies were noted to be less cohesive and less effective problem solvers than nuclear families, but most were not in the clinical range for these behaviors (e.g., Anderson & White, 1986; Bray, 1988; Peek, Bell, Waldren, & Sorell, 1988). Most studies in the 1980s were between-group designs that ignored stepfamily complexity, continuing a deficit-model approach and upholding the nuclear family as the ideal family structure.
By the 1990s, there was an increased focus on marital dynamics, quality, and stability among remarriage researchers (Coleman et al., 2000). Studies using larger samples as well as qualitative designs increased in number. Cohabitation before remarriage was found to be a common part of courtship, but research was lacking on decision-making processes regarding cohabitation and the effects of cohabitation on stepfamily life (Coleman et al., 2000). Few researchers examined the processes and mechanisms for building satisfying remarriages or for disrupting these relationships. However, researchers in the last decade of the 20th century showed more interest in remarriage dynamics than they had in earlier periods. They reported that decision making was perceived to be shared between spouses (e.g., Crosbie-Burnett & Giles-Sims, 1991; Pyke & Coltrane, 1996); that remarried couples expressed criticism, anger, and irritation more openly than first-married couples (Bray & Kelly, 1998; Hetherington, 1993); that men conceded more during marital conflicts than they did in their first marriages (Hobart, 1991); and that women had more power in financial decisions in remarriage than they had in their first marriages (e.g., Burgoyne & Morrison, 1997). Moreover, consensus-building behaviors were found to be important to stepfamily adjustment (Bray, Berger, & Boethel, 1994; Ganong & Coleman, 1994b), and expecting the remarriage and stepfamily to operate like a first-marriage family could be problematic (Bray & Kelly, 1998).
Research examining gay and lesbian stepfamilies expanded somewhat in recent years. These studies were characterized by use of qualitative methods and small samples (Lynch, 2000). The presence of a child was associated with increased relationship satisfaction and intimacy for lesbian couples (Koepke & Hare, 1992), and stepfamily satisfaction was associated with gay stepfathers’ feeling included and experiencing fewer problems with children’s movement between households (Crosbie-Burnett & Helmbrecht, 1993).
Early authors identified the major problem areas for stepparents as children, finances, and the ambiguity surrounding stepparenting (e.g., Bohannon, 1970; Duberman, 1975; Messinger, 1976). Most stepparenting studies have focused on stepfather families. Interestingly, an early study (Goldstein, 1974) suggested that being a stepmother was less difficult than being a stepfathera conclusion that has not received much support since then (e.g., Bray, 1988; MacDonald & DeMaris, 1996).
In general, a common conclusion from research in the 1980s and 1990s is that being a stepparent is more difficult than raising one’s own children, especially when the stepparent is a stepmother (e.g., MacDonald & DeMaris, 1996). Variations in role expectations and behaviors present problems for stepparents in deciding what their roles are or should be in the lives of stepchildren, especially early in remarriage (Bray & Kelly, 1998; Hetherington & Kelly, 2002), but typically there is couple consensus that involvement should increase over time and that stepparents should and do assume more parental responsibilities (Bray & Berger, 1993; Ganong & Coleman, 1994a). This pattern is characteristic also of gay- and lesbian-headed stepfamilies; however, Lynch (2000) found homosexual stepparents to be more flexible in adopting parenting roles than heterosexual stepfamilies.
Stepparents are provided greater latitude than biological parents in what is perceived as acceptable behavior. For example, stepfathers are expected to assume less responsibility for the care and control of stepchildren than are fathers (e.g., Fine, 1995; Hetherington & Kelly, 2002). However, if greater latitude is not coupled with realistic expectations, then stepparenting is difficult (Bray & Kelly, 1998).
Researchers have shown interest in identifying factors that affect stepparenting (e.g., age and sex of the child and sex of the stepparent). Stepparents have more conflicts with adolescent stepchildren than with younger stepchildren (Bray & Kelly, 1998; Golish & Caughlin, 2002; Hetherington & Kelly, 2002; MacDonald & DeMaris, 1996). Adolescents react negatively to attempts by stepparents to limit or control their behavior (Bray & Kelly, 1998; Golish & Caughlin, 2002; Hetherington & Kelly, 2002), even when the stepparent uses strategies characteristic of an authoritative parenting style (e.g., warmth, control). Parenting a stepdaughter is more challenging than parenting a stepson, as relationships between stepparents and stepdaughters are characterized by more conflict and negative interactions than are relationships between stepparents and stepsons (e.g., Hetherington & Clingempeel, 1992). In one of the few observational studies, Vuchinich, Hetherington, Vuchinich, and Clingempeel (1991) found that stepdaughters avoided conflicts and other interactions with stepfathers; however, when stepfathers initiated conflict, stepdaughters’ oppositional responses persisted. Such extended conflicts are consistent with reports in earlier studies of hostile, resentful withdrawal by stepdaughters (e.g., Bray, 1988; Hetherington, 1989).
When the stepparent is a resident stepmother, conflict and negativity are exacerbated (Bray, 1988; MacDonald & DeMaris, 1996). In part, this may stem from implicit expectations associated with women’s family roles and the structural complexity of stepmother families (Dunn, Davies, O’Connor, & Sturgess, 2000). For example, as stepmothers take on an inequitable share of family work, as they believe they should (Ceglian & Gardner, 2000; Orchard & Solberg, 1999), their relationships with stepchildren suffer (Ceglian & Gardner, 2000; Fine, Donnelly, & Voydanoff, 1991).
Stepfathers typically develop more positive relationships with stepchildren than stepmothers do, especially with stepsons over time. This may be because they make fewer attempts to monitor or control stepchildren’s behaviors early in remarriage (Bray & Kelly, 1998). Hetherington (1993) noted that stepfathers control their own negative feelings better than biological fathers do, and they attempt to establish a relationship with their stepchildren by self-disclosing and searching for common interests and experiences all in the face of aversive behavior from their stepchildren. There is no evidence that these strategies are used by stepmothers or, if used, are successful.
Recently, researchers have examined linkages among relationships within the family. For example, MacDonald and DeMaris (2002) found (a) that when the quality of the mother-child relationship was good, the stepfather-stepchild relationship benefited; and (b) that conflict between the biological parents was related negatively to the quality of the stepfather-stepchild relationship. Other researchers reported that mother-child relationship quality was related to the quality of the mother’s relationship with her partner and that the termination of that partnership was associated with poorer parenting quality (Thomson, Mosley, Hanson, & McLanahan, 2001).
We think that extant research findings do not adequately explain why stepparent-stepchild relationships are so challenging. Speculations offered by researchers suggest that stepparents’ problems stem from stepchildren’s perception that they are being treated differently than biological children (Clingempeel, Coylar, & Hetherington, 1994; Ganong & Coleman, 1994a). Findings indicate that (a) stepparents communicate more poorly with stepchildren than with their own children, (b) stepparents feel less warmth toward stepchildren and vice versa, (c) stepparents express fewer positive feelings and make fewer positive comments to stepchildren, and (d) stepparents provide less support and monitoring of stepchildren (Bray et al., 1994; Ganong & Coleman, 1994b; MacDonald & DeMaris, 1996; Mekos, Hetherington, & Reiss, 1996; Thomson, McLanahan, & Curtis, 1991). Poor communication, lack of warmth, fewer positive expressions, and less support and monitoring may combine to foster children’s feeling more disconnected and more resistant to change (parental remarriage) and agents of change (stepparents). It also has been suggested that part of this differential treatment results from the competition that occurs between stepparents and stepchildren as they vie for attention from the biological parents (Saint-Jacques, 1995). However, these speculations imply a unidirectional effect (i.e., stepparent behaviors affect children) rather than bidirectional effects (i.e., stepparents and stepchildren mutually influence each other’s behavior). Golish and Caughlin (2002) reported that adolescent stepchildren communicated infrequently with stepparents and avoided confiding in them as a means of reducing potential conflict, suggesting that stepchildren’s behavior plays a role in interactions with stepparents. Braithwaite, Olson, Golish, Soukup, and Turman (2001) found that failure to set boundaries and establish trust in the stepparent-stepchild relationship was associated with increased conflict, loss of hope, and lack of cohesion. Studies exploring bidirectional interactions (e.g., Vuchinich et al., 1991) are needed, especially addressing changes over time.
Studies are also needed that examine intergenerational relationships between older remarried parents, stepparents, and adult stepchildren. Although there have been some studies of intergenerational relationships (e.g., Ganong & Coleman, 1999; Spitze & Logan, 1992; Marks, 1995), particularly of transfers of resources (e.g., economic, informal care; Goldscheider, Thornton, & Li-Shou, 2001), more research is needed on stepfamilies across the life course.
Research on Child Outcomes
There has been tremendous growth in the number of studies in which the effects of remarriages and stepfamilies on children have been examined. Coleman and Ganong (1990) suggested that two paradigms, a problem-oriented perspective and a normative-adaptive perspective, characterized this research, with the former receiving the greatest attention. Typically, problem-oriented studies included between-group comparisons of psychological (e.g., self-esteem) and behavioral problems in children residing in stepfamily households, nuclear family households, and single-parent-headed households. Normative-adaptive studies tended to examine children’s outcomes and stepfamily processes like emotional closeness and conflict. By 2000, over one third of the studies “dealt with the effects on children of living with a remarried or cohabiting stepparent” (Coleman et al., 2000, p. 1292), and many used data from national, longitudinal studies (e.g., National Survey of Families and Households). Typical outcome measures included multiple indicators of academic performance, internalizing behaviors, and externalizing behaviors. The results of a meta-analysis (Amato, 1994) indicated that differences in children’s outcomes by family structure were small and that most stepchildren did not have psychological and behavior difficulties or school-related problems. Since the review by Coleman et al. (2000), researchers continue to report mixed findings regarding stepchildren’s outcomes: In some studies, there are no differences between stepchildren and other children (e.g., Ackerman, D’Eramo, Umylny, Schultz, & Izard, 2001); in other studies, stepchildren are at a deficit (e.g., Aquilino & Supple, 2001; Wills, Sandy, Yaeger, & Shinar, 2001); and in other studies, differences disappear when demographic controls are included (e.g., Hoffman, 2002).
A good deal is known about factors associated with children’s outcomes, such as demographic and other characteristics of children and their parents that predate the remarriage (Amato & Booth, 1996; Simons & Associates, 1996). Findings also show that multiple transitions and marital conflict are related to poor adjustment for stepchildren (Hanson & McLanahan, 1996; Kurdek, Fine, & Sinclair, 1995). However, few researchers have examined causal linkages among various factors and stepchildren’s outcomes (e.g., Buchanan, Maccoby, & Dornbusch, 1996).
Some studies have provided information on factors that may protect stepchildren from poor outcomes. Coughlin and Vuchinich (1996) found in a prospective study that good family problem solving protected boys in stepfamilies from increased risk for delinquency but that problem solving did not offer the same positive effect for children in two-parent nuclear families or single-mother families. Other research suggests that stepparents may buffer the negative outcomes of divorce on children through improving their economic status (McLanahan & Sandefur, 1994). Still other research (Bray & Kelly, 1998; Hetherington & Clingempeel, 1992) showed that a stepparent who assumes the role of monitor (e.g., camp counselor) is less likely to exacerbate child-related problems than are stepparents who quickly attempt to take control and to initiate many changes in household rules and routines. Parents’ and stepparents’ use of authoritative parenting involving warmth, high monitoring, low coerciveness, firm but responsive control and demands, and expectations for mature behavior has been associated with less externalizing and internalizing behavior and greater social and academic competence in children (Hetherington, 1993; Hetherington & Kelly, 2002).
It would be difficult to accurately predict which children resided in stepfamilies and which resided in first-marriage families on the basis of behavioral outcomes (e.g., academic achievement, aggression, anxiety, depression, delinquent acts). However, this is not to suggest that children in stepfamilies are problem free. Findings from nationally representative samples in both the United States and Britain suggest that about 20% of stepchildren are at risk for negative outcomes somewhat more than children living with both biological parents (Cherlin et al., 1991). This represents a sizable portion of children and deserves attention and concern (Bray, 1999).
We think it is essential that demographers and other researchers keep stepfamilies’ structural complexities at the forefront of new data collection efforts. A primary concern with available demographic data is that the number of stepfamilies is underestimated. Stepfamilies often are omitted from demographic studies when residential stepchildren are older than 18 and when stepchildren reside elsewhere (i.e., with their other parent, a grandparent, on their own). In addition, stepfamilies headed by cohabiting adults often are designated as single-parent families. The lack of accurate stepfamily demographic data makes it difficult for researchers to determine the representativeness of local samples. If researchers are to successfully study the complexity of stepfamilies, reliable demographic information is needed.
Remarriage and Stepfamily Interaction Research
In additional to the need for demographic information, we think that studies are needed that identify stepfamily processes and strategies related to making effective decisions, resolving problems, and effectively negotiating solutions, particularly involving issues related to children. Searching for similarities and differences in processes among varied stepfamily structures and within different cultural contexts is an essential activity. Some questions worthy of study include:
- Are there common strategies used in making decisions and resolving problems in stepfamilies? Do strategies vary in different structural variations of stepfamilies?
- How does the remarried stepfamily process differ from that in cohabiting stepfamilies and first-marriage families?
- How does the intersection of socio-economic status and race affect stepfamily processes?
- What influence, if any, do interactions from previous relationships have on remarriages?
- What factors contribute to remarital adjustment?
- How does the intersection of gender and parental status affect remarriage? How does this intersection affect the way in which conflict is resolved and finances are managed?
Although some researchers suggest that the greater instability of remarriages may be due in part to spouses’ personality characteristics, this has not been examined thoroughly (see Coleman et al., 2000, for a discussion). Important for future efforts are studies that tease out the nature, level, and processes through which personality characteristics interact with demographic characteristics and marital interaction patterns to explain marital interaction and stability. Because scholars often examine few explanatory variables (e.g., either personality characteristics or demographic characteristics), testing more comprehensive models of the etiology and outcomes of marital interaction is needed. Only by doing so will we understand developmental pathways and conditions that affect these pathways (e.g., redivorce).
Between-group designs can be used to address potential differences between first marriages and remarriages. For example, it would be worthwhile to identify the threshold under which the balance between exchanges of negative messages (e.g., criticism) and positive messages (e.g., concession; Gottman, 1994) is disrupted by husbands’ willingness to concede to wives during arguments (Hobart, 1991) and prompts movement toward marital dissatisfaction and instability.
There is limited research on the effects of multiple relationship transitions beyond those that are legal marriages. Future researchers need to examine the complexity of relationship transitions because many individuals experience multiple relationships between marriages, and others elect not to marry at all but experience a succession of cohabiting relationships.
Intersecting Systems and Stepfamily Research
Few scholars have addressed the ways in which various systems (e.g., family, work, school) affect remarriages and stepfamilies. Presser (2000), a notable exception, found that nonstandard work schedules (e.g., working evenings, nights, or rotating schedules and weekends) increased marital instability only for those with children and that remarried women married 5 or more years were more likely to divorce. Although this study represents an effort to connect work and stepfamilies, future research could address how nonstandard work schedules affect different types of stepfamilies, how stepfamilies are affected when only one or both partners have nonstandard work schedules, and the influence of competent child care on work-stepfamily intersections. Research on the intersections of work, income, and race should also be undertaken with stepfamilies to provide insight into work-stepfamily linkages.
Another systemic link of importance is between current and prior family relationships. For example, a recent study showed that for women in stepfamilies, economic obligations to stepchildren, jealousy of husbands’ former wives, and lack of support from friends and family were associated with less marital happiness, more thoughts of divorce, and regrets regarding marriage (Knox & Zusman, 2001). Although there are some studies regarding co-parenting after divorce (e.g., Stewart, 1999), little is known about co-parenting relationships after remarriage (Buehler & Ryan, 1994). For instance, there is no research on how co-parenting following remarriage affects children who reside in step-family households compared to those who live elsewhere and visit stepfamily households.
There is even less knowledge about the intersection between stepfamilies and extended kin, such as grandparents, aunts, and cousins. Although there have been are a few recent studies about extended kin relationships (e.g., Mills, Wakeman, & Fea, 2001), these studies have not shed much light on the dynamics of such relationships or their effects on stepfamily members. The dearth of information on these relationships might suggest that stepfamilies reside in isolation or are immune to the effects of relationships with extended family members; stepfamily researchers’ pragmatic decisions to limit data collection to household residents prevents scholars from exploring these relationships.
Child Outcome Research and Stepfamilies
We agree with other scholars who contend that stepfamily researchers persist in focusing on negative outcomes for stepchildren (Bray, 1999; Coleman et al., 2000). Moreover, most researchers focus on statistical significance and fail to address the magnitude of effects. Bray (1999) argued that many of the statistically significant differences between children from different family structures, including stepfamilies, are due to developmental changes rather than deficits resulting from residing in a particular family structure. We question the value of continuing to ask about differences in child outcomes by family structure unless there are greater attempts to employ conceptual models that address the processes and mechanisms by which certain outcomes result. In place of exploring between-group differences, greater attention must be directed toward making within-group comparisons in which the diversity of developmental pathways experienced by stepchildren and stepfamilies and the conditions that enhance or inhibit certain outcomes are examined.
Theoretical and Methodological Needs
Researchers have long called for the use of theories (e.g., systems, life course, exchange, and stress; Price-Bonham & Balswick, 1980) and for increased theory-building efforts in the study of stepfamilies (Coleman & Ganong, 1990). Some of these calls have been heeded. For example, Robila and Taylor (2001) noted that many studies on stepparent-adolescent relationships explicitly relied on theory (most often systems theory) to frame their studies. There has been greater use of life course theory (e.g., Vartanian & McNamara, 2002; Wilmoth & Koso, 2002) to examine stepfamilies, with a focus on multiple trajectories and social contexts. More use of social exchange and other economic theories has occurred (e.g., Elman & London, 2002; Sweeney, 2002). Since the 1990s, in particular, the use of qualitative approaches, such as phenomenology and grounded theory (e.g., Arnaut, Fromme, Stoll, & Felker, 2000), has increased.
Because family structure alone provides little, if any, explanatory power (e.g., Biblarz & Gottainer, 2000), inductive theory-building efforts have increased (e.g., Braithwaite et al., 2001; Lansford, Ceballo, & Stewart, 2001; Kim, Hetherington, & Reiss, 1999). The development and testing of inductive models, especially in concert with observational methods, have potential to provide insight into stepfamily dynamics.
Measures used to assess remarriage and stepfamily dynamics continue to be instruments that were designed and validated on first-marriage families. Thus, there is a need for the development of reliable and valid measures that are developed to assess the unique aspects of stepfamilies.
Further, few researchers have used mixed-methods design in studying remarriages and stepfamilies. The use of mixed-methods designs, observational methods, and sequential analyses would better inform us about the patterns of interaction that serve stepfamilies well.
Over the last several years, there has been an increased reliance on secondary analyses of data sets drawn from nationally representative samples (e.g., National Education Longitudinal Study [NELS]:88; the National Survey of Families and Households [NSFH], SIPP). For researchers, this means working within the confines of existing questions and measures that often were not created to study stepfamilies. Although these data provide fertile grounds for scholars to pursue valuable research questions, reliance on these data limits the ability to understand stepfamily processes.
Ultimately, research that provides insight into stepfamily processes is helpful to prevention and intervention efforts that can best serve stepfamilies in the future. We challenge researchers to continue to identify and discuss the application of their findings to the real-life needs of stepfamilies. Only then will research contribute to the improving the quality of life in these complex families.