Mothers and grandmothers: A multigenerational approach to adolescent parenting

Despite declining estimates of adolescent births over the past decade, the United States leads other industrialized countries in adolescent birthrates (The Alan Guttmacher Institute, 1999). An estimated 10% of US women aged 15–19 become pregnant annually, with approximately half giving birth (Child Trends, 2002). Residence for adolescent mothers has emerged as an important and controversial topic following passage of the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) of 1996. Mothers under the age of 18 are required to live in adult-supervised settings to qualify for public benefits. This mandate was based on the belief that living with an adult would promote educational and economic attainment for adolescent mothers (Kalil & Danziger, 2000).

Multigenerational households have been associated with both advantages and disadvantages for adolescents, infants, and grandmothers (Apfel & Seitz, 1991; Black et al., 2002; Brooks-Gunn, 1990; Chase-Landsdale, Brooks-Gunn, & Zamsky, 1994; Coley & Chase-Landsdale, 1998; Kalil & Danziger, 2000; Sands & Goldberg-Glen, 2000; SmithBattle, 1996). Evidence suggests that living at home prolongs adolescence by allowing mothers to complete their education, fulfill their own developmental needs, and become competent parents (Smith, 1975; Spieker & Bensley, 1994; Taylor, Chatters, & Jackson, 1993). Adolescents may receive the tangible and intangible support needed to increase the overall quality of childcare (Denham & Smith, 1989; Oyserman, Radin, & Saltz, 1994; SmithBattle, 1996). Coresidence also allows low-income families to manage limited resources and effectively develop an interdependent family network (Wilson, 1984).

Other research suggests that while coresidence may further educational goals, it may actually impede the development of adolescent parenting skills (Black & Nitz, 1996; Spieker & Bensley, 1994; Unger & Cooley, 1992). Role confusion experienced by multiple caregivers has been associated with low quality of childcare (Chase-Landsdale et al., 1994; Stevens, 1984). Finally, multigenerational households may increase the likelihood of conflict in the adolescent-grandmother relationship, as both adapt to new caregiver roles (Apfel & Seitz, 1991).

Four models of adolescent mother–grandmother relationships

In an effort to comprehend the complexities of three-generational relationships, Apfel and Seitz (1991) completed qualitative interviews with 119 urban African American grandmothers and adolescent mothers 18 months after the birth of their first child. Based on discussions of parenting practices, daily routines, and the division of childcare in their households, Apfel and Seitz (1991) proposed four models of adolescent mother–grandmother relationships.

In the Parental Replacement model, the grandmother assumes primary responsibility for childrearing, “I am raising your child for you” (Apfel & Seitz, 1991, p. 427). Approximately 10% of the families in their study fell into this category. In the Parental Supplement model the adolescent mother and grandmother, or other coresiding family members, share caregiving responsibilities, “We are all raising this child” (Apfel & Seitz, 1991, p. 427). More than 50% of their sample was described by this model. In the Supported Primary Parent model the adolescent mother assumes primary responsibility for caregiving, with only occasional assistance from the grandmother, “This is your child and he is your responsibility to raise” (Apfel & Seitz, 1991, p. 427). This model described approximately 20% of their sample.

Finally, in the Parental Apprentice model, the grandmother encourages the adolescent’s successful transition to parenthood, “I will act as your mentor as you learn how to raise your child” (Apfel & Seitz, 1991, p. 427). This model described nearly 10% of the sample. The Parental Apprentice model differs from the other three models in that it describes a mentoring process that occurs over time and may include aspects of the other models as the adolescent takes over the parenting role.

The research by Apfel and Seitz (1991) provided important insights into the daily routines of adolescents and grandmothers who are adjusting to an expanding family and new roles. The qualitative nature of their study afforded an impressive glimpse into the complementary and competing caregiver roles of urban family members through their own words. The models offered a useful structure, but they have not been investigated in the last 15 years and questions remain regarding the adjustment of adolescents to parenthood. Although 88% of the mothers in Apfel and Seitz’s study reported that they were not married or cohabitating with a partner, not all lived in multigenerational households. Their study was conducted prior to the 1996 passage of PRWORA, when coresidence with an adult was not required for the receipt of public benefits. In addition, the models are somewhat dependent on the quality of the relationship between the adolescent mother and grandmother, but quality of relationship was not assessed in Apfel and Seitz’s study.

The present study is an adaptation and extension of Apfel and Seitz’s (1991) models in an independent sample of African American adolescent mothers. Our study differed from theirs in six important ways. First, the current study is quantitative, compared to the qualitative nature of Apfel and Seitz’s (1991) study. Second, quality of the adolescent-grandmother relationship was explicitly assessed in the current study. Third, participants in the current sample were approximately 5 months younger at delivery than the women in Apfel and Seitz’s sample. Fourth, all of the mothers in the present sample were recruited from hospitals following delivery, while Apfel and Seitz’s sample was recruited from referrals made to a school-based program for pregnant students. Fifth, all mothers in this sample were living in multigenerational households, while those in Apfel and Seitz’s sample were not all coresiding with their mothers. Finally, measures in the present study were collected within the first 6 months postpartum, compared to 18 months postpartum in Apfel and Seitz’s sample.

This study examined four hypotheses based on the parenting models proposed by Apfel and Seitz (1991). The first test investigated the relationship between division of caregiving responsibilities and quality of relationship. Apfel and Seitz suggested that the role confusion associated with shared caregiving leads to conflict regarding parenting practices and discipline. Others have suggested that conflict arises out of differing parenting strategies, poor adolescent maternal self-efficacy, and grandmother stress in the caregiver role (Coley & Chase-Landsdale, 1998; Kalil & Danziger, 2000). We hypothesized that shared caregiving responsibilities would be associated with conflict and a negative adolescent-grandmother relationship.

The second test investigated the relationship between division of caregiving responsibilities and an adolescent mother’s sense of parenting competence. Apfel and Seitz (1991) argue that as parenting experience increases, adolescents feel more competent as parents. We hypothesized that adolescent mothers assuming primary responsibility for caregiving would report higher levels of parenting competence than adolescent mothers who were not primarily responsible for caregiving tasks.

We expanded on Apfel and Seitz’s (1991) research by assessing the adolescent-grandmother quality of relationship. A positive adolescent-grandmother relationship appears to be the mechanism whereby grandmothers transmit information regarding parenting and adolescent parents build parenting competence (Hess, Papas, & Black, 2002). We hypothesized that a positive adolescent-grandmother relationship would be associated with parenting competence among adolescent mothers.

We hypothesized that quality of the adolescent-grandmother relationship, as assessed through observation soon after delivery and self-report 6 months later, would be associated with adolescent mothers’ parenting competence, adjusting for the division of caregiving. Apfel and Seitz (1991) implicitly included the quality of the adolescent-grandmother relationship in their models, and our final hypothesis provides a test of the importance of that relationship in building parenting competence in adolescent mothers.

Method

Participants

The present study utilized data from a longitudinal randomized controlled trial of home intervention designed to promote parenting and adolescent development (Black, Siegel, Abel, & Bentley, 2001).

Eligibility criteria for mothers included age less than 18 years at delivery, first time delivery, African American, low income (defined as eligible for Supplemental Nutritional Services for Women, Infants, and Children [WIC]—family income under 185% of poverty level), and no chronic illnesses that would interfere with parenting or adolescent development. Infants of eligible mothers had to be full-term (≥37 weeks of gestation), with birth weight above 2,500 g, and no congenital problems or chronic illnesses. None of the infants experienced complications following delivery that required neonatal intensive care services.

Procedures

Adolescent mothers were recruited shortly after delivery from three urban hospitals in Baltimore, Maryland between June 1997 and September 1999. Mothers were given information regarding the investigation, and those who chose to participate were scheduled for baseline evaluations in their homes within 3 weeks of delivery. Over 83% of the eligible mothers agreed to participate and 181 completed the baseline evaluation. There were no differences in maternal age or education between those who completed the baseline evaluation and those who did not.

Baseline measures included family demographics, personal and mental health, mother–grandmother relationships, access to public services, and early adjustment to parenting. Mothers completed these measures on a laptop computer. The items were presented both visually on the computer screen and aurally through headphones. All responses were entered with a mouse. Mothers and grandmothers were also videotaped sorting childcare task cards and discussing topics of conflict. Research assistants set up the video equipment and then left the room during this task to provide as much privacy as possible.

Following the baseline evaluation, all mothers were given information on community resources for young mothers and their children. Using a randomization procedure stratified on maternal age and gender of the child, mothers were assigned to either an intervention or a control group. Mothers in the intervention group received home visits every other week over the next year. Mothers in the control group received no further contact until the follow-up evaluation visits.

Follow-up home evaluations were conducted when infants were 6, 13, and 24 months of age. The current study is based on data collected at baseline and from the 148 (82%) mothers and 107 grandmothers who completed the evaluation at 6 months (see Table 1). The mothers in this sample were an average of 5 months younger than those in Apfel and Seitz’s (1991) sample. Seventy two percent of this sample was enrolled in school, and 68% of Apfel and Seitz’s sample was either pursuing education or had graduated from high school. Twenty four percent of this sample was employed. Follow-up families did not differ in maternal age, maternal education, infant birth weight, infant gender, or intervention status from the 33 nonparticipating families.

Table 1 Sample characteristics (N = 148)
Full size table

The procedures were approved by Institutional Review Boards of the University of Maryland, School of Medicine and all participating hospitals. Mothers were compensated for baseline and follow-up evaluation visits.

Measures

The division of caregiving responsibilities was assessed at the 6 month follow-up evaluation by administering the Who Does What scale (WDW; Cowan et al., 1985) to adolescent mothers and grandmothers independently. The 19 items are scored on a 9-point scale, from Someone else does it all (1) to I do it all (9). Example items include “Preparing meals for my child” and “Dealing with the doctor regarding my child’s health.” Scale developers report Cronbach’s alpha for this measure greater than .92, and moderate to strong correlations between caregivers (Cowan et al., 1985). In this sample, Cronbach’s alpha was equal to .93 for both the adolescent and grandmother reports.

Adolescent mothers’ parenting competence at the 6 month follow-up was assessed with the Parenting Sense of Competence Scale (PSOC; Johnston & Mash, 1989). The responses to these 17 items range from Strongly agree (1) to Strongly disagree (6). Example items include “Being a good mother is rewarding” and “I have all the skills to be a good mother.” In this sample, the internal reliability for the scale was alpha = .76.

The quality of the adolescent-grandmother relationship at baseline was measured by the Scale of Intergenerational Relationship Quality (SIRQ; Wakschlag, Chase-Landsdale, & Brooks-Gunn, 1996), an observational assessment. Adolescent mothers and grandmothers chose topics on which they disagreed, and were videotaped discussing these topics for approximately 15 min. This measure, designed to be a reliable and valid assessment of quality of relationship between African American low-income mothers and grandmothers, was standardized in Baltimore. A member of Wakschlag’s research team trained the coders for this study and also established reliability. The four factors of this measure include Emotional Closeness (connectedness), Positive Affect (upbeat tone), Grandmother Directness (demandingness and clarity) and Individuation (autonomy and individuality). The internal consistency of these factors ranged from .68 to .78 in this sample, and the factors were moderately to strongly correlated among each other, similar to the correlations reported by Wakschlag et al. (1996).

The quality of the adolescent-grandmother relationship at 6 months was measured by The Network of Relationship Inventory (NRI; Furman & Buhrmester, 1985), a self-report measure completed by the adolescent. The measure assesses both support and conflict, with responses to 30 items ranging from Little or none (1) to The most (5). The two subscales include Support (affection, admiration, reliable alliance, intimacy, companionship, and instrumental help) and Conflict (negative interactions and antagonism). The Support scale includes items such as “How much does your mom treat you like you are good at many things?” and the Conflict scale includes items such as “How much do you and your mom get upset with or mad at each other?” This reliability and validity of the measure have been empirically supported in previous research (Furman & Buhrmester, 1985), and Cronbach’s alphas for the two factors in this sample were .82 (Support) and .88 (Conflict). The Support and Conflict scales were significantly negatively correlated in this sample (r = −.41, p < .01).

Data analysis

Twelve adolescents who were not residing with the grandmother at the 6-month follow-up were excluded from data analyses. Although this decision ignores adolescents who may have gained economic stability and moved out of their mother’s home, the interest in this study is the adolescent-grandmother relationship under conditions of coresidence. No adolescent mother–grandmother pair reported that “someone else” provided all childcare. Adolescent age was examined as a moderator of the relationship between shared caregiving, quality of relationship, and parenting outcomes.

There were no direct effects of the intervention on predictor or outcome variables and no moderated effects of the intervention on outcome values. To ensure that the findings were not confounded by intervention status, intervention was dummy coded and entered as a covariate prior to data analysis. We replicated all analyses with the subsample of adolescents in the control group, with similar results.

This sample was grouped according to Apfel and Seitz’s (1991) models. Agreement between the adolescent and grandmother reports of division of caregiving and quality of relationship were examined. Because there was relatively low agreement between adolescent mother and grandmother scores on division of caregiving (r = .06, p = .57) and quality of the adolescent-grandmother relationship (r = −.15, p = .13), we focused on adolescent reports in an effort to understand parenting from the mother’s perspective. This is in line with our prospective interest in examining how adolescent mothers make the transition to becoming primary caregivers.

For the first hypothesis, we examined the association between the division of caregiving and conflict in the adolescent-grandmother relationship using an ordinary least squares regression model. A measure of shared caregiving was created by computing the average absolute difference between the WDW score, ranging from (1) Someone else does it all to (9) I do it all, and 5, the midpoint on the 9-point scale. The transformed variable ranged from 0 to 4, with values close to 0 reflecting high levels of shared caregiving, and larger values reflecting less shared caregiving. Adolescent age and the interaction of caregiving and age were also entered into the regression to examine whether the relationship between division of caregiving and quality of relationship was moderated by adolescent age.

For the second hypothesis, we tested the relationship between caregiving responsibilities and parenting competence. Adolescent age and an interaction term of adolescent age and division of caregiving were included as predictors to investigate possible moderation by adolescent age.

For the third hypothesis, we investigated the relationship between the quality of adolescent-grandmother relationship and adolescents’ sense of parenting competence. Adolescent age and an interaction term of adolescent age and quality of relationship were included to investigate possible moderation by age.

Finally, we examined the association between the quality of adolescent-grandmother relationship and adolescents’ parenting competence above and beyond the division of childcare. First, the relationship between the observational and self-report measure of quality of relationship was estimated using a Pearson correlation. Since the correlation did not suggest multicollinearity, both measures were entered as predictors into the regression model. This hypothesis was tested through the incremental change in variance predicted in the sense of parenting competence by the inclusion of quality of relationship above and beyond the division of childcare.

Results

We categorized the sample by the division of childcare, similar to the groups described by Apfel and Seitz (1991) (Table 2). The Parental Apprentice model was not included in the present study because the construct of mentoring is not captured through the division of caregiving responsibilities. After excluding the Parental Apprentice model and adjusting Apfel and Seitz’s (1991) remaining categories proportionally, the categorization of this independent sample closely approximated their category distributions. In this sample, no adolescents (0%) reported scores between 1 and 3 on the 9-point WDW scale (i.e., Someone else does it all), while 12.5% of Apfel and Seitz’s (1991) sample was described with the Parental Replacement model. Another 66.2% of the present sample reported scores between 4 and 6 on the WDW scale (i.e., We both do this about equally), compared to 62.5% of Apfel and Seitz’s (1991) sample described by the Parental Supplemental model. Finally, 33.8% of this sample reported scores between 7 and 9 on the WDW scale (i.e., I do it all), compared to 25% of Apfel and Seitz’s (1991) sample categorized into the Supported Primary Parent model.

Table 2 Distribution of parenting models
Full size table

Scores on the division of caregiving were reduced from a 9-point scale to a 3-point scale, representing someone else has primary responsibility (1), shared responsibility (2), and primary responsibility (3). Results of a chi-square test of independence suggest that the two samples significantly differ in their category groupings (χ2(2, n = 267) = 36.17, p < .01). Fifty eight pairs (56.3%) of adolescents and grandmothers agreed that they share childcare tasks. Twenty six pairs (25.2%) disagreed somewhat, with the adolescent reporting that she provided most of the childcare and the grandmother reporting shared caregiving. Only 4 (3.8%) adolescent-grandmother pairs disagreed completely, with both reporting that they performed all caregiving. Sixty five percent of adolescents reported shared caregiving and 81% of grandmothers endorsed shared caregiving (Table 3).

Table 3 Adolescent-grandmother childcare agreement
Full size table

The first model examined whether shared caregiving was associated with conflict in the adolescent-grandmother relationship. Adolescent reports of conflict did not result in a significant association between shared caregiving and conflict (Table 4, Model 1).

Table 4 Regression analysis summary for three models predicting (1) conflict in the adolescent mother–grandmother relationship, (2) parenting sense of competence, (3) division of caregiving, (4) parenting sense of competence above and beyond the division of childcare
Full size table

The second model examined whether adolescents assuming caregiving responsibilities reported parenting competence. Division of caregiving was significantly associated with the parenting competence score, with adolescents assuming caregiving responsibilities also reporting parenting competence (Table 4, Model 2). This relationship did not differ by adolescent age.

The third model tested the association between supportive quality of the adolescent-grandmother relationship and the parenting sense of competence score. NRI support scores were significantly associated with adolescent parenting competence, with adolescents reporting supportive mother–grandmother relationships also reporting parenting competence (Table 4, Model 3). Results did not support moderation of this relationship by adolescent age.

The fourth model examined whether supportive quality of the adolescent mother–grandmother relationship was associated with adolescent parenting competence, adjusting for division of caregiving. The quality of the adolescent-grandmother relationship assessed by the 6-month NRI support scores, but not by the baseline observational measure, was associated with parenting competence, adjusting for caregiving responsibilities (Table 4, Model 4).

Discussion

More than a decade after the publication of Apfel and Seitz’s (1991) research, questions remain about the effects of coresidence and multigenerational households on adolescent parenting. Their models have not been empirically tested, and the quantitative tests provided in our sample supplement their qualitative work. Coresidence is even more common following the 1996 enactment of PRWORA than it was when Apfel and Seitz conducted their investigation in the early 1990s. Although both the present sample and Apfel and Seitz’s sample consisted of urban, low-income, African American first-time mothers, the two samples differed along six dimensions. First, we employed quantitative, rather than qualitative, methods to examine caregiving responsibilities from the perspectives of both the adolescent mothers and grandmothers. Second, quality of the adolescent-grandmother relationship was explicitly assessed through observation of and self-report from both the adolescents and grandmothers. Third, young women in the current sample were approximately 5 months younger at delivery than those in Apfel and Seitz’s sample. Fourth, participants in the present sample were recruited from hospitals following delivery, while Apfel and Seitz’s sample was recruited from referrals made to a school-based program for pregnant students. Fifth, all mothers in this sample were living in multigenerational households, while those in Apfel and Seitz’s sample were not all coresiding with their mothers. While this reduces the variability of possible living arrangements, it reflects the realistic situation for many adolescents following the passage of PRWORA. Finally, measures in the current study were collected shortly after delivery and 6 months postpartum, compared to 18 months postpartum in Apfel and Seitz’s sample.

Differences between these two samples may be influenced by the developmental aspect of parenting an infant versus a toddler. Six-month-old infants may be learning to sit, crawl, and self-feed, while 18-month-old toddlers are walking, talking, and striving to become autonomous (Sander, 1962; Tulkin & Cohler, 1973). The young women in the current sample had one year less parenting experience than those in Apfel and Seitz’s (1991) sample. At 6 months postpartum, mothers are somewhat new to parenting and the family may be adjusting to the caregiving routine. At 18 months postpartum, parenting is less novel and families may have settled into a more stable caregiving schedule.

The categorization of these families into the Parental Replacement, Parental Supplemental, and Supported Primary Parent models is similar to the groupings described by Apfel and Seitz (1991). Approximately two-thirds of both samples endorsed shared caregiving, illustrating that this is a common pattern of parenting among African American adolescent mothers and grandmothers (Pearson, Hunger, Cook, Ialongo, & Kellam, 1997; Pearson, Hunger, Ensminger, & Kellam, 1990). In the present study, mothers were more likely than grandmothers to report full caregiving responsibilities, perhaps reflecting their desire to be independent parents.

In contrast to Apfel and Seitz’s (1991) conclusion that families sharing caregiving tasks may experience role confusion, in this study there was no relationship between shared caregiving and reported conflict from the adolescent mother’s perspective. There are at least three possible explanations for the absence of a relationship between shared caregiving and conflict. First, in an environment with the socioenvironmental constraints of limited resources and required coresidence, African American families may be forced to reconcile differences regarding caregiving and discipline practices (Burton, 1990). Only 12 (8%) adolescent mothers had moved out of the grandmother’s household within 6 months of delivery, and they did not report more conflict than those mothers who remained in the multigenerational household. Second, shared caregiving may be culturally accepted in multigenerational African American households where marital rates are low and limited resources are most efficiently managed within the family network (Wilson, 1984). The PRWORA requirement that adolescent mothers must reside with a guardian to receive benefits may contribute to the cultural acceptance of shared caregiving. Third, most adolescents in the present study were still in school 6 months after delivery and not financially independent. They may have been less likely than older mothers to experience role confusion, as developmentally, they were still adolescents.

While the multigenerational families of African American adolescent mothers may experience some conflict during role transitions, grandmothers continue to be an important source of tangible and intangible support for young mothers (SmithBattle, 1996). Shared caregiving appears to be an integral part of this support, at least during the first 6 months of parenting when young mothers are building skills and competence. The conflict in adolescent-grandmother relationships described by Apfel and Seitz (1991) may occur after the apprentice phase, when mothers feel that they have acquired the necessary mentorship and are seeking increased autonomy. The search for increased autonomy is the hallmark of adolescence and early adulthood, and may be heightened in African American families during the second year of parenting as young women transition to motherhood and assume additional caregiving responsibilities (Cauce et al., 1996). Thus, shared grandmother-adolescent mother caregiving may take on a developmental perspective in African American families, with role support during the initial months of parenting, followed by role conflict during the second year of parenting.

Apfel and Seitz’s (1991) finding of an association between the division of caregiving and parenting competence was also discovered in this study, where those mothers reporting high levels of involvement in caregiving were also more likely to see themselves as competent parents. However, the correlational design of these analyses prevents the report of causal conclusions. Clear role definition may either enable young mothers to gain parenting experience, as suggested by Apfel and Seitz (1991), or competent parents may be more likely to engage in caregiving. The experience of caregiving is integral to the development of adaptive parenting strategies, and multigenerational households offer a unique environment for apprenticeship, where adolescents can practice and gain confidence in the parent role (SmithBattle, 1996).

These analyses did not support the moderation of quality of adolescent-grandmother relationships and parenting outcomes by adolescent age, perhaps as a result of the restricted age range. Mothers ranged from 13.5 to 17.9 years of age at the time of the initial interview, and most were 16 or 17 years old. This range may not have included sufficient variability to detect moderation by adolescent age. Previous research demonstrating moderation by adolescent age included an extended age range of young mothers from 13.3 to 25.5 years (Chase-Lansdale et al., 1994).

The quality of the adolescent mother–grandmother relationship was explicitly assessed in this study, and the findings confirm the association between the division of caregiving and positive quality of adolescent mother–grandmother relationship. The supportive quality of the adolescent-grandmother relationship was assessed by self-report and observation, which is a unique addition to the literature. The observational measure of quality of relationship was not associated with parenting competence, perhaps because it assessed quality of relationship prior to the transition to motherhood, as it was completed at delivery.

The division of childcare and the quality of the adolescent-grandmother relationship were associated with parenting sense of competence, indicating that mothers who are involved in caregiving and have a supportive relationship with their own mothers feel more competent as parents. This study provides support for Apfel and Seitz’s (1991) Parental Apprentice model. Adolescents engaging in caregiving tasks may successfully transition to parenthood. A supportive mother–grandmother relationship may allow for the mentoring described by Apfel and Seitz. More qualitative and quantitative research is needed to investigate the mentoring component of Apfel and Seitz’s Parental Apprentice model, which they propose to be the optimal model for infant, adolescent, and grandmother.

There are several limitations to this research. These data were gathered as part of a larger investigation addressing adolescent development and parenting, and the experience of being in the intervention may have changed the lives of the participants. However, after controlling for intervention group, reporting the absence of significant interactions driven by the intervention group, and replicating the analyses with the non-intervention group, the possibility of reporting misleading conclusions is reduced.

The self-report measures utilized in this study were collected either at delivery or at 6 months. Causal statements cannot be made regarding the association between the division of caregiving, quality of the adolescent-grandmother relationship, and adolescent mothers’ parenting competence. Although self-report measures are subject to social desirability, computerized data collection procedures were used to reduce the likelihood of misleading results (Schroder, Carey, & Vanable, 2003).

Many low-income African American adolescent mothers live with their mothers and, as long as public policy decisions require coresidence for adolescent mothers seeking public benefits, it is important to examine the potential advantages and disadvantages of parenting in multigenerational households. These findings suggest that Apfel and Seitz’s (1991) models of adolescent parenting describe the situation of many African American adolescent parents living in multigenerational households. However, coresidence is not static or exclusive; groupings of families at one point in time do not capture the complexity of multigenerational caregiving (Solomon & Marx, 1995).

An adolescent mother’s involvement with caregiving responsibilities and the supportive quality of the mother–grandmother relationship are central to the development of her parenting competence. These issues can be addressed over time, during the prenatal period, in the hospital following delivery, and through home visitation. Future interventions should consider targeting role definitions in the mother–grandmother relationship and address strategies for shared caregiving and adolescent involvement in caregiving in order to promote adaptive adolescent parenting.