| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
|
|
|
|||||||
|
|||||||||
Correspondence: Address correspondence to Dr. Jim Hinterlong, College of Social Work, Florida State University, Tallahassee, FL 32306-2570. E-mail: jhinterlong{at}fsu.edu
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Key Words: Grander families • Intergenerational households • Older parents
Literature Review
To establish a bridge with the extant research, we situate our study within the literature of kinship care with a focus on informal and formal care arrangements. Four areas of research relate to grander families created through adoption: kinship care, grandparents raising their grandchildren, kinship foster care, and kin adopting kin. Unfortunately, much of this literature, especially studies of foster care and adoption, blends together caregivers of all age groups, thus making it difficult to decipher the specific impact upon or contribution by adults aged 60 and older. Whenever possible, we highlight findings specifically related to older parents.
Kinship Care
The predominant focus on kinship care arrangements within the current literature is understandable. Kinship is the basic principle of organizing individuals into social groups and roles. Since antiquity, kin relationships have included obligations to care for one another. For a variety of reasons, such as the parents' death or inability to care adequately for their child, adult relatives have come forward to care for their younger kin. Kinship care occurs either as informal responsibility for children without legal endorsement or by means of formal avenues such as legal guardianship, foster care, or adoption. Older family members have always participated in care arrangements, but rising life expectancy means that multiple generations are "fellow life travelers" for unprecedented spans of their lives (Plath, 1980). This opens new possibilities for older family members to assume central roles in childrearing either in coresident or custodial capacities. Thus, kinship care serves as an important pathway to the creation of some grander families.
Children are placed away from their family of origin for a variety of reasons; issues of safety and child welfare often precipitate this action (U.S. Department of Health and Human Services, 2005), as does the incidence of HIV–AIDS among biological parents and children (Linsk & Mason, 2004; Winston, 2006). Many children enter the foster care system, 24% of whom are placed with relatives (U.S. Children's Bureau, 2006). The goal for half is reunification with their parents or previous caregiver (U.S. Children's Bureau). In fiscal year 2005, 109,000 children were waiting to be adopted through the public child welfare system, with nearly 20,000 residing with kin (U.S. Children's Bureau). Eleven percent of children exiting foster care continued to live with relatives, and a further 18% were adopted—including some by kin. These most recent data reveal that, of the approximately 51,000 children adopted from the public foster care system in 2005, an estimated 25% joined kinship families, making this one of the fastest growing adoptive family forms (U.S. Children's Bureau; Howard & Smith, 2003).
Grandparents Raising Their Grandchildren
Custodial grandparenting is a form of kinship care. Grandparents are identified generally by their relationship to the child, not by their age. Hayslip and Kaminiski (2005b) note that most custodial grandparents are younger than 65 years of age. Thus, many are not considered older adults under the prevailing use of that term in the United States. Analyzing 2000 U.S. Census data, Simmons and Dye (2003) found that approximately 4.5 million children live in 2.4 million grandparent-headed households, which is a 30% increase from 1990. These adults reported being "primarily responsible for basic needs" of the grandchild(ren) in their residence. For many, the commitment is long term: 39% had cared for their grandchildren for 5 or more years.
Triggers for custodial grandparenting vary. Jendrek (1993) found that emotional, mental, and drug or alcohol problems of the child's parents were the most often cited reasons. Other causes include abandonment, child abuse, death of the parent, divorce, illness, incarceration, military deployment, and teenage pregnancy (Hayslip & Kaminiski, 2005a, 2005b). Indeed, the role of custodial grandparenting can be fraught with complexities. The literature is replete with challenges faced by grandparents, including more depressive symptoms and functional limitations (Minkler & Fuller-Thompson, 1999; Szinovacz, DeViney, & Atkinson, 1999), negative social consequences (Spence, Black, Adams, & Crowther, 2001), and increased stress that is due to caring concurrently for one's own adult children (Burton & deVries, 1993).
Caring for children with intense needs induces greater strain. A recent study of children whose mothers died of HIV–AIDS found that 77% of the children were living with family (primarily grandparents; Cook et al., 2003). Fewer than 25% of these arrangements were formalized legally prior to the mothers' deaths. Despite the lack of formal planning or financial assistance, relatives (especially grandparents) kept families together, preventing these children from entering the child welfare system. Daphne (2000) explored the physical and emotional health of grandparents raising HIV-affected grandchildren. The findings generally underscored the emotional resiliency of older relatives, but they also showed that some experience emotional distress that may reveal vulnerabilities masked by a determination to provide care.
Custodial grandparents also face social and economic difficulties producing high levels of stress and role strain (Bowers & Myers, 1999; Kelley, Whitley, Sipe, & Yorker, 2000), which negatively affects well-being. The individuals most often make up female, single-parent households (Fuller-Thompson & Minkler, 2000; Szinovacz, 1998), and grandparenting is frequently added to a full plate of existing responsibilities. Grandparents frequently meet these obligations with limited legal and economic resources. For example, most grandparents have only informal custodial arrangements (Landry-Meyer & Newman, 2004). Custodial grandparents are also more likely to live in poverty, although some are able to receive temporary assistance to needy families (Minkler & Fuller-Thompson, 2005).
Despite the increased risk for negative outcomes, many grandparents report having a special relationship with their grandchild (Erhle & Day, 1994), or a renewed sense of purpose in life (Giarrusso, Silverstein, & Feng, 2000), and they find this role emotionally fulfilling (Cherlin & Furstenberg, 1992). Goodman and Silverstein (2002) noted that African American grandmothers experienced an enhanced sense of well-being. Longitudinal analyses of data from 12,872 grandparents between the ages of 50 and 80 years from the Health and Retirement Study reveal few health effects of grandparenting, net of grandparent characteristics and prior health (Hughes, Waite, LaPierre, & Luo, 2007). Deleterious outcomes emerged primarily during the initial period of providing custodial care in skipped-generation households, whereas grandparents who babysat showed improvements in health (Hughes et al.). Hayslip and Shore (2000) reported that 90% of grandparents said they would make the same choice to care for their grandchild again, which shows the perceived value of this role. Moreover, grandparents can serve as a source of stability in the child's life and act as a buffer against the effects of family disruption (Emick & Hayslip, 1999). A small body of empirical literature points to more positive behavior among children under their care (Hayslip & Kaminski, 2005a).
Kinship Foster Care
Kinship foster care shifts preexisting kin relationships, including but not limited to grandparents, toward a legally recognized parent–child relationship. Kin foster parenting has exploded in recent years as a practicable, and often preferable, alternative to traditional (i.e., nonrelative) placements (Gibbs & Müller, 2000). Nevertheless, children in kin foster care are more likely to stay in care longer and are less likely to be reunified with their birth parents or adopted (Barth, Courtney, Berrick, & Albert, 1994). Kinship caregivers also receive less monitoring and support than do traditional foster parents; they have significantly lower incomes and more children in the home (Berrick, Barth, & Needell, 1994). Conversely, kin foster care is associated with greater placement stability, increased likelihood of placement with siblings, and less out-of-home maltreatment (Scannapieco, Hegar, & McAlpine, 1997).
Some studies have found that children placed with kin have fewer mental health and behavioral problems than those placed with nonrelatives (Berrick et al., 1994). Additionally, these children showed a lower level of risk for maltreatment while in care (Zuravin, Benedict, & Somerfield, 1993). In a geographically diverse sample of children in kin and nonkin foster homes, researchers found that kinship foster children had behavioral scores (as measured by the Child Behavior Checklist) comparable with those of children in the general population; however, those children in nonkin foster homes were more often reported to have lower levels of competence and higher levels of problematic behaviors, with substantial numbers scoring in the clinical range (Keller et al., 2001). Another study utilizing teacher reports of child behaviors found no differences between children in kin and nonkin foster care placements (Shore, Sim, Le Prohn, & Keller, 2002). Although there are mixed findings, most published studies illustrate that there are positive potential benefits of being placed with relatives. As a result, federal legislation gives preference to kinship care whenever possible.
Kin Adoption
Federal legislation also strongly encourages adoption for children in foster care (Gibbs & Müller, 2000). As the number of children placed in kinship foster care has grown, so too has the incidence of kin adoption (U.S. Children's Bureau, 2006). However, there is a dearth of literature specifically focusing on such adoption forms; much of what is assumed about kinship adoptions comes from research on kinship foster care. Although kin relationships may be captured in extant adoption research, few studies appear to focus specifically on the outcomes for kin-adopted children and their adoptive parents, and none examines adoptive families headed by older adults.
In a descriptive comparison of kin and nonkin adoptions, Magruder (1994) found that 27.3% of children were placed for adoption with a relative. These children spent more time in foster care, but had fewer placements prior to adoption. Billing and colleagues (Billing, Ehrle, & Kortenkamp, 2002) compared children cared for by kin and those residing with birth parent(s). Those in kinship care were significantly worse off on almost every measure (school activity and involvement, health, and caregiver relationship). Although the investigators did not consider caregiver age, their work reveals the vulnerability of these at-risk children.
In a study of 799 adoptions of all types, Rosenthal and Groze (1992) identified 73 kinship adoptions. They compared kin adoptive parents' assessments of the adoption's impact on their family. Grandparents adopting grandchildren rated the adoption as "very positive" most frequently (73%), whereas other kin adoptive parents (i.e., aunts, uncles, cousins, and other relatives) did so less often. However, the study did not examine how adoption by an older adult, in combination with other variables, contributes to outcomes.
In the most extensive study that we know of on kinship adoptions to date, Howard and Smith (2003) compared kin adopters (n = 523), previous foster parent adopters (n = 589), and previously unknown matched adopters (n = 183) of special needs children. They found that kin adoptive parents were more likely to be single parents and older than their counterparts (52 years old on average), and they had significantly less income. Children were placed into kin adoptive homes significantly earlier than their counterparts, but they waited longer for adoption finalization. Although this is the most comprehensive picture of kinship adoptions to date, it does not explore how preadoptive kinship or parental age impacts adoption outcomes.
We find limited research on the significance of preadoptive kinship to adoption outcomes and no previous study specifically focused on kin adoptions by adults aged 60 and older. To address these gaps, we pose the following research questions. First, what are the characteristics of older kin and older nonkin adoptive parents? Second, what are the characteristics of the children adopted by older kin and older nonkin parents? Third, what are the parent–child relationship characteristics exhibited by older kin and older nonkin adoptive families? Fourth, does preadoptive kinship among older adoptive families predict current parent–child relationship quality, family functioning, or adoption outcomes after one controls for various household, parent, and child factors?
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
The sampling frame was the current adoption subsidy role, which included 9,170 parents and 14,746 children. We excluded those parents electing to be removed from the survey or with incorrect addresses, as well as children ineligible for the study (i.e., over the age of 18), leaving 6,782 families with 10,923 children. Of the remaining parents, 1,694 (25%) returned surveys for 2,382 children (21.8%). Parents indicated if and in what capacity they had known the child prior to adoption. We classify as "kin" those reporting to be a relative of the child. Further information about the nature of that relationship was not sought.
We subset the sample to children with at least one adoptive parent aged 60 and older, and we divided this group on the basis of preadoptive kinship status between the parent(s) and adopted child. This resulted two family groups: older nonkin (n = 310) and older kin (n = 98). Subscript notations identify group statistics: We use OK for older kin and ONK for older nonkin.
Research Instruments
The survey was developed in partnership with an advisory panel to obtain a broad range of data from adoptive families. The panel was composed of persons with personal or professional adoption experience, including four internationally known adoption scholars, the head adoption official for the state of Florida at the time, three district-level adoption professionals, four adoptive parents, and three adult adopted persons. Each panelist provided comments on the inclusion or exclusion of questions or topics, as well as the appropriateness of terminology, readability, and overall survey structure.
Missing Values
Item nonresponse is expected in survey designs. Preliminary analyses revealed less than 10% missing data among the analysis variables. We used multiple random imputation (MRI) to address this problem. MRI is a gold-standard approach to representing missing values, because it is not deterministic and incorporates uncertainty into the final imputed values (Allison, 2002). MRI enables the imputed data to be analyzed with any technique appropriate to complete data (Schafer & Graham, 2002). MRI also preserves statistical power while not affecting the validity of analyses when the sample size is large yet the missing data cannot be considered missing at random or deviate significantly from multivariate normality (Wang, Sedransk, & Jinn, 1992). All analysis variables in this study, as well as auxiliary variables theoretically or statistically related to the analysis variables, were used in the imputation procedure (Schafer & Graham).
Analysis
We describe the study sample and use chi-square and t-test methods to compare the subsamples. We then use ordinary least squares regression to estimate the effects of preadoptive kin status on five outcomes net of other factors. Some families reported adopting more than one child. To assess whether the clustering of adopted children within families impacted our analyses, we fit null hierarchical linear models using the unique family identifier as the second-level factor. The results showed no significant clustering effect for any model. We also tested for but did not find problematic multicollinearity among our independent measures.
Control Variables
We include several control measures that have been found important in prior empirical studies of adoption outcomes. These include household income and size (Rosenthal, 1993), child needs severity and externalizing behavior (Brodzinsky, 1993; Rosenthal & Groze, 1991), finalization delay, and current child age (Rosenthal).
Traditionally, parents with higher human and financial capital have been preferred for adoptive placements even though "low income and education, minority ethnicity, single-parent family structure ... may be modest predictors of increased success in special needs adoption" (Rosenthal, 1993, p. 82). Parental education represents the highest level of formal education achieved by the parent(s), and it ranges from less than high school (1) to postgraduate (6). We dichotomized parental race as White versus non-White, with mixed-race couples coded as non-White (n = 13). We also include continuous measures of household income and household size to adjust for demands on family resources.
Participants were asked whether their child exhibited behavioral, emotional, physical, or educational needs at the time of adoption, and they were asked to rate the severity of each need type from 1 (mild) to 10 (severe). We created a composite need measure from these responses. Need severity is the child's mean z score (in relation to the entire sample) across the four severity measures. The observed range was –1.9 to 2.1, with higher scores indicating a greater need for support services.
The Child Behavior Checklist (CBCL) is the most widely used scale of child behavior (Achenbach & Rescorla, 2001) and is widely used in domestic and international adoption research (Ryan, Tyson, & Teasley, 2008). In contrast to internalizing behavior (e.g., somatic complaints, social withdrawal, and depression), research suggests that externalizing behavior, such as breaking rules, running away, fighting, and exhibiting threatening behavior, is most disconcerting to adoptive parents and is associated with placement instability and adoption disruptions (Brodzinsky, 1993; Brodzinsky, Smith, & Brodzinsky, 1998). Commentary reviews (Freeark et al., 2005) and empirical research (Gibbs, Barth, & Houts, 2005) reveal that adoptive parents are most concerned about externalizing disorders (Slap, Goodman, & Huang, 2001; Stukes Chipungu & Bent-Goodley, 2004). We use the 21-item Externalizing Behaviors scale from the CBCL. Reliability testing with this sample yielded Cronbach's alpha of 
= 0.94.
Time is a critical factor in adoption success. Nearly 10% to 15% of older child placements disrupt, and adoption delays contribute to poorer family and adoption outcomes (Rosenthal, 1993). We control for both the child's current age and finalization delay (the difference in years between a child's placement with the family and adoption finalization).
Dependent Variables
Three dependent variables are single-item measures. Parents were asked following: How satisfied are you with your overall adoption experience with this child?; If you had it to do over again, would you adopt this child?; and Overall, the impact of this adopted child's placement on your family has been __? Responses range from very negative (–2) to very positive (2).
To assess parent–child relationship quality, we use a six-item version of a previously developed scale based on social control theory and related to parent–child attachment (Groze, 1996). This scale assesses how well the parent and child communicate, how much time they spend together, how much trust and respect the parent has for the child, and how emotionally close they are. Responses to each item were based on a 4-point scale, with lower scores representing less parent–child relationship satisfaction, as reported by the primary respondent. Reliability testing with this sample produced a value of = 0.88.
Respondents completed the Family Function Style Scale (FFSS; Trivette, Dunst, Deal, Hamby, & Sexton, 1994). The FFSS is based on the strengths perspective and measures the extent to which the respondent believes his or her family exhibits different strengths and capabilities. It consists of 26 items rated on a 5-point scale ranging from 1 (not at all like my family) to 3 (sometimes like my family) to 5 (almost always like my family). We use the mean score as a global assessment of family functioning. Higher scores reflect more positive family functioning. The coefficient alpha for the scale with this sample is = 0.92.
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
|
Parent–Child Relationship
Adoptive parents expressed different attitudes toward the adoption (see Table 3). Both parent groups reported similar overall satisfaction with the adoption (MOK = 0.83, MONK = 0.81), assessment of family impact (MOK = 1.23, MONK = 1.32), and parent–child relationship quality (MOK = 1.54, MONK = 1.55). Kin were more likely to indicate a willingness to adopt the child again, if given the opportunity (MOK = 0.94, MONK = 0.88; t = –2.00, p <.05); however, they also reported significantly less positive family functioning (MOK = 4.01, MONK = 4.27; t = –2.79, p <.01).
|
|
R2 = 0.02, p <.01). Only externalizing behavior was significant across the five models, exhibiting a negative relationship with each outcome. In addition, higher current child age predicted a lower likelihood that a parent would adopt again, a more negative assessment of family impact, and poorer parent–child relationship satisfaction. These outcomes were unaffected by parental education or race. |
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This study also exhibits several important strengths. First, to our knowledge, this is the first large-scale investigation of adoptions within families headed by older parents. Second, the data are drawn from one of the largest adoption-focused samples in the literature. This allows us to compare subsets of older adoptive parents. Third, the sample exhibits a similar kin distribution to national prevalence data (U.S. Children's Bureau, 2006).
We identified descriptive and outcome differences between older kin and nonkin adoptive families. We find that older kin families are smaller in overall size, contain fewer children, and report an average of $516 less monthly household income than do nonkin families. Children adopted into both family types experienced an average of two prior placements and lived with their adoptive family for an average of 2.8 years prior to finalizing the adoption.
Given the special needs criterion for study inclusion, these children entered adoptive families with a high rate and variety of physical, behavioral, emotional, and educational needs requiring additional services and parental assistance. Children adopted by kin had less severe needs. Externalizing behavior scores reported were similar to those reported in other adoption studies, with no significant differences noted for the older adoptive sample or either subgroup (e.g., Groza & Ryan, 2002).
Strength in Kinship
Our analyses reveal a negative relationship between preadoptive kinship and perceived impact of the adoption on the family, net of household, parent, and child characteristics. Kin parents rated the adoption's impact on their family as less positive and reported that their families exhibited relatively lower family functioning. Nevertheless, older kin families are clearly committed to their children—as shown by their greater propensity to indicate a willingness to adopt the same child again. This mirrors previous research on custodial grandparents (Hayslip & Shore, 2000), and it illuminates the strength of the kinship bond in the face of challenges.
We did not find that preadoptive kinship affects parent–child relationship quality, satisfaction with the adoption, or willingness to adopt the child again (even though earlier bivariate analyses revealed a difference on this variable). Future research should explore factors leading to this disparity, including ongoing contact with the child's birth parent(s) and strains on household resources.
The adoption process is challenging for families under the best circumstances, and it may differ significantly for grandparents and other kin attempting to adopt children already in their care. Festinger and Pratt (2003) examined information from callers to an adoption hotline. When asked about general barriers to adopting a child from foster care, the primary concerns identified were the children's problems (21%), length of the process (18%), and the role of birth families (15%). Further study is needed to determine whether these concerns inhibit adoptive placements by older kin and older nonkin, and what, if any, unique difficulties these groups face.
Similarly, future investigations should consider how adoption by older adults affects marital relationships, household financial status, and public assistance use. For example, prior research shows that older adults, especially grandmothers, caring for related children are at increased risk of poverty (Park, 2006). Longitudinal analyses would enable more careful assessment of how undertaking the parenting role affects the older adult's health, functioning, and quality of life. As implied by previous research, these outcomes may vary according to the other demands faced by the older adult, such as preexisting caregiving responsibilities, the disruption of established patterns of social and productive activities (Hinterlong, 2006; Spence et al., 2001), or other preexisting or contextual factors (Hughes et al., 2007).
Similarly, we have left unanswered the question of whether adoption by an older adult affects the child's developmental trajectory and well-being differently than other forms of adoption: How do family, household, and other parental characteristics, for example, shape this relationship? In general, adoptive parents who see their child as more like themselves also see the child as less problematic, specifically in regard to externalizing behavior (Phillips, 2003). These behaviors, as noted previously, are those most frequently cited as causing great stress for adoptive parents. This connection (parent–child similarity) may be even more prevalent in adoptive placements with grandparents, where this connection predates the placement. Future research could explore how such feelings impact the success of the adoptive placements for both the older adult and the child.
Strategies to Promote and Support Grander Families
We conclude that grander families created through adoption are generally positive for the children and families, but they are not without challenges. Our findings suggest that adults who assume the parenting role later in life may benefit from targeted supports. Family functioning processes and the child's behavior shape daily interactions and appear to be more important to outcomes than the child's preadoptive history. This understanding can assist practitioners to work more effectively with older adoptive families and better prepare and support grander parents regarding the impact the adoption may have on their other relationships and family functioning. In particular, practitioners can assist older parents to develop strategies to manage and decrease behavior problems while at the same time change their reaction and expectations (i.e., perceptions) about their children (see Hayslip & Kaminski, 2005b for an excellent discussion of these strategies). These approaches could be blended into existing preadoption and postadoption services that contribute to overall adoption success, especially given these families' limited resources (Main, Macomber, & Geen, 2006).
We propose that older adults should be more closely considered as a resource within the child welfare system. Although there are no upper age restrictions for persons wishing to adopt from the child welfare system, the low rate of involvement of older adults as adoptive parents may indicate that targeted recruitment strategies are necessary to encourage them to adopt. Goughler and Trunzo (2005) outline a similar approach that could be modified to recruit and support older adoptive parents. First, specific recruitment strategies aimed at older persons must be created. Associations and other connections must be targeted, with media illustrating the potential benefits the older person may experience from the adoptive relationship. Second, support and education must be in place to ensure the stability and success of such family forms. This may include, but is not limited to, respite care, food and housing assistance, and specific training on managing the complex details of parenting in today's society.
Lastly, although this and other studies demonstrate positive outcomes for kin and older adoptive families, ageism among child welfare workers may be a barrier to such placements. Other studies have found that worker views regarding nontraditional placements, despite empirical evidence to the contrary, can inhibit such placements (Rodriguez & Meyer, 1990; Ryan, 2000). Although we explore only public adoptions in which there is no age limit for adoptive parents, private domestic and international adoptions often have upper age restrictions. A recent review of Holt International Children's Services website (www.holtintl.org), the largest international adoption child-placing agency, notes that there are currently 12 countries open for international adoptions. Three countries allowed persons to adopt up to the age of 55 years, with the remainder having younger age limits (the lowest, Korea, was 42 years). The average age limit is 50.42 years of age (SD = 3.55). These formal policies remove older adults from the pool of potential parents and support discriminatory practices, even though these grander parents appear to form generally positive relationships with children. Given the explicit goal of the Adoption and Safe Families Act to promote timely permanency through adoption, this potential barrier should be addressed. Although it is clear that no one has the "right" to adopt a child, children do have the right to an appropriate, loving family within which to grow and develop. The findings presented herein suggest that such grander adoptive family forms meet this criterion.
Irrespective of whether such changes are made, healthy longevity and changing family structures will continue to increase the prevalence of grander families. This will alter the trajectory and timing of familial roles across the life course. Therefore, we hope this work sparks greater attention to the challenges and opportunities presented by grander families.
|
Footnotes |
|---|
1 College of Social Work, Florida State University, Tallahassee.
Decision Editor: William J. McAuley, PhD
Received for publication May 16, 2007. Accepted for publication September 7, 2007.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| All GSA journals | Journals of Gerontology Series A: Biological Sciences and Medical Sciences | Journals of Gerontology Series B: Psychological Sciences and Social Sciences |
