Health and Well-Being of Children in Kinship Care: Findings from the National Survey of Children in Nonparental Care

Abstract

This study uses nationally representative survey data to describe differences in characteristics, adverse family experiences, and child well-being among children in kinship care with varying levels of involvement with the child welfare system. Well-being is examined in the domains of physical and mental health, education, and permanency. Comparisons provide insight on kinship care arrangements inside and outside the child welfare system, as well as the variability among nonfoster kinship care arrangements.

INTRODUCTION

Once largely separate from the child welfare system, in recent decades kinship care has become an integral part of child welfare practice and is often used as a preventive alternative to foster care (i.e., voluntary kinship care), as a foster care placement (public kinship care), or as an exit destination (permanent kinship care). This complexity exists, in part, due to layers of federal statute that have evolved over time and are not entirely consistent. Since the 1996 Personal Responsibility and Work Opportunity Reconciliation Act, child welfare agencies have been directed to consider giving preference to an adult relative over an unrelated caregiver when placing a child in foster care, provided the relative caregiver meets all relevant child protection standards (42 U.S.C. 671(a)(19)). In addition, since the 2008 Fostering Connections to Success and Increasing Adoptions Act, agencies have been required to notify adult relatives when a child is placed in foster care (42 U.S.C. 671(a)(29)). Federal law has also since 2008 provided agencies with the option to establish kinship Guardianship Assistance Programs with partial federal funding under title IV-E of the Social Security Act (42 U.S.C. 673(d); Testa, Snyder, Wu, Rolock, & Liao, 2015). Federal policy creates an inherent tension, however, as the preference for initial, temporary foster care placement with relatives is replaced by a hierarchy of permanency preferences that, if reunification with parents is not feasible, prioritizes first adoption, then placement with relatives (42 U.S.C. 675(1)). By 2014, nearly one-third of children in foster care (29%) and adopted from foster care (32%) nationally were cared for by relatives (U.S. Department of Health and Human Services, 2015). Cassanueva and colleagues found that of children living outside their parents’ home following a maltreatment investigation, an estimated 48% resided in private or voluntary kinship care (calculated from Exhibit 1 of Casanueva, Tueller, Dolan, Smith, & Ringeisen, 2012).

The tension between some kinship caregivers’ reluctance to consider adoption and the policy preference for adoption as a permanency outcome has prompted a number of policy and practice responses. As of March 2016, 33 states and 6 Indian Tribes have made guardianship subsidies a component of their title IV-E permanency programs (Administration for Children and Families, 2016). Child welfare agencies have created initiatives to encourage relatives to adopt the children in their care and train caseworkers on speaking with relatives about adoption and guardianship options (Pasztor, Mayers, Petras & Rainey, 2013). And child welfare agencies, relatives, and legal advisors have parsed the advantages and disadvantages of adoption versus guardianship (Saisan, Smith and Segal, 2016).

Although placement preference is given to relatives, it remains unclear what nonparental living situations best support children’s development (Winokur, Holtan, & Batchelder, 2014). Children in kinship foster care have shown more positive behavioral development, mental health, and placement stability than children in nonkin foster care (Wu et al., 2015, Winokur et al., 2014). Yet, children in nonkin foster care may fare better in accessing needed services and achieving adoption (Winokur et al., 2014).

Less is known about the well-being of children in private kinship care than about children in public kinship care (Littlewood, 2015). The living arrangements of children in public kinship care and nonkin foster care are similar in that both are monitored by caseworkers and in administrative databases from which children in informal kin care are absent (Stein et al., 2014). Researchers have lamented the lack of research on informal relative care following a Child Protective Services (CPS) investigation (Stein et al., 2014). Little is known about whether children’s situations improve if they are diverted to voluntary kinship care without oversight of the child welfare agency following a CPS investigation.

This study describes differences in characteristics, adverse family experiences, and well-being among children in kinship care in the following subgroups: (1) children in public kinship care; (2) children in voluntary kinship care for whom there is a current or past open CPS case; (3) children in voluntary kinship care without an open case but the relative reports other CPS involvement (e.g., the child welfare agency facilitated the placement); and (4) children in private kinship care with no current or past CPS involvement. We conceptualize kinship care as a continuum of arrangements, arrayed here according to decreasing level of child protective services involvement, and expect higher intensity of CPS involvement among children most at risk. However, state and local policies and practices may also influence the level of CPS involvement in less formal care arrangements.

As far as we are aware, the 2013 National Survey of Children in Nonparental Care (NSCNC) is the first population-based, nationally-representative survey of all children in nonparental care, the majority of whom are in relative care. Most surveys do not include enough cases to generate sufficient sample sizes in this rare subpopulation. While large surveys or Census data may include sufficient sample, they typically do not include relevant topical content (such as child well-being), nor the detail necessary to identify specific care types (foster, grandparent, other) or involvement with the child welfare system. Administrative foster care data cannot be used to compare foster children to children in other living arrangements and includes scant data on child well-being. Analysis of administrative data from CPS would similarly suffer from the inability to compare outcomes for children with varying levels of involvement with CPS (particularly those with none). NSCNC meets all these requirements.

METHODS

Data

Data were drawn from two national surveys conducted by the National Center for Health Statistics (NCHS): the 2011–2012 National Survey of Children’s Health (NSCH), a nationally-representative survey of households with children, and the 2013 NSCNC, which re-interviewed almost 1,300 households identified as nonparental care households in the NSCH, including foster care, grandparent care, and other households with no parents present. Both surveys were modules of NCHS’ State and Local Area Integrated Telephone Survey. NSCH was sponsored by the Health Resources and Services Administration’s Maternal and Child Health Bureau; NSCNC was sponsored by the Department of Health and Human Services’ Office of the Assistant Secretary for Planning and Evaluation, with supplemental funding from the Annie E. Casey Foundation.

NSCH was a random-digit-dial landline and cell telephone survey that interviewed 95,677 households with children throughout the U.S. The NSCH sample is nationally representative of noninstitutionalized children aged 0 to 17 years in the United States in 2011–2012.

NSCNC was a follow-back survey 1–2 years after the NSCH for children who lived in households with no parents present and were ages 0–16 when the NSCH was administered. Interviews were conducted with a current caregiver of the child, in some cases a parent who had reunited with the child since the NSCH interview. To distinguish among relative and nonrelative foster care and informal relative care situations, respondents who identified as foster parents were asked whether they were related to the child and respondents who identified as relatives were asked whether they were the child’s foster parent.

NSCH had a 23% overall response rate (partly due to the inclusion of a cell-phone sample to maximize coverage of the population), but this does not mean that three‐quarters of eligible households refused to participate in the survey. The response rate is low in part because it includes phone numbers that ring with no answer and for whom eligibility cannot be determined, especially among cell phone numbers. The NSCH cooperation rate among eligible households, or interview completion rate, was 51.4%. NSCNC had a 52% interview completion rate among eligible households 1–2 years later. Weighting adjustments were applied such that the population estimated by the sample of completed interviews matched that of the pool of eligible households demographically. This dramatically reduced estimated nonresponse bias; remaining bias in weighted estimates was smaller than sampling error. More information about NSCH and NSCNC may be found at: http://www.cdc.gov/nchs/slaits.htm or by referring to the associated documentation (CDC, 2013; CDC, 2014).

Statistical Analysis

Weighted estimates were calculated using SUDAAN (RTI, 2008) to account for complex sample design. Four subgroups of children were compared on well-being outcomes: those in public kinship care (1), and three subgroups of children in nonpublic kinship care: 2) those in voluntary kinship care for whom there had ever been an open CPS case; 3) those in voluntary kinship care without an open case but with other CPS involvement in the placement (a background check or home visit or CPS had arranged for the child’s placement); and 4) children in private kinship care with no CPS involvement. Excluding from the analytic sample children not in relative care, the sample size of 1,298 is reduced to 1,122.

Demographic and socioeconomic characteristics, adverse family experiences, and child well-being (health, academics, and permanence) were compared between children in public and nonpublic kinship care. Linear and curvilinear (quadratic) trend tests were performed among the nonpublic children across the three ordinal categories of CPS involvement. The significance of public/nonpublic differences or trends across categories of CPS involvement was evaluated at the 0.05 level.

Measures

In addition to child demographics (age, sex and race/ethnicity), sociodemographic characteristics examined include household income relative to Federal Poverty Level and caregiver age.

We examined seven adverse family experiences (AFEs): whether the child had (1) experienced a parent’s death, (2) experienced parents’ divorce or separation, (3) experienced parental incarceration, (4) witnessed violence in the home, (5) experienced or witnessed violence in the neighborhood, (6) lived with a mentally ill person, or (7) lived with a substance abuser.

Child health is measured by indicators of: overall health (excellent/very good versus good/fair/poor); whether the child has any mental health conditions (ADHD, learning disability, depression, anxiety, behavior/conduct disorder, autism, developmental delay, intellectual disability, cerebral palsy, speech problems, or Tourette’s syndrome) or any physical conditions (asthma, diabetes, epilepsy, hearing problems, vision problems, bone/joint/muscle problems, or brain injury); whether the child received mental health care in the previous year; and whether the child has special health care needs (SHCN) (any of five health care consequences resulting from a chronic health condition: (1) need for prescription medications; (2) need for specialized therapies; (3) need for more health care services than most children the same age; (4) treatment for a behavioral, developmental or emotional problem; and/or (5) activity limitation).

Academic well-being indicators include whether the child has an Individualized Family Service Plan (children under age 6) or an Individualized Education Program (ages 6+) (IFSP/IEP); whether the child is engaged in school (i.e., the child cares about school and does all required homework); whether the child repeated any grades; and math and reading/writing performance (excellent/very good versus good/fair/poor).

Indicators of permanence include whether the child lived with the caregiver since birth, whether the child lives all/most of the time with the caregiver, whether the caregiver feels the child is likely to live with them until grown, whether the caregiver has legal custody, and whether the caregiver has or intends to adopt the child.

Child sex, race/ethnicity, overall health, chronic conditions, SHCN, AFEs, IFSP/IEP, school engagement and grade repetition are drawn from NSCH. Values may have changed between surveys – e.g., a child may have lost a diagnosis and no longer be considered to have a health condition – and events that occurred between surveys would not be included – e.g., a grade repeated after the NSCH interview would not be identified. Academic measures from NSCH are not available for children younger than 6 in 2011–2012. The remaining covariates are drawn from NSCNC.

RESULTS

Of children in nonparental relative care, an estimated 11.1% were in public kinship care; 21.1% were in voluntary kinship care and had ever had an open CPS case (“Open case”); 19% were in voluntary kinship care without an open case but with other CPS involvement in the child’s placement (“Other CPS”); and 49% were in private kinship care without CPS involvement (“No CPS”).

Table 1 shows demographic and socioeconomic characteristics. Children in public kinship care were less likely than children in nonpublic kinship care to be ages 9–12, in the highest income category, or to have a caregiver aged 70+ years. Quadratic trend tests show parabolic associations in which children in the Other-CPS group are least likely to be Hispanic, most likely to be non-Hispanic black, and more likely to have caregivers aged 55–59 and less likely under age 55.

Table 1.

Demographic and Socioeconomic Characteristics by Kin Care Type and Child Protective Services (CPS) Involvement

Characteristic	Public Kinship Care	Informal (Nonpublic) Kinship Care
		All Nonpublic Kinship Care	Voluntary kinship care, ever an Open CPS case	Voluntary kinship care, no open case, but other CPS involvement	Private Kinship Care, No CPS involvement
Child	Percent (standard error)
Age: 1–8 years	34.5 (9.79)	28.8 (3.70)	33.9 (7.09)	25.9 (5.46)	27.8 (5.39)
9–12	12.8 (4.41)	26.2 (2.97)*	26.3 (6.65)	27.4 (5.72)	25.7 (4.15)
13–17	52.7 (10.39)	45.0 (4.02)	39.8 (8.61)	46.7 (7.19)	46.5 (5.68)
Sex: male	59.5 (9.95)	47.9 (4.03)	40.5 (7.57)	47.7 (7.02)	51.1 (5.60)
Female	40.5 (9.95)	52.2 (4.03)	59.5 (7.57)	52.3 (7.02)	48.9 (5.60)
Race/ethnicity: Hispanic	16.1 (8.38)	20.1 (3.86)	25.3 (9.69)±	8.8 (2.68)±	22.3 (5.95)±
Non-Hispanic white	40.3 (9.67)	34.7 (3.14)	44.2 (7.68)	29.5 (5.56)	32.7 (4.36)
Non-Hispanic black	32.4 (10.37)	36.3 (3.64)	21.0 (5.90)±	55.1 (6.66)±	35.6 (5.52)±
Non-Hispanic other	11.1 (4.12)	8.9 (1.99)	9.5 (4.39)	6.7 (1.82)	9.4 (3.01)
Household
Federal Poverty Level: 0–50%	23.9 (9.72)	17.6 (3.67)	18.6 (10.13)	30.2 (8.44)	13.2 (3.22)
50–100%	19.5 (9.08)	18.0 (2.68)	16.1 (4.80)	18.2 (4.70)	19.5 (4.22)
100–200%	38.1 (10.05)	30.9 (3.90)	23.2 (5.10)	28.1 (5.92)	36.6 (5.91)
200–400%	16.0 (6.24)	20.5 (3.13)	21.3 (5.63)	16.3 (3.86)	22.6 (5.01)
>400%	1.8 (1.00)	10.7 (2.33)*	20.8 (7.92)	7.2 (2.90)	8.2 (1.96)
Caregiver
Age: <55 years	36.4 (9.90)	34.8 (4.17)	40.8 (8.57)±	20.0 (5.04)±	38.3 (6.02)±
55–59	21.0 (7.28)	17.8 (2.80)	15.4 (5.01)±	36.6 (7.97)±	11.8 (2.11)±
60–64	19.1 (8.99)	17.4 (3.22)	13.9 (4.01)	15.1 (4.38)	20.1 (5.12)
65–69	18.0 (9.08)	17.1 (3.05)	19.2 (9.29)	16.4 (4.40)	16.7 (3.22)
70+	4.2 (2.05)	12.2 (1.94)*	10.7 (3.47)	12.0 (3.01)	13.2 (3.12)

Source: National Survey of Children’s Health 2011–2012 & National Survey of Children in Nonparental Care 2013;

^*Estimate differs at the 0.05 level from that of Public Kinship Care;

^±Quadratic trend by level of CPS involvement is significant at the 0.05 level.

Table 2 shows AFEs that may have precipitated the child’s entry into nonparental care. None showed a significant difference between foster and nonfoster children, but 5 of the 7 showed a significant linear trend among the private and voluntary kinship care groups in which higher prevalence is associated with more CPS involvement. Although the linear trend was not significant and differences were smaller, the remaining adverse experiences (parental divorce/separation and parental death) also showed the highest prevalence for children with the most CPS involvement.

Table 2.

Adverse Family Experiences by Kin Care Type and Child Protective Services (CPS) Involvement

Characteristic	Public Kinship Care	Informal (Nonpublic) Kinship Care
		All Nonpublic Kinship Care	Voluntary kinship care, ever an Open CPS case	Voluntary kinship care, no open case, but other CPS involvement	Private Kinship Care, No CPS involvement
Adverse Family Experiences	Percent (standard error)
Ever lived with a parent who …
… got divorced or separated	43.0 (10.16)	47.7 (3.98)	53.2 (8.33)	45.5 (6.81)	46.1 (5.59)
… died	11.7 (4.15)	18.3 (2.57)	19.6 (5.89)	18.3 (4.79)	17.8 (3.41)
… was incarcerated	49.6 (10.58)	37.8 (3.77)	62.9 (7.06)†	32.9 (6.22)†	28.9 (5.35)†
… was mentally ill	24.2 (7.42)	25.4 (3.55)	43.8 (7.80)†	24.3 (5.33)†	17.9 (5.04)†
… had drug/alcohol problems	50.0 (10.56)	47.2 (4.03)	70.2 (9.25)†	42.9 (6.63)†	38.9 (5.47)†
Ever witnessed violence in home	25.1 (7.25)	31.3 (3.79)	54.4 (8.27)†	20.8 (5.00)†	25.4 (5.23)†
Ever the victim of violence or witnessed violence in neighborhood	20.6 (6.87)	23.2 (3.02)	39.2 (7.69)†	24.5 (6.94)†	15.7 (3.23)†

Source: National Survey of Children’s Health 2011–2012 & National Survey of Children in Nonparental Care 2013;

^†Linear trend by level of CPS involvement is significant at the 0.05 level.

Some Table 2 estimates and differences are large. Seventy percent of Open case children had lived with a substance abuser, and 63% experienced parental incarceration. The incidences of incarceration and witnessing violence in the home were roughly twice as likely among Open case children as among other nonpublic groups. Other adverse experiences with significant linear trends showed similar patterns of differences almost as large.

Table 3 shows health conditions, special health care needs, and receipt of mental health care. Physical and mental health conditions show a significant pattern of more conditions with more CPS involvement, or a greater likelihood of zero conditions with less CPS involvement. SHCN and mental health care also follow a pattern of higher estimates with more CPS involvement. SHCN and mental health care were particularly high among Open case children, who had SHCN at twice or more than twice the rate of foster children and voluntary kinship care children, respectively. More than half of Open case children had received mental health care and more than 60% had SHCN.

Table 3.

Health Characteristics by Kin Care Type and Child Protective Services (CPS) Involvement

Characteristic	Public Kinship Care	Informal (Nonpublic) Kinship Care
		All Nonpublic Kinship Care	Voluntary kinship care, ever an Open CPS case	Voluntary kinship care, no open case, but other CPS involvement	Private Kinship Care, No CPS involvement
	Percent (standard error)
Overall health Excellent/very good	70.7 (10.60)	81.7 (2.25)	75.9 (5.91)	76.4 (5.41)	86.3 (2.65)
No mental health conditions	71.3 (9.85)	74.3 (3.41)	56.6 (8.88)†	68.2 (6.84)†	84.4 (2.56)†
1 mental condition	11.5 (8.50)	10.9 (2.92)	18.6 (9.35)	12.2 (5.88)	7.0 (1.66)
2+ mental conditions	17.1 (6.84)	14.8 (2.05)	24.8 (5.94)†	19.6 (5.00)†	8.6 (1.74)†
No physical conditions	75.1 (9.90)	72.9 (3.81)	59.2 (9.32)†	73.1 (6.51)†	78.8 (3.96)†
1 physical condition	22.2 (9.86)	22.2 (3.70)	31.5 (9.70)	20.2 (6.39)	19.0 (3.92)
2+ physical conditions	2.8 (2.06)	4.9 (1.27)	9.3 (4.43)	6.7 (2.48)	2.3 (0.70)
Special health care needs	30.6 (9.58)	38.5 (4.06)	61.3 (7.85)†	46.9 (7.16)†	25.4 (4.30)†
Received mental health care in previous year	29.1 (9.33)	27.2 (3.54)	51.5 (8.64)†	29.4 (6.20)†	15.8 (2.77)†

Source: National Survey of Children’s Health 2011–2012 & National Survey of Children in Nonparental Care 2013;

^†Linear trend by level of CPS involvement is significant at the 0.05 level.

Table 4 shows academics and permanence. There were few significant public/nonpublic differences but many indicators showed a linear trend across nonpublic categories. Significant academic measures (IFSP/IEP and grade repetition) showed a pattern of more favorable outcomes for children with less CPS involvement – and although the trends were not significant at the conventional level, most other measures of school engagement and academic performance were consistent with this pattern. These findings suggest that children in private kinship care tend to have better academic outcomes than children in other types of informal care. But while their academic outcomes may have been better, their permanency outcomes were worse: those with less CPS involvement were less likely to have been adopted or to have caregivers with adoption plans. Quadratic trend tests suggest that the likelihood that the child will live with the caregiver until grown or that the caregiver has custody are L-shaped associations in which children with no CPS involvement are much less likely than the other two groups to achieve these permanence indicators. Only 57% of children in voluntary kinship care had caregivers with legal custody, compared with about 90% of the other nonfoster groups.

Table 4.

Academic and Permanence Characteristics by Kin Care Type and Child Protective Services (CPS) Involvement

Characteristic	Public Kinship Care	Informal (Nonpublic) Kinship Care
		All Nonpublic Kinship Care	Voluntary kinship care, ever an Open CPS case	Voluntary kinship care, no open case, but other CPS involvement	Private Kinship Care, No CPS involvement
Academic	Percent (standard error)
Has IFSP (ages 0–5)/IEP (6–17)	10.2 (4.76)	15.5 (2.41)	27.7 (7.38)†	20.1 (5.45)†	8.4 (1.70)†
Is engaged in school (6–17)	76.9 (10.64)	66.0 (4.54)	55.5 (11.61)	54.1 (8.13)	74.9 (4.80)
Cares about school (6–17)	89.7 (4.48)	75.6 (4.38)*	59.8 (12.08)	76.8 (7.61)	81.2 (4.57)
Does all homework (6–17)	80.7 (10.68)	75.6 (3.73)	74.9 (8.83)	65.8 (7.82)	79.9 (4.51)
Repeated any grades (6–17)	29.2 (12.04)	22.2 (4.20)	39.2 (11.98)†	28.1 (7.93)†	13.1 (3.41)†
Excellent/Very good reading/writing	64.0 (9.65)	62.0 (4.12)	48.1 (8.85)	59.6 (7.42)	68.7 (5.84)
Excellent/Very good math	58.5 (10.64)	46.1 (3.94)	32.8 (7.10)	47.1 (7.26)	51.3 (6.03)
Permanence
Lived with caregiver since birth	17.5 (8.66)	17.8 (2.64)	10.8 (4.56)±	32.1 (7.05)±	15.3 (2.91)±
Lives all/most of time here	99.9 (0.13)	96.5 (0.96)*	97.9 (1.10)±	99.2 (0.60)±	94.7 (1.68)±
Will live here until grown	72.5 (11.23)	87.5 (3.63)	97.5 (1.20)±	97.3 (1.44)±	79.2 (6.10)±
Caregiver has formal/legal custody	75.1 (10.41)	73.5 (3.80)	91.4 (3.19)±	88.8 (5.31)±	57.4 (5.91)±
Child was adopted between surveys	7.4 (4.03)	8.1 (1.28)	11.5 (3.19)†	13.5 (3.51)†	3.9 (1.20)†
Caregiver has plans to adopt	31.0 (9.68)	23.6 (4.11)	35.5 (9.97)†	31.1 (7.65)†	14.3 (4.07)†
Not adopted, no plans to adopt	61.6 (9.91)	68.3 (4.05)	53.0 (8.99)†	55.1 (7.19)†	81.8 (4.18)†

Source: National Survey of Children’s Health 2011–2012 & National Survey of Children in Nonparental Care 2013;

^*Estimate differs at the 0.05 level from that of Public Kinship Care;

^†Linear trend by level of CPS involvement is significant at the 0.05 level;

^±Quadratic trend by level of CPS involvement is significant at the 0.05 level.

DISCUSSION

When children do not live with parents a connection critical to healthy development has been disrupted. Other family members frequently step in to fill the parental role, independently or in partnership with a child welfare agency. In recent decades perceptions and policies have been evolving about the appropriate role and level of child welfare involvement when relatives care for children due to parents’ absence or incapacity (Hegar and Scannapieco, 2005; Allen, DeVooght and Geen, 2008). This study advances this discussion by comparing the well-being of children in kinship care with varying levels of child welfare involvement.

We found that children with current or former open CPS cases, but who were not in foster care at the time of the survey, had particularly high rates of SHCN and mental health care compared with children in nonfoster relative care with less CPS involvement. It is possible that some children with health problems have open cases in part because the caregiver needed additional support for the child’s needs (Child Welfare Information Gateway, 2014).

Many adverse family experiences were increasingly likely with greater CPS involvement: violence in the home or neighborhood or having lived with the mentally ill, substance abusers, or parents who were incarcerated. Most were almost twice as prevalent among the open case children as among other nonfoster groups.

Because CPS cases could have occurred at any time, it is likely that some open case children were investigated by CPS, spent time in foster care and were later discharged to relatives. Nationally, 16% of children exiting foster care in 2014 lived with relatives or guardians following discharge (U.S. DHHS, 2015). Discharge from foster care to private kinship care typically involves cessation of monitoring by CPS. Children who had an open CPS case may be particularly vulnerable given the frequency with which they have SHCN and receive mental health care. While post-reunification services for parents have received considerable attention (Child Welfare Information Gateway, 2012) and post-adoption services have also been discussed (Smith, 2014; Zosky, Howard, Smith, Howard & Shelvin, 2005), there has been less discussion in the child welfare field about post-permanency services for kinship caregivers.

Just under half of children in kinship care are not involved with the child welfare system at all, according to their caregivers. These children are absent from child welfare administrative databases and are invisible in most surveys. NSCNC affords an opportunity to examine well-being and permanency outcomes in a comparative analysis including this subgroup. We found that these children tended to have better health and academic outcomes than other nonfoster groups; but they also tended to have poorer prospects for permanence.

The concept of “permanence” is complex and is viewed differently by various parties within and outside the child welfare system. Thompson and Greeson (2015) differentiate between legal permanence, i.e., the attainment of a court-sanctioned legal status according to the hierarchy defined in federal law, and relational permanence, meaning the subjective experience of a long-term emotional and social connection to one or more caring adults. Many relative caregivers think of the child as “already family” and report that as a reason for not considering adoption (Berrick, Barth, & Needell, 1994; Bramlett & Radel, 2016). Legal custody may not be seen as necessary given the family bond. However, children without CPS involvement did not just show lower rates of adoption or custody – they also had lower caregivers’ expectations that the living situation would last throughout childhood, perhaps indicating a hope for reunification with parents. Researchers have suggested that children in relative care need definitions of permanence that work for their particular situation and strategies to achieve permanence should be differentially targeted to specific subgroups (Yampolskaya, Sharrock, Armstrong, Strozier, & Swanke, 2014).

For informal relative caregivers for whom adoption and custody are less likely, accessing support services for the child can be difficult. Grandparent caregivers are often ineligible for certain financial supports because they lack status as foster parents or formal caregivers (Fruhauf, Pevney, & Bundy-Fazioli, 2015), but might be eligible for subsidies and Medicaid benefits if they adopted or established a subsidized guardianship. Health care for the child can be complicated when there is, “difficulty identifying who has the authority to consent for health care on behalf of the child” (Szilagyi et al., 2015: p. e1133). There is often a poor fit between services and families’ needs because policies were designed without consideration for nonstandard family status (Annie E. Casey Foundation, 2012). Not surprisingly, then, “kinship caregivers receive far fewer support services than nonkinship foster caregivers” (Stein et al., 2014: p. 560).

Limitations

Findings from NSCH and NSCNC are based on caregivers’ experiences and perceptions. Information provided about health status and health care was not verified with health professionals. Information about CPS involvement was not verified with child welfare agencies. Sampling weights were adjusted to minimize nonresponse bias and evidence suggests that remaining estimated biases tend to be smaller than sampling error (CDC, 2014), but because bias can only be estimated, the low overall response rate means that bias resulting from nonresponse cannot be completely ruled out.

That 75% of children in public kinship care were reported to be in the legal custody of their relative foster parents, when formal foster care usually means that the state retains formal custody, might indicate that custody was poorly understood by respondents – most of whom are grandparents – as distinct from having the child placed in their care. The question asked about a “formal or legal agreement about custody or guardianship for [the child]” and help text was provided (if necessary) to define custody as the legal right to make decisions for the child and to indicate that it may be conferred on a relative by a court. Custody status was not verified with child welfare agencies. It is possible that for some, custody was transferred between surveys. However, this limitation is less likely to affect comparisons among the nonpublic kinship care groups, and other measures of permanence, such as adoption and expectations for the future, showed similar patterns as for custody.

Despite these limitations, the authors know of no other data source that includes a population-based national sample of children in relative care, can identify the subgroups analyzed here, and includes survey content directly relevant to this population.

CONCLUSION

Relative care is an essential component of the safety net for children whose parents cannot care for them. Relatives are frequently sought by child welfare agencies as placement resources, particularly since the implementation of the 2008 Fostering Connections to Success and Increasing Adoptions Act’s requirement that agencies notify relatives of children’s placement in foster care. We found that just under half of kinship care is arranged privately among parents and relatives, i.e., it occurs outside the context of formal foster care, without child protective services involvement. This does not mean, however, that children in voluntary kinship care have not faced serious adversity; a sizeable proportion have experienced disruption and adverse family experiences.

Children with current or former CPS cases tended to have poorer health and academic outcomes than other children in nonpublic kinship care and may be particularly vulnerable given the frequency with which they have special health care needs. Aftercare services for children and youth discharged to relatives have received less attention than post-permanency services to support reunification and adoption. Understanding the long term well-being of children who exit foster care across discharge destinations could benefit from further research.

That more than 40% of caregivers without CPS involvement lack custody may indicate a vulnerability regarding legal permanence. Such children do not have institutional advocates for permanency since they are not engaged with child welfare agencies or courts and relative caregivers may see little need for legal permanence. Yet the lack of legal guardianship may leave children vulnerable in the long term either from a troubled parent who retains legal custody or from instability if the current caregiver cannot provide a stable home or adequate access to supports and services.