Search Type Title Author/Investigator Search Advanced Search

Table of Contents | Previous | Next

7. Conclusions

We have provided in the preceding chapters a description of the NSCAW OYFC children and families and their experiences in the child welfare system. The sample of children and caregivers obtained in the one year in foster care (OYFC) component of the study appears representative of the entire population of children who have been in foster care for one year. With only these preliminary analyses for this sample component completed, it is not yet possible to draw conclusions sufficient for strong policy or practice recommendations. However, with the data and analyses from the future waves of NSCAW, the project team and the research community will have more solid evidence on which to base such recommendations. The following is a summary of findings and provisional implications for child welfare services.

7.1 Child Characteristics

This report presents substantial evidence that the children who remain in foster care for approximately one year have been seriously abused and neglected, and that they have widespread and substantial developmental disadvantages. Although some of these disadvantages might partially reflect developmental outcomes more frequent among children growing up in families struggling with poverty and substance abuse, the range and magnitude of the developmental risks suggests that the future for these children has been seriously threatened by their experiences to date. Although it is possible that, by the time we assessed the children, their disadvantages might already have been partially diminished as a result of the year of child welfare, health, and mental health services they had received (Horwitz, Balestracci, and Simms, 2001), the finding of such a high prevalence of developmental problems for these children has important implications for the provision of child welfare and allied services.

Child Demographics

The ages of the OYFC children in care are distributed relatively evenly among groups of those who are about 1 to 2 years old (24%), 3 to 5 (17%), 6 to 10 (32%), and 11 to 15 years old (27%). Yet this seemingly even distribution belies the fact that the proportion of children in the sample who are 1 or 2 years old is about twice what it is for children of any other age. This is consistent with other studies of the characteristics of children in foster care and calls for more intensive efforts to address the reasons why such young children enter, and remain in, foster care at such high rates. The population is split evenly between males and females.

The largest racial group for the OYFC children is black (45%); 31% are white, and the remainder are American Indian/Alaskan Native (6%), Asian (2%), or other (7%). These percentages differ significantly from the proportion of children of each of these racial groups in the general population, a finding consistent with all preceding descriptions of children in foster care. Hispanic children constitute 17% of the OYFC children; the races of children identified as ethnically Hispanic are 51% white, 6% black, and 43% from the other, smaller racial categories. These proportions resemble those in the Adoption and Foster Care Analysis and Reporting System (AFCARS), which also includes children with shorter and longer stays (Table 7-1).

Table 7-1. Comparison of Children in AFCARS and NSCAW OYFC, by Race

	AFCARS^ (%)	NSCAW OYFC (%)
White/non-Hispanic	34	31
Black/non-Hispanic	39	45
Hispanic	17	17
AI/AN non-Hispanic	2
Asian/PI non-Hispanic	1
Unknown/unable to determine	7
Other^^		7

^ Data for Adoption and Foster Care Analysis and Reporting System (AFCARS) represent all children in foster care on September 30, 1999: http://www.acf.hhs.gov/programs/cb/dis/afcars/cwstats.htm, Oct. 6, 2001. ^^ ”Other” race/ethnicity is an NSCAW category and includes all children not identified as white, black, or Hispanic.

 

Experiences of Maltreatment Prior to Entering Out-of-Home Care

The available research on populations of children in foster care provides little information about the type, seriousness, and duration of the maltreatment children experience prior to entering foster care. The NSCAW data indicate that the typical child still in foster care at one year has experienced serious, multiple forms of abuse, as well as enduring abuse. Failure to provide typically involved the inadequate receipt of medical, dental, and/or mental health care (about 40% of these failure-to-provide cases involved the failure to seek or comply with medical treatment for potentially life-threatening illness or injury), a lack of adequate food, and a lack of adequate shelter. The high proportion identified as having service needs (rather than more material needs) suggests that the parent’s ability to obtain and use necessary services is critical to resolving these foster care episodes. The child welfare workers report that these services are available, suggesting that the opportunity exists for clients to obtain them.

For those in the failure-to-supervise category, the types of incidents that most often brought children into care to stay for one year include being left unsupervised for various amounts of time, abandonment, the failure to provide adequate substitute care (meaning the parents had left their children with caregivers who posed a risk to the children), and the failure to ensure that a child is playing in a safe area. The service implications for the more serious types of caregiving lapses are unclear, although the one-third of cases that involve inadequate or unsafe care seem most amenable to being addressed by responsive community-based services.

Children who were in care because of physical or sexual maltreatment had, in most cases, experienced serious harm or risk of harm, including a hit or kick to the face, head, or neck or some other body part; burns or scalding; violent handling (pushing, shoving, throwing, pulling, dragging); choking/smothering; and shaking.

Among the children still in foster care at one year for reason of sexual abuse, many had been exposed to very serious sexual abuse. Digital penetration of the vagina or anus, vaginal or anal intercourse, oral copulation, and masturbation with the perpetrator accounted for about 60% of the incidents. For well over one-third (39%) of the children in the sexual maltreatment category, the most serious subtype is fondling or molestation (without genital contact) or another, less severe type of sexual maltreatment, such as exposure to sex or pornography.

We also found multiple types of abuse in well more than half of the cases. Failure to supervise and failure to provide are often accompanying types of abuse. Half of the children with a most serious abuse type of failure to provide also had caretakers who failed to supervise them in some way, while 21% of the children with a most serious abuse type of failure to supervise were also victims of some type of failure to provide. Of children with a most serious abuse type of physical maltreatment, 33% also had caretakers who failed to supervise and 31% had caretakers who failed to provide. Children with a most serious abuse type of sexual maltreatment were especially likely to also experience another of the main types of abuse, as nearly two-thirds were not adequately supervised, 33% were also physically abused, and 22% were not adequately provided for in some way.

We examined the severity and duration (since the onset of the maltreatment) of the most serious abuse both separately and together. We found no significant correlation between the severity score alone and either the proportion of a child’s clinical scores or the distribution of standardized scores among the children in our study (as represented by a z-score). Nor was there a significant correlation between the duration since onset of abuse alone and the overall distribution of clinical problems, although there was a significant relationship between the duration of abuse and the proportion of a child’s clinical scores. Consistent with other findings that the severity and duration since onset of abuse make a combined contribution to maladaptive behavior, the combination of severity and duration has a significant correlation with both the proportion of scores children have in the clinical range and with the overall distribution of clinical problems.

We also considered the relationship between severity and duration of abuse and scores on individual developmental measures by type of abuse. The data show a gradient, with children who were placed into foster care for reasons of failure to provide having the fewest strong correlations with problem scores. The children who had experienced failure to supervise experience a stronger link between the severity and duration of their exposure and their measured developmental outcomes; the children who had experienced physical abuse of the longest severity and duration had the highest clinical scores. (It is unclear where sexual abuse fit on the gradient because of the limited number of cases that had scores on the measures for younger children.)

These findings present a compelling case that foster children are most often there because of prior dire circumstances and experiences in their homes. The multiplicity of abuse experiences, the severity of these types of abuse exposure, and the high levels of problem behavior that they experience identify the significant challenges that foster care providers, child welfare agencies, judges, and policy makers face. The children who have been sexually abused, although few in number, have especially high levels of serious forms of sexual abuse and involvement in other types of abuse. Neglected children, including those who experience failure to thrive and failure to supervise, are the least likely to experience other forms of maltreatment. This may be partly attributable to their younger age. Although researchers do not know how these children differ from children who never entered care or who went home from care, the data do indicate that the overall picture represents children who have experienced broad and pervasive exposure to maltreatment. These are questions that analyses of the NSCAW CPS sample component data will illuminate.

Type of Placement

Overall, the NSCAW OYFC cohort seems to fit squarely within the expected parameters for out-of-home care dynamics. One-quarter of OYFC children who were sampled had already gone home by the time of the interview. This is perhaps attributable to a timely one-year case review hearing followed by reunification. About half of all children either never had a reunification plan or had no current reunification plan. Among all children still in care (i.e., omitting the 25% who had gone home), 58% were in non-kin foster homes, 32% in kinship foster homes, and 9% in group homes. These proportions for kinship and non-kin foster care are identical to those reported in AFCARS for all children in out-of-home care. The proportion in AFCARS reported as in group or institutional care is higher (18%) than in NSCAW, although this could be because their data include children in the later teens; among our older adolescents the proportion in group care is about 18%.

The age characteristics of this population are consistent with those reported in many other studies (e.g., Barth, Courtney, Berrick, & Albert, 1996; Wulczyn, Hislop, & Goerge, 2000), showing that children at all ages are most likely to be in foster homes but that the older children have a greater likelihood than the younger ones of being in group care. Unlike other studies, we found no differences in the presence in kinship care for white children and black children; the proportion of Hispanic children in kinship care was the highest. The similar proportion of white and black children in kinship care could be attributable to the relatively short period that these children had been in care. Previous research would suggest that over time more of the children in kinship care will be black, because their stays in kinship care have been the longest (Courtney, Barth, Berrick, Brooks, Needell, & Park, 1996). White children are more likely than black or Hispanic children to be in group care, which is also typical of other findings (e.g., Barth, Courtney, Berrick, & Albert, 1994; Wulczyn, Brunner, & Goerge, 1999). Children of other races are most likely to be in group care, a finding that bears further investigation with the larger NSCAW CPS component data. The findings do not address the contribution that race may make in determining the outcome of remaining in foster care at one year—developing such a finding would require a different research design, one that follows a cohort of children from admission into foster care rather than sampling at one year in care.

Children’s Functioning

Data collection for NSCAW includes the most extensive battery of developmental measures ever completed with children in the child welfare system. The power of these findings arises, in great part, from the range of topics covered and by the array of sources of information in the measures. Analyses in this report focused on the standardized measures, which indicate that children in out-of-home care for one year are at serious social and cognitive disadvantage when compared with national norms. Only 34% of the children have no clinical scores on the potential 4 to 6 developmental measures (depending on their age) that they completed, but 35% of the children have 2 or more clinical scores. This is an extraordinarily high prevalence of social or cognitive problems. These findings are consistent with those of other investigators who have also pinpointed high levels of behavior problems among foster children (e.g., Clausen, Landsverk, Ganger, Chadwick, & Litrownik, 1998; Courtney & Zinn, 1996).

Children in out-of-home care for one year tend to fall marginally below the norm compared with the general population on nearly every measure, including those for cognitive capacities, language development, and academic achievement. These apparent cognitive vulnerabilities are consistent with other research (e.g., Bolger and Patterson, 2001) and, because of the magnitude of the difference from the norms, not fully explainable by previous findings that somewhat lower scores are overrepresented among poor children (Duncan and Brooks-Gunn, 2000; Parke, 2000). OYFC children have particularly low social skills, with almost two-fifths rated as having “fewer social skills.” Many also have less than adequate daily living skills, with almost one-fifth rating low. The poor performance on social and cognitive measures is pervasive and is demonstrated by children in kinship care, foster care, or group care. Although children in group care tend to have a higher proportion of social and cognitive problems, as do children with sexual abuse as the most serious type of abuse, this may be attributable to the older ages of the children who have been sexually abused or in group care rather than to a decision making process that funnels more difficult children into group care for treatment. At younger ages, equally difficult children are in all forms of out-of-home care.

We cannot determine with the OYFC data whether the generally poor social and cognitive functioning of these children differs at the time of the assessment from what it was at the time the children entered foster care. The limited number of studies of foster care that do have developmental measures suggests that children in foster care generally receive some developmental benefits (see, e.g., Fanshel and Shinn, 1974; Wald, Carlsmith, and Leiderman, 1988; Taussig, Landsverk, and Clyman, 2000; Horwitz, Balestracci, and Simms, 2001), which argues that the developmental status of these children is not likely to have worsened since entering care. To assess changes in performance for children in foster care, we will examine at entry and at 18-month follow-up those children who entered foster care from the NSCAW CPS sample component.

The transition from child abuse to delinquency has been described by several investigators (e.g., Jonson-Reid and Barth, 2000; Maxfield and Widom, 1996), although none has had measures of delinquent behavior during foster care. In this study, caregiver-reported and self-reported delinquency levels vary by placement type. Non-kinship and group home caregivers report clinical/borderline delinquent behavior in approximately 50% of the children compared with only 12% of kinship caregivers. Self-reports of clinical/borderline delinquent behavior are lower than caregiver reports, in general, but are consistent with the patterns reported by caregivers: they are highest for children in group home care (27%) compared to children in non-kinship and kinship foster care (10% and 4%, respectively).

These results also yielded some potentially positive findings. Contrary to our predictions that these children would have high self-reported depression scores due to the upheaval in their lives, this is not necessarily the case. Clinical levels of depression varied with the instrument used, with rates reported on the CDI being somewhat lower and rates on the YSR somewhat higher than national norms. This positive finding may be attributable to the insensitivity of the CDI to children’s sense of well-being in a more protective environment, to the difference in age ranges for the two measures, to a desire of youth to indicate how well-adjusted they are, or to some other unmeasured contributors. As predicted by the those who argue that children in kinship care experience greater comfort than children in non-kinship care (e.g., Needell and Gilbert, 1997), children in kinship foster care reported less depression than children in non-kinship foster care on the CDI. On the YSR Depression subscale, there was not a significant association between depression and placement type. Additionally, on the YSR, children reported more depression than the normative sample.

7.2 Current Caregivers Characteristics

Current caregivers of OYFC children are substantially older than American parents in general. Thirty percent are aged 50 or over. Most strikingly, a total of 11% are aged 60 or over, and another 19% are between 50 and 59 years old. In contrast, in the Urban Institute’s National Survey of America’s Families, only 2% of parents in all households are older than 54. Also in contrast with the general population, a relatively smaller proportion (24%) of caregivers of OYFC children are younger than 35; among the general population of America’s families, the proportion younger than age 35 is 41%. Group home caregivers are much younger than other caregivers as more than three-quarters (76%) of the group caregivers are under 40. Of the kinship and non-kinship foster caregivers, older caregivers reported better mental health, while younger caregivers reported better physical health. Many of the kinship caregivers had only been caring for this child for one year. Other findings suggest that the role of fulltime caregiving can become quite demanding after a time (Bowers & Myers, 1999; Caputo, 2001), so some changes in this level of well-being might be expected in the next wave of data collection.

Non-kinship foster caregivers tend to have more experience as foster parents than kinship caregivers. More than half of kin caregivers have a year or less experience, whereas more than one-third of non-kin foster caregivers have six or more years of experience. This evinces the potential value of accessible support and training for kinship caregivers.

Compared to all parents, foster caregiving and kinship caregiving parents are far less likely to be white. Only 36% of foster care and kinship caregivers are white. The largest group of foster care and kinship caregivers is black. Another 15% of current caregivers are Hispanic; the remaining 7% are of other races or ethnicities. The racial and ethnic makeup of group caregivers is different from foster and kinship caregivers. Almost half (49%) of group care staff interviewed in this study are white, 33% are black, 10% are Hispanic, and the remaining 8% are of other races or ethnicities. Most kinship and non-kinship caregivers are women (90% and 97%, respectively) compared to 75% of group caregivers.

The majority of kinship and non-kinship caregivers are married (55% and 73%, respectively). Kin are far more likely to have never been married (18% vs. 9%) or to have been divorced, widowed, or separated (28% vs. 18%). In their marital status, group caregivers are different from their foster care counterparts—the largest percentage of group caregivers (56%) have never been married; with 31% currently married, 4% widowed, and 9% divorced. Thus group caregivers are younger and less likely than foster care providers to have experienced marital family life.

Kinship and non-kinship foster parents generally have low educational attainment. In particular, higher proportions of non-kin caregivers have a GED or higher; about twice as many kin caregivers than non-kinship caregivers have no degree. More than half of kinship and non-kinship foster parents work full- or part-time, with no significant difference in employment status between kin and non-kin caregivers. Group caregivers are, generally, more educated than foster or kinship foster parents, with almost two-thirds (63%) holding a bachelor’s degree or higher. In making comparisons to educational levels in households nationally, group home caregivers differed from other households more than kinship or non-kinship caregiving households in that they almost always have at least a high school diploma or GED and usually have more than a high school education. Kinship and non-kinship households are also at lower income levels than most U.S. households. Fifty percent of families in the U.S. earn $50,000 or more annually (U.S. Census Bureau, 2001), while only approximately one-third of kinship and non-kinship households earn this amount. More kinship caregivers tend to have lower incomes than non-kinship caregivers, with 41% having incomes under $25,000, compared with only 21% of non-kinship households.

The low education and income levels of kinship foster parents suggest that the growing concerns about the educational underachievement of foster children are not going to be easily remedied, since so many foster children live in homes with relatively few educational or other resources. The relatively high educational levels of the group caregivers—although very likely not fully representative of all group care providers, because the key respondent is likely to have been a leader in the facility—suggest that there may be a way to draw on the strengths of these personnel to partner with other less educated caregivers of children in out-of-home care.

Children in non-kin foster homes are likely to live in households with more family members than American children typically do—about one in three children in non-kinship care lives in a household with five or more children. Non-kin foster homes with five or more children are, however, the most likely of all caregiving households in the study to have two parents, and almost half of the mothers in these larger foster homes remain at home full-time. The high proportion of children living in foster care with five or more children in the household is striking and not previously reported. Although nearly half of these larger households have a “stay-at-home mom,” this is still something that needs more understanding, because these homes are now so different from those of the typical family. Perhaps these homes are a form of small group home, and child welfare services are developing a hybrid between foster homes and group homes. This possibility, and its implications for children’s care, deserves further review.

7.3 Children’s Experiences in Foster Care

Environment the Children Live In

Assessments of the physical environment of kinship and non-kinship homes, using the HOME-SF observation items, show that the difference between kin and non-kin caregivers is not significant for any individual physical environment indicator. The total mean physical environment score is, however, significantly higher among kin caregivers. In addition, kin appear to rate higher on emotional responsiveness and are not very different from non-kin on cognitive stimulation or in punitiveness. The finding that kinship environments are at least as positive as foster home environments, and possibly more positive, is a significant finding that runs counter to some conventional wisdom. It contrasts too with the finding that kin have somewhat lower incomes.

In examining the association between caregiver characteristics and caregiver behaviors, as measured using the HOME-SF, black caregivers obtain significantly lower scores on cognitive stimulation and emotional responsiveness. This finding is likely to be more attributable to the greater poverty in which African-Americans live than to cultural characteristics of black caregivers, although both may be influences (Bradley, Corwyn, Burchinal, McAdoo & Cool, 2001). Caregiver age is not consistently related to the provision of cognitive stimulation or emotional responsiveness.

Caregivers also reported on the environments surrounding the residences in which OYFC children live. In general, the community environment does not differ among types of placement, although comparisons show that kin caregivers have significantly more concern about the safety of their neighborhood and how their neighborhood compares with others than non-kin caregivers; group care providers have the most positive view of their surrounding community.

The findings, based on HOME-SF observations and children’s self-report, show that children and youth report relatively positive and close relationships with kinship foster parents and non-kin foster parents that have some punitive elements but are generally perceived—by the children—as caring. Some of the findings slightly favor the conclusion that children prefer kinship care to non-kin foster care. The results appear less positive for group care, although some comparisons were impossible because of the relatively small sample of children in group care. All of these findings must be understood as being somewhat speculative because children are not randomly assigned to kinship, non-kinship, and group care.

Analysis of children’s exposure to maltreatment and violence by their current caregiver indicates that a small proportion of children may be experiencing harsh discipline but that most are experiencing more routine forms of care in non-kin foster homes and kinship care settings: For example, 10% to 14% have seen someone get yelled at or have been yelled at by their current caregiver; and approximately 4% or fewer have observed a theft, an adult point a knife or gun at someone, an arrest, or a shooting. Children in kinship and non-kinship foster care experienced similar levels of these incidents.

Although the statistical analysis endeavors to control for important background characteristics like age and most serious type of abuse, there are very possibly other unmeasured and uncontrolled differences in children who enter each type of setting that could explain the differential response to kinship care, non-kinship care, and group care. These findings challenge child welfare services to look further for alternatives to our conventional ways of providing out-of-home care.

Children’s Perceptions and Expectations

Children appear to find some basis for optimism in their new caregiving environments. More than half report that their new schools and neighborhoods are better than before. Half think they would like their current home as a permanent home and the majority report their belief that their family circumstances will improve.

Children in group care are the least likely to hold these optimistic beliefs and least likely to have positive responses to their current living arrangement. This may occur because the group care children tend to be older and to have more serious behavior problems, or it may be due to the nature of group care—or both. Further analysis of the interplay between social and cognitive problems and attitudes about out-of-home care settings would be helpful.

Older children gave their views about their current living situation, where they would live in the future, and their biological parents. Most children (60%) reported seeing their biological mothers twice per month or less, and most children desired more contact with their mothers. Almost three quarters (73%) reported seeing their biological fathers twice per month or less; 60% desired more contact with their fathers. Most children (77%) reported wanting more contact with siblings and frequently missing their family. Over half (56%) of the children saw their siblings less than once per month. Very few children reported avoiding family visits and most were quite happy following their visits. Approximately one-third of children reported that visits were frequently cancelled.

High endorsements of items related to permanency suggest that children in foster care do seem to feel some sense of permanence even in the midst of enormous upheaval. Most children appear to view their living situations positively—they like living with the people they are living with and are positive about their new neighborhoods and schools—even while hoping for reunification with their families. Children in group care did have a somewhat lower mean score on the permanency measure and appear to be different in several ways. They are far more likely than those in kinship care to report that they do not like the people with whom they are living. They are more likely to report never seeing their biological father or mother. Children in group care are also more likely to report visits being cancelled.

Compared with children in non-kin foster care, children in kinship foster care placements tend to report more parental emotional support and somewhat more structure. One primary factor in this difference, however, is the secondary caregiver. Overall, OYFC children tend to feel at least slightly closer to their secondary caregiver, who is most frequently male (foster father, uncle, or grandfather). Children in foster care for one year do tend to report feeling less close to their caregivers than children in the nationally representative Adolescent Health (Add Health) sample. Yet OYFC children tend to work on school projects and talk about personal problems and school issues with their primary caregivers more often than do adolescents in the general population.

In general, children in out-of-home care appear to be troubling to their caregivers and teachers, but do not report being particularly troubled about their own lives. Their internalizing scores on the Child Behavior Checklist (CBCL) and Youth Self Report (YSR) are not very high, on average; their depression scores on the CDI are rarely in the clinical range, and their reports of comfort with their living arrangements are very high. On the placement satisfaction scale, OYFC children indicate on average that their contentment with their current living situation is 8.99 on a 10-point scale. These findings are consistent with those of other investigators who report that the experience of being taken into out-of-home care is often very difficult for them, but the experience of living in foster care is generally positive (Barth, 1990; Johnson & Voss, 1995).

The one exception to the relatively positive reports on their residence comes from those children living in group home care—who are also older than the other children. This may be partly attributable to the group care environment having fewer negatives but fewer positives as well. There are not options for long-term stays—that is, making this a permanent home—and few of the positive experiences of family life, such as feeling close to a few people whom you get to know and care for over a long period of time. The negative findings about the experiences of group care—including high rates of running away—call into question why group care continues to be such a commonly used approach in providing out-of-home care.

7.4 Service Needs and Receipt

Risk Factors and Initial Services

Although the data available for this report contained only a limited set of risk factors (e.g., parental substance abuse was not included), analysis of family risk factors present prior to placement show that these risk factors are not significantly related to the race of the child, most serious type of abuse, or placement type. However, low social support, a history of abuse by the primary caregiver, and prior reports of maltreatment are significantly related to child age. Older children (aged 11 and over) appear more likely to have had prior reports of maltreatment and to have not had a second supportive caregiver present; families with young children (1 to 2 years old) are most likely to have had low social support.

Child welfare workers indicated substantial unmet general needs for biological families, with over three-quarters of families needing income assistance, Medicaid, mental health, alcohol or drug treatment, or day care services. Mental health and alcohol and drug treatment services were the two services most likely to still be needed even though there had been a referral for them. This was primarily because parents refused the service. Child welfare workers rarely reported other logistical reasons for failure to use a referred service. Nonetheless, the vast majority of those perceived to have a drug or alcohol problem did obtain a formal assessment. Among those primary caregivers who obtained a formal assessment for either a drug or alcohol problem, 83% were found to have a serious or moderate impairment.

Many biological families had received child welfare services to prevent the placement of this child. Fifteen percent of families received intensive family-preservation services. Another 16% received other home- or community-based services, and 52% received non-intensive monitoring at home. Relatively few (less than 20%) received respite, parent aide, home management, and home repair services. The most commonly needed general services for which permanent primary caregivers were referred, prior to placement, were income assistance, housing, substance abuse, and mental health services.

Reunification and Adoption

Child welfare workers were asked if children then in an out-of-home placement had a current case plan to be reunified with someone in their family. There is a current plan for about one-quarter of the children who have been in foster care for one year. About one in twelve children had never had such a plan, but the majority had plans that had since ceased.

The youngest children appear to have fewer reunification plans than the children in the older age groups. This is consistent with foster care caseload dynamics research, indicating that younger children are much more likely to have reunification plans changed to adoption (Barth, 1996). Because the interviews occurred after one year, many children had already completed their permanency hearings, thus some reunification efforts would have ceased. Similarities in the presence of reunification plans for black, white, and Hispanic children are consistent with recent evidence that black children are beginning to have more equal access to reunification (Wulczyn and Brunner, 2002). There are no major differences between children in kinship and non-kinship care related to the proportion with reunification plans (although children in group care have the highest proportion of reunification plans). Children with a most serious abuse type of failure to supervise are more likely than children with a most serious abuse type of failure to provide to have a current reunification plan (or completed reunification).

We examined the risks present in the child’s home at the time of placement to see if these appeared to be related to whether the child currently has a reunification plan (or completed reunification). The percentage of children with a current reunification plan or completed reunification does not appear to vary depending on the risks present, with the possible exceptions of those children whose caregivers did not cooperate with the authorities and those children whose permanent caregiver was a victim of abuse. These children may have somewhat lower rates of reunification.

These findings are the first from a national study following the passage of the Adoption and Safe Families Act (ASFA). They indicate that most children continue to get reunification services, which often are multi-pronged, reaching out (at least at first) to multiple family members. That only 25% of children had a reunification plan at the time of the child welfare worker interview suggests that the one-year time limit on reunification is observed. Having so few extensions of reunification plans beyond the expected 12-month time frame is in keeping with the allowance and intent of the law.

Children in group home care have different reunification patterns than other children. These children and their families are different in other ways as well; the children are more likely to be white, to have been alleged to have emotional maltreatment, to have experienced sexual abuse, and to have substantial mental health problems. It is intriguing that children in group care are the most likely to have a reunification plan. This might well indicate that these children are in group care for different reasons than children who are in non-kin foster homes or kinship foster care.

Although all of the children in group care were identified by child welfare agencies as having entered care for reasons of abuse and neglect (and not simply for mental health treatment), there seems to be little involvement of these children with adoption or other alternative permanent plans. The young adolescents in our sample, whether in group care or not, tend to have different characteristics (including a greater extent of clinical problems) and patterns of care. Evidence from other sources (e.g., Wulczyn, Hislop, & Goerge, 2000) suggests that young adolescents are among the largest groups of children entering child welfare supervised out-of-home care, and this is particularly true in rural areas. The findings here provide some confirmation that adolescents who enter child welfare supervised out-of-home care are not uncommonly from a different population of children than other children receiving child welfare services.

Although questions for the caseworkers regarding adoption plans for OYFC children were reserved for the second wave of data collection, this first wave of data collection did include questions for foster parents regarding their feelings about adopting the children in their care. The majority of foster parents (68%) have considered adopting the child if that becomes an option in the future. The most common factors contributing to the foster parents’ desire to adopt are their feelings of love and affection for the child and the feeling that the child is part of their family. The most common factor discouraging those foster parents who have considered adopting the child from wanting to do so is that the biological parents’ rights have not been relinquished or terminated; however, almost one-fifth of these foster parents identify no factors discouraging them from adopting. The majority (85%) of foster parents who have considered adopting the child have spoken to their caseworker about this, and 89% of those who have spoken to their caseworker indicate the caseworker’s response to be encouraging or strongly encouraging.

Services to Children

Child welfare workers report referring children in need of various services from 40% to 100% of the time. When children in need were not referred, the most commonly cited reason is that the children were already receiving these services, although a sizable proportion had some “other unspecified reason.” Rarely does the child welfare worker indicate that a child was not referred because services were unavailable in a particular geographic area, could not be financed, or because the child was ineligible for the service.

Child welfare agencies provide a range of direct services to families, in concert with mental health and other service partners, to try to prevent placement and achieve the reunification plans described above. For many children and families, services begin well before this involvement with out-of-home care. Prior to coming to the placement that they were in at the time of this survey, children had significant involvement with other children’s services. About 23% had received some form of specialty mental health services in the past 12 months, including outpatient therapy, day treatment, and hospitalization. Significantly more children with a borderline or clinical score on the CBCL had received specialty mental health services than children without such a score, but no more than 43% of children with a clinical score had received such services in the past 12 months. Receipt of specialty mental health services was clearly and systematically associated with the children’s proportion of clinical scores. Multivariate modeling indicated that white children, children in group care, and children with a higher proportion of clinical scores were far more likely to receive specialty mental health services.

Over one-third of the children who have been in foster care for one year and who have a clinical score on at least one of the examined measures have begun receipt of special education services. Bivariate analyses reveal that significantly fewer children between the ages of 3 and 5 with at least one clinical score have received special education services than children in older age groups with at least one clinical score. In addition, bivariate analyses indicate that significantly more children with a most serious abuse type of physical maltreatment have received special education services than children with a most serious abuse type of failure to provide. A logistic regression that modeled school-aged children receiving special education services and controlled for age, gender, race/ethnicity, and placement type indicated that children six and older in group care are significantly more likely than children six and older in non-kinship foster care or kinship care to receive such services. When looking at young (aged 1 to 5) children, children in the 1- to 2-year old category are significantly more likely than children aged 3 to 5 to receive special education services. In addition, young Hispanic children are significantly more likely than young black children to receive such services.

A broader analysis, which included receipt of supplementary education services (i.e., services like assessment, tutoring, and counseling) along with special education services, revealed an exceptionally high level of such services among children with a clinical score who have been in foster care for one year—92% of this population has received supplementary and/or special education services. Logistic regression performed on the subpopulation of school-aged children indicated that females six and older are far less likely than males six and older to receive supplementary and/or special education services. For young children residing in kinship care, the likelihood of getting supplementary or special education services is far lower than it is for young children in non-kinship foster care.

Child Welfare Workers and System Characteristics

Several specific descriptors of child welfare service providers were collected. Most child welfare workers are relatively young, with more than half under 40 years old. About 46% are white, 32% are black, 11% are of Hispanic ethnicity, and 12% identify themselves as other. Child welfare workers carrying these cases have a wide variety of training, but fewer than half have bachelor’s or master’s training in social work. Child welfare workers are more likely to have a Bachelor’s degree in something other than social work than any other type of degree. Given the difficulties that this report has identified among children in foster care and the variety of caregiving characteristics and configurations that must be accommodated, the lack of advanced education and training received by the child welfare workers of these children is troubling.

Child welfare workers and advocates often suggest that matching the race of children to that of the workers serving them will result in services that better take into account the cultural norms of the children and families who are served. Black children were served by black child welfare workers in almost half of the cases, white children were served by a white child welfare worker 60% of the time, but Hispanic children were served by a Hispanic caseworker only 22% of the time. Although there appears to be consensus that culturally competent services are important, as is advanced training, there has been little discussion of the merits of the current configuration of the child welfare workforce. Perhaps a more complete analysis of the plentiful data on child welfare worker characteristics, experiences, and attitudes (largely unanalyzed in this report) could contribute to that discussion. Child welfare services are delivered locally, and these locations may be significantly related to service configurations. Using information collected in interviews with child welfare managers, the researchers examined a variety of questions about the relationship between agency and primary sampling unit (PSU) characteristics and service delivery, including the following:

• administration of the child welfare agency (state or county),

• PSU size (small/medium or large),

• urbanicity (rural or urban), and

• poverty level (poor or non-poor).

We examined many service delivery characteristics across these types of agency settings but found only a few significant differences. With regard to the number of risks present in the household at the time of placement, children in rural PSUs had significantly more risks than did those in urban PSUs. Yet children in non-poor PSUs had significantly more risks than those in poor PSUs. This suggests a different threshold for placement in non-poor and rural PSUs — perhaps cases have to be more severe to get placed in rural areas (where there are fewer resources) and receive more preplacement preventive services in non-poor communities, which can more readily afford them. The other two variables with notable differences were both caseworker characteristics—their highest educational degree and their race. County-administered agencies appear to employ caseworkers with more advanced degrees than do state-administered agencies. Non-poor PSUs have significantly more caseworkers with MSWs and significantly more with other bachelor’s degrees; poor counties have significantly more caseworkers with other master’s degrees. Small or medium, rural, and non-poor PSUs are all less diverse than their counterparts with regard to the race of their caseworkers. In the data, these small or medium, rural, and non-poor PSUs also have less ethnically and racially diverse clientele.

7.5 Summary

Overall, this study finds a child welfare system that seems to be operating roughly in accordance with timelines and caring for quite troubled children and youth, with caregivers and child welfare workers who are diverse but lack much formal training or education in the work they are doing. Out-of-home placements—either with kin or non-kin—are experienced by the children and youth as benign, with the exception of group home care, which is not favored. For children in foster care at one year, the picture is dominated by infants and adolescents who came into care because they were neglected. Reunification plans for younger children have largely stopped by this one-year mark; for adolescents, however, they have not.

This snapshot suggests the importance of continuing to follow the life course of these children to better understand whether these threats to their well-being and high levels of developmental difficulties will manifest themselves in greater problems or will be mitigated. Taken together, children one year in out-of-home care have substantial social and cognitive impairments—in each type of setting that they are in. Although this study cannot assess the capacity of out-of-home care providers to meet the needs of their children, the relatively large family sizes, low income and education, and older age of the caregivers suggest that these homes will require considerable support in order to help generate these important gains. These findings call for additional commitments to developing intervention approaches that help children in out-of-home care to achieve normative developmental outcomes (e.g., Fisher, Gunnar, Chamberlain, and Reid, 2000). The needs of these children and the fact that one year in foster care has not brought them even close to alignment with normative expectations is compelling evidence that their developmental needs are an appropriate focus for additional planning and services.

 

 

Table of Contents | Previous | Next