Search Type Title Author/Investigator Search Advanced Search

Table of Contents | Previous | Next

11. Summary of Findings

Our findings are consistent with earlier evidence that most children who come to the attention of child welfare services and are the subject of an investigation will not, at least at this point in time, receive ongoing child welfare services. The investigation is the primary “service” that approximately 2.4 million children, and their families, will receive from their new involvement with child welfare services, although their case remains open. Our NSCAW findings generally corroborate and extend those of other investigators in indicating that the vast majority of children who have a child welfare investigation will have the case closed at home with no ongoing services. About 1.5 million children each year will have their cases closed following the investigation even though their families are often experiencing substantial difficulties in providing safe care for children who have a substantial level of developmental problems.

Contrary to popular conception, child welfare agencies do not exclusively or even mostly provide foster care services. Child welfare service agencies are sometimes branded as the “foster care agency.” Yet foster care is not usually provided to children reported for abuse or neglect. Among the small proportion of children reported for abuse and neglect who do receive ongoing child welfare services, twice as many families will have their case opened with services at home than will have their child(ren) placed in out-of-home care (about 575,000 vs. 272,000). Nearly 90% of children whose cases are investigated will continue to reside at home with their families.

Children are most often investigated for reasons of neglect, with physical maltreatment seen as the most serious abuse type for over one-quarter of the children, and a larger share for older children. These proportions, and how they related to the ages of the children, are similar to those reported by NCANDS (U.S. DHHS, 2003). The most serious abuse type does not appear to be associated with whether children are in out-of-home placements, whether in-home children receive services, or in what level of care out-of-home children are placed. In some cases, however, the subtype of abuse is associated with the decision to provide services at home or in placement. For instance, children in out-of-home care are more likely to have been abandoned (a subtype of failure to supervise), while children remaining at home are more likely to have experienced the least severe forms of sexual maltreatment. The impact of these maltreatment types on service dynamics are likely to emerge over time—given previously developed evidence that physically and sexually abused children are far less likely to have longer stays in foster care (DHHS, 2003)—but they are not readily apparent at entry into CWS.

Children remaining at home were more likely to have experienced less severe maltreatment and to have had shorter times since onset of abuse than children in out-of-home care. For children remaining at home, those receiving services have experienced maltreatment for a longer time than those with no services; for children remaining at home with the abuse types of physical maltreatment or failure to provide, provision of services was more likely for children in more severe categories. Decisions about children experiencing other maltreatment types seemed to have less sensitivity to severity—perhaps because failure to supervise has fewer gradations and because a wider range of acts of sexual maltreatment are considered egregious. These findings are also confounded, somewhat, by the inclusion of cases that are re-entering care or are entering care for reasons that may be related to older children’s mental health problems.

In both of these types of cases, severity of maltreatment is not related to the decision to place children into out-of-home care.

If children are in out-of-home care, the children may be living with relatives in “kinship care,” with non-relatives in nonkinship care, or in group care. Children in out-of-home care have higher levels of problem behavior and are more likely to obtain special education and specialty mental health services than the opened or closed in-home cases.

When children are placed in out-of-home care, nonkinship foster care and kinship foster care are used in similar proportions, and at higher rates than group care, the latter of which is used most often for children age 11 and older. Children age 3 to 5 are far and away the least likely to be placed into out-of-home care.

We detected no differences based on age, race/ethnicity, gender, or most serious abuse type between the in-home children who receive child welfare services and those who do not receive such services. Problem levels for children are higher, however, for families receiving in-home services than they are for those who are not. The receipt of ongoing child welfare services following an investigation does not appear to result in a profusion of additional services for children. There are virtually no differences in the receipt of mental health or special education services between in-home open and in-home closed cases. This is consistent with the finding that a sizable proportion of the families we had identified as receiving in-home child welfare services had very limited contact with child welfare agency personnel. Yet, taken in concert with findings of substantial developmental disadvantage for children in this sample, this finding indicates the continued need to boost the coordination between child welfare services and other child serving entities.

We do, however, see significant increases in parental use of mental health and substance abuse services among the opened in-home service cases. The opening of a child welfare case is also strongly associated with the likelihood that parents will obtain other services, a finding that provides plausible evidence that the case management role that child welfare workers have increasingly assumed may function to the benefit of parents.

11.1 Family Risk at Investigation

The families of children who come to the attention of child welfare agencies on any given day have very often been there before. The family issues that led them to be investigated at this point appear to be persistent. More than half of all children/families have had prior reports of maltreatment to the agency, and 30% have prior CWS history (not including investigations).

The child welfare worker’s description of case risk factors offers the opportunity to identify the case characteristics on which the service decisions turn. “Family” risk factors—including no other supportive caregiver in the household, high stress in the family, low social support, and when the family has trouble paying basic necessities—are the most common but not the most consequential. Those children whose families had more parenting or substance abuse risk factors were less likely to be at home with no services. More generally, the more child and family risk factors, the more intensive the level of services. Supporting the idea that the circumstances of placements of children into kinship care may differ from those of children receiving other services, several of the risk factors—the drug abuse of the primary caregiver, a recent arrest history of the primary caregiver, and a serious mental health problem of the caregiver—were higher for children who were placed into kinship care following the investigation. The race/ethnicity of the child does not appear to be significantly related to critical factors in decision-making.

The severity of the maltreatment that children experience has an important relationship to their service receipt. Severity appears more predictive of service levels than the maltreatment type. Yet most research on child welfare services includes maltreatment type but no estimate of severity. In order to better understand the efficiency and equity of evolving child welfare services delivery, routine data collection on maltreatment severity, perhaps building on the Maltreatment Classification System, must become part of child welfare agency practice. Having a record of maltreatment severity will help us to understand the process by which children become re-involved with child welfare services—that is, whether or not the re-involvement was preventative and followed a low severity exposure to maltreatment. At the same time, nearly one in five children enter out-of-home care with low severity ratings—and these children are disproportionately those who are re-entering foster care or who have evidence of prior psychiatric hospitalizations or serious mental health disorders.

Risk assessment analyses found that agencies are very concerned about active substance abuse and serious mental health problems. Still, poor parenting—and the related concepts of motivation to change and cooperation with CWS—is the most significant factor influencing placement decisions. Poor parenting risk scores predicted placement in out-of-home care or receipt of services at home over having a case closed at home. In addition, high substance abuse risk scores predicted placement in out-of-home care as opposed to remaining at home with no services.

• Families are experiencing many contacts with CWS, and the longer the involvement with CWS the greater the likelihood of receiving higher levels of child welfare intervention. The impact on children of consistently living in such a way that triggers child abuse reports, but few services, cannot augur well for children. The impact of receipt of services will be better captured in the forthcoming analyses using the longitudinal data.

• When making decisions about the level of intervention required to protect children, child welfare workers are paying attention to family patterns of behavior over time. Previous reports and case openings are cited as influential in deciding the level of services that is needed. Decisions should be improved, then, when there is greater continuity of information about family case histories.

• The findings generate a range of recommendations for routine child welfare services data collection and analysis. Most prominently, such data collection should include the severity of the maltreatment and more differentiated forms of child neglect in order to allow greater understanding of what is occurring in child welfare decision-making and service provision. The field is well beyond the period when a single neglect category can be sufficient—especially because this label covers so many children with such diverse conditions. Further, because so many cases are investigated despite having relatively low severity and because severity ranges substantially, tallies of reports and re-reports that ignore severity are insufficiently precise to be used as determinants of the quality of service provision. Without such changes, reopening cases so that serious maltreatment can be preventive will routinely be viewed as service failure rather than as appropriate, preemptive care.

11.2 Children’s Development, Functioning, and Behavior

Our current findings offer long-awaited information about the functioning of children who first enter child welfare services. Most studies of children involved with child welfare have assessed children while they are in foster care or thereafter, often finding that they are not doing well and sometimes concluding that this is a result of service provision (e.g., McDonald, Allen, Westerfelt, & Piliavin, 1996). These children are quite troubled, whether remaining at home or, especially, going into out-of-home care or group care. In general, the children in kinship care are somewhat less troubled than those in nonkinship placements. Similarly, children with closed cases at home are less troubled than children at home with an open case. Children in group care have higher levels or equivalent levels of problems than children in any other setting.

The NSCAW data collection enlists an abundance of standardized measures of children's functioning, yet the findings do not allow for blithe comparisons to published norms because the children in our sample are poorer and less populated by White children than the standardization groups used for the measures.

With that caveat in mind, the evidence is persuasive that children involved with the child welfare system—whether in-home or out-of-home—score below the average for the general population of children the same age on physical, cognitive, emotional, and skill-based domains. Although not all of these differences—especially those on intelligence—are significantly different from the norms, the breadth and consistency of the underperformance is striking. Only 30% of children do not have any measures in the clinical or high-risk range. This study will, eventually, provide the best estimates ever obtained of how these children are faring over time and according to the services received. Other investigators have provided substantial reason to think that there are a range of deleterious effects experienced by maltreated children that may well last into adolescence (Lansford, et. al., 2002) and beyond (Dube, et. al., 2001). The minimal level of intervention that these children and families receive seems unbalanced against the likelihood of large long-term risks that they face.

Although the children who entered foster care often scored lower than children who remained at home, the low levels of performance of all the children is the most vivid finding. That is, the children who remained at home had proportions of scores in the clinical range that were more like those of the children who went into out-of-home care than they were like the children on whom the tests were normed. This finding indicates that the child welfare system attends most to its primary objective: to protect the safety of children from inadequate parenting. Decisions about the services that families should receive do not appear to hinge, fundamentally, on a child’s general cognitive or social functioning. Evidence presented in Chapter 4 suggests that decisions about the level of placement do vary systematically based on the assessment of risk done by the child welfare worker and that the evidence is focused, for the most part, on parental risks and supports (although a few child factors are included).

The safety and well-being of children, and their developmental futures, build on many factors, beginning with their physical well-being. The children in this study are very often not of normal weight; although the seriousness of these deviations from the norm cannot be determined immediately, they may be part of a set of neglect experiences that are predictive of substantial developmental delay (Dubowitz et al., 2002).

The development of children involved with CWS is not uniform and varies, in some ways, by the demographic characteristics of the children. Older children, whether or not they are in out-of-home placement, appear to have more behavior problems by report of their caregivers, and teachers and by self-report. In general, the greatest differences in problem behavior rates are between group care children and other children involved with CWS. Although some of the children who are in group care have been there long enough to have to have been influenced by others’ problem behavior—and, potentially, had their problems worsen as a result (Dishion, McCord, & Poulin, 1999)—this is unlikely to explain the pervasive pattern of problem behavior among children in group care. Children coming into group care almost certainly enter with worse behavior than children entering lower levels of care. This does not mean that treatment-focused foster care services could not provide an alternative to group care for many of these children.

Among the children in out-of-home care, the children in kinship care, in contrast to children in foster care or group care, have scores more like the children remaining at home. This partially confirms findings from other investigations that children in kinship care have fewer problems (e.g., Benedict, Zuravin & Stallings, 1996; Berrick, Barth & Needell, 1994; Keller et al., 2001) but may also be somewhat attributable to more critical evaluations of behavior by nonkinship caregivers (Shore, Sim & Le Prohn, 2002). The possibility that kinship caregivers are less problem-oriented in their ratings of the children in their care is partially born out by the findings that older children in kinship care rated themselves as having behavior that was marginally worse than children in nonkinship care. Yet understanding the possible impact of rater bias is obfuscated by teacher reports that tend to agree with the kinship caregivers in their ratings of the behavior of children in kinship care (Shore, Sim & Le Prohn, 2002).

The findings of compromised learning, social skills, and behavior among children alleged to be maltreated is not surprising given previous local and more circumscribed investigations that have drawn a similar picture about children in foster care over the last two decades. (See Landsverk, 1997, for a summary.) More recent work indicating the pervasive threat that child maltreatment represents to the cognitive and emotional well-being of children also predicts such findings (e.g., Glaser, 2000; Teicher, 2002). The major finding of this study is that the children who are alleged to have experienced maltreatment, even when their cases are unsubstantiated or closed, have quite high levels of problem behavior and concerning levels of social and cognitive deficits compared with the population norms. In essence, we know that child maltreatment has untoward effects on children’s development, but the child welfare system has little to offer many maltreated children. Children who do come to be recipients of child welfare services and receive them at home do not appear to have an increase in direct services to address their needs. Their primary means of benefiting is likely to be indirectly—through the additional services that their parents receive.

• Children involved with CWS, no matter what setting they are in, tend to have more physical problems and very often have substantially more cognitive, behavioral, and social difficulties than children in the general population. Although child welfare appears to be focusing appropriately on its primary foci of child safety and permanency, our developmental findings suggest that many of these children have such substantial developmental needs that they risk becoming children who are difficult to provide with safety and permanency. These developmental problems start very young. Renewed efforts are needed to coordinate the receipt of developmental services for the youngest children.

• Some very young children (younger than 11) are residing in group care. Although they are almost universally rated as having clinically significant problems by their group care providers, group care is not considered a developmentally appropriate setting for young children. A better understanding of the circumstances under which these very young children enter group care is needed. Policy responses to make alternative forms of family care more available for such young children may be warranted.

• Children in kinship foster care appear to have fewer problems than other children in out-of-home care and to receive fewer specialty services. These findings about kinship care should be added to the rich mixture of evidence from administrative and survey data to contribute to the diversification of child welfare and other human services for kinship caregivers.

11.3 Current Caregiver Characteristics

Although communities, peers, and services have an impact on children, differences in household and caregiver characteristics and caregiver relationships to children are likely to have the most immediate impact on children’s safety, permanency, and well-being. The children in our study are living with caregivers with substantial problems in living; overall more than three quarters of in-home caregivers have at least one of the following conditions: a history of exposure to domestic violence, a substance abuse problem, a serious mental illness, or household resources that place them below 100% of poverty. A remarkable 10% of the in-home caregivers have experienced a psychiatric hospitalization at some time in their life.

Caregivers of children involved with the child welfare system (both in-home and out-of-home) are substantially more likely to have less education and live below the poverty level than caregivers in the general population. Only 20% of all families providing care for CWS-involved children have incomes at or above 200% of poverty compared with 60% of households nationally. Families receiving in-home services are very often desperately poor. Out-of-home caregivers are less poor, but kinship caregivers are almost three times as likely as nonkinship caregivers to have incomes below the poverty level (40% vs. 16%). Caregivers across out-of-home care settings generally report average mental and physical health.

Generally, the race/ethnicity of caregivers and children is shared, as 78% of all children in foster care live with a caregiver with shared racial/ethnic identity. The levels of child and caregiver matches are greatest for Black and for White children. Hispanic caregivers and children and children identified as other race/ethnicity are matched considerably less often.

Among all the children in foster care or kinship care, about one-third of children live with both parents, and for most children, these are the only adults in the home. But for the many children living at home with only one parent (or step-parent), a sizable proportion of the households have at least one other adult also living in the home. The goal of keeping siblings together, which has become a practice standard across the country, appears to be being achieved to a substantial degree.

Differences between in-home and out-of-home caregivers also exist for household characteristics. The children living in nonkinship foster care live in the largest households, by far. Older children residing at home are more likely to live in larger households and households with more children overall, although for children in out-of-home care this trend is reversed—homes in which infants are placed have more children than those in which children age 11 and older are placed.

• Sizeable challenges are faced by many caregivers involved in the care of maltreated children—whether or not they are the children’s biological parents. They attempt the extreme challenge of raising healthy and successful children with few tools and many of their own troubles or impairments. Although involvement with child welfare appears to connect biological parents with specialized and general human services, it appears that foster caregivers may also be in need of these supports, and the biological and kinship caregivers may need increased efforts in this area.

11.3.1 Exposure to Violence

Children entering child welfare services have experienced substantial amounts of severe violence during their lifetimes—especially those children who enter out-of-home care. Children who are residing at home with no ongoing child welfare services have the lowest lifetime exposure to severe violence, although about one-quarter of children remaining at home have experienced severe physical assault. Overall, children in group care have the highest lifetime levels of exposure to violence. Contrary to the plausible argument that older children would have experienced more severe discipline and violence simply because they have had more time to be exposed, older age does not seem to be a critical determinant of lifetime prevalence. One possibility is that the older children may come into care from a different subgroup of children, a group that did not experience rates of assault similar to those of the younger children.

When we look at recent exposure to violence, the picture changes and the children residing at home have substantially greater exposure than the children in placement. Taken together, out-of-home children have more violent pasts but in-home children experience greater violence in their current living environment. Data from the 18-month follow-up will be better able to ascertain the relationship between exposure to violence and the receipt of child welfare services.

• Because children involved with CWS have such high likelihoods of witnessing or experiencing violence, CWS should audit its screening, assessment, and referral procedures to be sure that potentially traumatizing or violence-inducing sequelae of violence are addressed.

11.3.2 Children’s Relationships

Involvement with the child welfare system may, but typically does not, create disruptions in a child’s caregiving and educational circumstances. Children generally report a positive sense of relatedness with caregivers, though children in nonkinship foster care tend to feel less close to their caregivers than children remaining in the home and not receiving child welfare services. Children involved with child welfare services, whether remaining in the home or living in out-of-home care, report similar levels of activities with their caregivers to children in the general population.

Children in different service settings report similar levels of satisfaction with peer relationships and have low levels of loneliness and social dissatisfaction. Boys tend to report lower school engagement as well as more homework completion and school discipline problems—they may be in particular need of early and supportive educational interventions.

Children living in out-of-home care have various experiences and feelings about their situation. Most children desire more contact than they have currently with their biological parents and siblings and report enjoying that contact, albeit not without some ambivalence. Children are about evenly split between those who would like their current placement to become a permanent placement and those wishing to be reunited with their biological parents. Children living in group home care are significantly less positive about their living situation than children in foster or kinship care.

Some demographic differences do exist among children living in out-of-home care. Children from all other racial/ethnic groups are more likely than Whites to run away from a placement. Males are more likely to not feel like a part of the family with whom they are living.

• Although child welfare services research has not emphasized gender differences in response to out-of-home care, there is evidence in these data, and elsewhere (e.g., Jonson-Reid, 2003; Jonson-Reid & Barth, 2000b), that this would be a profitable direction for research and service design.

• Even though children appear satisfied with their out-of-home caregivers, they still report wanting more contact with all members of their biological family. These findings should encourage agencies to find creative ways of helping children maintain contact with their biological parents and to help child welfare workers and foster parents be supportive of this continued connection.

• Our findings suggest that boys may be in need of supportive educational intervention aimed at increasing their engagement in school.

11.3.3 Parental Substance Abuse, Mental Illness, and Other Risks

Estimates of the level of risk factors experienced by parents are higher than those found in the general population, but the extent of this discrepancy depends on the source of information. Using the CIDI-SF, a standardized self-report instrument to report on substance use and dependency, in-home caregivers reported significant levels of alcohol and drug use, with just over 9% reporting consuming four or more drinks in one day, 15% abusing prescription drugs, and 10% using illegal drugs in the previous 12 months. Self-reported alcohol dependence (2.1%) and drug dependence (2.8%) rates were only slightly higher than those reported in the general population and were much lower than rates of substance abuse problems often cited in the child welfare literature.

Even when child welfare workers were reporting on parents who had their children removed, the rates were not as high as previously described and now part of child welfare lore (e.g., U.S. DHHS, 1999). Among caregivers of children remaining at home, about 13% were identified by child welfare workers as needing services for an alcohol or drug problem. Child welfare workers indicated that a total of 10% of the in-home caregivers had an active alcohol or drug problem. (Among all families in the study, the proportion with either problem was 14%). Only 4% of those with open in-home cases were currently receiving alcohol and drug treatment services despite the conclusions by child welfare workers that 20% had substance abuse as a risk factor at the time of the investigation.

Parental mental illness and child maltreatment is less often discussed by child welfare scholars than is parental substance abuse, but this study finds that the issue is very salient. Almost one-fourth (23%) of the in-home caregivers reported experiencing major depression in the past 12 months. The fact that 2.5% of in-home caregivers had been hospitalized in the past year and that 10% of in-home caregivers had used inpatient mental health services, at some time, suggests that adult mental illness is a substantial contributor to the problems in parenting that child welfare services attempts to address (Famularo, Kinscherff, & Fenton, 1992). Relatively few (8%) caregivers reported currently receiving mental health services, while 12% had received these services in the past year. Although we do not have direct measures of the mental health status of caregivers whose children went into out-of-home care, we would expect from child welfare worker reports and from emerging research (e.g., Bellis et. al., 2001; Kotch, 1999) that rates of caregiver mental illness are much higher among parents of children who were placed.

Self-reported rates of domestic violence reported by these in-home caregivers were considerably higher (45%) than the national estimates of victimization in the general public (Tjaden & Thoennes, 2000). These caregiver-reported rates exceed what child welfare workers report as part of the risk assessment; child welfare workers report that almost one-third of caregivers had a history of domestic violence (a substantial underestimation when contrasted with the 45% self-reported rate), and that over one-tenth of caregivers were experiencing domestic violence at the time of the investigation (which again appears to be an underestimation, given the self-reported 18% having experienced severe violence in the last year). Based on child welfare worker reports, rates of domestic violence against the caregiver are significantly higher among caregivers of children living out of the home than among caregivers of children living at home. Child welfare workers appear to take the caregivers’ history of domestic violence victimization or abuse or neglect victimization seriously in case planning, as these rates were much higher for caregivers of children receiving in-home services than caregivers of children with closed cases.

Victimization by a caregiver in childhood has been associated with involvement with child welfare services as a parent (Straus, Gelles, & Smith, 1990). Although most alleged perpetrators did not report childhood victimization, those who did had a far higher likelihood of receiving ongoing CWS services than having their cases closed without services.

The role of parental arrest in placements of children into foster care has been largely overlooked in the analysis of child welfare services dynamics. The finding of a strong association between recent parental arrest and the level of child welfare intervention calls for a more penetrating examination of the relationship between child welfare, police, and correctional services, when an arrest occurs. Given the substantial overrepresentation of Black families among those involved with recent arrests, in our data, the nexus between child welfare, police, and corrections seems like a particularly promising area for the development of services that might help reduce the disproportionate placement of Black children in out-of-home care.

The poverty rates for the CWS-involved households are exceptionally high. More than half of all households had an income below the federal poverty threshold, and more than one-in-five had an income at 50% of the poverty level. Over and above 80% of study families (whose children remained at home) had incomes below 200% of the poverty level. Receipt of TANF services among in-home caregivers was high, although there appear to have been many families with very low incomes not receiving TANF. Although 61% of caregivers reported ever having received these services in their lifetime, only 21% of caregivers whose children were living at home were currently receiving TANF. The findings may partially confirm that child welfare services are primarily a response to destitute poverty, but the study design does not allow us to understand why so many families in similar economic circumstances do not become involved with child welfare services.

• Substance abuse is a major contributor to child welfare involvement, as it has historically been; still it is important to recognize that this is not as pervasive a factor in the referral of families to child welfare services as sometimes estimated. Maternal mental illness may, instead, be underestimated as a contributing cause of child maltreatment. The past decade has witnessed substantial efforts at strengthening links between child welfare and substance abuse service providers and develop in new service models. The same level of effort could profitably be made to link child welfare and adult mental health services.

• The current concern about developing policies to address the overlap between domestic violence and child maltreatment finds buttressing in these data—the overlap is substantial. Family involvement with domestic violence may need to become an additional element that is routinely reported in administrative data about child maltreatment reports and their disposition. When cases involve exposure to domestic violence or endangerment from living in a home with domestic violence, they appear to be coded under neglect, failure to supervise, or emotional maltreatment. A clear alternative should be provided.

• Poverty, alone, is not often identified as a key concern in child welfare decision-making, but this belies the finding that the biological parents in this study are exceptionally poor. Poor housing, a signal of underlying poverty, is identified as an important contributor to case planning. Impaired parenting and parental substance abuse are the key factors cited for service decisions for children across the age span.

• The discrepant child welfare worker and client reports of these important risks to safe parenting, indicate the possibility that child welfare workers lack the time or skill needed for accurately assessing active substance abuse, mental illness and domestic violence. Although there are plausible methodological reasons for some of these discrepancies, there is also substantial reason to believe that the current approach to intake does not generate an acceptable amount of information. If this information is critical to decisions about which service path a family should follow, then serious errors could occur. Further research into this question is critical since these areas directly impact a caregiver’s ability to parent adequately.

• Although decisions about which child and family receive which level of child welfare services is complicated, in general, families with children who are removed from the home have higher levels of problems than families with children remaining at home. A significant problem is parental arrest—caregivers with recent arrests are especially likely to have their children placed into foster care. Child welfare and justice agencies should work in concert to minimize unnecessary placements that may follow parental arrests.

11.4 In-Home Caregivers’ Relationships with Child Welfare Workers

One of the most compelling findings on child welfare services is the sizable proportion of families that we believed to be receiving in-home services but who had had no contact with their child welfare worker even though their case had been open for several months. Although there are a variety of legitimate reasons for this to occur, there are still a sizable proportion of cases that should have had prompt contact with child welfare services that they did not receive. We suspect that child welfare agencies are simply not able to meet all of their obligations to families and children.

When caregivers do have contact with families, they report that they generally perceive their relationships with their child welfare workers to be of moderately high quality. This likelihood was increased when there was consistent and frequent contact. Caregiver race/ethnicity, caregiver age, and racial match between child welfare worker and caregiver do not seem to be associated with caregivers’ perceptions of satisfaction. Overall, caregivers appear far more satisfied with the relationship they have with their child welfare workers than they are with the adequacy of services that have been provided to them.

• In-home child welfare agency services appear to be quite small interventions endeavoring to address some very large shortcomings in family and child functioning. Many families report sporadic or nonexistent contact with child welfare workers and services that they find unsatisfactory, although this result could be predicted from child welfare services’ lack of success in generating evidence-based interventions to help families living at home (e.g., Littel & Schuerman, 2002). The need is great to improve the package of services available to the in-home service cases, as they represent the vast majority of families that come into contact with CWS.

11.5 Related Children’s Services

Child welfare services are intended to be a mechanism to address safety and permanency issues for children and to provide an opportunity to promote children’s well-being. Caregivers’ reports of their child’s overall health status indicate that children were in good health and were receiving preventive care such as immunizations and dental and vision examinations. About one-quarter of caregivers reported that their child had a chronic health problem, but only one in six of these children were also identified as being in poor or fair health.

Placement into out-of-home care appears to be associated with better child safety. Caregiver reports of illnesses, injuries, and accidents showed that children who remained in their home of origin had been to the emergency room since the close of the investigation (37%),which significantly exceeded the rates for children in out-of-home care (in our study). 35

Despite exceptionally high rates of behavior problems and previous experience with inpatient mental health services, outpatient mental health services were currently being used by a relatively small proportion (11%) of children involved with the child welfare system, with only 7% receiving care from a mental health specialist. Children in group care settings were far more likely to have previously utilized residential care (60%) than those in foster (11%) or kinship (4%) care. This appears not to be strictly age-related, although older children are more likely to have experienced residential care and psychiatric hospitalization.

Caregiver report of special education services showed that 17% of children had been tested for learning problems since the investigation date. This represents a substantial proportion of all the children who came to CWS (since about 15% were already receiving early intervention or special education services). This action would seem to be consistent with the desires of caregivers, as about one out of every five caregivers of children who had not been tested for special education reported that their child needed this service.

• The high number of children receiving care in emergency rooms may indicate higher levels of injury or an unmet need for access to primary health care providers. Future research is needed to determine the source of this issue.

• Caregivers appear to be asking for educational intervention for the children in their care. This argues for more coordination with school systems in order to properly educate these youngsters.

11.5.1 Child Welfare Worker Characteristics

No national survey of child welfare workers has ever been undertaken, and NSCAW is not such a survey. The child welfare workers in this study were not randomly selected from all child welfare workers; still, they are a reasonable approximation of a random sample of child welfare workers who become involved with children and families at intake into the child welfare system.

Child welfare workers are a diverse group in terms of age, race/ethnicity, and education. The average length of experience for the child welfare workers is about 7 years. Nearly three-quarters of White children have a child welfare worker of the same race/ethnicity; only one-quarter to one-third of Hispanic and Black children have child welfare workers of their same race/ethnicity—this reflects the overall predominance of White child welfare workers.

With regard to education, only 12% of child welfare workers had a master’s of social work (MSW), with an additional 14% holding another master’s degree or a doctorate (which could be in social work and in addition to the MSW). Most (97%) had at least a college degree. However, the relative disconnect between a worker’s educational preparation and the proportion of workers holding degrees that might prepare them for child welfare work is significant. Our findings would support policy efforts aimed at increasing the numbers of workers with relevant professional preparation prior to practice.

• Child welfare workers have considerably more experience than they have training in the provision of child welfare services. Many MSW programs offer curriculum and internships expressly designed to provide child welfare workers with advanced child welfare practice and policy knowledge. These appear to be much needed, given the lack of specialized education and training of child welfare workers.

11.5.2 Developmental Themes

These analyses cannot answer many gripping questions about the short- and long-term impact of child welfare services involvement. Although the levels of child, family, and community risk factors are high—and these cumulative risks do not augur well (Deater-Deckard et al., 1998; Herrenkohl et al., 2000)—there are other countervailing forces that may help these young people to succeed. Many mediators may buffer the likelihood that children with very high levels of cumulative risk will show antisocial behavior. Indeed, some evidence argues that the children with the greatest risk are those most likely to benefit from such buffering effects (Pollard, Hawkins, and Arthur, 1999). These data are not yet longitudinal and cannot tell us whether involvement with child welfare services will buffer the impact of the great levels of risk seen at baseline. This will have to wait until the next wave of data is analyzed.

Interpreting the developmental data about our incoming children is difficult and is made more so by the finding that, across the age groups, a relatively stable proportion of children have school problems but the behavior problems of children entering CWS worsen with the increasing ages of children. There appears to be a different selection process for older children entering CWS. Children’s behavior problems may be more readily recognized among older children and considered a central reason for child welfare involvement (despite the silence of child welfare policy in allowing such a reason for child welfare involvement).

Considerable research is showing the significance of early intervention for high-risk children, and the children in this study certainly qualify, in general, as the highest risk in our society. More than one-third of the children in this age group have two or more clinical scores, but only 10% of children of this age are receiving specialty education or mental health services.

More generally, we witness a steady increase in the use of mental health or special education services across all the age groups, from 3% of the infants to 35% of the 11- to 15-year-olds. The vast majority of children in each age group, however, are not receiving either form of specialty service. This lack of specialty service provision leaves child welfare services to carry the primary burden of meeting the needs of maltreated children—a responsibility that should be more broadly shared if their needs will be met (Simms, Dubowitz, & Szilagyi, 2000).

The children and families becoming involved with child welfare services every day are extraordinarily varied in their backgrounds and experiences. Few service systems have the mandate to provide care for children without regard for age, developmental standing, health status, and type of injurious event. Our findings support those of the GAO (2003) in showing that many of the youth who are involved with child welfare services have also had psychiatric placements. Although we could not unequivocally determine whether or not they became involved with CWS primarily for reasons of addressing their mental health disorders, we have conducted additional analyses that suggest that this is not infrequently the case (Barth, Wildfire, & Green, 2003). Further research into this phenomenon would make an important contribution. In addition, NSCAW offers sufficient evidence to argue strongly for the generation of improved and additional mechanisms for serving maltreated children involved with child welfare services.

• Greater recognition is needed of the extent of developmental problems in the children who are maltreated—even among those who are remaining at home with no services, a group little studied, before.

• A large group of children exists who are underserved by specialty mental health and special education. Very young children, children with closed cases at home, and children in kinship care are most prone to being underserved.

• More information is needed on the reasons for children’s placements and how CWS respond to those reasons. Innovations in parent training and other forms of work with parents need development to provide a better correspondence to the developmental and behavioral needs of children.

11.6 Conclusions

This report provides unprecedented amounts of information about the safety and well-being of children entering the child welfare system. Overall, the findings show that the children who are placed into out-of-home care have significantly more family risks, greater exposure to violence, and more serious levels of maltreatment than children who receive services at home. These findings go a long way to vanquish the arguments of those who would argue that children are placed into child welfare services for reasons of poverty alone (Pelton, 1989) or following a decision-making process that is largely random (e.g., Lindsey, 1992; 1994) or that is fundamentally determined by the race of the child (Roberts, 2002). Although our findings cannot show that individual case characteristics are always weighed the same in each decision or that there are no errors in making the best decisions for individual children, we had few analyses that indicated that the races of the children were determinant in child welfare decision making, when other background factors were controlled. Ages of children are, on the other hand, consistently related to service and placement decisions, with the youngest and the oldest children having the highest rates of placement.

This finding of general consistency between family risks and child welfare decisions may support the argument that child welfare workers share a common scale with varying thresholds for making placement decisions (Schuerman, Rossi & Budde, 1999; p. 616). These threshold differences may be partly ascribed to urban/nonurban differences, as nonurban PSUs have a lower proportion of children entering placements, but those who do have significantly higher numbers of risks and higher CBCL scores. Although there is unarguably a need for better training and more service options, in order to better match children with child welfare services, this report provides reassurance of general attentiveness to child welfare risk in making placement decisions.

The findings also support the argument that the child welfare system must maintain some inconsistency if it is to make the right decisions under somewhat unique circumstances (Schuerman, Rossi, & Budde, 1999). Our data show that apparently anomalous decisions may have merit—for example, those cases that involve the placements of children who have recently had low-severity maltreatment or who have not had a significant history of child maltreatment because of other mental health needs of the child and caregivers. The complexity of child welfare decision making is immense because child welfare service providers are making decisions during very distinct developmental states for children (Berrick et al., 1998), at points in a family’s life course with different trajectories (Elder, 1998), with widely varying indicators of risk and protective factors (Thomlison, in press), and with access to a significantly varying degree of resources (Mitchell et al., in press). This is further complicated because child welfare services play such a central role in providing resources to children with mental health and juvenile justice needs (U.S. GAO, 2003). Our data suggest that the underlying rationale for decision making differs substantially by age group and that improvements in precision of decision making will require better articulation of age group–based differences in reasons for placement.

Given these widely varying circumstances under which children enter child welfare services, the challenge of isolating the unique contributions of services to the current well-being of children is great. At this point in the children’s experiences with CWS services, this simply cannot be done. Given the high rates of prior exposure to child welfare services among the involved families, efforts to determine how much this spell of services contributed to child well-being will always be vexing. Even understanding the contribution of maltreatment to the poor developmental outcomes that we witness is substantial because we have so little comparable information about families that are similarly situated but not alleged to be maltreated. We do need a precise measure of this impact to understand the great risk that these children are experiencing. Still, the sample size in this study and the longitudinal design will eventually allow us to bring substantial power to understanding what happens to the development of maltreated children exposed to different services and family settings.

Our data provide far less ambiguous findings about services. There is no doubt that most of the children and families who come to the attention of child welfare agencies receive very little direct service from the agency. The typical child who is investigated for maltreatment will not receive any ongoing child welfare services and is not currently receiving any specialty mental health or special education services. Few cases are opened for ongoing services, and those that are opened at home very often have had only a modicum of services at the point of their assessment for this study—86% had their investigation open for at least 2 months at the time that they reported on their services receipt.

The extraordinary level of prior child welfare involvement among the families and children in this study is also an arresting finding. Although we cannot tell at this stage in the study how many children who received child welfare services will not come back into the child welfare system, we can say that many children who have previously had child welfare services and an open child welfare case are again involved with child welfare. If they previously had a child welfare case opening or placement, they are more likely to receive a similarly high level of care as a result of this event. More broadly, there is no doubt that child welfare decision making relies on previous reports and assessments for making current assessments, as relatively few children who receive higher levels of service do so without previously having had some contact with child welfare services. This fact should not be lost on those states and municipalities that are reducing record keeping about child welfare services involvement related to the implementation of multiple response systems or privacy initiatives.

Across the many domains of child and caregiver characteristics analyzed in this report, a common, if not absolutely consistent, finding emerges that the children and families who receive services at home are more likely, at baseline, to have a greater level of parent and family risk and child problems than those who remain at home without services and a lesser level of problems than children who are placed into out-of-home care. The children placed into kinship care fit this pattern at times, but on other dimensions look more like the children who remain at home with open cases. Overall, the amount of in-home services provided directly by child welfare is negligible although these services do seem to be supported by the added value of the addition of allied services provided to the parents. Additional services that come to children, associated with having an open child welfare case, appear minimal. However, these services are still evolving.

No other service program in our society, with the possible exception of the public education system, is required to do so much on behalf of children. The mandates of child welfare are far greater than those of developmental disabilities, early intervention, special education, juvenile justice, or mental health. Child welfare is, further, virtually the only other system of care that must, as part of its central mission, address the needs of the child’s family as well as the child—other systems of care may have voluntarily adopted such a family focus but are not mandated to account for and try to address the great variety of family needs that influence children’s safety, well-being, and permanence.

The diversity of child and family characteristics coming to the attention of child welfare services, the urgency with which decisions must be made, and the high stakes for children and parents who become involved with child welfare services make planning extremely difficult. As a result, these systems tend to respond in a highly routinized manner that focuses on identifying and addressing minimal standards of parental care and child safety. Calls for diversity of approaches to address the diversity of family needs have become stronger in recent years. These calls envision a system in which multiple assessors could divert potential child welfare cases into other systems, theoretically leading to greater specificity in the services. Yet the breadth of family and child problems, strengths, and circumstances that child welfare cases contain is almost astonishing. The evidence in this study unmistakably supports the development of greater diversification of child welfare services, yet the high level of developmental need shown in this study argues that diversification must not come at the expense of further growth of the proportion of maltreated children who do not receive services that they apparently need. The next generation of services should address underutilization as a primary objective.

---

 

Table of Contents | Previous | Next