Table of Contents | Previous | Next |
3. Implications
This report provides a relatively coarse picture of the characteristics of child welfare services agencies. Because of the small sample sizes, the percentages and rates of occurrence presented here must be understood to be rough estimates (how rough the estimates are is shown in the report through the confidence intervals and standard errors). Still, only the most definitive findings are included in this section.
3.1 Child Welfare Services Agency Structure
The majority of child welfare agencies, about two-thirds, are units within larger agencies rather than freestanding units. Child welfare agencies are highly collaborative, having organizational linkages to TANF services in almost all counties and to substance abuse treatment, mental health, and juvenile justice services in about 40% of the counties. Local agency directors reported, about two-thirds of the time, that they had substantial control over how child welfare dollars were spent in their agency.
3.2 Service Delivery Mechanisms
About 40% of agencies had developed new initiatives in the past 12 months, including specialized units of service, multidisciplinary teams, and additional community-based branch offices. Most agencies reported having completed the development of their concurrent planning mechanisms. Less common were innovations like having one caseworker assigned to cases from beginning to end and the use of voluntary foster care placements. Structured risk assessment was identified as being in place in nearly every agency, but its application varied widely. Most often, agencies indicated using structured decision making or risk assessment only at the outset of the case.
3.3 Staffing and Training
Agencies almost universally require a college education for their child welfare services investigators. About 10% of the agencies had no degree requirement for workers who were not CWS investigators. Training of child welfare personnel also occurs before or during the job. The vast majority of agencies require new workers to have four or more days of pre-service training, but at least one quarter (and possibly as high as three-quarters) of the agencies require more than two workweeks of pre-service training. Annual in-service training requirements were typically less than one day (51%) or none at all (20%).
Caregivers are also expected to receive training, though this does not universally apply to relatives providing foster care. The training amount was typically less than one day (between 40% and 70% of agencies so indicated). In as many as one-third of agencies, training was routinely less than four days. Caregivers are almost never required to obtain one full day or more of annual in-service training.
Child welfare worker turnover was lower than expected in the reports from agencies. The average turnover rate appeared to be just 13%, but because only 38 agencies did the complicated math required to compute turnover rate, the confidence interval allows the outside possibility that the turnover rate was really as high as 86%. This does, at least, exclude the likelihood that turnover is generally 100%, even if that is true in a few agencies.
The agencies did not report large growth during the year prior to the interviews, which is consistent with other survey data indicating modest growth from 1996 to 1998 (Bess, Leos-Urbel & Green, 2001). About a quarter of authorized positions were CPS workers, and a third of authorized positions were direct service workers. Supervisors represented about one in eight positions. Most authorized positions were filled, and the vacancy rate had not increased sharply in the prior year. When new positions were added, the greatest area of increase was in noninvestigative direct service workers. Smaller and more rural counties had a much higher proportion of CPS workers to direct services workers, as will be discussed in more detail below.
An element of child welfare services staffing has to do with the use of staff from other agencies to provide child welfare services. Family preservation/in-home services were the most commonly referred subcontracted service, with between one-third and three-quarters of agencies subcontracting them. Residential treatment was also commonly provided by subcontractors. Family reunification services and conventional foster care were far less likely. Investigations services were almost never contracted out.
3.4 Service Dynamics and Special Initiatives
On average, agencies received about five times as many reports of children in poverty as of those not in poverty. About two-thirds of these reports were referred for investigation. About 1 child per every 100 in the county received family preservation or family support services, although this number was considerably higher among poor families.
Study researchers estimated that foster care expenditures accounted for almost half (45%) of all child welfare expenditures, with an average out-of-home placement cost of $7,283 (not including the child welfare worker or administration time). Because there are fairly definitive data from other sources indicating that kinship foster care and treatment foster care are a growing source of placements (e.g., Berrick, Needell, Barth, & Jonson-Reid, 1998), this growth was not assessed. Kinship placement and specialized foster care appear to be more common in large urban areas. Voluntary placement of children was rarely done—agencies averaged only about four voluntary custody placements a year; less than 1% of children investigated for child abuse and neglect later received a voluntary placement. The overall out-of-home placement rate was about 1 child per 100 children in the county/PSU and 3 children per poor family in the county/PSU.
3.5 Response to Child Welfare Policy Changes
In the year prior to the survey, agencies reported adopting about one-quarter to one-third of the children with a goal of adoption and about three times that many of the children who had already been freed for adoption. State-administered programs seem to have much greater success in accomplishing adoptions, as the discussion below will attest.
Between one in seven and one in four agencies had concerns about the over- or under-representation of African American children in foster care. A slightly larger group initiated training efforts. This was strongly associated with the type of county.
The agencies reported spending about $6,000,000 on average (from all sources) during the most recent fiscal year. This represents an average per-child expenditure rate of about $8,000 (with a standard error of about $2,000). The largest proportion of funds went to placement services (about half) with about one-quarter going to CPS investigations and about one-sixth going to in-home services. Family support services, adoption, and independent living accounted for relatively small amounts of the total.
The largest recent increase in state funding was for child abuse prevention services, followed by adoption services and CPS. In-home services and placement received somewhat lesser increases. About one-quarter of agencies reported greater flexibility in the use of funds.
The effects of TANF, the landmark reform of the nation’s core public assistance effort, on the child welfare caseloads or service approaches were relatively modest. The creation of multiprogram teams that included child welfare and TANF staff was the most common change cited. There was no indication at this time of widespread changes in caseload that were identified by respondents as related to the impact of TANF.
The implementation of ASFA resulted in more significant changes. For about 60% of agencies, ASFA brought about a greater emphasis on ensuring safety and, for almost all, shortened the time frames for decision making. For between 53% and 88% of agencies, ASFA increased the emphasis on adoption for children living in kinship foster care. An estimated 28% of agencies (with an upper bound of 53%) indicated that they would increase the number of families who would not get reunification services. There was uniform agreement that agency regulations and paperwork had increased and general agreement that the number of hours spent on a case had increased with no corresponding decrease in the number of cases. This may be because workers are concentrating more effort into a shorter span of time, although the data are not definitive about this.
Effects of MEPA were far less evident in these data. Although an estimated 29% had increased training, 77% identified no increase in the proportion of transracial foster care or adoption placements. Only 8% of agencies, generally those in large urban areas, saw the creation of new recruitment resources.
Although both federal and local pressures to assess child welfare performance have increased, only a few agencies indicated initiating performance measures in their agencies, and those that did were more often rural agencies. The use of performance measures seemed to be consistent across the child welfare services functions and programs (e.g., investigations to independent living).
3.6 Configuration of Child Welfare Services According to Service Context
Another way to view the information is according to the contrasts made between the characteristics of child welfare agencies in large vs. other counties, poorer vs. nonpoor counties, urban vs. nonurban counties, and state- vs. county-administered child welfare programs.
3.6.1 Large vs. Other CountiesLarge counties appeared to differ substantially from small counties regarding the delivery of child welfare services, employing a significantly higher proportion of direct service workers, compared to CPS workers, than did other counties. Subcontracting for some of those direct services—especially family reunification service—was also more common in large counties.
Changes in the recent reconfiguration of child welfare services were associated with county size, with large counties being more likely than others to have developed specialized units, blended teams, or community-based offices. The use of fictive kin as a placement resource was more likely in large and urban counties than in other counties.
3.6.2 Poor vs. Nonpoor Counties
Nonpoor counties had a significantly greater—about four times higher—average per-child child welfare expenditure ($10,739) than did poor counties ($2,689). They also had higher expenditures of CPS dollars relative to the total number of children investigated. It follows, then, that in poor counties, having a smaller portion of their resources allocated to services, a higher rate of reports were investigated than in nonpoor counties, and there was a lower proportion of families receiving family preservation services.
Yet, poor counties also provided more training for their new child welfare workers: the great majority of agencies serving poorer counties required two weeks or more of pre-service training, whereas fewer than half of agencies serving nonpoor counties required this much training. (This could be partially attributable to the greater federal participation in training, which makes it more affordable than service provision for poor counties.)
The proportion of investigated reports relative to the number of poor families with children in the county differed by county poverty levels: nonpoor counties investigated a lower proportion of reports relative to the number of poor families in the county than did poor counties. Nonpoor counties had an overall higher rate of reports relative to the number of poor families in the county. Thus, in nonpoor counties, poor children are more likely to be reported but not more likely to be investigated. This may affect service provision to poor families in nonpoor counties who need services that might follow an investigation. The findings also indicate that poorer counties had lower adoption rates than did nonpoor counties.
3.6.3 Urban vs. Nonurban Counties
In general, urban counties are reorganizing their services more rapidly than nonurban counties. Urban counties were more likely than rural counties to have begun offering neighborhood services or opened satellite offices than county-administered states. Urban counties were also more likely than rural counties to have created multidisciplinary teams with the TANF program to better serve their mutual clientele. Yet, urban counties were less likely to have completed implementation of concurrent planning than other counties.
Provision of a foster care payment plus normal licensing requirements for kinship foster care was more common in urban counties. The rate of voluntary placements was also associated with urbanicity: urban counties reported a higher average rate than rural counties; overall, however, voluntary placements were rarely used and accounted for less than 1% of all foster care placements.
Concerns about over- or under-representation of minority children (African American) were associated with county size and urbanicity: large counties were significantly more likely than other counties and urban counties were more likely than rural ones to have such concerns. Yet, urban counties had higher adoption rates than other counties, partially because they were much more likely to have developed specialized recruitment resources. Further, rural counties were more likely than urban counties to have no changes in agency services as a result of MEPA and IEP. ASFA may help to even out these differences, as there is strong evidence that rural counties were more likely than urban counties to have increased their emphasis on adoption—especially of children in kinship foster care.
Subcontracting for services was much more common in urban areas. This included a greater likelihood of contracting for family reunification services, private foster care services, residential treatment services, and adoptive recruitment and placement services. Urban counties also had a lower proportion of authorized CPS positions than did nonurban counties.
3.6.4 State- vs. County-Administered Agencies
State-administered child welfare systems appear to have a more structured approach to risk assessment, licensing of kinship homes, and training of child welfare workers and caregivers. State-administered, rather than county-administered, agencies appeared more likely to require the use of a structured risk assessment approach when deciding whether a case was substantiated and whether to reunify a child once placed.
State-administered agencies also were far more likely to require licensing for all foster care placements and to provide a foster care payment to relatives. On the other hand, proportionately more foster care homes from county-administered agencies received specialized (higher) payments than did foster care homes from state-administered agencies. There was also evidence that county-administered agencies provided more training and supervision for their child welfare workers.
State-administered agencies were more likely than county-administered ones to have concerns about representation of minority children. Accordingly, participation in special training initiatives to address over- or under- representation of minority children was more likely in state-administered agencies. These agencies also had a higher rate of adoption than county-administered states. This finding is consonant with the evidence that state-administered agencies were more likely than county-administered agencies to have an increased number of families who would not get reunification services after ASFA. Another contributing reason could be the significantly greater likelihood of increased adoption resources in state-administered agencies. Lastly, there is also evidence that state-administered agencies were more likely to have changed their adoption activities following the passage of the MEPA and IEP.
Child welfare agencies within state-administered systems are apparently more likely than agencies within county-administered systems to be part of a larger agency. State-administered agencies also appear to innovate in different ways from county organizations, as state-administered agencies were more likely to have developed specialized service units than were county-administered agencies. State-administered agencies were also more likely to have begun offering neighborhood services or opened satellite offices than were county-administered agencies. The type of agency administration was related both to the creation of multiprogram teams with both TANF and CPS caseworkers: state-administered agencies were more likely than county-administered agencies to create such teams. Likewise, state-administered agencies were more likely than county-administered ones to routinely make referrals from TANF to child welfare services when clients were sanctioned.
State-administered agencies were more likely than county-administered ones to subcontract recruitment services for foster homes and adoptive homes but were otherwise no more or less likely to subcontract with private agencies for services. County-administered agencies had a significantly higher ratio of CPS dollars spent relative to the total number of children investigated than did state-administered agencies.
| Table of Contents | Previous | Next |