Children's Bureau Website - Indiana - Flexible Funding

Indiana

Demonstration Type:	Flexible Funding - Phase II¹
Approval Date:	June 30, 2005
Implementation Date:	July 1, 2005
Expected Completion Date:	June 30, 2010
Interim Evaluation Report Received:	January 14, 2008
Final Evaluation Report Expected:	January 1, 2011

Background

Indiana's original flexible funding waiver demonstration was completed in January 2003 and continued under several short-term extensions through June 30, 2005. For its five-year (Phase II) waiver extension, the State is continuing its demonstration of the flexible use of title IV-E funds and seeks to improve on the process and outcome findings reported for its original waiver demonstration. In particular, the State hopes to promote the utilization of waiver dollars by a greater number of counties in light of the finding from its original demonstration that only 25 of 90 participating counties made significant use of waiver funds.

Target Population

The target population for the Phase II demonstration includes title IV-E and non-IV-E-eligible children at risk of or currently in out-of-home placement, as well as their parents or caregivers. In 2006, the State modified its criteria for referring cases to the waiver demonstration; the new referral protocol more narrowly defined cases eligible for the demonstration. Specifically, "service cases" (i.e., families with a substantiated maltreatment report but no previous CPS history and no recommendation for CPS involvement) were phased out beginning September 1, 2006 and were not eligible for waiver assignment after April 2007.

Jurisdiction

All 92 counties in Indiana are eligible to participate in the Phase II waiver demonstration.

Intervention

Under its waiver extension, Indiana counties continue to develop and implement innovative child welfare services, including community-based wraparound services and home-based alternatives to out-of-home placement. As in the original demonstration, each participating county receives a certain number of waiver "slots" in which eligible children may be placed. A capitated payment of $9,000 is allocated to each slot, which is used by a county to provide targeted community and home-based services appropriate for the needs and circumstances of the child and his or her family. The State allocates slots to participating counties based on selected demographic variables, including population size and poverty rates. Statewide, no more than 4,000 waiver slots are available at any given time.

A new feature of Indiana's Phase II demonstration involves the institution of a statewide system of waiver "champions" to serve as experts and consultants on the IV-E waiver demonstration. The champions include family case managers, bookkeepers, child welfare supervisors, and county directors.

Evaluation Design

The evaluation includes process, outcome, and cost-effectiveness components. Using a matched case comparison group design, the evaluation tests the hypothesis that the flexible use of title IV-E funds for wraparound services and home-based placement alternatives will (1) prevent out-of-home placements, particularly in restrictive institutional settings; (2) reduce lengths of stay in out-of-home care; (3) decrease the incidence and recurrence of child maltreatment; and (4) enhance child and family well-being.

To implement the matched case comparison design, the State's evaluation contractor uses a computer algorithm that selects the best possible match for each experimental group child from the pool of children who have not been assigned to the waiver. This method ensures that the IV-E status of the experimental group child matches that of the comparison child at the time of the match and that the case type of the experimental group child (e.g., delinquency, CHINS) matches that of the comparison child. Other matching variables include (1) county of the case, (2) opening date of the case, (3) age of the child, (4) sex of the child, (5) removal and placement status, (6) number of previous removals and placements, (7) number of days in previous placement, (8) type of substantiated child abuse or neglect, and (9) maltreatment risk level.

Data Collection

For evaluation purposes, the State's evaluation contractor maintains a database that consists of monthly file extracts from the Indiana Child Welfare Information System (ICWIS). These extracts are cumulative from 1997 to the present and include data on all children ever assigned to the waiver experimental group; all other children currently in or who have been in out-of-home placement and their siblings; all children assigned to court custody but not removed from their homes; and all other children with an open child welfare case. Newly assigned experimental group children and matching comparison group children are added to the database over time.

Process Evaluation

The State's process evaluation describes how the demonstration was implemented in each participating county and identifies differences in the services received by experimental and matched comparison cases. Specific research questions addressed through the process evaluation include the following:

To what extent did counties utilize their allocated waiver slots? What factors accounted for differences among counties in their utilization of waiver slots?
What differences emerge among counties in terms of the populations targeted for waiver slots (e.g., CHINS cases, delinquency cases)?
What are the demographic characteristics (e.g., sex, race/ethnicity, poverty rates) of children assigned to waiver slots?
What differences emerge among counties in terms of the types of services provided using allocated waiver funds? To what extent did children and families receive services and supports that they would not have received in the absence of the waiver?
What are the attitudes of child welfare caseworkers, child welfare agency administrators, juvenile court judges, probation officers, and community stakeholders (e.g., school administrators, local child and family service organizations) toward the waiver?
To what extent were community stakeholders (e.g., juvenile courts, probation officers, school personnel, county officials) involved in planning and implementing the waiver?

Outcome Evaluation

The State's outcome evaluation compares the experimental and matched comparison groups for significant differences in the following outcome measures:

The proportion of children with an alleged or substantiated disposition of child abuse or neglect who enter out-of-home placement;
The proportion of children who enter placement in restrictive institutional settings, including out-of-state facilities;
Of all children who enter out-of-home placement, the proportion exiting to reunification, adoption, or guardianship;
The time from foster care entry to foster care exit for each permanency outcome;
Placement duration in restrictive institutional settings, including out-of-state facilities;
Of all children who exit to each permanency outcome, the proportion experiencing a subsequent substantiated report of abuse or neglect;
Of all children who exit to reunification, the proportion who re-enter out-of-home care; and
Changes in child and family well-being and functioning.

Evaluation Findings

Process Findings

The State's Interim Evaluation Report covering the period of July 1, 2005 through December 31, 2007 noted that waiver usage has steadily increased over time as indicated by the number of active waiver cases. For the first 6-month period of the waiver extension, the average monthly number of active waiver cases was 846; however, that number increased to 1,828 by September 2007. In examining waiver usage, the evaluation team distinguished 36 counties as "program counties" (i.e., those counties that adhere most closely to the original vision for the waiver demonstration and actively use their waiver slots). This represents an increase of 11 counties (40 percent) over the original demonstration, in which only 25 counties were identified as active waiver users. These counties represent 55.4 percent of the State's general population and 69.5 percent of children assigned to the waiver.

Between July 1, 2005 and September 30, 2007, 4,236 children were newly assigned to the experimental group. Of these, 1,918 children are title IV-E eligible while 2,318 are non-IV-E eligible. In addition, 824 children were carried over from the bridge extension period, for a total of 5,060 children assigned to the experimental group during the long-term extension as of January 2008.

With regard to service provision, children in the experimental group who were not placed in out-of-home care were significantly more likely to receive services to prevent removal than children in the control group (88.7 percent vs. 73.6 percent, p<.05). Of the 34 types of services noted through case-specific sample surveys and family surveys, workers reported a larger percentage of experimental cases receiving 23 of the 34 services. Of these services, 8 were provided to experimental group families at significantly higher levels than to control group families (p<.05), including services related to household needs (e.g., utility and rent payments), homemaker services, basic child needs (e.g., clothing and school supplies), transportation, housing assistance, money management, life-skills training, and childcare.

Outcome Evaluation

Major outcome findings contained in the State's Interim Evaluation Report are summarized below:

Children with access to waiver-funded services are less likely to enter out-of-home placement. Of the 1,956 experimental group children and 1,565 control group children not in placement at the time of their assignment to the project, 412 (21.1 percent) children in the experimental group were subsequently removed and placed in out-of-home care compared with 468 (29.9 percent) control group children, a statistically significant difference. Placement rates in "program" counties were similar to the statewide analysis, with 21.6 percent of children in the experimental group placed in out-of-home care compared with 30.5 percent in the control group, also a statistically significant difference.
Overall, experimental group children were reunified at higher rates than matched comparison children. Specifically, the State's analysis found that 57 percent of experimental group children had been reunified compared with 44.1 percent of control children, a statistically significant difference. In contrast, control group children were much more likely to be adopted, with 22.1 percent exiting to adoption compared with 7.7 percent of experimental group children, a statistically significant difference. Guardianships occurred more frequently in the experimental group than in the control group (12.8 percent versus 10.3 percent), although this difference was not significant.
On average, children with access to waiver-funded services spend less time in out-of-home placement. Among children reunited, adopted, or placed with a guardian, those in the experimental group averaged 346 days in placement compared with 508 days for children in the control group, a statistically significant difference. When these data were analyzed by placement outcome, reunified children were responsible for much of this difference. Although adopted children in the experimental group spent less time in placement than control group children (885.6 days versus 959.5 days), this difference was not statistically significant.
Children with access to waiver-funded services avoid maltreatment recurrence for longer periods. Survival analyses indicate a greater delay in a new substantiated report after original case closure for children in the experimental group (371 days) compared to children in the control group (254 days), a statistically significant difference.
Children with access to waiver-funded services may avoid foster care re-entry at higher rates and for longer periods. According to the State's Interim Evaluation Report, 13.9 percent of children in the experimental group who were previously reunified reentered placement compared with 18.4 percent of children in the control group, a difference that fell just short of statistical significance at p = .054. However, survival analyses reveal that experimental group children avoided placement reentry longer than matched control children, with an average of 136 days before reentry compared with 147 days for control group children, a statistically significant difference.
Self-report surveys of families revealed positive well-being trends in favor of the experimental group in several domains, including relationships with other adults; relationships with their child(ren); overall well-being of their child(ren); respondents' general well-being; economic or financial outlook; current job or job prospects; home life; and life in general. However, the mean difference in survey scores was statistically significant for only three domains: current job or job prospects, home life, and life and general.

Additional findings will become available as implementation continues.

¹ Based on information submitted by the State as of January 2008. Back

Back to Table of Contents