Children's Bureau Website - North Carolina - Flexible Funding/Assisted Guardianship

North Carolina

Demonstration Type:	Flexible Funding/Assisted Guardianship - Phase I¹
Approval Date:	November 14, 1996
Implementation Date:	July 1, 1997
Completion Date:	June 30, 2004²
Interim Evaluation Report Date:	June 30, 2002
Final Evaluation Report Date:	November 2002

Target Population

The eligible population consisted of children residing in experimental group counties who were at imminent risk of placement or who were already in placement. Each participating county, however, could choose to implement initiatives that affected some or all of these children.

Jurisdiction

Nineteen of the 100 counties in the State participated in the demonstration.

Intervention

Each county was able to develop its own initiatives with approval of the State. Counties differed in both the number and type of initiatives developed for the demonstration:

13 counties used flexible funds to meet needs on a case-by-case basis;
11 counties engaged in various collaborative activities;
17 counties used funds to support organizational changes;
15 counties used funds to support court reform activities;
16 counties developed contracts for new services;
9 counties developed new services in-house;
8 counties provided enhanced support for resource families; and
8 counties provided an assisted guardianship option.

The number of activities or services implemented ranged from two new service areas (in two counties) to seven new service areas (in five counties). Further, as summarized below, some counties chose to enter into contracts with private providers for services, while others chose to develop service delivery capabilities in-house.

Type of Service	Number of Counties
Type of Service	Entered Contracts	Developing In-House Services
Family Support	10	7
Assessment	9	5
Adoption	3	4
Post Adoption Placement, Post Finalization Services	4	4
Substance Abuse Services	6	3
Mental Health Services	9	4
Family Reunification	5	4
Legal Services for TPR/Adoption	9	5

Financial Structure

The demonstration would be deemed cost-neutral if the rate of growth in expenditures of title IV-E foster care and title IV-E administrative funds by the experimental group was equal to or less than the rate of growth over the baseline of those same expenditures by the comparison group. (Local agencies in the experimental group were given broad flexibility in using IV-E funds to prevent children from entering care, to help children exit care sooner, and to prevent children from re-entering care.)

In addition, the State established local trust accounts for each of the 19 counties that volunteered to participate in the demonstration. Unexpended State funds, which were budgeted for the cost of care for non-IV-E-eligible children, were placed in those trust funds for use by the individual demonstration counties. These were 100 percent State dollars, and demonstration counties could use these funds, matched with IV-E administrative dollars, for innovative efforts that targeted one or more of the three goals of the demonstration. Individual counties could access these funds if they had achieved cost neutrality and had a reinvestment plan approved by the State. In the final year of the demonstration, counties that had not been cost neutral were also allowed to use the funds in their trust accounts.

Evaluation Design

The evaluation consisted of process, outcome, and cost-effectiveness components. To evaluate the demonstration, the State used a comparison group design. The 19 comparison counties were selected based on size, demographics, the number of title IV-E-eligible children, and socioeconomic status of families. The State compared the experiences of successive cohorts of children reported as abused or neglected and/or who entered out-of-home care. The final report divided these cohorts into four groups for comparative analysis: (1) active waiver counties³, (2) less active waiver counties, (3) comparison counties, and (4) other counties in the State. North Carolina analyzed data for the following outcomes: rate of initial entry into foster care, time spent in out-of-home care, and rate of re-entry.

Evaluation Findings - Capped IV-E Allocations and Flexibility to Local Agencies

Process Evaluation

The State reported that the necessity to include local and county fiscal staff in the planning and procedural implementation of the demonstration was a key lesson learned. Although the demonstration began operations in 1997, and experimental counties had the ability to access funds in local trust accounts since 1998, counties initially appeared reluctant to use these funds. Counties increased their use of flexible funds over time as they became more familiar with the demonstration, procedures for accessing flexible funds, and cost neutrality requirements.

Outcome Evaluation

To assess the impact of the demonstration, it was necessary to control for several factors: (1) the presence of other child welfare reform initiatives in both experimental and comparison counties, (2) changes in population characteristics in experimental and comparison counties that could affect children's degree of risk for maltreatment and subsequent foster care placement, and (3) differences in the level and types of initiatives instituted in the waiver counties. Multivariate analyses incorporating measures of these factors provided the basis for findings presented in the final evaluation report.

Probability of out-of-home placement: The probability of placement for 175,190 children who experienced an initial substantiated incident of abuse and/or neglect between State Fiscal Year 1994 (SFY94) and SFY01 was calculated using data in the State Child Abuse and Neglect Registry. Findings indicated that among children with a substantiated report of abuse or neglect from 1997 to 2001, the probability of placement for children in experimental counties declined more than for children in the comparison counties, or for children in other counties in the State. No significant differences were found between more active and less active experimental counties.⁴
Length of stay in foster care: The State's evaluators developed a longitudinal database to track the experiences of 41,585 children who initially entered placement from SFY94 through SFY01. Two indicators of length of stay were used: (1) the likelihood of exiting placement, and (2) the likelihood of exiting placement for children who remained in care two years after initial entry. Findings indicated that children entering placement during the demonstration were more likely to exit placement than those who entered in pre-waiver years. This trend was true for all groups of counties.

However, an analysis of vital statistical data indicated that the risk profile for children entering care in the experimental counties became more serious over the term of the demonstration.

Therefore, the evaluation concluded that the experimental counties were able to reduce lengths of stay even though the seriousness of risks for children entering placement increased after the demonstration was implemented. Also, more active experimental counties showed greater rates of decline in length of stay in foster care. However, when exit rates for children who remained in foster care two years after initial entry were examined, no significant differences were found among children in experimental counties, comparison counties, or other counties in the State.
Re-entry into foster care: The probability of re-entry among children who achieved permanency at the end of their first out-of-home placement was examined across all county groups. Two subgroups were defined in order to test the impact of the intervention: (1) active waiver counties were those that initiated four or more new services or began accessing trust fund resources prior to June 30, 2000, and (2) other waiver counties were those that did not.

Analysis showed a consistently lower rate of re-entry in waiver counties compared to baseline SFY93. The decrease in re-entry is somewhat larger in the active waiver counties in the early waiver years. Only in the most recent two years for which data were available did other waiver counties surpass active experimental counties in improvements to re-entry rates. Re-entry rates for children exiting placement in comparison counties were stable in the years immediately preceding waiver implementation. These rates increased until SFY00, when the likelihood of re-entry was about 10 percent less than seen in the baseline year.

Evaluation Findings - Assisted Guardianship

The 19 experimental counties also had the option of developing assisted guardianship; however, only eight counties utilized this option. A total of 38 assisted guardianships were established, with one county having established 17.

Initially, the assisted guardianship payment was $250 per month, which was less than the standard foster care payment of $315 to $415 per month (based on the age of the child). During the first three years of demonstration, none of the counties used the guardianship option. In October 2000, the payment was increased so that it was equal to the foster care maintenance payment.

Although the assisted guardianship option was intended for children for whom efforts at reunification or adoption were unsuccessful, counties did not appear to use the option to achieve permanency in backlogged cases. Only 12 of the children with assisted guardianships were initially placed in foster care between SFY95 and SFY96; 22 had been placed between SFY97 and SFY99; and 10 had not entered foster care until SFY00.

Evaluators and county staff discussed reasons why so few assisted guardianships were established early in the demonstration. In addition, they conducted a survey of 16 counties that never or rarely used assisted guardianship in order to determine the barriers to guardianship. The most frequently reported reason was the financial risk to the counties of continuing guardianship payments after the end of the demonstration period. Another frequently cited issue concerned the agency staff's beliefs about the appropriateness of guardianship arrangements. In initial discussions, and later in the 16-county survey, several staff noted that their primary goal was to place children in adoptive homes.

During site visits in SFY99 and SFY00, staff in seven counties also expressed "confusion about specific assisted guardianship rules." Even in counties that had established guardianships, staff noted that they had been confused about State support for assisted guardianship, waiver rules, and regulations for guardianship, including Medicaid eligibility and receipt of Supplemental Security Income (SSI) payments.

Based on a focus group that included the staff of the county agency that had established the most assisted guardianships (n=17), the Final Evaluation Report noted that, "assisted guardianship met the needs of some African-American adolescents who resisted the idea of Termination of Parental Rights." They also noted that "...DSS (Department of Social Services) staff experience with assisted guardianship led to a change in agency norms for the use of guardianship in general. Staff began to value guardianship, subsidized or not, as a way of expediting permanency without eliminating future options for reunification or adoption."⁵

¹ Based on information submitted by the State as of November 2002. Back

² North Carolina's original completion date was June 30, 2002. The State received four short-term bridge extensions. Back

³ The State uses the term "waiver counties" to refer to experimental group counties. Back

⁴ Thirteen of 19 experimental counties were identified as active. Active counties were those that initiated 4 or more new services or began accessing available resources in the trust fund prior to June 30, 2002. Back

⁵ It is important to note that the rationale for guardianship was quite different from other States where assisted guardianship is used only when adoption and reunification are not viable options. Back

Back to Table of Contents