Home > Programs & Funding > Profiles of the Child Welfare Waiver Demonstration Projects > Ohio - Flexible Funding - Phase I

Ohio

Demonstration Type:	Flexible Funding – Phase I1
Approved:	February 14, 1997
Implemented:	October 1, 1997
Completed:	September 30, 20042
Interim Evaluation Report Date:	November 5, 2000
Final Evaluation Report Date:	June 19, 2003

Target Population

All children in the State’s experimental counties who are at risk of entering placement or who have already been placed were eligible to participate in Ohio's title IV-E Waiver demonstration.

Jurisdiction

Fourteen counties were given flexibility under the demonstration to develop managed care strategies with the goal of improving outcomes for children and families and controlling foster care expenditures.

Intervention

The State granted 14 counties flexible use of capped allocations of title IV-E funds.  The underlying theory behind the demonstration was that, by employing managed care strategies, counties would be more efficient and effective in serving children and families.  The 14 counties were responsible for achieving desired safety, permanency, and well-being outcomes with a fixed amount of funds.  Counties were obligated to make available all services necessary to meet established outcomes, either directly or through contracts with community-based service providers.  Each county developed and implemented a variety of managed care strategies to suit its objectives.

At the beginning of the demonstration, counties focused on retraining staff and supervisors.  In the initial years of the demonstration, the State provided support and assistance to the experimental counties via conferences and training on managed care techniques and financing models.  Counties developed one or more strategies to improve services and control costs.  Common strategies included expanding the array of services, creating capitated contracts for services, improving case management and coordination, increasing competition among providers, establishing utilization review mechanisms, and developing quality assurance systems.  Throughout the project, experimental counties received ongoing support and assistance from the State through bimonthly meetings and monthly comprehensive data reports. 

The State paid participating counties a capped allocation based on historical and projected numbers of days in foster care, costs of care, and the percentage of children in care who were title IV-E eligible.  Counties, in turn, negotiated financial and risk-sharing agreements with private providers.  Counties that achieved title IV-E savings could reinvest the funds in other child welfare services.

Evaluation Design

The evaluation consisted of process, outcome, and cost-effectiveness components.  All Ohio counties had the option to participate in the demonstration.  After 14 counties volunteered, 14 other counties with similar characteristics were selected to serve as a comparison group.

Ohio examined the following measures:  changes in county child welfare agency operations, use of managed care strategies, shifts in expenditure patterns, changes in caseload size and composition, changes in patterns of first placement into out-of-home care, changes in destination for children leaving their first placement episode, and changes in length of time for children to exit from care to different destinations.

The evaluation of the five-year project consisted of four related studies, which assessed the demonstration's hypothesis from different perspectives:

Process Implementation Study:  This study used site visits and other primary data collection methods to document the Waiver's effects.  In year five of the evaluation, this study analyzed prevention initiatives, mental health services, the relationship between the child welfare agency and the juvenile court, and interagency collaboration.  Telephone interviews were conducted in select counties and a survey of child-serving agencies was conducted in each participating county. 

Community Impact Study:  This study observed how changes over time in demonstration and comparison counties affected the infrastructure and dynamics of the larger community.  In year five of the evaluation, the Community Impact Study was merged into the Process Implementation Study.

Fiscal Outcomes Study:  This study examined whether spending patterns were changed under the demonstration and, if so, how spending changed.  It included an analysis of State and county-level child welfare expenditures in demonstration and comparison counties from two years prior to the demonstration (1996) through the final year of the demonstration (2002).

Participant Outcomes Study:  This study compared outcomes for participants over the five-year demonstration period by county group and county size.  In year four, all active caseworkers were surveyed about their backgrounds, attitudes, and one randomly-selected open case.  In year five, survival analysis was used to determine how child and family outcomes would have differed in the absence of the title IV-E waiver.  The focus was on differences in the length of first placement by specific exit outcomes.

Evaluation Findings

Process Findings

The State's Fourth Semi-Annual Report (July 2002-December 2002)provided a summary of initial implementation issues through December 2002.  The changes observed in each of the managed care areas (service array, financing methods/capitation and risk, case management/care coordination, provider competition, utilization review, and quality assurance) reveal a pattern.  Both experimental and comparison counties increased their overall use of managed care strategies from year two to year four.  Experimental counties continued to increase their overall involvement in these managed care activities beyond year four, while comparison counties made significant changes in the way they offered services by increasing their use of managed care strategies.

Ohio's final report explains that experimental counties differed from comparison counties in five areas:  service array; targeting of services; quality assurance and data management; overall use of managed care strategies; and interagency collaboration.  Specifically, the demonstration appears to have led to several changes in experimental counties that were not matched by comparison counties: 

• New prevention activities were targeted to service areas that had been identified as insufficient.

• More targeting of new initiatives to particular populations occurred.

• Outcome information was more often systematically gathered, shared with staff, and used in management decisions.

• In both Year Two and Year Four of the demonstration, managed care strategies were used more often.  In Year Two, the differences between experimental and comparison counties were statistically significant.

• Some differences in aspects of interagency collaboration were evident.  Specifically, the evaluation revealed a moderate difference between experimental and comparison counties in pooling or sharing funds, with experimental counties somewhat more likely to adopt joint funding mechanisms.

However, the Process Study component identified six areas in which systematic differences between demonstration and comparison counties were not observed:

• Case management:  No significant differences emerged between experimental and comparison counties in terms of family involvement in case decision-making, team conferencing, and screening processes.

• Financing:  Although some demonstration counties made greater use of managed care contracting mechanisms, too few counties participated in managed care to identify a systematic effect across all experimental counties.

• Competition:  Both experimental and comparison counties have sought to increase competition by expanding agency foster homes through increased per diem payments and other methods.

• Utilization review and quality assurance:  Experimental counties were only slightly more likely to conduct formal reviews of children entering placement or already in placement.

• Services array:  Few differences emerged between the two groups in terms of improved service availability, the nature of new services, or timely access to services.

• Interagency collaboration:  Both experimental and comparison counties have developed strong relations with child welfare organizations, juvenile courts, and mental health agencies.   

Study Limitations

In its final report, the State identified several key factors that diminished measurable effects of the waiver: 

• Attempts to analyze the effects of initiatives at the county level involved too small a sample (14 experimental sites and 14 comparison sites) to produce significant findings. 

• Many of the counties involved in the demonstration had small populations and small foster care caseloads, which led to a limited data set available for analysis.

• Many counties submitted imprecise or incomplete data for the Fiscal Study.

• County fiscal administrators operated separately from program administrators and were therefore unable to track progress toward desired outcomes throughout the demonstration and change course as necessary.  In addition, fiscal staff did not necessarily connect the title IV-E funding available through the Waiver with any particular outcomes.

• Evaluators for each of the four studies used somewhat different standards; therefore, in some cases, findings between studies are not comparable.

• In the context of public funding for child welfare services in Ohio, title IV-E is only one of several major funding streams.  Local tax levies, which account for more than half of the budgets of local child welfare agencies, could be used flexibly for new programs and services in comparison counties.  Further diminishing the effects of the title IV-E changes were substantial cutbacks in State support in the later years of the demonstration.

• Finally, although the waiver created a fiscal incentive to reduce foster care expenditures, the financial risk to counties for not reducing foster care spending was limited.

Cost Neutrality Findings

Overall, no significant differences in child welfare spending were found between experimental and comparison counties.  Growth in paid placement days and in the average daily cost of foster care was experienced by both experimental and comparison counties, and neither group significantly changed the percentage of placement days in residential settings. 

However, variances were noted in growth in foster care spending, which suggests that experimental counties were able to contain foster care growth more than comparison sites.  Differences were also found in growth in non-foster care expenditures, which suggests that experimental counties took advantage of the waiver’s flexibility to expand activities into new areas.  These observed patterns of change were close to achieving statistical significance.

Outcome Findings

Overall, Ohio’s final report presented mixed results.  Although some changes were noted as a result of the demonstration, they were not strong enough to reform the State's child welfare system fundamentally.  In addition, observed changes were neither large nor targeted enough to create statistically significant differences in foster care expenditures or child and family outcomes. 

In addition, the waiver did not significantly affect the following:

• Permanency:   The waiver had no significant affects on reunification rates, adoption rates, or median length of stay prior to reunification or adoption.³ 

• Placement stability:   Both experimental and comparison counties increased the percentage of children who made no moves during their first placement and decreased the percentage who made five or more moves.  Experimental counties were no more successful than comparison counties in moving children to less restrictive settings. 

• Use of relatives for placements:  Only four experimental counties significantly increased the use of relatives for first placements, whereas three experimental counties significantly decreased relative placements. 

• Proportion of children served in-home:  Both experimental and comparison counties maintained a pattern of serving approximately three-fourths of cases in-home.

• Safety of children returned home:  Results indicate that the safety of children in experimental counties who were returned home was maintained at the same level as that experienced in comparison counties.⁴

In addition to the outcomes study, the Ohio evaluation integrated findings from 6 of the 14 demonstration counties⁵ into a Case Study Findings section in the State’s final report.  Two of the six case study counties (Lorain and Muskingum) experienced the most positive effects of the Waiver.  Both were able to reduce placement, increase the number of children served in-home or referred to community agencies, and reduce payment utilization during the demonstration.   These two counties had in common an early and ongoing commitment to expand resources for child welfare activities, along with clearly defined programmatic reform efforts which were supported by spending shifts.

The other four case study counties (Fairfield, Franklin, Stark, and Clark) experienced mixed results.  All of the sites were able to leverage funds from other sources to expand non-foster care activities, and they succeeded in changing common patterns of intervention with new cases.  However, these counties continued to struggle with the challenge of serving the population of children already in foster care at the start of the demonstration.

 

 

¹Based on information submitted by the State as of June 2003. Back

²The demonstration was originally expected to end (after one bridge extension) in October 2003.  HHS has granted the State a second bridge extension through March 2004 and a third bridge extension through September 30, 2004.  The State has requested a long-term extension, during which they may implement the demonstration statewide. Back

³Adoptions increased substantially in both experimental and comparison counties during the demonstration; however, it cannot be ascertained whether this increase was associated with the effects of the waiver. Back

⁴This finding may alleviate concerns that the waiver could result in children returning home too soon.  However, the State recommends in its final evaluation that further research on child safety outcomes be completed, as its findings were constrained by data limitations. Back

⁵The subset represents those counties that experienced the most significant changes regarding children exiting their first out-of-home placements. Back

Back to Table of Contents