Child Welfare: New Financing and Service Strategies Hold Promise, but
Effects Unknown (Testimony, 07/20/2000, GAO/T-HEHS-00-158).

Pursuant to a congressional request, GAO discussed the progress made by
states and localities as they develop new financing, service-delivery,
and accountability strategies for child welfare programs, focusing on:
(1) the financial and service-delivery changes states and localities
have made in their managed care initiatives; (2) how they are measuring
the initiatives' outcomes; and (3) what is known about the effect of
these changes on children and families.

GAO noted that: (1) states and localities that are implementing child
welfare managed care initiatives are moving away from a traditional
fee-for-service reimbursement approach to one that funds a single
provider in advance under a capitated payment; (2) this allows the
single provider--now assuming greater responsibility for case planning
and providing needed services-- the flexibility to package and manage an
array of child and family services; (3) under these new arrangements,
states and localities are taking steps toward becoming more
performance-based and results-oriented as they implement child welfare
managed care initiatives; (4) GAO found that the state and local
agencies operating these initiatives are beginning to identify measures
associated with five child and family outcome categories--child safety,
a permanent home for the child, child and family well-being, the
stability of out-of-home placements, and clients' satisfaction with the
services they received; (5) these agencies are using such strategies as
setting performance standards and incorporating financial incentives in
contracts with service providers to hold them accountable for their
performance and ensure that desired results are achieved; (6) however,
GAO found that many of the state and local agencies operating these
initiatives do not have appropriate data systems in place to store,
analyze, and retrieve information on client outcomes; (7) most state and
local officials GAO talked with who were responsible for the initiatives
are encouraged by the changes occurring in child and family outcomes;
(8) while controlling costs was seen as a potential benefit of managed
care, an equally if not more important goal was improved services for
children and families; (9) in fact in some cases, overall spending has
increased; (10) whether any outcome changes associated with these
initiatives can be attributed to the new strategies is still largely
unknown because they have not been rigorously evaluated; and (11)
planned evaluations under the federal waiver demonstration program
will--in the future--yield additional information about the
effectiveness of child welfare managed care arrangements.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  T-HEHS-00-158
     TITLE:  Child Welfare: New Financing and Service Strategies Hold
	     Promise, but Effects Unknown
      DATE:  07/20/2000
   SUBJECT:  Foster children
	     Public assistance programs
	     Cost control
	     Accountability
	     State-administered programs
	     Performance measures
	     Management information systems
	     Families
IDENTIFIER:  HHS Child Welfare System
	     HHS Child Welfare Services Program
	     Title IV-E Foster Care Program

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Testimony.                                               **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************

GAO/T-HEHS-00-158

   * For Release on Delivery
     Expected at 1:00 p.m.

Thursday, July 20, 2000

GAO/T-HEHS-00-158

CHILD WELFARE

New Financing and Service Strategies Hold Promise, but Effects Unknown

        Statement of Cynthia M. Fagnoni, Director

Education, Workforce, and Income Security Issues

Health, Education, and Human Services Division

Testimony

Before the Subcommittee on Human Resources, Committee on Ways and Means,
House of Representatives

United States General Accounting Office

GAO

Child Welfare: New Financing and Service Strategies Hold Promise, but
Effects Unknown

Madam Chairman and Members of the Subcommittee:

I am pleased to be here today to discuss the progress made by states and
localities as they develop new financing, service-delivery, and
accountability strategies for their child welfare programs. In the
mid-1980s, child welfare agencies faced a poorly integrated patchwork of
services for children and families accompanied by escalating costs. As we
reported to this Subcommittee in October 1998, a number of states have
incorporated or are considering incorporating some of the principles of
managed care into their family preservation, foster care, and adoption
programs. Under a managed care approach, states and localities prospectively
pay fixed, capitated amounts to providers to coordinate and meet all the
service needs of referred children and families. The officials responsible
for these new managed care initiatives saw this approach as a strategy both
to improve the quality of care children and families in the child welfare
system received and to control the rising costs of delivering services while
holding all the partners in the system accountable.

Now that many of these initiatives have been in operation for 3 or more
years, you asked us to report on their progress. As you requested, I will
focus my remarks on (1) the financial and service-delivery changes states
and localities have made in their managed care initiatives, (2) how they are
measuring the initiatives' outcomes, and (3) what is known about the effect
of these changes on children and families. My testimony is based on our past
and ongoing work on 27 state and local initiatives that have been in
operation since January 1998 or earlier.

In summary, states and localities that are implementing child welfare
managed care initiatives are moving away from a traditional fee-for-service
reimbursement approach to one that funds a single provider in advance under
a capitated payment. This allows the single provider-now assuming greater
responsibility for case planning and providing needed services-the
flexibility to package and manage an array of child and family services.
Under these new arrangements, states and localities are taking steps toward
becoming more performance-based and results-oriented as they implement child
welfare managed care initiatives. We found that the state and local agencies
operating these initiatives are beginning to identify measures associated
with five child and family outcome categories-child safety, a permanent home
for the child, child and family well-being, the stability of out-of-home
placements, and clients' satisfaction with the services they received. In
addition, these agencies are using such strategies as setting performance
standards and incorporating financial incentives in contracts with service
providers to hold them accountable for their performance and ensure that
desired results are achieved. However, we found that many of the state and
local agencies operating these initiatives do not have appropriate data
systems in place to store, analyze, and retrieve information on client
outcomes. Most state and local officials we talked with who were responsible
for the initiatives are encouraged by the changes occurring in child and
family outcomes. While controlling costs was seen as a potential benefit of
managed care, an equally if not more important goal was improved services
for children and families. In fact, in some cases, overall spending has
increased. Whether any outcome changes associated with these initiatives can
be attributed to the new strategies is still largely unknown because they
have not been rigorously evaluated. Planned evaluations under the federal
waiver demonstration program will-in the future-yield additional information
about the effectiveness of child welfare managed care arrangements.

Background

In our 1998 report, we concluded that initiatives in which principles of
managed care were being implemented were still in the early stages of
program development and, as a result, were largely untested. We found that,
for these initiatives to mature and meet officials' program expectations,
state and local agencies needed to resolve three important issues. The first
was to address cash flow problems in a new environment of funding services
prospectively under a capitated payment system while seeking reimbursement
for the federal share of costs only after services are delivered. In
addition, state and local agencies stood a better chance of reducing or
eliminating the service access problems often associated with different
eligibility requirements in categorical funding streams if there was funding
flexibility. The second issue facing state and local agencies was to
continue to improve their capacity to collect, analyze, and report client
and service data. Such data are paramount for state and local agencies to
set reasonable and appropriate payment rates and performance standards, make
additional programmatic changes or give service providers feedback, and
improve policies and procedures for serving children and families. The third
issue requiring resolution was that state and local agencies needed to
continue to develop and refine strategies to hold both themselves and their
private partners accountable for achieving desired outcomes. Moreover, these
agencies needed to develop the capacity to continuously measure and report
their progress toward meeting performance goals. Outcome measurement and
performance management were new areas of focus for the child welfare system.

States and Localities Implement New Financing and Service-Delivery
Strategies

Capitated Payments Provide Flexibility

To further increase service flexibility, some states and localities are
funding capitated payment arrangements by pooling individual state funding
streams that support different services that children and families in the
child welfare system need. Because of restrictions on eligibility and
prohibitions on certain uses of funds, public and private child welfare
caseworkers often encounter problems accessing needed services for clients.
By pooling or blending funds from various sources, these states and
localities seek to reduce service access problems sometimes associated with
categorical programs and increase flexibility in the use of funds. In
Colorado, for example, the state blended funds from several child welfare
and child care budget line items and allocated a fixed level of
funding-equivalent to a block grant-to its counties. Block-granting state
dollars in this way loosened the restrictions on the use of these typically
categorical funds and increased counties' flexibility. Boulder County
further pooled its child welfare block grant with funding from the mental
health agency and youth corrections agency to finance its Integrated Managed
Partnership for Adolescent Community Treatment (IMPACT) initiative, serving
adolescents at imminent risk of placement in group or residential care.

Service-Delivery Changes Are Designed to Improve Access to Care

States and localities have also shifted more case management
responsibilities-much of which public agency workers had performed-to
private contractors as part of their new role as care coordinators. In an
effort to better match services with client needs, the primary contractor in
many of the 27 initiatives included in our study uses a team approach to
managing its caseload of children and families. This approach is designed to
avoid the duplication, time delays, and fragmentation that often result
under traditional case management, when different service systems and the
many providers involved in a child's care are not part of the treatment
planning and decision-making process. In some initiatives, the treatment
team consists of those individuals who are regularly in direct contact with
the child, including the case manager, therapist, parents or guardians,
school officials, and other service providers. In other initiatives, case
management teams include representatives from multiple agencies, such as
child welfare, mental health, and juvenile justice agencies.

In most of the 27 initiatives, states and localities have contracted both
the management and the coordination of care for children who have been or
are at risk of being abused and neglected. However, not all aspects of the
child welfare system have been contracted to private entities. States and
localities have retained certain functions that officials believe are
critical to meeting their legal responsibility for the safety and well-being
of children in the child welfare system. In every initiative, the state or
locality continues to conduct all child protection functions related to
investigating reports of child maltreatment and recommending to the courts
whether a child needs to enter the child welfare system for protective or
any other services. A child enters the managed care system on the basis of a
referral from the state or locality to the managed care entity. In some
initiatives, the state or locality also maintains its presence by retaining
the authority to approve contractors' decisions related to reducing a
child's level of care, such as moving a child from residential care to
family foster care.

States and Localities Are Taking Steps Toward a More Performance-Based and
Results-Oriented Approach

Agencies Are Beginning to Identify Various Child and Family Outcome Measures

Table 1: Examples of Child and Family Outcome Measures
 Category      Outcome             Measure

 Safety        Children are safe   Confirmed reports of abuse and neglect
               from maltreatment   in the general population
                                   Recurrence of abuse or neglect while
                                   children are receiving in-home services
                                   Reports of abuse or neglect while the
                                   children are in out-of-home care
                                   Recurrence of physical abuse, sexual
                                   abuse, or neglect after children have
                                   left care
               Children are placed Children who are returned to their
 Permanency    in a permanent home parents or relatives within a specified
               in a timely manner  time
                                   Finalized adoptions
                                   Children who achieve permanency within a
                                   specified time
                                   Average length of stay in out-of-home
                                   care
                                   Children who are maintained in their
                                   home and do not enter out-of-home care
               Children maintain
               the permanent       Children who reenter care within a
               placement           specified time
               Children function
                                   Children's emotional and behavior crises
 Well-being    adequately in their that result in hospital use or police
               families and
               communities         calls
                                   Children's behaviors related to sexual
                                   misconduct, running away, and suicide
                                   Children's scores on standardized tests
                                   of childhood functioning
                                   Children's movement to less restrictive
                                   placement settings
                                   Youths discharged from care who have
                                   completed high school, have obtained a
                                   general equivalency diploma, or are
                                   participating in an educational or job
                                   training program
               Families function
               adequately in their Families' adaptation to caregiving
               communities
               Children experience
 Stability     a minimum number of Number of placements while in
               placements          out-of-home care
               Children maintain
               contact with their  Children placed with at least one
               family and communitysibling
                                   Children placed within their home or
                                   contiguous county
                                   Children placed out-of-state
               Clients are         Youths who reported satisfaction with
 Satisfaction  satisfied with      services, as measured by the Client
               services            Satisfaction Survey
                                   Children who reported satisfaction with
                                   their foster care placement, based on an
                                   exit interview
                                   Families who reported that the
                                   initiative provided them a valuable
                                   service

Source: GAO analysis of interview data.

We also found that states and localities are measuring different outcomes,
depending on the population served by the initiative and the states' or
localities' goals. For example, El Paso County's initiative in Colorado
encompasses all children and services in the county's child welfare system;
as a result, the county established a broad safety outcome and is measuring
child abuse and neglect rates among the general population. In contrast,
Massachusetts targets older children in residential care for its Commonworks
initiative in which the lead contractors only serve children, while the
state serves the family and decides when a child can return home. Instead of
monitoring the recurrence of maltreatment, the state measures outcomes
related to children's movement to less restrictive settings and reentry into
residential care. One of Illinois' goals for its performance contracting
initiative is to find foster children a permanent home in a timely manner
while minimizing multiple out-of-home placements. To monitor progress toward
this goal, the state established several outcome measures, including average
length of stay and the number of placements in different foster homes while
children are in out-of-home care.

Agencies Are Attempting to Hold Service Providers Accountable

States and localities responsible for these child welfare initiatives are
using their outcome measures to establish performance standards for both
public and private service providers. By doing so, they are trying to hold
all the parties in the initiative accountable for results. States and
localities have established performance standards for 11 of the 27
initiatives we reviewed. Most performance standards are expressed as a
specified level of outcome to be attained. For its Multi-Agency Team for
Children (MATCH) initiative for seriously emotionally disturbed children,
for example, Georgia has included standards that 40 percent of the children
will improve their functioning and be discharged to a less restrictive
placement setting, and that a 20-percent decrease will occur in the
frequency with which children harm others.

As states and localities gain more experience with managed care, officials
expect to adjust existing standards or introduce new ones. For example, in
Kansas' foster care initiative, state officials realized that their
first-year performance expectations for the lead contractors were in all
likelihood unrealistic because the standards were not based on past program
performance. As a result, Kansas officials expected to and did adjust
performance standards annually as more current information was collected. In
contrast, Massachusetts took a more incremental approach for its Commonworks
initiative. The state did not introduce performance standards in the lead
agencies' contracts until the third year of operation, after sufficient
information had been collected to establish a baseline from which to set
standards.

Another strategy to hold managed care providers accountable for their
performance and achieving desired results is to link financial rewards and
penalties to outcomes. In some initiatives, the state or local agency offers
bonuses as a financial incentive for the managed care entity to meet
performance standards and penalties for poor performance. In the TrueCare
partnership initiative in Hamilton County, Ohio, for example, the managed
care contractor can earn bonuses when it meets individual performance
indicators related to (1) child and family outcomes, such as ensuring
children's safety and reducing the risk of harm, and (2) management
services, including maintaining a competent provider network and maximizing
revenues. Similarly, the contractor can incur financial penalties when it
fails to meet the performance indicators. Massachusetts offers bonuses to
the lead contractors for achieving interim or successful outcomes. In
Massachusetts' Commonworks initiative, a lead contractor can earn bonuses at
three different intervals-when a child transitions to a less costly level of
care, when a child leaves placement, and when a child does not re-enter the
lead contractor's care within 6 months of discharge. In addition, in the
Massachusetts initiative as well as others, poor performers risk not having
their contracts renewed. However, even satisfactory performers may lose
their contracts because of other factors. For Illinois' performance
contracting initiative, foster care providers that met performance standards
but were not the top performers lost their contracts when the successful
outcome of a declining child welfare population resulted in a need for fewer
providers.

Data Systems Are Needed to Manage Information on Outcomes

Although agencies are taking steps to identify and measure outcomes, many
have done so without appropriate information systems in place. In many
instances, private service providers and states and localities are working
with multiple, incompatible, or manual systems. While these systems may
yield information on child and family outcomes, they are inefficient. For
example, the lead contractor for the managed care initiative in Sarasota
County, Florida, uses three separate, unintegrated data systems to track
client and service data, and must enter duplicate information into each
system and physically locate the three computer terminals side-by-side to
ensure consistent data. For some initiatives in other states, agency staff
manually collected outcome data because information systems had yet to be
developed.

In several locations, data systems were developed specifically for the child
welfare managed care initiative. In both Massachusetts' Commonworks and the
Hamilton County, Ohio, TrueCare Partnership initiatives, the state or local
agency required one of its managed care contractors to develop a data system
specifically for the managed care initiative at the same time that new
financial and service-delivery arrangements were implemented. These systems
were not integrated with the state or local agencies' child welfare
information systems at the time of our study, but may be in the future.

States and localities have not used federally supported statewide data
systems to implement, monitor, or manage their child welfare managed care
initiatives. Among the 12 initiatives we contacted about this issue, none of
the state or local agencies are using their state's Statewide Automated
Child Welfare Information Systems (SACWIS) to manage information on their
initiatives' clients, services, or outcomes. Whether the state's SACWIS was
operational or still under development, officials for some initiatives told
us they hoped to either link their initiatives' data system to SACWIS or
incorporate SACWIS into their initiative in the future.

Effectiveness of Managed Care Initiatives Is Largely Unknown

Officials Report Improved Child and Family Outcomes

State and local agencies used their outcome measures to track their
initiatives' progress in several ways. For some initiatives, outcomes were
reported as change that occurred during the initiative. The Colorado example
on the reduced incidence of abuse and neglect used the previous year as a
comparison. For other initiatives, agencies reported outcomes without any
indication of change-sometimes because comparisons had not been made.
Initiatives that targeted the hard to serve and most costly children-those
in need of placement in residential treatment centers-were considered to
have had a positive outcome when children successfully transitioned to less
restrictive, less costly placement settings. For Georgia's MATCH initiative,
for example, officials reported that 41 percent of the program participants
improved functioning and were discharged from a more restrictive residential
setting to a less restrictive placement, such as a group home, treatment
foster home, or their own home.

While the potential to control costs attracted state and local child welfare
agencies to managed care, their primary objective was not necessarily to
reduce spending. Instead, some officials expressed a desire to reduce
certain types of costs-such as the living expenses for out-of-home
placements-or to use existing funds more efficiently and reinvest any
savings into services. For some initiatives, officials reported that overall
spending has actually increased as a result of additional administrative
costs associated with private entities assuming responsibility for managing
clients' care and the state or locality overseeing contracts. For example,
Massachusetts reported that its Commonworks initiative is costing more,
overall, despite realizing savings in some specific areas. Out-of-home
placement costs averaged 3 percent less than the lead contractors' capitated
payment rate. Although spending for in-home or aftercare services increased
80 percent as more children moved from residential treatment to less
restrictive settings, the net effect was a cost reduction in spending for
out-of-home and in-home services combined. Both the state and its lead
contractors have reinvested the service-cost savings into program
development. However, the state has incurred additional costs for an
administrative services organization (ASO) to provide management services,
lead contractors to manage their respective service-provider networks, and
the state's oversight and management of the ASO and six lead contracts.

Lack of Rigorous Evaluation Leaves Initiatives' Effects Unknown

While the 27 initiatives included in our study have had limited evaluation,
planned evaluations under the federal title IV-E waiver demonstration
program will yield additional information about the effectiveness of child
welfare managed care arrangements. By law, states receiving this waiver must
have an independent evaluation of the initiative that, at a minimum,
compares and assesses child and family outcomes, methods of service
delivery, and fiscal consequences. According to HHS officials, evaluations
for the 12 waiver states that are testing managed care principles for child
welfare services should be completed within the next 5 years. To date, one
ongoing evaluation-of Ohio's demonstration of child welfare managed care in
several counties-has compiled baseline information on child and family
outcomes. Evaluation results are not yet available from any of the waiver
states.

Madam Chairman, this concludes my prepared statement. I will be happy to
respond to any questions that you or other Members of the Subcommittee may
have.

GAO Contact And Acknowledgments

Appendix

Child Welfare Managed Care Initiatives' Outcomes to Date

Table 3 includes the 27 managed care initiatives about which we collected
information regarding documented child and family outcomes, as of April
2000. In particular, we list quantitative results in the outcome areas of
child safety, permanent homes, child and family well-being, out-of-home
placement stability, and clients' satisfaction with the services that they
received. Preinitiative baseline data were generally not available. We
indicate changes and describe cost savings where data were available. In
some cases, results were not reported for individual initiatives but were
aggregated across multiple initiatives in a single state. Unless otherwise
noted, the combined outcomes are shown for (1) the three district
initiatives in Florida and (2) Champaign and Madison Counties in Ohio.

Table 3: Child and Family Outcomes for 27 Ongoing Child Welfare Managed Care
Initiatives, as of April 2000
                           Managed care
 Location and project name modela and        Child and family outcomes
                           project
                           description
 State-level initiatives
                                             Fiscal year 1998-99 results

                                             Children's behavior
                                             improved-incidents of negative
                                             behavior, such as aggression,
                                             self abuse, and property
                                             damage, decreased 21 percent
                                             between the 6-month and
                                             12-month evaluations for
                                             children admitted during 1998,
                                             and decreased 35 percent
                                             between the 6-month and
                                             discharge evaluations for
                           Public model      children discharged during
                                             1998

 Georgia                   Statewide         41 percent of the children
                           residential
                                             were either discharged from
 Multi-Agency              treatment         the project or stepped down to
                           services for
                                             a less restrictive setting
 Team for Children (MATCH) severely          during 1999
                           emotionally
                           disturbed
                           children          66 percent of the children who
                                             were discharged from the
                                             project were still in a less
                                             restrictive setting 6 months
                                             after discharge during 1999

                                             42 percent of the children who
                                             had progressed to a less
                                             restrictive setting were still
                                             in a less restrictive setting
                                             6 months after their transfer
                                             in 1999

                                             All children were placed
                                             within the state during 1999
                                             1998-99 results

                                             Permanency rate in 1999
                                             increased 149 percent over the
                                             previous year in Cook County's
                                             Home of Relative Foster Care
                                             program

                                             Number of 1999 adoptions
                           Public model      increased 70 percent over 1998
                                             and 228 percent over 1997
 Illinois
                           Relative and
                                             3,660 children achieved
 Performance Contracting   traditional       permanency through subsidized
                           foster care
                           statewide         guardianship between 1997 and
                                             1999

                                             Number of reunited families
                                             increased 12 percent between
                                             1997 and 1999

                                             Movement of children to more
                                             restrictive placement settings
                                             fell by more than half
                                             statewide
                           Managed care
                           organization
                           model

                           Wraparound
                           services for
                           seriously
 Indiana                   emotionally
                           disturbed
                                             Outcomes were not provided
 The Dawn Project          children, aged 5
                           to 17, who have
                           been impaired for
                           more than 6
                           months and
                           involved with
                           multiple service
                           systems in Marion
                           County
                                             Year 3 evaluation results,
                                             Jan.-Sept. 1999

                                             99 percent of the children did
                                             not experience abuse or
                                             neglect while in out-of-home
                                             placement (consistent with
                                             years 1 and 2)

                                             97 percent of the children did
                                             not experience abuse or
                                             neglect within 12 months of
                                             reuniting with their families
                                             (same as year 2)

                                             27 percent of the children
                                             placed in out-of-home care
                                             were returned to their
                                             families within 6 months
                                             (consistent with year 2;
                                             decrease of 49 percent from
                                             year 1)

                                             41 percent of the children
                                             placed in out-of-home care
                                             were returned to their
                                             families or achieved other
                                             permanency within 12 months
                                             (increase of 24 percent from
                                             year 2)

                                             74 percent of the children who
                                             returned to their families did
                                             not reenter out-of-home care
                           Lead agency model within 12 months of returning
                                             home (increase of 9 percent
 Kansas                                      from year 2)
                           Statewide foster
 Foster Care Privatization care services to  81 percent of youths, who were
                           children in state
                           custody           aged 16 and over and released
                                             from the state's custody, had
                                             completed high school,
                                             obtained a general equivalency
                                             diploma, or were participating
                                             in an educational or job
                                             training program (increase of
                                             8 percent from year 2 and 53
                                             percent from year 1)

                                             99 percent of the children
                                             experienced no more than three
                                             placement moves while in
                                             out-of-home care (consistent
                                             with years 1 and 2)

                                             71 percent of all children
                                             were placed with at least one
                                             sibling (decrease of 9 percent
                                             from year 2; consistent with
                                             year 1)

                                             78 percent of the children
                                             were placed within their
                                             regional boundaries
                                             (consistent with year 2;
                                             decrease of 9 percent from
                                             year 1)

                                             47 percent of the adults and
                                             70 percent of the youths (aged
                                             14 and over) reported
                                             satisfaction with services
                                             (decrease of 6 percent for the
                                             adults and consistent for the
                                             youth from year 2)
                                             Administrative services
                                             organization report, 1999

                                             More children moved from
                                             residential treatment to less
                                             restrictive settings-the use
                                             of group homes, specialized
                                             foster care, and independent
                                             living increased 73 percent
                                             from July 1997 to June 1999

                                             Children's placement in less
                                             restrictive settings was
                                             supported by increased
                                             provision of aftercare
                                             services-expenditures for
                           Administrative    aftercare services increased
                           services          80 percent, and the monthly
                           organization with average number of clients
                           lead agency model receiving aftercare services
                                             increased 51 percent over 1998
 Massachusetts
                           Statewide foster
                                             Recidivism rate of 6 percent
 Commonworks               care for          for youths who had a planned
                           adolescents
                           needing group     discharge, such as return to
                           care or           home; 17 percent for all
                           residential       youths discharged, including
                           treatment         unplanned discharges such as
                                             running away from foster care
                                             placement (Jan. 1997-Sept.
                                             1998)

                                             Savings achieved for
                                             out-of-home and aftercare
                                             services in 1999 (excludes
                                             administrative costs
                                             associated with the
                                             administrative service
                                             organization, lead
                                             contractors' management of
                                             provider networks, and state
                                             oversight)-lead contractors'
                                             monthly client placement costs
                                             averaged 3 percent less than
                                             the capitated case rate
                                             Descriptive evaluation
                                             results, 1998

                                             Child abuse and neglect rate
                                             of 9 percent during families'
                                             involvement in MIFPI (compared
                                             with the rate for all children
                                             in the state of 8.4 per 1,000
                                             in 1996)

                                             Child abuse and neglect rate
                                             of 2 percent after families'
                                             involvement in MIFPI

                                             Out-of-home placement rate
                                             decreased 38 percent during
                           Lead agency model involvement with MIFPI for
                                             children who were in a
                           Wraparound        placement setting at the time
 Michigan                  services for      they entered the project;
                           seriously         decrease of 39 percent for
 Interagency Family        emotionally       children who were not in a
 Preservation Initiative   disturbed         placement setting at the time
 (MIFPI)                   children involved they entered the project
                           with multiple
                           service systems   Children and families
                           at selected sites improved, on average, on all
                                             scales of well-being and
                                             functioning, such as family
                                             and peer relationships,
                                             community involvement,
                                             behavior, school experiences,
                                             and family's adaptation to
                                             caregiving, with the greatest
                                             improvement in lowering
                                             detentions and increasing the
                                             family's adaptation to
                                             caregiving

                                             94 percent of the parents
                                             involved in MIFPI reported
                                             satisfaction with the services
                                             they received
                           Public model
                                             Annual report, July 1998-June
                                             1999
 Tennessee                 Statewide foster
                           care for older
                                             59 percent of the children
 Continuum of Care         children with     discharged were discharged to
 Contracts                 moderate to       their own family, an adoptive
                           severe emotional
                           and behavioral    family, or a less restrictive
                           problems          setting
                           Lead agency model

 Wisconsin                 Family
                           preservation
                                             Outcomes were not provided
 Safety Services Program   services for
                           noncourt families
                           in Milwaukee
                           County
 Local-level initiatives
                                             Evaluation results, 1999

                                             Project children were at least
                                             as safe as children in the
                                             comparison group on risk
                                             indexes such as alcohol and
                                             drug use, abuse against other
                                             children, medical emergencies,
                                             and running away

                                             75 percent of the project
                                             children were residing in a
                                             less restrictive setting; 25
                                             percent of the children could
                                             not be maintained in less
                                             restrictive settings
                                             (comparison data were not
                                             available)

                                             No significant difference in
                                             improvement in children's
                           Lead agency model mental health between the
                                             project and comparison groups
                           Foster care for
                                             Academic performance of
 Alameda County, Calif.    seriously         project children was
                           emotionally
                                             comparable to the comparison
 Project Destiny           disturbed         group on three measures-school
                           children in
                           residential       attendance, conduct reports,
                           treatment in the  and academic improvement;
                           county            however, project children's
                                             academic performance relative
                                             to grade level declined
                                             significantly over time while
                                             the comparison group improved
                                             on this measure

                                             Reduced levels of placement
                                             were not stable for a majority
                                             of the project children-60
                                             percent of the children
                                             experienced two to eight
                                             additional changes in
                                             placement (comparison data
                                             were not available)

                                             Project very nearly reached
                                             its goal of revenue neutrality
                                             by the end of the second year;
                                             between 1997 and 1999, the
                                             project realized a net gain of
                                             2 percent of its capitated
                                             rate
                                             State managed care report,
                                             1999

                                             Confirmed reports of abuse and
                           Public model      neglect decreased 23 percent
                                             over 1998 baseline
 Boulder County, Colo.
                           Foster care for
                                             Finalized adoptions increased
 Integrated Managed        adolescents       13 percent over 1998 baseline
 Partnership for           needing group
 Adolescent Community      care or           Savings were reinvested in
 Treatment (IMPACT)        residential       child welfare services-the
                           treatment in the
                           county            county realized a savings of
                                             less than 1 percent of its
                                             capped allocation from the
                                             state in 1998; the dollar
                                             amount of savings increased
                                             128 percent in 1999
                                             State managed care report,
                           Administrative    1999
                           services
                           organization with Confirmed reports of abuse and
                                             neglect decreased 19 percent
 El Paso County, Colo.     lead agency model over 1998 baseline

 Child Placement Agency    Foster care for   Finalized adoptions increased
 Pilot                     children placed   84 percent over 1998 baseline
                           by Child
                           Placement
                           Agencies in the   Savings were reinvested in
                           county            child welfare services-the
                                             county realized $1.3 million
                                             in savings in 1999
                                             State managed care report,
                                             1999

                                             Confirmed reports of abuse and
                                             neglect decreased 18 percent
                           Public model      over 1998 baseline
 Jefferson County, Colo.
                           All child welfare Finalized adoptions decreased
 Child Welfare Pilot       services in the   23 percent and family
                           county            reunification increased 21
                                             percent over 1998 baseline

                                             Savings were reinvested in
                                             child welfare services-the
                                             county accrued no savings in
                                             1998 and $175,000 in 1999
                                             State managed care report,
                                             1999

                                             Confirmed reports of abuse and
                                             neglect decreased 20 percent
                                             over 1998 baseline
                           Public model
 Mesa County, Colo.                          Finalized adoptions increased
                           All child welfare 118 percent over 1998 baseline
 Child Welfare Pilot       services in the
                           county            Savings were placed in a
                                             reserve account in 1998 and
                                             reinvested in child welfare
                                             services in 1999-the dollar
                                             amount of savings to the
                                             county increased 50 percent
                                             between 1998 and 1999
                                             Outcome evaluation report
                                             covering all Florida
                                             initiatives, 1998-99
                           Administrative
                           services          Placement rate was 69 percent
                           organization with more than the comparison site
                                             (specific to District 4)
 District 4, Fla.          lead agency model
                                             Length of stay was 66 percent
 Privatization Pilot       Foster care and   less than the comparison site
                           independent
                           living services   (specific to District 4)
                           for adolescents
                           in the district   73 percent of the families
                                             served in Districts 4, 8, and
                                             13 combined were satisfied
                                             with the care they received
                                             (similar to comparison sites)
                                             Outcome evaluation report
                                             covering all Florida
                                             initiatives, 1998-99

                                             86 percent of the cases in
                                             Districts 8 and 13 combined
                                             were closed in 1997-98 without
                                             reported recurrence of abuse
                                             or neglect within 1 year of
                                             case closure (similar to
                                             statewide rate)

                                             Placement rate in Districts 8
                                             and 13 combined was 29 percent
                                             less than the comparison sites

                                             Cases were closed at a faster
                                             rate than the public agency
                                             had before the initiative
                                             (specific to District 8)

                                             Average length of stay in
                                             Districts 8 and 13 combined
                                             was 111 days (similar to
                           Lead agency model comparison sites)

                           All children      20 percent of the children in
 District 8, Fla.          needing           Districts 8 and 13 combined
                           protective        were placed with a parent,
 Sarasota County           services, foster  guardian, or relative within
 Privatization Pilot       care, and         15 months of the date of
                           adoption services removal from their home (43
                           in Sarasota and   percent less than the
                           Manatee Counties  comparison sites)

                                             40 percent or more of the
                                             children legally available for
                                             adoption were adopted
                                             (specific to District 8)

                                             77 percent of the children in
                                             Districts 8 and 13 combined
                                             were still in foster care 15
                                             months after removal from
                                             their home (51 percent more
                                             than the comparison sites)

                                             73 percent of the families
                                             served in Districts 4, 8, and
                                             13 combined were satisfied
                                             with the care they received
                                             (similar to comparison sites)

                                             Average case cost in 1997-98
                                             was about 10 percent less than
                                             what the public agency spent
                                             before the initiative
                                             (specific to District 8)
                                             Outcome evaluation report
                                             covering all Florida
                                             initiatives, 1998-99

                                             86 percent of the cases in
                                             Districts 8 and 13 combined
                                             were closed in 1997-98 without
                                             reported recurrence of abuse
                                             or neglect within 1 year of
                                             case closure (similar to the
                                             statewide rate)

                                             Placement rate in Districts 8
                                             and 13 combined was 29 percent
                                             less than the comparison sites

                                             Average length of stay in
                           Lead agency model Districts 8 and 13 combined
                                             was 111 days (similar to
 District 13, Fla.         Children needing  comparison sites)
                           foster care and
 Bridges Program           adoption services 20 percent of the children in
                           in Lake and       Districts 8 and 13 combined
                           Sumter Counties   were placed with a parent,
                                             guardian, or relative within
                                             15 months of the date of
                                             removal from their homes (43
                                             percent less than the
                                             comparison site)

                                             77 percent of the children in
                                             Districts 8 and 13 combined
                                             were still in foster care 15
                                             months after removal from
                                             their home (51 percent more
                                             than the comparison sites)

                                             73 percent of the families
                                             served in Districts 4, 8, and
                                             13 combined were satisfied
                                             with the care they received
                                             (similar to comparison sites)
                           Public model
 Albany County, N.Y.
                           Children needing
                                             Outcomes were not available
 Preventive Services       preventive
                           services in the
                           county
                           Lead agency model
                                             Outcomes were not available.
 Broome County,                              The pilot project has been
                           Children needing
                                             discontinued because of
 N.Y.                      family            problems with implementing new
                           preservation,
                                             financial and service-delivery
 Child Welfare Care        foster care, and  arrangements in accordance
 Management                independent       with federal and state
                           living services
                           at one site       regulations.
                           Lead agency model

                           Wraparound
 Oneida County,            services for
                           seriously
 N.Y.                      emotionally       Outcomes were not available
                           disturbed
 Kids Oneida               children in the
                           county in or at
                           risk of
                           out-of-home
                           placement
                                             Outcome data, 1994-98

                                             Foster care days were
                                             reduced-children admitted and
                                             discharged from the program
                                             avoided staying in foster care
                                             246,834 days since 1994

                                             Children were discharged from
                                             foster care more quickly-76
                                             percent of the children were
                                             discharged from foster care;
                                             half the children who were
                                             placed in foster care since
                                             1994 were discharged in 79
                                             days (decrease of 77 percent
                                             from 1992, 78 percent from
                                             1991, and 75 percent from
                                             1990)

                           Public model      56 percent of the children who
 Onondaga County, N.Y.                       were discharged returned to
                           Children needing  their parents and 33 percent
 Family Support Center     emergency foster  were released to relatives
 Program                   care services in
                           the county        Children had early contact
                                             with their families, where
                                             appropriate-70 percent of the
                                             children visited with a family
                                             member within 72 hours of
                                             placement and 41 percent
                                             visited within 24 hours

                                             88 percent of the children
                                             with siblings were initially
                                             placed with their siblings

                                             20 percent of the children
                                             discharged from foster care
                                             were readmitted (14 percent
                                             less than the overall county
                                             rate)

                                             Educational continuity was
                                             maintained-all school-aged
                                             children attended their home
                                             schools
                                             Program review, 1997

                                             86 percent of the youths were
                                             free of legal involvement,
                                             such as arrests

                                             All families improved their
                                             functioning, such as parenting
                                             skills

                                             All youths improved their
                                             ability to control violent and
                           Lead agency       impulsive behaviors both
                                             inside and outside the home
 Tompkins County,          Wraparound
                           services for      Little effect in reducing
 N.Y.                      youth in          youths' involvement with drugs
                           residential or    and alcohol-17 percent of the
 Youth Advocate Program    institutional     youths improved on this
                           placements in the measure
                           county
                                             School performance varied
                                             among participants but
                                             improved for 55 percent of the
                                             youths

                                             60 percent of the youths had
                                             successful reports from their
                                             employers

                                             88 percent of the youths
                                             improved in their community
                                             involvement
                                             Outcomes report covering
                                             Champaign and Madison
                                             Counties, 1999
                           Public model
                                             All children discharged from
 Champaign County, Ohio    Foster care for   managed care in Champaign and
                           children needing  Madison Counties were
 Human Services/Adriel     out-of-home       discharged to a less
 School                    placement with a  restrictive setting
                           nonrelative in
                           the county        In 1999, 63 percent of the
                                             children who were placed
                                             through managed care did not
                                             reenter a managed care
                                             placement within 12 months
                           Lead agency model

                           Foster care for
                           children placed
 Crawford County, Ohio     outside the
                           county in         Outcomes were not available
 Out-of-County Placement   therapeutic
                           family foster
                           home, group care,
                           or residential
                           treatment
                           Managed care
                           organization      Managed care entity report,
                           model             1998
                                             62 percent of the children who
 Hamilton County, Ohio     Foster care and   had been in a more restrictive
                           independent
                                             setting-such as residential
 TrueCare Partnership      living services   treatment, group home,
                           for children in
                           outpatient mental treatment foster care, or day
                           health and        treatment-were able to remain
                           therapeutic       in a stable, less restrictive
                           placements.       setting after 6 months
                                             Outcomes report covering
                                             Champaign and Madison
                                             Counties, 1999
                           Lead agency model
                                             All children discharged from
 Madison County, Ohio      Foster care for   managed care in Champaign and
                           children in the   Madison Counties were
 Adriel Out-of-Home Care   county needing    discharged to a less
 Placements                nonrelative,      restrictive setting
                           out-of-home
                           placement         In 1999, 63 percent of the
                                             children who were placed
                                             through managed care did not
                                             reenter a managed care
                                             placement within 12 months
                                             Outcome report, Aug. 1997-Aug.
                                             1999

                                             Status offenders-youths with
                                             delinquent behaviors such as
                                             disorderly conduct, fighting,
                                             truancy, possession of
                                             marijuana, and curfew
                                             violation-had fewer contacts
                                             with the courts after
                                             participating in the
                                             initiative: 70 percent of the
                                             youths had no contact, 25
                                             percent had one to five
                                             contacts, and 5 percent had
                                             six or more contacts compared
                                             with before the initiative,
                           Lead agency model when 42 percent had no
                                             contact, 44 percent had one to
                           Wraparound        five contacts, and 14 percent
 Dodge County, Wis.b       services for      had six or more contacts
                           adolescents in
 Family Partnership        child care        Criminal offenders-those
 Initiative                institutions or   youths with delinquent
                           juvenile          behaviors such as theft,
                           corrections in 10 criminal damage to property,
                           counties          burglary, bomb threat,
                                             battery, sexual assault,
                                             receiving stolen property,
                                             possession of a firearm, and
                                             auto theft-similarly had fewer
                                             contacts with the courts: 75
                                             percent had no contact and 25
                                             percent had up to five
                                             contacts compared with before
                                             the initiative, when 72
                                             percent had no contact, 43
                                             percent had up to five
                                             contacts, and 15 percent had
                                             six or more contacts

                                             Truancy rate improved from 25
                                             percent before the initiative
                                             to 15 percent after the
                                             initiative
                                             Quality assurance/improvement
                                             and utilization review report,
                                             second quarter of 1999

                                             Youths spent less time in
                                             residential care during their
                                             first year in the program-the
                                             percentage of days youths were
                                             in residential care decreased
                                             29 percent

                                             Youths spent more time with
                                             their parent-the percentage of
                                             days youths were with their
                                             biological parent increased 32
                                             percent

                                             Children experienced an
                           Public model      overall improvement of 21
                                             percent on measures of
                           Wraparound        behavioral change-such as
 Milwaukee County, Wis.    services for      symptoms of depression,
                           children in the   anxiety, withdrawal, social
 Wraparound Milwaukee      county in or at   problems, delinquency, and
                           risk of           aggressive behavior-at 12
                           residential       months after entry into the
                           treatment         program

                                             Children experienced an
                                             overall improvement of 34
                                             percent on scales of
                                             child-adolescent functioning,
                                             such as their ability to
                                             function adequately at home,
                                             in the community, and at
                                             school; their behavior toward
                                             others; emotional problems;
                                             self-harmful behavior; and
                                             substance abuse, at 12 months
                                             after entry into the program

                                             Average monthly cost of
                                             providing services decreased 8
                                             percent between the first and
                                             second quarters of 1999

aOrganizational arrangements among public and private entities generally
fell into one of the following managed care models: (1) public model, which
maintains the traditional management and service-delivery structure while
the public agency incorporates managed care elements into its own practices
and existing contracts with service providers; (2) lead agency model, where
the public agency contracts with a private entity that is responsible for
coordinating and providing all necessary services-either directly itself or
by subcontracting with a network of service providers-for a defined
population of children and families; (3) administrative services
organization model, where the public agency contracts with a private
organization for administrative services only, and direct services are
structured as in the lead agency or public models; and (4) managed care
organization (MCO) model, where the public agency contracts with a private
organization as in the lead agency model, but the MCO arranges for the
delivery of all necessary services by subcontracting with other service
providers and does not itself provide direct services.

bTen-county initiative includes Columbia, Dodge, Green Lake, Jefferson,
Manitowoc, Ozaukee, Sauk, Sheboygan, Washington, and Winnebago Counties.

(116039)

        Orders by Internet

For information on how to access GAO reports on the Internet, send an e-mail
message with "info" in the body to:

[email protected]

or visit GAO's World Wide Web home page at:

http://www.gao.gov

        Web site: http://www.gao.gov/fraudnet/fraudnet.htm

E-mail: [email protected]

1-800-424-5454 (automated answering system)
  
*** End of document. ***