Statewide Evaluation of Florida’sCommunity-Based Care:

2004 Final Report

Prepared by:Mary Armstrong, Ph.D.

Neil Jordan, Ph.D.Mary Ann Kershaw, B.S.

Amy C. Vargo, M.A.Frances Wallace, M.P.H

Svetlana Yampolskaya, Ph.D.

Submitted to theFlorida Department of Children and Families

June 2004

ii

ACKNOWLEDGEMENTSThe authors gratefully acknowledge the assistance provided by:Deborah AdamsBarbara AlexanderRobert AndersonKaren ArdntSusan ArnettPeter BalitsarisJeanette BakerRobert BarkerKathryn BeasleyAnn BernardMarcie BiddlemanPam BlumenthalLinda BriggsMaggie BurkhalterChris CardSusan ChaseWalter ChildsThomas ChristianJim CoatesCarolyn CrockerPhil DiazPamela DunstonJean ElderBeth EnglanderBob FaginDavid FairbanksKarin FlositzMary Ann FontainePatricia GlassDon HallBen HarrisPeter HerbertLinda HerlockerRoger HinoteSharon HodgesRobert IsraelLee JohnsonMeresa JonesMaxine KaminChristy KaneMonica KingDonna KrauserAmelia LeeTheresa LeslieBecky LyonsJohn LyonsKaren Maziarz

Elwood McElhaneyPam MenendezChristie MooreBarbara MossLuanne PanacekChris PanzarinoC.J. PatelKeith PerlmanDon PolicellaRobyn PoppickAmy PoseyLynn RichardKevin RobertsPat RobinsonGayle RosenkjarDorothy SchwabLisa SchwadronCheri ShefferRay SmithTommy St. AmantStewart StearnsPaul StilesLisa WhiteLarry WilmsSandy WilsonJane WiseHarry YatesChuck YoungColeman Zuber

Additional Research Team Members:Jill JonesRoxann McNeishStephen Roggenbaum

iii

TABLE OF CONTENTS

ACKNOWLEDGEMENTS............................................................................................... iiLIST OF FIGURES......................................................................................................... viLIST OF TABLES ......................................................................................................... viiEXECUTIVE SUMMARY ................................................................................................ 1POLICY RECOMMENDATIONS..................................................................................... 3INTRODUCTION............................................................................................................. 5BACKGROUND.............................................................................................................. 5 Privatization Initiatives Across the United States ........................................... 5RESEARCH QUESTIONS .............................................................................................. 8ROLL OUT MAP........................................................................................................... 11METHODS…................................................................................................................. 13 Organizational Analysis................................................................................... 13 Programmatic Outcomes................................................................................. 13

Quality ............................................................................................................... 14 Cost ................................................................................................................... 14RESEARCH QUESTION 1 ........................................................................................... 15 Introduction ...................................................................................................... 15

Methods ............................................................................................................ 15 Organizational Structure ................................................................................. 16

Complexity ............................................................................................. 16Structure of Provider Networks....................................................................... 23

Utilization of Emerging Promising Approaches to Supervisory Practice and Training...................................................................................................... 29 Child and Family Connection, Inc- Mentoring System ............................. 30

ChildNet- Performance Appraisal ............................................................ 31FamiliesFirst Network (FFN)- Structured Quality Improvement. .............. 32

Heartland for Children (Heartland) -Community Training Needs Assessment ............................................................................................. 32

Hillsborough Kids, Inc. (HKI)- Wraparound Supervision Training ........... 33 Partners for Community Based Care (PCBC)- Supervisory Training....... 33

Recognition and Reduction of Conflicts of Interest...................................... 34Financial Flexibility .......................................................................................... 37Limitations ........................................................................................................ 37Summary of Research Question 1 .................................................................. 38

RESEARCH QUESTION 2 ........................................................................................... 39Introduction....................................................................................................... 39

Methods ............................................................................................................ 39Community as Services Providers.................................................................. 40Limitations ........................................................................................................ 41

Community as Governance Partners ............................................................ 41 Community Alliances .................................................................................... 42

Community Alliance Membership .......................................................... 43 Community Alliance Mission & Discussion of Child Welfare .................. 45

Ability of Alliances to Make suggestions Regarding Lead Agencies ....... 48

iv

Facilities and Barriers to Communication Process with DCF and Lead Agencies.......................................................................................... 100 Other Community Stakeholder Groups.......................................................... 50 Local Revenue Sources................................................................................... 52 Summary of Research Question 2 .................................................................. 53RESEARCH QUESTION 3 ........................................................................................... 54 Introduction ...................................................................................................... 54 Lead Agency Case Staffing Structures .......................................................... 57 Methods.................................................................................................. 57 Programmatic Outcomes.................................................................................. 61

Indicators Calculated for Current Evaluation .......................................... 62 Source of Data ....................................................................................... 62 Methods................................................................................................... 62

Limitations ........................................................................................................ 64 Reentry into Out-of-Home Care....................................................................... 64

Description of the Indicator ...................................................................... 64 Results ............................................................................................ 65 Proportion of Children with Recurrence of Maltreatment ............................ 66 Description of the Indicator..................................................................... 66

Results............................................................................................. 67Proportion of Children Who Exited Out-of-Home Care Within 12Months............................................................................................................... 69

Description of the Indicator ...................................................................... 69 Results............................................................................................. 70

Proportion of Children Returned to Parents or Legal Guardians and Placement with Relatives After Exiting Out-of-Home Care .................... 72

Description of the Indicator ...................................................................... 72 Results: Children Returned to Parents After Exiting Out-of- Home

Care ................................................................................................. 73 Results: Children Placed With Relatives After Exiting Out-of- Home Care ...................................................................................... 76

Children With Adoption Finalized After Exiting Out-of-Home Care…................................................................................................................ 77

Description of the Indicator ....................................................................... 77 Results............................................................................................. 77

Summary of Research Question 3 ................................................................. 79RESEARCH QUESTION 4 ........................................................................................... 80 Introduction ..................................................................................................... 80 Analysis of CBC Expenditure for Direct Child Protective Services ............ 81 Methods................................................................................................... 81 Findings................................................................................................... 83

Conclusions and Limitations .................................................................... 87Analysis of Lead Agency Appropriations and Expenditures........................ 89 Methods................................................................................................... 89 Findings ................................................................................................... 89

v

Conclusions and Limitations ...................................................................... 91 Financial Risk .................................................................................................. 92 Risk Sharing .............................................................................................. 93 Summary of Research Question 4 .................................................................. 96CONCLUSION: RESEARCH QUESTION 5 ................................................................. 97SUMMARY.................................................................................................................... 97 Additional Lessons Learned ........................................................................... 99POLICY RECOMMENDATIONS................................................................................. 100NEXT STEPS.............................................................................................................. 101REFERENCES… ........................................................................................................ 105

APPENDIX 1: Review of Privatization Initiatives across the United States.......... 107APPENDIX 2: Lead Agency CEO Protocol .............................................................. 121APPENDIX 3: Sample Letters to Agencies Soliciting Promising Approach toSupervisory Training/practice 124APPENDIX 4: Community Alliance Chair Protocol................................................. 126APPENDIX 5: Community Alliances: Responsibilities, Duties, Members and FocusOutcome Areas.......................................................................................................... 127APPENDIX 6: Type of Child Protection Staffings .................................................. 128APPENDIX 7: Statistical Results.............................................................................. 129APPENDIX 8: Direct Services and OCA’s by Services Category (FY02-03)Out-of-Home Services 135APPENDIX 9: Total Expenditures for district Specific Projects by District andFiscal Year ................................................................................................................. 139APPENDIX 10: Proportion of Each Study Year Data were Allocated to CBC Status,by CBC Site ............................................................................................................... 140

vi

LIST OF FIGURESFigure 1. Status of CBC Implementation as of April 2004 ............................................ 11Figure 2. Example of Low Vertical Differentiation: Heartland for Children OrganizationalChart ............................................................................................................................. 19Figure 3. Example of High Vertical Differentiation: ChildNet Organizational Chart....... 20Figure 4. Model of Provider Structure with Parent/Partner Organizations .................... 24Figure 5. Model of Provider Structure without Parent/Partner Organizations ............... 25Figure 6. Model of Provider Structure with Service Centers ......................................... 26Figure 7. Percentage of Children who Exited Out-of-Home Care in FY01-02and Reentered Within 12 Months After their Exit by Lead Agency Service ContractCategory ...................................................................................................................... 66Figure 8. Percentage of Children who Experienced Recurrence of Maltreatment Within12 Months after their First Episode Based on FY01-02 Cohort by Service ContractCategory........................................................................................................................ 69Figure 9. Percentage of Children who Entered Out-of-Home Care in FY01-02 andExited Within 12 Months by Service Contract Category ................................................ 72Figure 10. Percentage of Children Reunified with Parents, Placed with Relatives andwith Finalized Adoptions Based on FY02-03 Cohorts by Lead Agency......................... 74Figure 11. Percentage of Children Reunified with Parents After Exiting Out-of-HomeCare by Site, by Time and Site by Time ........................................................................ 75Figure 12. Percentage of Children Placed with Relatives After Exiting Out-of-HomeCare by Site, by Time and Site by Time ........................................................................ 77Figure 13. Percentage of Children with Adoptions Finalized After Exiting Out-of-HomeCare by Site, by time and Site by Time ........................................................................ 78Figure 14. Average Expenditures For Direct Child Protective Services Per ChildServed: CBC Sites vs. Non-CBC Sites.......................................................................... 84Figure 15. Average Expenditure for Direct Child Protective Services Per Child Day inSystem: CBC Sites vs. Non-CBC Sites ......................................................................... 85Figure 16. Average Expenditure for Direct Child Protective Services Per Capita: CBCSites vs. Non-CBC Sites ............................................................................................... 86Figure 17. Direct Child Protective Services Expenditure by Type of Service (FY02-02)-CBC Sites and Non-CBC Sites ..................................................................................... 87

vii

LIST OF TABLESTable 1. Facilitators Common to the Implementation of PrivatizationInitiatives Across the United States ................................................................................. 7Table 2. Frequent Challenges to the Implementation of PrivatizationInitiatives Across the United States ................................................................................. 7Table 3. Research Questions ......................................................................................... 9Table 4. Lead Agencies and Counties Included in the Evaluation ................................ 12Table 5. Research Question 1: Evaluation Questions, Indicators/Analyses,and Sources ................................................................................................................. 15Table 6. Characteristics of Organizational Structures: Strengths.................................. 21Table 7. Lead Agency and Network Provider Communication ...................................... 28Table 8. Paradigm Shift in Supervision ......................................................................... 33Table 9. Board Membership by Organizational Affiliation ............................................. 35Table 10.Types of Potential Conflict Situations............................................................. 36Table 11. Research Question 2: Evaluation Questions, Indicators/Analyses,and Sources .................................................................................................................. 39Table 12. Average and Range of Providers and Contacts by Length ofLead Agency Service Contract ...................................................................................... 41Table 13. Community Alliance Membership Professional Roles ................................... 44Table 14. Facilitations and Barriers to Communication................................................. 50Table 15. Other Community Stakeholder Groups ......................................................... 51Table 16. Local Revenue Sources ............................................................................... 53Table 17. Research Question 3: Evaluation Questions, Indicators/Analyses,and Sources .................................................................................................................. 55Table 18. Lead agency Case Staffing Structures (Spring 2004) ................................... 57Table 19. Florida DCF Reports Based on HSn ............................................................. 61Table 20. Lead Agencies by Length of Service Contract .............................................. 63Table 21. Percentage of Children who Exited Out-of Home Care inFY01-02 and Reentered Within 12 Months by Lead Agency......................................... 65Table 22. Percentage of Children who Experienced Recurrence of MaltreatmentWithin 12 Months After Their First Episode Based on FY01-02 Cohort byLead Agency ................................................................................................................. 67Table 23. Percentage of Children Who Entered Out-of-Home Care inFY01-02 and Exited Within 12 Months by Lead Agency ............................................... 71Table 24. Research Question 4: Evaluation Questions, Indicators/Analyses,and Sources .................................................................................................................. 80Table 25. Lead Agency Budget Vs. Actuals (FY02-03)................................................. 89Table 26. Lead Agency Budget Vs. Actuals by Funding Source-FCP(FY02-03) ...................................................................................................................... 90Table 27. Lead Agency Budget Vs. Actuals by Funding Source-FFN(FY02-03) ...................................................................................................................... 90Table 28. Lead Agency Budget Vs. Actuals by Funding Source-HKI(FY02-03) ...................................................................................................................... 90Table 29. Lead Agency Budget Vs. Actuals by Funding Source- PCBC(FY02-03) ...................................................................................................................... 91Table 30. Lead Agency Budget Vs. Actuals by Funding Source- Sarasota YMCA

viii

(FY02-03) ..................................................................................................................... 91Table 31. Lead Agency Budget Vs. Actuals by Funding Source-All Lead Agencies Combined (FY02-03) ....................................................................... 91Table 32. Financing Issues Faced During National Privatization Initiatives .................. 94Table 33. Funding Challenges Faced by Lead Agencies.............................................. 95Table 34. Self Reported Lead Agency Outcomes ....................................................... 100

1

EXECUTIVE SUMMARY

Throughout our country, states and communities are struggling with the complexchallenges of ensuring the safety and well-being of children in the child welfare system,and many states, including Florida, are experimenting with various types of privatizationof child welfare services. This evaluation report examines the status of Community-Based Care (CBC) in Florida, with a special focus on quality, cost and child and familyoutcomes. The time period covered by this evaluation (FY03-04) was a year of rapidgrowth for CBC, resulting in 11 lead agencies in 28 counties serving over half of thechildren in Florida’s child welfare system, and additional counties in the start-up phaseas the year progressed.

The evaluation is organized around a set of research questions regarding theeffectiveness of Community-Based Care. The first area of CBC examined was theeffectiveness of lead agencies at designing and improving systems and servicesfor child protection. This question was addressed through an analysis of theorganizational structure of lead agencies, their provider networks, and boards ofdirectors, and an inventory of promising approaches in supervision training and practice.An examination of organizational complexity revealed that all lead agencies haddeveloped 4-5 similar/comparable departments or divisions, indicating consensusregarding the types of specializations needed to administer Community-Based Care.The degree of vertical differentiation varied, with 4 lead agencies averaging 3 personsfrom the lowest level of the organization to the CEO, and 3 lead agencies averaging 4.6persons. Greater vertical differentiation indicates a need for more process andcommunication protocols and standards.

The analysis of provider networks revealed three models of organizational structures: astructure that includes parent or partner organizations, a more traditional providernetwork without parent/partner organizations, and a model that uses geographic servicecenters. Many examples of re-design were evident in the early implementation stages atboth the lead agency and provider network levels, indicating that the CBCs were willingto revise initial plans when there were indicators that systems and services were notoptimal. Potential conflicts of interest through interlocking boards were indicated forseveral lead agencies by the inclusion of contracted providers on boards of directors. Asecond potential problem was the lack of diversity in board membership, with underrepresentation from community members, foster parents, and nonprofit organizationsbeyond contracted providers. This lack of diversity could result in a lack of communityinvolvement and ownership, and may prevent infusion of new local revenues and/orinnovative services.

Six of the 11 lead agencies submitted promising supervisory approaches, indicating thatseveral sites are trying to incorporate innovation in supervisory training and practice.There was also great variability in the types of approaches across sites, anotherindicator of individualized efforts to design effective systems and services.

2

The effectiveness of CBC at involving the community as service partners andresource contributors was the second area of study for the report. Lead agenciesoperating 18 months or more had a smaller average number and range in number ofproviders and contracts than agencies operating for shorter periods. This reduction inproviders and contracts may be related to a trend towards stabilization and maturity ofthe service system, although this area will continue to be tracked over time. Leadagencies reported relatively little success in generating local revenues, especially afterthe start-up phase.

In addition to the lead agency boards of directors, Community Alliances representanother potential community governance partnership for lead agencies. Of the 7Community Alliances who responded to the study’s survey, only 3 had representativesfrom mental health or the business community; only 2 included local foundations orfoster parents, and only 1 Community Alliance had substance abuse or consumerrepresentatives. The majority of Community Alliances reported that child welfare is astanding agenda item but several felt limited by their “advisory role” and expresseddesire for more involvement in local system of care development. An unresolved issuefor lead agencies is clarification of the governance roles of boards of directors,Community Alliances, and other community stakeholder groups.

The third area of CBC effectiveness reviewed was the ability to meet the outcomes ofchild safety, permanency, and well-being. A review of the various staffing modelsused by the lead agencies indicated the high importance placed on permanency.Permanency staffings typically were held frequently. In addition, many examplesillustrate unique staffing practices that allow lead agencies to deal with routine andemergent issues while maintaining a focus on safety, permanency, and child well-being.

Two indicators were examined for child safety and well-being: the rate of recurrence ofmaltreatment and re-entry into out-of-home care. The highest probability of re-entry intoout-of-home care was found in counties where CBC had been operating for 7-17months; this rate was significantly lower for CBC counties operating for 18 months ormore. Rates of recurrence of maltreatment were lowest for counties just prior to CBCimplementation; no differences were found between the categories of 7-17 months, and18 months or more of CBC operation.

Four indicators were used to examine the permanency outcome. Findings were mixedacross the 3 CBC length of service contract categories for the proportion of childrenexiting out-of-home care in less than 12 months, and no significant differences werefound. For the proportion of children returned to parents, there was a decrease forcounties with 18 months or more of operation, and a slight increase for the 7-17 monthscategory. For the indicator of percent of children placed with relatives, the proportionincreased over time, with a significant difference in the increase for CBC counties with18 months or more of operation. The highest increase in the proportion of children withadoption finalized was in the counties just prior to CBC implementation, indicating thatthe lead agencies have challenges with meeting the permanency goal of adoption.

3

The evaluation also examined the effectiveness of Community-Based Care atmanaging resources and costs efficiently. A comparison of CBC and non-CBCcounties found that the average spending per child for child protection services hasbeen lower for CBC counties than for non-CBC counties for the past three years. Onereason may be attributed to differences in the service mix. During FY02-03, CBCsdevoted a smaller proportion (54%) of resources to out-of-home care than non-CBCs(59%).

A comparison was made of actual lead agency expenditures with lead agencyappropriation amounts, and whether the differences varied by funding source. DuringFY02-03, every lead agency with a service contract spent within 3% of the total dollarsappropriated. There was considerable variation by funding source, with all 5 leadagencies spending fewer state dollars than appropriated, but spending more Title IV-Efunds than appropriated. Several additional funding challenges were reported by leadagencies, including inadequate statewide funds for adoption subsidies and the Road toIndependence scholarships, the funding allocation formula versus actual casecomposition, and overall inadequate funding levels.

POLICY RECOMMENDATIONS

• Several study findings indicate that lead agencies may need additional supportand technical assistance during the mid-implementation phase (7-17 months) sothat the lead agencies do not lose their start-up momentum and communitysupport as they struggle to solidify their organization’s infrastructure.

• Lead agencies should examine the composition of their boards of directors inrelationship to interlocking boards and lack of diversity (e.g., foster parents,children who have aged out of the foster care system, and other communitymembers).

• Community Alliances are potential community governance partners, butclarification is needed regarding their authority, and their role vis-à-vis the leadagency boards of directors and other community stakeholder groups.Legislatively appropriated incentives for Alliance membership and engagement inchild welfare issues should be considered.

• More direct communication is encouraged between DCF Central Office and theCommunity Alliances, for example, by legislative updates on child welfare relatedbills, so that the Alliance members can voice their opinion before items arepassed/vetoed.

• The lead agencies and DCF partnership should continue to address fundingissues, including adequacy of funding levels, and finding appropriate allocationmixes.

4

• Lead agencies may benefit from consultation and technical assistance regardingevidence-based and promising practices for the permanency goal of adoption.

• Lead agencies should be provided with support to advance their emergingpromising approaches to supervisory practice through technical assistance andthrough peer review of their processes.

• Lead agencies should be encouraged to continue developing individualizedpatterns of staffings so long as they meet statutory requirement. In addition tomaintaining a focus on child well-being, staffings also create opportunities forsupervision and inclusion of community stakeholders at the child and family level.

5

INTRODUCTION

In fiscal year (FY) 2003-04, the Department of Children and Families (DCF) contractedwith the University of South Florida Louis de la Parte Florida Mental Health Institute(FMHI) to conduct an evaluation of the 11 lead agencies and 28 counties in whichCommunity-Based Care (CBC) was operational including: Child and FamilyConnections (CFC), ChildNet, Inc., FamiliesFirst Network (FFN), Family ContinuityPrograms (FCP), Family Support Services of North Florida, Inc. (FSS), Heartland forChildren, Inc. (Heartland), Hillsborough Kids, Inc. (HKI), Partnership for Families, Inc.(PFF), Partners for Community-Based Care (PCBC), YMCA Children, Youth & FamilyServices, Inc. (Sarasota YMCA), and United for Families, Inc. (UFF). The CBCEvaluation Plan has been designed in keeping with the legislative intent for CBC(s. 409.1671, F.S.), which is to assess quality of service, programmatic outcomes, andcost-efficiency of the CBC sites.

Florida’s CBC initiative is still quite new in many counties. This evaluation focuses onearly outcomes and the strengths and challenges of privatization as it has beenimplemented in Florida thus far. This report’s goal is to provide policymakers withconcrete information and recommendations about next steps and mid-coursecorrections.

BACKGROUND

Privatization Initiatives Across the United States

Ensuring the safety and well-being of children in the child welfare system has proven tobe a long-term and complicated national problem, as noted in a recent national reportby the Urban Institute (Malm, Bess, Leos-Urbel, Geen, & Markowitz, 2001). Ongoingissues identified in the report included: (1) escalating costs; (2) more families comingunder care with more severe problems (e.g., substance abuse, family violence, and anincreased severity of abuse and neglect); (3) increased public scrutiny of child welfareagencies; and (4) litigation against the system. Fragmented and uncoordinated serviceswith little accountability to the state and the local community were also commonplace,as well as a chronic under-funding of most state systems.

The history of privatization efforts includes both major successes and major failures(Gibelman & Demone, 1998). While there has been an ongoing broadening of the kindsof services being privatized, the privatization of human services represents a relativelysmall percentage of all privatization efforts but has proven to be highly problematic.Unfortunately, there have been very few empirical examinations of these privatizationefforts, which make the evaluation of such programs even more important (Gibelman &Demone, 1998). Furthermore, a recent study of child welfare privatization activities bythe Child Welfare League of America showed that privatization has not led to lowercosts (CWLA, 2000). Part of the reason for this may be that child welfare systems ingeneral are chronically under-funded so there is little, if any, room for cost savings.Local circumstances and the implementation processes (the issues detailed in the

6

policy framework described below) have been key factors in the success and failure ofsuch efforts.

The following information describes themes found in child welfare privatization initiativesin Kansas, Arizona, Colorado, Michigan, Missouri, South Carolina and Ohio (for a morethorough description see Appendix 1). Although many states have long beencontracting out for some of their child welfare services, this has traditionally been basedon a fee-for-service model, which has been criticized as misaligning incentives andleading to prolonged lengths of stay for children. Under new child welfare adaptations ofmanaged care strategies and contracts with private, non-profit, community-basedorganizations, the goal is creative reimbursement strategies that would improveoutcomes for children and families. However, new problems with financial risk, clarity ofroles, and shared accountability have often resulted.

As part of this trend towards privatization and CBC, states and localities are increasinglyturning to a model where a single agency is charged with coordinating and providing allservices. The intent of this model is to reduce the need for families to negotiate a mazeof individual agencies, improve the likelihood that there is a match between needs andservices, increase access to services, and assume that families will be more acceptingand trusting of local community agencies than services run by the state (McCullough &Schmitt, 2000; U.S. General Accounting Office, 2000). This lead agency design hasbeen the most common approach of state governments in the child welfare field(McCullough & Schmitt, 2000).

Although each initiative is unique to its own local context and circumstances, there areseveral key implementation themes present throughout (see Tables 1 and 2), such asthe popularity of a lead agency design, importance of start-up funds and a phasedapproach to implementation, as well as the critical nature of community involvement inthe planning and quality improvement processes. A shared vision, agreement onperformance standards, and consistency in leadership were all listed as facilitators toprivatization efforts, while problems with reimbursement systems and tracking data werenot unique to any one agency.

7

Table 1. Facilitators Common to the Implementation of Privatization InitiativesAcross the United States

Facilitators to Implementation• Lead agency model, with some variations• Commitment and consistency in both

leadership and mission• Role and responsibility clarification• Start-up time and funding in order to build

agency infrastructure and capacity• Staged or gradual implementation• Contracts designed to encourage competitive

bidding• Monitoring visits and activities are combined

with other audits• Community and client involvement throughout

planning and implementation

Table 2. Frequent Challenges to the Implementation of Privatization InitiativesAcross the United States

Barriers to Implementation• Cost efficiency and effectiveness are not yet

well-established outcomes• Cost savings are rarely achieved• It remains difficult to define and agree on

indicators, as well as to measure actualperformance

• Inadequate information systems for datatracking and program improvement

• Service capacity in more rural areas• Contract requirements that prevent competition• Monitoring that is either overdone or underdone• Most fiscal issues (e.g., determining case rates,

trying to protect agencies from deficits)

Throughout many of these initiatives there is a common desire to hold agenciesaccountable for their performance, and the resulting outcomes for children and families.This is being accomplished through both standards of performance and care, as well asfinancial incentives (Sells, 2001). Many of the agencies are reimbursed by theirrespective states via a case rate or capitated system in which fees are predeterminedand not increased based on length of stay or severity of case. The rationale is that if aprivate agency can improve outcomes, decrease number of children in care anddecrease length of stay in out of home care, increase permanency of placements, as

8

well as ensure successful adoptions, the agency can retain the savings. Likewise, if anagency does not do well, they will soon be operating at a deficit (Geen & Tumlin, 1999).

An increased amount of contact and communication between privatized agencies andtheir clients has also been noted, as well as the advantage of identifying risks andpreventing further problems that come with increased engagement of families (Geen &Tumlin, 1999). This is particularly encouraging given that the foster care population hasincreased significantly across the country (specifically, an increase of 25,000 childrenentering foster care from 1996 to 1999) due to legislation passed in many states thatincreased awareness and reporting of child abuse (Snell, 2000).

While some child welfare privatization efforts did originate as Legislative mandates(e.g., Florida and Colorado), and others were prompted by positive feedback from theKansas and Florida initiatives, many simply sought to relocate services andaccountability into local communities in the hope of improved outcomes (Geen &Tumlin, 1999). It is clear, however, that efforts at privatization are a response to publicchild welfare systems that have for years been criticized as failing children in their care.Nationwide, many child welfare constituents are encouraging privatization to someextent. For example, Sells states, “After reviewing the various governmental failures atimproving the quality of children’s lives through child welfare services and attemptedreform efforts, only one conclusion can be drawn: that extensive governmentinvolvement is not alleviating the problems, it is causing them” (2001: p.15). Whetherprivatization will make a long-term difference in the lives of children can only bedetermined after longitudinal study of various initiatives and the performance data thatlead agencies are only beginning to track.

RESEARCH QUESTIONS

The following research questions are the guiding framework for the evaluation andserves as the foundation for the organization of the report:

How effective is Community-Based Care at:

1. Designing and improving systems and services for child protection?2. Involving the community in child protection, both as service

partners and as resource contributors?3. Meeting the Adoption and Safe Families Act (AFSA) outcome

requirements for child safety, permanency, and well-being?4. Managing all resources and costs efficiently?5. Identifying and meeting the needs of the families and children who

have been maltreated?

The following table (Table 3) details the first four research questions in this evaluation,as well as the evaluation questions related to these research questions. The indicators

9

that were used to answer these questions and the main data sources for each indicatorcan be found in an expanded version of the table under each research question. Thefifth research question serves as the concluding question of the report and includes asummary of the data presented in the first four questions.

Table 3. Research Questions

Research Question Evaluation Question

How are the lead agencies organized?

How are the provider networks structured and what interaction takes placebetween the lead agency and provider?

Are there qualitatively different approaches to supervision that impact thequality of service delivery to children and families?

How effective isCommunity-BasedCare at designingand improvingsystems andservices for childprotection?

Do the lead agencies have financial flexibility?

How are local community resources being developed and utilized?

Are community services being accessed by the lead agencies?

How effective isCommunity-BasedCare at involving thecommunity in childprotection both asservice partners andresourcecontributors?

What types of community governance boards support the lead agency?

Are there mechanisms in place at the lead agency level to assurecontinued focus on child safety, permanency and well-being?

What proportion of children exit out-of-home care during first 12 monthsafter entry?

Is there a significant difference between the proportion of children exitingout-of-home care for CBC by lead agency length of service contractcategory?*What proportion of children re-enter out-of-home care within 12 monthsafter exiting their first episode of out-of-home care?

How effective isCommunity-BasedCare at meetingAFSA outcomerequirements forchild safety,permanency, andwell-being?

Is there a significant difference between the proportion of childrenreentering out-of-home care in lead agency length of service contractcategory? *

10

What proportion of children experience reoccurrence of maltreatmentwithin 12 months of their first episode of maltreatment?

Is there a significant difference between the proportion of children withrecurrence of maltreatment by lead agency length of service contractcategory? *What is the proportion of children in out-of-home care who were reunifiedwith their parents?

Is there a significant difference between the proportion of children reunifiedwith parents by site, time (FY01-02 and FY02-03) and site x time?

What proportion of children are placed with relatives for out-of-home care?

Is there a significant difference between the proportion of children placedwith relatives for out-of-home care by site, time (FY01-02 and FY02-03)and site x time?

What proportion of children have a finalized adoption?

Is there a significant difference between the proportion of children withfinalized adoptions by site, time (FY01-02 and FY02-03) and site x time?

How do average expenditures per child for direct child protective servicesin the CBC sites pre- and post-CBC compare with expenditures in the non-CBC counties?

How do actual lead agency expenditures compare with lead agencybudget amounts? Does the difference between expenditures and budgetamount (i.e., variance) differ by funding source?

How effective isCommunity-BasedCare at managing allresources and costsefficiently?

How do lead agencies handle financing issues and financial risk?

*Note: See page 12 for lead agency length of service contract categories.

When comparing the results of the previous evaluations to the current report, it isimportant to remember that the pace of implementing CBC has accelerateddramatically. As illustrated in Figure 1, every district is now involved in some phase ofthe CBC implementation process. This creates a very different environment in whichpolicy and jurisdictional issues need to be addressed, compared to prior years whenonly a few lead agencies were operational.

11

ROLL-OUT MAP

Figure 1. Status of CBC Implementation as of April 2004

Available online at: http://www.dcf.state.fl.us/cbc/docs/cbcstatusmap.pdf

In previous years’ evaluations it has been noted that although more CBC sites were invarious phases of operation and were seen as very important by DCF, they were stillonly a small part of the agency’s overall operation. Consequently, policies, procedures,and structures have in the past evolved as the need arose through a negotiated process(Paulson, Armstrong, Fitzpatrick, Jordan, Kershaw, Reyes, Vargo, and Yampolskaya2003). However, during FY03-04 there has been a major transformation in DCF’sapproach to the implementation of CBC, and a more organized “roll-out” plan suited tosuch a statewide effort has been created.

The following table (Table 4) lists the lead agencies (and counties) included in thisevaluation along with the number of maltreatment reports and children served by eachlead agency.

Wakulla

Leon

Franklin

Liberty

GadsdenJackson

CalhounBay

WashingtonHolmes 2

Flagler

Volusia 12

DuvalBaker

Nassau

Clay St.Johns

4

Lake

Marion

SumterHernando13

Palm Beach

Broward

Hendry

Glades

Lee

Collier8Dade

Monroe

Hardee

Polk

Highlands

147Osceola Br

evard

Martin

Okeechobee

Indian River

St.Lucie

Escambia

WaltonOkaloosaSanta

Rosa

15

910

11

3

Putnam

Columbia

AlachuaDixieGilchrist

LafayetteBradford

Levy

Suwannee

Union

Hamilton

Manatee

Pasco

Pinellas

De SotoSarasota

Hillsborough

SunCoastRegion

Service Contracts

SunCoast Region District 14 (Phasing in services)District 01 District 8 (Phasing in services)District 12 St. Johns CountyDuval County Nassau CountyDistrict 10 Clay/Baker CountiesDistrict 15 Orange/Osceola CountiesDistrict 9 (Phasing in services) District 13 (Phasing in services)District 2a (Phasing in services) District 2B (Phasing in services)

Start-Up ContractsSeminole CountyBrevard CountyDistrict 3District 11

12

Table 4. Lead Agencies and Counties Included in the EvaluationLead Agency(counties served)

Number of MaltreatmentReports FY02-03 (duplicatedcount)

Number of ChildrenServed FY02-03(duplicated count)

Service Contract held for 18 months or morePartners forCommunity-BasedCare(Volusia, Flagler) 12,995 2,900Sarasota YMCA(Sarasota, DeSoto,Manatee) 13,045 2,088Family ContinuityPrograms(Pasco, Pinellas)

27,781 5,784

Service Contract held 7-17 monthsHillsborough Kids,Inc.(Hillsborough) 23,345 6,437FamiliesFirst Network(Escambia, Walton,Okaloosa, Santa Rosa) 17,349 5,089

Service Contract held 6 months or less*Child and FamilyConnections (PalmBeach) 20,449 3,737ChildNet(Broward) 20,618 6,532Family SupportServices of NorthFlorida(Duval) 21,585 4,618Heartland for Children(Polk, Hardee, Highlands) 17,744 5,260Partnership forFamilies(Holmes, Washington, Bay,Gulf, Calhoun,Jackson)

8,610 1,763

United for Families(Okeechobee, St. Lucie,Indian River, Martin) 9,429 2,957*Note: These lead agencies did not begin the implementation of Community-Based Care during FY02-03but had service contracts for 6 months or less as of January 2004.

13

METHODS

The following is a summary of the methods used for the organizational analysis,programmatic outcomes, quality of services, and cost data presented in this report. Adetailed explanation of methods used can be found under each research question.

Organizational Analysis

For the organizational component, 11 lead agency CEOs were asked to complete a 30question survey covering: (1) community governance (board of director issues andcommunity participation), (2) financial risk (shared risk arrangements), (3) differences inlead agency service systems, and (4) lessons learned in CBC implementation (seeprotocol, Appendix 2). The CEOs were also asked to provide any available visualrepresentations and documents related to the above areas. In addition, the project team(a) asked each Alliance Chair and Vice Chair to complete a survey regarding examplesof connectedness to lead agencies and Community-Based Care, and (b) Alliancemembership to illustrate the relationships between the Community Alliance and the leadagency.

Programmatic Outcomes

In keeping with AFSA standards of safety, permanency and well-being, six quantitativeindicators were used in this evaluation to measure child safety, permanency and well-being:

(c) Proportion of children exiting out-of-home care,(a) Rate of re-entry into out-of-home care,(b) Rate of recurrence of maltreatment,(c) Rate of reunification with parents,(d) Rate of custodial placement with relatives, and(e) Proportion of children with finalized adoption.

Analysis of these indicators utilized entry or exit cohorts of children as opposed togroups of children who were in care at a specific point-in-time thus reducing selectionbias. The above mentioned indicators were calculated using information fromHomeSafenet (HSn) for FY01-02 and FY02-03 to collect data for lead agencies andcompare lead agency length of service contract categories (based on the length of timea lead agency has held their service contract) during this time period. Lead agencyservice contract categories include:

1) Lead agencies that have held a service contract for 18 months or more,2) Lead agencies that have held a service contract for 7-17 months, and,3) Lead agencies that have held a service contract for 6 months or less.

14

Statistical methods included survival analysis of time-to-event indicators and logisticregression for point-in-time indicators.

Quality

For this evaluation, quality of services was examined in two areas: (1) recognition andutilization of promising approaches to supervisory practice and training, and (2) leadagency case staffing structures. The CEOs of 11 lead agencies were contacted to solicitpromising approaches to supervisory training and practice with follow-up by theevaluators for clarification purposes. To gather data concerning case staffing structures,six lead agencies were contacted and asked to complete the Types of Child ProtectionStaffings table (Appendix 6) as related to safety and permanency. (Note: Five leadagencies had previously been contacted during the first half of FY03-04.)

Cost

For this evaluation, cost data were collected for the following analyses: (1) directservices expenditures, and (2) comparison of lead agency budget amounts with leadagency expenditures for FY02-03. Direct services expenditures were analyzed in eachDCF district and included those expenditures at the district level for out-of-home care,in-home services, services related to adoption and other child protective services thatdo not fit in the previous categories. DCF accounting data from the Florida Accountingand Information Resource (FLAIR) were used to calculate child protective servicesexpenditures, with the Office of Revenue Management generating spreadsheetscontaining expenditures by District and Other Cost Accumulator (OCA) for BudgetEntities 60600700, 60910303 and 60910304.

15

RESEARCH QUESTION 1:How effective is Community-Based Care at designing and improving systems andservices for child protection?

Introduction

In order to examine the effectiveness of CBC at designing and improving systems andservices for child protection, this research question includes the following evaluationquestions, indicators and data sources (Table 5):

Table 5. Research Question 1: Evaluation Questions, Indicators/Analyses, andSource

Evaluation Questions Indicator/Analyses Source

How are the lead agenciesorganized?

-Analyses of organizationalcharts of CBC lead agencies

-Reported implementationsuccess/failures attributableto organizational structure

-Description of potentialconflicts of interest

Lead AgencyDocumentation

CEO Survey

How are the provider networksstructured and what interaction takesplace between the lead agency andprovider?

- Description of organizationalstructure of provider network

- Analysis of communicationand interaction patternsbetween lead agency andproviders

CEO Survey

Lead AgencyDocumentation

DistrictAdministratorMaterials

Are there qualitatively differentapproaches to supervision that impactthe quality of service delivery tochildren and families?

- Description of promisingapproaches to supervisorypractices and training asidentified by lead agencies

Lead AgencyCommunication

How effectiveisCommunity-Based Careat designingand improvingsystems andservices forchildprotection?

Do the lead agencies have financialflexibility?

- Description of fundingissues CEO Survey

Methods

For the organizational component, team members began by contacting the 11 leadagency CEOs to explain the purpose of the email survey. The CEOs were then e-maileda 30 question survey that covered: (1) community governance (board of director issues

16

and community participation); (2) financial risk (shared risk arrangements); (3)differences in lead agency service systems; and (4) lessons learned in CBCimplementation. As part of this protocol (see Appendix 2), lead agency CEOs wereasked to identify their Boards of Directors members and to describe their roles as wellas any potential conflicts of interest encountered in the board member selectionprocess.

Each lead agency CEO was also asked for a visual representation of communitygovernance, financial risk, and their service delivery model, or any documents theyalready had on these topics that they were willing to share. The project team used thisin conjunction with the survey responses. Organizational models and charts previouslyreceived from lead agencies and Central Office staff were also included in the analysis.In addition, a request was made for a copy of each lead agency’s ITN, a list of providernetwork contracts, and a sample of those contracts.

Content analysis of the documents and open-ended survey questions was used toanalyze the qualitative data collected for this study. Content analysis involves reviewingqualitative data to identify common themes and trends. The primary goal of contentanalysis is to condense a large amount of qualitative data into a list of variables that canbe examined for correlations, patterns and themes.

Organizational Structure

Overall, organizational structure is the platform for all the organizational activities anddecision-making, as well as the framework that determines how well the organizationalgoals and outcomes are met (Hall, 1996). Understanding the structure of anorganization allows for a better picture of the “daily” practices and procedures of anorganization and the barriers or facilitators that influence those practices andprocedures. Analysis of an organization begins with a description of the structure itselfand the process/patterns that make up the organization. For this report, the preliminaryanalysis of the lead agencies begins with a high level description of their organizationalstructure, as depicted through their organizational charts and responses to the CEOsurvey. As the organizational analysis of the lead agencies continues and expands,deeper levels of analysis will be examined to create a holistic picture of the leadagencies with a focus on themes and patterns across agencies.

Complexity

While there are multiple components used to describe organizational structure(complexity, formalization, and centralization), this report will address the component ofcomplexity. Knowing the level of complexity of an organization is important because itcan dictate how communication and interaction occur throughout the organization, theamount of effort, administration, and standardization needed to perform the activities ofthe organization, the behavior of the employees, and the organization’s relationship toexternal environments. Complexity is positively correlated with the size of the

17

organization, the number of position titles, and the number of departments or sections inan organization (Hall, 1996).

The degree of complexity in an organization is measured by the amount of horizontaldifferentiation, vertical differentiation, and spatial dispersion (Fitzgerald, 2002) (seeFocus On: Organizational Structure Definitions). In terms of horizontal differentiation, alllead agency organizational charts examined for this evaluation (seven total) showed 4or 5 different departments/divisions across their organizations as indicated by a distinctpersonnel title (e.g., Operations and Finance). In all, approximately eight different titlesappeared across the organizations that represent the varied divisions of the leadagencies and reflect distinct areas of specialization, including:

• Operations• Finance or Chief Financial Officer• Quality Assurance• Community Relation/Liaisons• Network Development• Client Services/Case Management• Technology• Child/Family Services

Focus On: Organizational Structure Definitions

Complexity: refers to how much difference exists in the various tasks, proceduresand practices in the organization (Fitzgerald, 2002). The degree of complexity in anorganization is measured by the amount of horizontal differentiation, verticaldifferentiation, and spatial dispersion (Hall, 1996; Fitzgerald, 2002).

Horizontal Differentiation: indicates subdivided responsibilities and activities oftenrepresented by the number of various positions and specializations across theorganization or the number of divisions and departments that segment theorganization (Fitzgerald, 2002). The level of job training/education for aspecialization is an indicator of the level of horizontal differentiation and thereforethe complexity: the greater the number of jobs in an organization that require specialskills, the more complex the organization will be (Robbins, 1987).

Vertical Differentiation: refers to the number of employees, or hierarchical levels,from the very top level of the organization to the lowest level and represents thedegree of this arrangement in the organization (Hall, 1996; Robbins, 1987).

Spatial Dispersion: refers to the number of offices not located in the immediatepresence of the “main” operations of the organization (Robbins, 1987).

18

The consistency in number of divisions across the lead agencies suggests that thereare agreed upon specializations/training required for the activities of lead agencies anda similar level of horizontal differentiation.

While horizontal differentiation was consistent across the lead agencies, analysis of thelead agency organizational charts illustrates varying amounts of vertical differentiation.Four lead agencies, including Heartland for Children (Heartland), Family SupportServices of North Florida (FSS), United for Families (UFF) and FamiliesFirstNetwork (FFN), had an average of 3 persons between the lowest and highest levels ofthe organization (see Figure 2 for Heartland’s organizational chart; inserted for visualpurposes only). In contrast, the remaining agencies, Hillsborough Kids, Inc. (HKI),Child and Family Connections (CFC), and ChildNet, Inc., had an average of 4.6persons between the lowest and highest level of the organization (see Figure 3 forChildNet’s organizational chart; inserted for visual purposes only). While the differencemay seem negligible, the group with the higher vertical differentiation has almost twomore people between themselves and the top-level administration. The hierarchicalincrease would require more process and communication standardization than thatneeded in the less vertically differentiated agencies.

The last measure of complexity, spatial dispersion, affects those lead agencies withoffices or service centers not located in the same geographic location as the “main”office. Those lead agencies serving multiple counties, with or without service centers,will have greater structural complexity simply due to geography. For example, ChildNetand FFN have greater spatial dispersion because their services are spread across thecounty or counties they serve through service centers.

Overall, all of these measures can be placed on a continuum from low to high so thatmixed variations of the measures can exist across the lead agencies or across thedivisions or departments of the lead agency (Hall, 1996). Specific measures can alsoimpact one another. For example, an organization with high spatial dispersion such asFFN, may then require high vertical differentiation to account for multiple personnelpositions needed at each individual service center.

When considering the descriptions and measures of how complex lead agencies arebased on their organizational charts and the responses from the CEO survey, it isimportant to note that this is only the first step in describing and understanding eachagency’s organizational structure. Understanding the complexity of their lead agencywill allow administrators to recognize the needs of the organization based on structure(e.g., increase workforce as complexity increases). Lead agencies may also find thattheir structure is too complex, or not complex enough, to adequately perform theactivities of their organization. Future reports will continue to examine the structure ofthe lead agencies and include an analysis of the communication and interactionprocesses across the organization and the level of formalization and centralization inplace.

19

Organizations are always changing in order to accommodate and assimilate incominginformation from various channels. Given the varied environments and contexts oforganizations, there cannot be one organizational model that can be universally appliedto any one “type” of organization such as child welfare organizations (Hall, 1996).Because each lead agency operates within different contexts and environments, itwould be premature to suggest that one “type” of organizational structure would bepreferable over another, or that a structure is appropriate or inappropriate for aparticular agency.

Figure 2. Example of Low Vertical Differentiation: Heartland for ChildrenOrganizational Chart

Heartland for ChildrenOrganizational Chart - Services

Devereux Foundation

Heartland for Children Board of Directors

Network Development and Community

Relations Director

Financial and Data Management

Director

Operatons Director Quality Management

Director

Data Services Supervisor

UM Unit3 FTEs

Executive Director

Heartland for ChildrenLocal Advisory and

Resource DevelopmentBoard

ContractManagers

3 FTEs

TrainingSpecialist

Court Liaison

Intake Supervisor

SPOA

Intake/Placement5 FTEs

Admin SupportTBA

Vouchering Clerk

Executive Secretary

ProgramSpecialist3 FTEs

Quality AssuranceSpecialist

Heartland Community Alliance

D14 Operations Workgroup

District 14 Prevention Workgroup

CMO and Contract Providers Network

Foster Parent Associations

3/22/04 HFC/mb

Devereux FloridaAdministrative

ServicesContractCase

Management Organizations

Service Providers

Assistant Director HR Coordinator

Finance Clerk

Staffing Masters

John Brown Vanessa McCottryNicolas Moughis

Prevention Director

Community Resource Staffing Coordinator

Assistant Director (TBA)

Admin Support

Specialists

2 FTEs

File Clerks

4 FTEs

20

Figure 3. Example of High Vertical Differentiation: ChildNet Organizational Chart

C h ild N e tB o a rd o f D ir e c to rs

V P F in a n c e & A d m in is t ra t io n /C F O

8 F T E s 5 F T E s 4 F T E s

P re s id e n t /C E O

V P C l ie n t S e rv ic e s V P S y s t e m Q u a l ity& P ro f D e v

V P N e tw o rkD e v e lo p m e n t

D ire c to r o f L e g a l S e r v ic e s

E x e c u t iveA s s is ta n t

7 F T E sD ir . In d e p e n d e n t L iv in g

S ite D ir e c to rC e n t ra l

A d m in . A s s is t . C l ie n t S vc s

D ir . P e rm a n e n c y& T P R

S ite D ir e c to rN o r th

S ite D ir e c to r S o u th

S ite D ir e c to rS a fe P la c e

D a ta S u p e r v is o r

R e v M a x S u p e rv is o r

M e d ic a l L ia is o nS u v.

2 M e d C o o r. A s s is t .

C o u rt L ia is o nA s s is t . S ite D ir .

S a fe P la c e E S S L ia is o nA d m in A s s is t .

S a fe P la c e

S e c re t a ry S a fe P la c e

R e c e p t io n is tC l ie n t S vs

3 IP A A s s is ta n ts

C h ild A d vo c a te A s s is ta n t

C h ild A d vo c a te A s s is ta n t

C h ild A d vo c a te A s s is ta n t

C h ild A d vo c a te A s s is ta n t

C h ild A d vo c a te A s s is ta n t

C h ild A d vo c a te A s s is ta n t

C h ild A d vo c a te A s s is ta n t

C h ild A d vo c a te A s s is ta n t

C h ild A d vo c a te A s s is ta n tC h ild A d vo c a te

A s s is ta n t

D a ta S p e c ia l is t

C o u r t L ia is o n

C A C a s e In v e s t ig a t o r

In ta k e D a ta S p e c ia l is t

C A C a s e In v e s t ig a t o r

In ta k e A d vo c a teS u v. U n it 2

5 In ta k e A d v o c a t eS p e c ia l is t

In ta k e A d vo c a teS u v. U n it 1

5 In ta k e A d v o c a t eS p e c ia l is t

The CBC lead agency CEOs also had an opportunity in the CEO survey to respond toquestions concerning the strengths and challenges of their organizational structure.Table 6 delineates the various strengths described by the CEOs that could be attributedto their organizational structure. An assortment of weaknesses were noted by the

21

CEOs, including insufficient board membership from community-based individuals whoare unaffiliated with the partner organizations and significant funding shortages in onearea adversely impacting service delivery in other areas. The need for modification ofexisting organizational policies and procedures to meet differing needs of CBC, as wellas strain on Administrative Services as a result of both adapting to a new line ofbusiness and associated increase in workload, were also noted as weaknessesattributable to organizational structure.

Table 6. Characteristics of Organizational Structures: Strengths

Strengths

• Founding partners/agencies that can guarantee lines of credit and take sharedresponsibility for debt should it occur

• Capability to coordinate management of varied funding sources and providers tointegrate a comprehensive array of behavioral health and child welfare services

• Maximizes functional oversight for the system of care• Highly flexible• Communication remains fluid among departments• The mix of utilization management with several levels of staffings – including

staffings designed to prevent families from moving into more intensive levels ofcare – serves to maximize resources and minimize risk

• The structure allows the provider network to connect with the organization atseveral points so that the lead agency works at all times in true partnership withnetwork members

• There are several venues for providers, consumers, and advocates to bringissues to the attention of workgroups for discussion, review, and action

• The partner organizations that have representatives on the board and alsoprovide case management services have a vested interest in the success of thelead agency and work hard to enable the CBC lead agency to succeed

• Board of Directors is very supportive of CBC• Attempting to be proactive rather than reactive

• The majority of services are contracted out to a diverse network of local providers

22

Focus On: Heartland for Children’s organizational structure

Heartland for Children made several changes to their organizational structure after the agency began providingservices. As described by Heartland’s CEO, the original organizational structure for Heartland included fiveDirectors, as follows:

• Director of Case Management• Director of Quality Management• Director of Finance and Data Management• Director of Network Development and Community Relations• Director of Utilization Management

After Heartland for Children assumed services, it became clear that the Director of Case Management was amisnomer since all case management functions were subcontracted. In that sense, the direct line of responsibilityfor case management, then, fell to the contract manager, who reports to the Director of Network Development andCommunity Relations. The Director of Case Management position was changed to Director of Intake andPlacements. The Intake and Placements division already had a supervisor position. When the Director of CaseManagement position was redefined with a more limited scope, it seemed that the position was no longer needed,and was subsequently dissolved. The Intake and Placements function was moved under the Utilization Managementumbrella, and the Utilization Management Director was renamed the Director of Operations.

The funding for the Director of Case Management/Director of Intake and Placements position was reallocated tosupport an Assistant Director for Finance and Data Management. With the corporate emphasis on finance and riskmanagement, it seemed that the Director should be able to focus more fully on those aspects of the job. TheAssistant Director position was created to assume most of the burden for the Data Management functions within thedivision. The Assistant Director will also be tasked with a range of collateral functions such as supervision of thedata entry staff, providing data reports as required, I.T. Help Desk, inventory control, and building and facilities.

As the time for transitioning services grew closer and Heartland for Children spent an increased amount of timediscussing their system of care, it became clear that because of the system’s emphasis on prevention, there neededto be someone at the Director level to manage those responsibilities. A position for a Director of Prevention was thusdeveloped.

Throughout the contract start-up phase, the Special Projects Coordinator assumed responsibility for coordinating allcontract deliverables, keeping the staff on task according to an action plan, and serving as liaison with the contractmanager at the district office. When the start-up phase and the action plan were complete, the person in thatposition was promoted to the position of Assistant Director of Quality Management. The assumption was that the QMDepartment would need someone to manage the accreditation process in much the same way that start-up wasmanaged. The establishment of the Assistant Directorship to serve that purpose, as well as to assist in themanagement of the rest of the Quality Assurance/Quality Improvement functions, seemed to make good sense.

After transition, the Director of Network Development and Community Relations found that the responsibility formanagement of all the contracts was an all-consuming job, leaving little time for “network development andcommunity relations.” With that in mind, the assistant director position was moved from the Quality Managementdepartment to the Network Development and Community Relations department, and the Assistant QualityManagement Director became the Assistant Director of Network Development and Community Relations. This moveallowed the Network Development and Community Relations staff to focus more fully on all aspects of thedepartment’s responsibilities. Additionally, it was determined that the quality background of the Assistant Directorwould serve the other department well in accomplishing contract compliance. It was also determined that skills andattention to detail that the Assistant Director had honed while serving as Special Projects Coordinator during start-upwould be a strong asset in the world of contract management.

Therefore, Heartland’s current organizational structure includes five Directors and two Assistant Directors, asfollows:

• Director of Prevention• Director of Quality Management• Director of Finance and Data Management

° Assistant Director of Finance and Data Management• Director of Network Development and Community Relations

° Assistant Director of Network Development and Community Relations• Director of Operations

23

Structure of Provider Networks

An analysis of organizational structure would not be complete without considering theinter-organizational, or interagency, relationships that organizations develop within theircommunity; for the CBC lead agencies, an example of important interagencyrelationships is with the provider network. Overall, the analysis identifiedthree models ofprovider network structures that can be used to visually depict the provider networkconfigurations and their relationship to the lead agency including: (1) a providerstructure that maintains parent or partner organizations (Figure 4); (2) a more traditionalprovider model that excludes parent/partner organizations (Figure 5); and (3) a modelthat depicts the use of service centers in the provider structure (Figure 6).

Heartland, utilizing a provider structure that involves parent or partner organizations(Figure 4), has a two-tiered provider network that includes network providers andcommunity providers. In this approach, the network providers are four organizations thatprovide case management services, including the Devereux Foundation, Heartland’sparent organization. The community providers that make up the second tier ofHeartland’s provider network are those community-based organizations providingservices related to residential care, prevention, and wraparound, including foster care,Family Builders and parenting services. Non-contracted community providers are alsoused to provide services through a “community resources staffing process” that refersfamilies to the Heartland Community Resource Staffing Master, who using a familyconferencing model, engages families with representatives from the non-contractedcommunity providers.

Along with Heartland, HKI, CFC, and UFF (Figure 4) also have parent or partnerorganizations that provide either all, or part, of the services related to casemanagement, foster care, adoption, and crisis intervention, in addition to contractingwith community-based organizations for the provision of services. HKI has five partnerorganizations that provide services and share financial risk, with three of theseorganizations (Children’s Home Inc., Children’s Home Society of Florida, and NorthsideMental Health Center, Inc.) each operating a care center with geographic responsibilityfor child protection services. Camelot Care Center, the organization responsible foradoptive and foster parents, serves as the coordinator for the Foster Home andAdoption Network (FHAN), which is composed of seven partner and contractedcommunity-based organizations. UFF has four community partners who also sharefinancial risk, including Children’s Home Society of Florida, Family PreservationServices of Florida, Exchange Club CASTLE and New Horizons of the Treasure Coast.

24

Board of Directors

Lead Agency

Structured and Reoccurring

Provider/Lead agencyMeeting

Provider /Community Network.Some providers are co-located as Indicated by the solid connector line.These providers may also be responsiblefor all services in a defined geographicarea

Parent /Partner Organizations(Typically providing services such as casemanagement.)

Meetings occur with various combinations of providers/partners.May include parent org. /partners orprovider network separately or jointmeetings.

Some providers may havea person serving on BoardOf Directors.

Figure 4: Model of Provider Structure with Parent/Partner Organizations

25

FSS utilizes a more traditional provider network that does not include parent or providerorganizations (Figure 5). To maintain natural neighborhood boundaries and build on thesupports provided by the individual neighborhoods, FSS uses 8 main providers acrossDuval County that are responsible for services in the neighborhood in which they arelocated. In addition, FSS has residential services providers.

Board of Directors

Lead Agency

Structured andReoccurring

Provider/Lead AgencyMeeting

Provider Network. Some providers are co-located asindicated by the solid connector line. These providers may also be responsible for all services in a defined geographic area.

The providers and lead agency meet at a organized meeting.

ChildNet, like FFN, utilizes service centers across their county/counties to provideservices to the defined geographic area (Figure 6). ChildNet uses four service centersacross Broward County while FFN utilizes six service centers across the four countieswhere FFN has lead agency responsibility.

Figure 5: Model of Provider Structure without Parent/Partner Organizations

26

Board of Directors May or may nothave Parent orPartner Orgs

ServiceCenter

ServiceCenter

ServiceCenter

Service Center

Service Center

Some service centers arespread across one county and serve different geographic areas.

Others are acrossmultiple counties(connected together by the solid line) andmay access the sameproviders for servicesas indicated by the dashline.

Structured andReocurring

Provider/Lead agencyMeeting

Lead Agency

The Lead Agencycontracts with community-based providers to servethe service centers.

Provides services and supports for all service centers.

Provides services forCounty.

Figure 6: Model of Provider Structure with Service Centers

27

In regard to provider network/lead agency interaction, every lead agency responding tothe CEO survey indicated a pattern of ongoing communication between the lead agencyand their providers as presented in the Table 7.

Focus On: FamiliesFirst Network’s provider structure

FFN structures their provider network differently than the other lead agencies by allowingproviders to serve families across District One’s four counties. FamiliesFirst Network is a divisionof Lakeview Center, Inc. Six service centers have been developed (two each for Santa Rosa andOkaloosa Counties and one each for Escambia and Walton Counties) to provide services to thesurrounding community. Through these services centers, local community-based providersprovide services through contractual relationships with Lakeview Center, of which half are for theprovision of intensive home services. These contracted providers can serve the entire District, onecounty or one service area.

Contracted services include non-residential services (e.g., adoption support, sibling visits,visitation centers, family resource and support, and private practitioners), and residential services(e.g., group shelter, group foster care, therapeutic foster care and SIPP).

Non-contracted community resources include (Domestic Violence Coalition/Shelters, drug court,school readiness coalition, guardian ad litem program, health departments in each county, BigBrothers/Big Sisters, and local food banks).

Currently, FFN is redesigning their non-residential service system through a collaborative effortwith DCF, community stakeholders, and those providers already providing non-residentialservices. The goals of the redesign effort are greater flexibility and accessibility for the familiesbeing served by FFN.

28

Table 7. Lead Agency and Network Provider Communication

Lead Agency Communication/Interaction Description

Heartland forChildren

Communication between Heartland and the providernetwork is facilitated by a weekly case managementorganization meeting, as well as monthly meetings withcommunity-based providers and bi-weekly workgroups thatinclude DCF, case management organizations, childwelfare legal services (CWLS), and community providers.

Hillsborough Kids,Inc.

HKI facilitates communication among the provideragencies by having agencies co-located, networked oncommon MIS and communication systems, involved withthe Leadership Council (HKI’s policy board), and throughparticipation at Board of Directors meetings.

Child and FamilyConnections

CFC partners providing case management services areco-located and share training across the organizations. Inaddition, CFC holds formal, community-based providermeetings for training and information exchange forcommon issues.

FamiliesFirstNetwork

FFN has services co-located in their service centers aswell as quarterly provider forums. Provider networkworkgroups also meet frequently to discuss and plan forthe redesign of the non-residential service system.

Family SupportServices of NorthFlorida

FSS meets monthly with the executive directors from eachprovider, and weekly meetings are held at each servicecenter.

ChildNet, Inc. ChildNet is currently working towards having someproviders/services co-located and utilizes monthly and/orcontract meetings to engage providers.

United For Families UFF meets twice a month with the case managementagencies, foster parent support agency, and theDepartment to discuss systems issues. UFF personnelalso meet monthly with those agencies providing casemanagement services, and personnel attend all fosterparent association meetings.

29

Utilization of Emerging Promising Approaches to Supervisory Practice andTraining

An issue that has repeatedly arisen during this evaluation is the importance of the roleof supervisors in assuring the success of any child protective system. The June 2003evaluation report made specific recommendations about: (1) the need to adapt trainingand supervision for case managers to meet the demands of community-based care, and(2) changes in training for case management supervisors to support the changesexpected in frontline practice (Paulson, et.al. 2003).

Across the country, researchers and practitioners alike are focusing their attention onthe need to “train child welfare supervisors as child welfare supervisors” (Maximus &Florida State University [FSU], 2003). Poor supervisory relationships are pointed to asbeing one of the primary factors leading to employee dissatisfaction and turnover(Martone & Solek, 2001). Other factors include: unmanageable duties (perceived), peerconflicts, poor training and preparation, lack of advancement opportunities, and less-than-desirable salary.

In their review of “best practices” Maximus and FSU (2003) identified:• Community/university partnerships to improve training (e.g., Los Angeles County

Inter-University Consortium for Child Welfare);• Specialized training for supervisors and interactive review of practice by

supervisors and case workers (e.g., Pennsylvania);• The development of career and educational paths for supervisors.

The Center for the Study of Social Policy (2003) in their implementation of CommunityPartnerships for Protecting Children (the forerunner of Florida’s NeighborhoodInitiatives), emphasizes the importance of supervisors to implementing communitypartnerships. They identify several important lessons learned:

• Get frontline supervisors involved in the partnership planning from the verybeginning.

• Engage supervisors in the “vision“ of changed practice by ensuring that theyparticipate in training to improve the skills needed to engage families, assessstrengths and needs, and include families (through the use of family teammeetings) in the planning process.

• Give supervisors time to work with their staff – being readily available to providecase consultation “spur of the moment” as well as in regularly scheduled one-to-one reviews.

• Supervisors who enjoy working with families and children are good role modelsfor workers. In some community partnerships, supervisors “share” very difficultcases with workers as teaching, learning, and supportive functions.

The Walter Trieschman Center (a division of the Child Welfare League of America) isdevoted to the development and delivery of an “Effective Supervisory Practice” program

30

that is based on the premise that “effective state of the art treatment and serviceprovision for children, youths and families requires highly competent, well-trainedsupervisors and a commitment from leadership to the development of competentsupervisors” (CWLA, 2004). Their training programs include opportunities foradvancement of case managers to senior case management positions after completionof the training (Martone & Solek 2001).

For the identification of emerging promising approaches to supervisory practice andtraining, the Executive Directors/CEOs of the 11 lead agencies with service contracts asof January 1, 2004 were contacted (see Appendix 3). The purpose was to solicit whatthis group believed to be emerging promising approaches to supervisory practices ortraining. Once evaluators received the approaches, the lead agencies were contacted togain a further understanding of the approach. It is of interest to note that several of theapproaches were preliminary, but were believed to hold promise (e.g., “Although wehave great plans I think our current practice has been limited” [personal communicationwith lead agency representative, February 20, 2004]). Also, the identified practicesmirror what is found around the country at the current time. The following lead agenciessubmitted emerging promising approaches:

1. Child & Family Connection, Inc. (CFC)2. ChildNet, Inc. (ChildNet)3. FamiliesFirst Network (FFN)4. Heartland for Children (Heartland)5. Hillsborough Kids, Inc. (HKI)6. Partners for Community-Based Care (PCBC)

Child and Family Connections, Inc. – Mentoring System

CFC, along with partnering agencies, is introducing a number of ideas they believe willenhance the performance of their supervisors. These ideas include:

• Single case manager model - one supervisor to seven staff members.• Case assignment made based on the knowledge and strengths of team

members.• Bi-weekly individual supervision that is to be presented and viewed as “protected

supervision time”. Supervision will be used to assist staff in reviewing andprioritizing case tasks and reinforcing the quality of decision-making throughoutcasework.

• Bi-weekly group supervision meetings with supervisors and their team. Thesemeetings allow for the introduction and review of new procedures beingimplemented within the program. In the near future, staff members will beexpected to do case presentations on cases that have either been a challenge ora success. In addition to receiving feedback from their peers, this will allowteam members to become more familiar with their co-workers’ cases.

31

• Each unit has court calendars and other pertinent information posted inaccessible areas such as supervisors’ offices and unit bulletin boards.

Perhaps the most exciting feature is CFC’s mentoring system. The plan is forsupervisors to connect each senior staff member with a less senior staff member. Thementor will serve as a point of reference/support to new workers. A mentor may assistin demonstrating use of program forms, inform new worker of community resources, ormodel job function. At this time, the agency is not able to financially reinforce the workof the mentor, rather it is viewed as a way “to foster a supportive environment and ateam approach” (personal communication with CFC, March 9, 2004) in their currentsystem. This approach is similar to that highlighted by the Trieschman Center, in whichcareer ladders and staff development activities are combined (Martone & Solek, 2001).As identified by CFC, the mentor system allows them to develop the skills of newerworkers, while also providing advanced responsibilities (and hopefully in the future,advanced salary) for experienced workers.

ChildNet – Performance Appraisals

ChildNet is in the process of determining training and development programs for thecurrent year. They reported that they have identified some practices “to enhancecommunication and foster productive supervisory relationships” (personalcommunication with ChildNet, February 19, 2004). These practices include:

• Setting up the physical environment for supervisors to be in close proximity tostaff.

• Encouraging informal and formal meetings between supervisors and staff.• Offering training with supervisors and child advocates in the same classes.• Conducting performance appraisal classes for supervisors.

The performance appraisals offer a means for ongoing supervision and identification ofstaff training needs. In addition to training in the completion of the appraisals, ChildNethas offered follow-up sessions for supervisors with the Director of ProfessionalDevelopment to develop specific parts of their appraisals in a lab environment.Supervisors have been offered one-on-one coaching on ways to give constructivefeedback, establish performance expectations, follow-up and provide recognition. Toadd a personal development aspect to their professional development, supervisors arealso coached on their own actual work-related issues, and they role play givingperformance feedback to staff members.

These appraisals appear to be similar in nature to those identified by Maximus and FSU(2003) in the Commonwealth of Pennsylvania. In Pennsylvania’s comprehensiveapproach to training and education in public child welfare, they have implementedIndividual Training Needs Assessments (ITNAs) that allow supervisors to work annuallywith staff to measure each worker’s training needs.

32

FamiliesFirst Network (FFN) – Structured Quality Improvement

FFN is moving to enhance the performance through the implementation of a three-stepsystem of structured quality improvement. The general framework is Monitoring -Review – Improvement. Each service unit is reviewed at a regular interval, allowingthem to roll case-specific details up to the unit level, then to the service center level,then to their overall system. This “roll up” allows them to review practice and implementpractice improvement strategies.

In support of their quality improvement model, supervisors meet with their FamilyServices Counselors for monthly supervision in which they review ALL families on theircaseloads. The supervisors are also encouraged to identify and share best practiceswith their supervisory colleagues. Finally, supervisors receive training on how to captureand interpret consistent management data that will allow them to review the quality ofservice provision within their units.

Heartland for Children (Heartland) – Community Training Needs Assessment

Heartland is developing an extensive training plan that is designed to meet the needsof supervisors, staff, and community providers (e.g., subcontracted case managers).Heartland conducted a community training needs assessment of what types of trainingcommunity providers need and what training community providers offer. “The goal is tocapitalize on the expertise already resident in the area, and to give providers ownershipover the training program” (personal communication with Heartland, February 16, 2004).This community ownership is already becoming evident based on preliminaryidentification of training needs. Some suggestions to date include:

• Resource training for supervisors, counselors and caseworkers presentingformal and informal resources available in the community. Their plan is toinclude community organizations such as churches, clothing banks, and foodbanks in the training process.

• Use of contracted providers to explain the types of services available in thecommunities through their agencies.

• Asking the local Sheriff’s Department to present relevant topics, such as types ofparaphernalia and drugs, gang issues, and domestic violence from a legalstandpoint.

Strategic objectives included in their plan for supervisors include retention (throughpersonal and professional development training) and enhancement of professional skills(through specialized trainings offered by community provider network and local trainingprograms). Throughout their training plan is the goal of capitalizing on and building uponlocal resources, involving the community throughout their system of care.

33

Hillsborough Kids, Inc. (HKI) – Wraparound Supervision Training

HKI, in partnership with community consultants, is implementing a wraparoundsupervision training that is part of their supervisory orientation. The training is based onthe principles of wraparound practice, the core clinical feature of the HKI system of care.Wraparound supervision is viewed as being strength-based, staff-centered andindividualized. The role of the supervisor is to help staff identify strengths and prioritizetheir needs as they strive to identify and meet the needs of the children and familiesthey serve. The work of the Community Partnerships has found, not surprisingly, that“supervisors who model respect and empathy in their treatment of workers are morelikely to have workers who adopt these attitudes and practices with families.” A samplefrom the training presents the targeted paradigm shift for supervisory practice (Table 8):

Table 8. Paradigm Shift in Supervision

From: To:• Focus on corrective action • Focus on strengths and skills• “Planning for or with” • Consultation• Diagnosing and seeking

professional supports/services• Emphasis on tapping into and

developing natural supports• Schedule of supervision • Supervision and support is

scheduled and flexible to staffneeds

To date, HKI has developed a training module and a trainer’s manual including trainernotes and handouts. The Director of Operations is currently the primary trainer for thismodule. All new care management supervisors receive the training during theirorientation. Although the training is not yet mandatory, most supervisors havecompleted the training. In the future, this training will be expanded to include care teamcoordinators. The training is presented in a four-hour session with exercises and role-playing activities.

Partners for Community Based Care (PCBC) – Supervisory Training

The Maximus Report identified the need to “train child welfare supervisors as childwelfare supervisors”. This would include a dual emphasis on practical skill sets thatsupervisors can put into practice on a day-to-day basis as well as the child welfareissues confronted by supervisors each day. PCBC is developing such a system.Personal communication (March 9, 2004) with PCBC indicates, “Supervisors clearlyneed to develop their leadership skills in the field where good case managers getpromoted even though they may not always have good management techniques. Whilesome skills in working with families can translate to working with employees, many moredomains of leadership qualities are needed to motivate, encourage, and monitor casemanagers on the front lines of child welfare.”

34

PCBC is developing a comprehensive training program that includes input from thecommunity. They have contracted with a full-time trainer who is an employee of TheLearning Institute, a division of Children’s Home Society that deals specifically with childwelfare education issues. This trainer conducts in-service training for case managementand supervisory employees, as well as providing pre-service status training to traineesduring their time at the Professional Development Center. At other times, the trainerfacilitates training that is being offered by an expert in the community. Supervisory staffhas input into what training is developed. Supervisors have requested training topicssuch as conflict resolution, team building skills, and employee motivation. There is aformal workgroup meeting held quarterly with supervisor involvement, althoughsupervisors are encouraged to request general training, unit-specific training or one-on-one sessions for staff at any time it is needed. Additionally, the Learning Institute trainerreceives feedback from the Community Alliance, DCF and other service providers aboutthe perceived training needs of CBC staff.

Recognition and Reduction of Conflicts of Interest

Designing and improving CBC systems also means recognizing and reducing thepotential for conflicts of interest that could lead to problems for CBC lead agencies. Twotypes of conflict related to CBC lead agencies include:

1. Interlocking boards;2. Lack of diversity in board membership

The phenomena of interlocking boards of directors, when the board members from oneorganization sit on the board of another organization, is an example of a relationshipbetween boards that may be considered a conflict of interest, especially when thoseboards are in contractual relationships (Hall, 1996). An illustration of interlocking boardswould be that demonstrated by many lead agencies when representatives fromcontracted provider organizations, or partner organizations, serve as lead agencyboard members. Either through lead agency partner demand or lack of available andknowledgeable board members, several lead agencies have contracted providers asboard members. All lead agencies but one (involving a contracted provider as a boardmember) responded that they did not have any conflicts of interest presently on theirboard.

The lack of diversity of members on the board of directors can represent a type ofconflict for the community being served. An analysis of lead agency board member listsindicates that six of the eight lead agency boards of directors have more contractedproviders, state and local government personnel and private for-profit members servingon their boards than community members, foster parents, and representatives fromother non-profit organizations (see Table 9). Only two agencies, FSS and HKI, hadmore equal participation by community representatives, foster parents and nonprofitorganizations represented on their board. Agencies with boards containing manycontracted providers, state and local government, and private business sector

35

representatives may not be as “in touch” with the needs of the community, families andchildren served. The lack of community involvement may also prevent the infusion ofnew organizations, new services and resources, into the provider network fold.

Table 9. Board Membership by Organizational Affiliation(excludes lead agency CEO/ED)

Lead Agency PrivateSector

County orState Rep.

ContractedProviders/

ParentOrgs.

CommunityRep.

/AdvisoryBoards

Non-profitOrgs

FosterParents

FSS 5 5 0 4 3 2

ChildNet 7 1 3 0 0 1

PCBC 3 0 3 0 0 0

FFN 8 1 1 0 3 0

HKI 7 0 Partnerschoose reps

fromcommunity

to sit onboard

3 3 0

Heartland 0 0 4 0 0 0

UFF 0 0 8 1 0 0

CFC 1 0 6 0 1 0

One of the potential outcomes of a lack of diversity on the boards and interlockingboards can be a lack of fairness and openness when provider agencies are competingfor services. For example, if it is a requirement of a contracted service provider orpartner organization to be the sole provider of a service this can result in a lack ofopportunity for other providers and the lead agency.

Table 10 delineates the types of situations where conflicts of interest can occur, thefactors that sustain these situations, and the problems that can arise due to thesesituations.

36

Table 10. Types of Potential Conflict SituationsPotentialConflict

Factors that Sustain the Situation Problems Associated with the Situation

Lack ofmemberdiversity onBoard ofDirectors/InterlockingBoards

1) May be requirement of contractedproviders to maintain a seat on the leadagency board for their best interest.

2) A new lead agency needs time todevelop a more rounded board; difficultyfinding relevant board members.

3) The lead agency does not have anindividual board; the parentorganization’s board is used.

4) May increase the power an agencyhas in the community by placingrespected, highly visible individualsalready involved with children andfamilies on the board.

5) Interlocking boards of directors.

1) Board can become laden with members that havemore face value and potential than actual expertiseand action.

2) An established parent organization board maynot be close enough to the lead agency to provideproper direction and/or influence decision-making.

3) Lack of community membership on board mayreduce the board’s ability to recognize communityissues and respond to families and children in thecommunity.

4) The lack of client representatives (e.g., fosterparent) on the board may reduce recognition ofclient related issues.

5) Representatives from contracted providers mayhave difficulty remaining free from conflicts ofinterest.

Lack ofOpen/FairProcurementfor Services

1) Lack of providers/services in the“geographic market” necessitates theuse of available providers.

2) Reduces risk by contracting withknown providers.

3) May be requirement of partneragencies to ensure viability of the leadagency and reduce risk.

4) Political statement. “Restrictedmarket entry” to those organizationsidentifying with same communitypolitics.

5) Certain types of services do not lendthemselves easily to a competitiveprocess, and multiple changes inproviders require budget increases.

6) Interlocking boards of directors.

1) Non-profit networks (and provider networks) canlimit the number of providers allowed access to thenetwork, thus creating an organized monopoly forservices.

2) Forced dependence on providers that may not beperforming successfully.

3) Failure of an organization that has been the soleprovider of a service.

5) Potential to drive up costs of services if providersof certain specializations “collectively organize.”

6) Interlocking boards of directors can increase thelikelihood that the market will be closed to “outside”organizations.

Focus On: HKI’s Board Member Selection

HKI utilizes a unique approach to ensure that the potential for conflict of interest in the boardof directors is reduced. HKI’s five partner organizations choose individuals from thecommunity to serve as their representatives on the board of directors. In addition, thesecontracted/partner representatives will at no time account for more than one-third of the totalboard membership.

37

Financial Flexibility

The ability to design, improve, and alter service systems for children also means havingthe ability to be financially flexible. Many lead agency respondents felt that the degree offinancial flexibility they had hoped for was not present due to many of the Federalrequirements. In some instances the lead agencies’ allocation between different Federalrevenue sources was problematic (e.g., IV-E) because the lead agency had drawndown all the funds in one category but could not spend all of the money in anothercategory. This issue is a result of Federal regulations and statutes. While a more flexiblesystem of matching a lead agency’s needs with the purpose of the fund, or “trading”categorical funds between sites would be extremely helpful, the State’s options arelimited because of the Federal requirements.

However, lead agencies have tried to meet the funding challenge. One lead agencyraised private funds for special projects such as a staff member to do faith-basedrecruitment, as well as computer resources for youth. Although District 1 is one of themore under-funded Districts in the State, FFN has used Lakeview Center for financialsupport while pursuing equity and adequacy in funding from the State. The diversity ofthe agency, and its long history of community service, allows this flexibility to bepossible. Another lead agency stakeholder commented, “There is nothing that may beneeded that we cannot purchase; there are simply limits on the amount of money wehave available.”

New lead agencies offer another perspective on financial flexibility. As was reported inprevious reports, it can be quite challenging for lead agency staff to gain access tovarious financial resources (Paulson, et. al. 2003). In addition, the letter of credit neededor the requirement to have 60 days cash on hand has been an obstacle for some newlead agencies. As one lead agency stakeholder explained, “It is hard to get financingfrom the bank without a history.” Overall, lead agency respondents felt that their agencyhad been able to provide for additional testing of creative and flexible strategies forfinancing child welfare services to enhance individualized treatment and/or child-leveloutcomes.

Limitations

A limitation for this research question was the lack of responsiveness from leadagencies and Community Alliances. A few declined to complete the survey, and ofthose that did complete it there was inconsistency in the supporting documentation thatthey submitted. Upon initial distribution of the surveys, lead agencies and CommunityAlliance members were contacted via e-mail and through follow-up phone calls to checkthe status of their responses. Based upon these follow-ups, respondents provided avariety of reasons for declining to complete the survey such as recent changes inleadership, a lack of historical and contextual background as leadership was in the earlystages of transition, a deferment to information previously submitted to evaluationproject staff, and time constraints preventing the completion of the survey.

38

Summary of Research Question 1

Analysis of the organizational structures of the lead agencies demonstrates a differencein the level of complexity across CBC lead agencies. While the agencies wereconsistent in the amount of horizontal differentiation, they varied on their level of verticaldifferentiation. Understanding the complexity of their lead agency can allowadministrators to respond to the needs of their organization and recognize strengthsand weakness that can be attributed to their structure. In regard to provider networkstructure, three models can be used to visually depict the way in which lead agenciesare arranging their provider network (with slight variations across lead agencies). Thesemodels represent those agencies with partner/parent organizations involved, the moretraditional perspective that does not involve partner/parent organizations, and a modelthat includes the use of service centers. The differences in provider network structureswould indicate that CBC lead agencies are developing their provider networks based onthe availability of resources in their individual communities while creating ways to reachall of the children and families in their service area.

In regard to emerging promising supervisory practices, six of the 11 lead agenciessubmitted approaches, indicating that several sites are trying to incorporate innovationin supervisory training and practice. There was also great variability in the types ofapproaches across sites, another indicator of individualized efforts to design effectivesystems and services.

Overall, lead agency boards of directors are utilizing more contracted providers, privatesector, and state and local government representatives on their boards than communitymembers, foster parent and non-profit organization representatives. However, leadagencies indicated that they do not currently have problems with conflicts of interests ontheir boards of directors.

39

RESEARCH QUESTION 2:How effective is Community-Based Care at involving the community in childprotection both as service partners and resource contributors?

Introduction

In order to examine the effectiveness of CBC at involving the community in childprotection, both as service partners and resource contributors, this research questionincludes the following evaluation questions, indicators and data sources (Table 11):

Table 11. Research Question 2: Evaluation Questions, Indicator/Analysis, andSource

Evaluation Questions Indicator/Analysis Source

Are community services beingaccessed by the lead agencies?

-Average number and rangeof contracts and providers forlead agencies

CEO Survey

DCFDocumentation

What types of community governanceboards support the lead agency?

- Description of CommunityAlliances and theirrelationship with leadagencies

- Description of othercommunity boards utilized bylead agencies

CEO Survey

CommunityAllianceChair/ViceChair Survey

How effectiveisCommunity-Based Careat involvingthecommunity inchildprotectionboth asservicepartners andresourcecontributors?

How are local community resourcesbeing developed and utilized?

- Description of local revenuesources CEO Survey

Methods

Community Alliances are comprised of key stakeholders actively involved in thecommunity. The project team contacted each Alliance Chair and Vice Chair in order tointroduce the email survey. The Community Alliance Chairs and Vice Chairs received abrief set of questions via e-mail (see Appendix 4). The questions were designed within astrength-based framework to elicit examples of connectedness to lead agencies andCommunity-Based Care, as well as basic membership information. Community AllianceChairs and Vice Chairs were also asked to submit pre-existing documents such asrecent meeting minutes to show type and evidence of interaction with lead agencies.Comparison of survey responses illuminated the variability across geographical areas in

40

the degree to which there was extensive communication and collaboration between theAlliances and lead agencies.

Community as Service Providers

To get an indication of the extent to which the community is being utilized as serviceproviders, the number of providers and contracts for each lead agency was obtainedthrough the CEO survey and communication with DCF district administrators (DAs). Inparticular, CEOs and DAs were asked, “How many contracts with providers does yourlead agency (or the name of the agency) currently maintain?” For this report, and thefollowing discussion concerning contracts and providers, the numbers of contracts andproviders used are those responses from the DA or CEO (e.g., “27 contracts” or a listingof contracts). If a number or list was not provided, the lead agency was not included inthe analysis.

Table 12 shows the average number of providers and contracts utilized by the leadagencies as well as the range in the number of providers and contracts across the leadagencies by the length of time the lead agency has had their service contract. From thistable it is clear that agencies with service contracts for 6 months or less, or between 7and 17 months, have a greater number of community providers and contracts thanthose agencies with service contracts 18 months or more. The variation in the numberof providers and contracts held across lead agencies is quite considerable in thisanalysis. However, the variation may be more a function of how the contracts arestructured (such as why one service would be an attachment while another iscontracted for the same provider) or reflect the addition of DCF contracts for theagencies not yet fully responsible for all services. In addition, those agencies that haveheld a service contract for 18 months or more may experience a stabilization of theirservice environment that may result in fewer contracts. Further analysis of contractedprovider services is needed to determine the cause of the variation.

41

Table 12. Average and Range of Providers and Contracts by Length of LeadAgency Service Contract

Average and Range of Number of Providers and Contracts by Length of TimeLead Agency had a Service Contract

SC held 6 monthsor less

SC held 7-17months

SC held 18 monthsor more

Average # ofProviders

17.3

(3 of 6 DAs or CEOsreporting)

19

( All DAs or CEOsreporting)

12

( All DAs or CEOsreporting)

Range in # ofProviders

9-29 13-25 10-14

Average # ofContracts

36.5

(4 of 6 DAs or CEOsreporting)

30

(All DAs or CEOsreporting)

20.5

(All DAs or CEOsreporting)

Range in# of Contracts 20-52 21-40 14-27

Limitations

In regard to provider contracting, due to the different ways lead agencies develop theircontracts, for future reports lead agencies will be asked to explain their contractingprocedures with questions expanded to include their contract attachments, rateagreements, etc. The number of services, which may not be represented by the numberof contracts, would be a better estimate of community involvement and would includecontracts, attachments, rate agreement, etc. The analysis also will be expanded toinclude an examination of the relationships between the number of counties andchildren served by the lead agencies and their number of providers and services.

Community as Governance Partners

“Of importance is the acceptance and realization that children are a communityresponsibility. Their protection must be a collective concern.” (Barter, 2001: 271)

42

“…government alone cannot be expected to achieve the results that communitieswant for their families and children.” (Center for the Study of Social Policy (CSSP),1998: 13)

Focusing on the community in terms of leadership and governance is one of the vitalcomponents of CBC. Community building, which involves bringing together thosecommunity stakeholders involved in children’s issues, and child protection, communityofficials and citizens, etc., allows for the community to guide decision-making for betteroutcomes for children and families. With the use of innovative, empowering, collectivecommunity partnerships, communities can better understand the needs of their ownchildren and families and create community goals for child protection that will generatepositive results (Barter, 2001).

One platform for organizing community partnerships that can have significant influenceon child and family outcomes is the development of a governance partnership (in theform of a board or a created independent entity), which for CBC has become theCommunity Alliance. Through these cooperative partnerships, initiatives and programsfor children and families can be planned more effectively, monitored more efficiently,and have the capacity for change as the needs of the community change.

However, making such an impact on the lives of children and families in the communityrequires power. The governance boards or independent entities created within thecommunity must have the power and backing to be heard and respected in thecommunity. To do this, partnerships must establish their legitimacy early on and becredible in the eyes of the community. For many governance partnerships’ this includesgovernment mandates that support the formation and continuation of the joint venture,or, ensuring that high profile community individuals serve in the partnership (CSSP,1998).

Community Alliances

As mentioned previously, one governance partnership model for CBC is the CommunityAlliance. The intended responsibilities, duties, initial members and possible outcomefocus areas for the Community Alliances are listed in Appendix 5.

The Community Alliances were mandated to “provide a focal point of communityparticipation and governance of community-based services“ (s. 20.19(6)(a), F.S.). TheAlliances, although unique to each community, were designed to consist of a broadspectrum of community stakeholders. The Alliances’ duties were to include needsassessment, setting priorities, planning for resource utilization, determining locally-driven outcomes to supplement state–required outcomes, and community education.The scope of the Community Alliances was designed to include Community-Based Careissues, in addition to broader human service areas.

43

Community Alliance Membership

The evaluation team sent surveys to 17 Community Alliance Chairs and Vice Chairs.They were contacted via e-mail and through follow-up phone calls to assist in surveycompletion. Nine responses were received from Chairs and Vice Chairs, submitting avariety of supporting documentation. Alliances generally reported that they containedthose members specified in Statute, in addition to members at large from each countywithin the Alliance’s domain. Examples of members’ professional roles include: DCF,county government, juvenile welfare, school district, court system, United Way, and theSheriff’s Office. See Table 13 for a breakdown of types of members on each Alliance.

44

Table 13. Community Alliance Membership Professional Roles (continues on nextpage)

1 Roles listed in bold are required by s. 20.19 (6) (d) F.S. A voting versus attending member will beclarified in the next evaluation report.

Roles

Shared ServicesAlliance ofOkeechobee and theTreasure Coast, inIndian River,Okeechobee, St.Lucie and MartinCounties

NorthwestFloridaCommunityAlliance, inHolmes,Jackson,Washingtonand CalhounCounties

ChildrenandFamiliesCommunityAlliance ofNortheastFlorida, inDuval,Baker, Clayand St.JohnsCounties

PinellasCountyCommunityAlliance

CommunityAlliance ofSarasotaCounty

CommunityAlliance ofPalm BeachCounty

DCF1 X X X X X XCountyGovernment

X X X X X

Education X X X X X XUnited Way X X X X X XSheriff’sOffice

X X X X X X

Courts X X X X X XHealth (e.g.,county dept.,Healthy Start)

X X X X X

JuvenileWelfare

X X X X

CriminalJustice

X X X

SalvationArmy/ Faith-BasedCommunity

X X X

Office ofHousing &CommunityDevelopment/CommunityPlanning

X X

SubstanceAbuse

X

MentalHealth

X X X

Stakeholder/CommunityAdvisoryGroup

X X

45

Roles

Shared ServicesAlliance ofOkeechobeeand theTreasure Coast,in Indian River,Okeechobee, St.Lucie andMartin Counties

NorthwestFloridaCommunityAlliance, inHolmes,Jackson,WashingtonandCalhounCounties

Children andFamiliesCommunityAlliance ofNortheastFlorida, inDuval, Baker,Clay and St.JohnsCounties

PinellasCountyCommunity Alliance

CommunityAlliance ofSarasotaCounty

CommunityAlliance ofPalm BeachCounty

OtherCommunityProviders(e.g., homelessproviders,aging services,developmentaldisabilities)

X X X X

Other LocalFoundations

X X

BusinessCommunity

X X X

OtherCommunityReps (e.g.,NAACP,neighborhoodassociations,retired,unknownaffiliation)

X X X X

Client Reps(e.g., NAMI)

X

Foster Parent X X

Community Alliance Mission and Discussion of Child Welfare

The following section reports on Community Alliance mission, discussion of childwelfare issues at meetings, and the Alliances’ relationships with DCF and leadagencies. Information is grouped under each Community Alliance.

Heart of Florida Community Alliance. The Heart of Florida Community Alliance meetsmonthly, with the District Administrator (DA) and representatives from the lead agency,who provide updates, reports, and progress information. They describe ongoing efforts,staff training, service updates, etc. They also bring in different specialists/speakers tothe meetings.

46

The Alliance recently changed its name from the CEO Roundtable to the Heart ofFlorida Community Alliance. The significance of the name change is demonstrated inthe new membership structure (5 members per county), which provides morerepresentation by having a total of 19 members for the tri-county area plus 4 membersat-large. Another primary reason for the change is that the Roundtable’s CEOmembership was dominated by court, law enforcement and government leaders.Additional areas being considered are local coalitions, early childhood services, mentalhealth/substance abuse, YMCA, and the Chamber of Commerce.

The Executive Director of Heartland for Children attends monthly meetings and is partof a 4 member committee that has been instrumental in redesigning the Alliance duringthe past several months. The new Community Alliance is focused on broader outcomesand its oversight functions are more oriented toward community performance goals.

Shared Services Alliance of Okeechobee and the Treasure Coast. This Alliancereported that they played an active role in local child welfare issues. The Allianceestablished a community outcomes committee, was instrumental in selecting the leadagency, and insured that the outcomes were part of the contract between the leadagency and DCF. The Alliance is active in Child Watch programs on an annual basis. Inaddition, the Alliance also discusses juvenile justice, mental health and substanceabuse issues on a regular basis.

The Shared Services Alliance of Okeechobee and the Treasure Coast includes the DAas a member of the Alliance and a member of the Executive Committee. The DA reportsany actions that occurred during the previous month between the Department and LeadAgency. The DA also keeps the Alliance abreast of actions impacting the DCForganization and issues raised by the State Office that may have an impact on childrenand their welfare. Often the Lead Agency is asked to have a representative present tonot only respond to the committee's inquiries but also to insure the agency is preparedfor the Alliance meeting.

Northwest Florida Community Alliance. This Alliance also routinely discusses childwelfare issues at their quarterly meetings. The Alliance recently approved the leadagency, Partnership for Families, and a December 2003 vote was taken to approve thetransition to become effective February 2004. The Alliance has received a transitionspreadsheet showing when the services will begin in the various areas covered by thecontract, and discussion has been focused on how the transition will work.

The District Office provides regular updates at each quarterly Northwest FloridaCommunity Alliance meeting regarding issues of mutual concern, and has a standingreport on the agenda of each Alliance quarterly meeting. In addition, the District uses e-mail updates to communicate with Alliance members on an as-needed basis. Alliancemembers are provided with updates on significant events that are time-sensitive in thismanner.

47

Community Alliance of Palm Beach County. This Alliance also reported muchactivity around child welfare issues. The CFC CEO attends all Alliance ExecutiveCommittee and regular meetings. The Alliance has advocated for increased funding forthe lead agency and for more community involvement on the CFC Board of Directors.

District One Community Alliance. Established in 2002, the District One CommunityAlliance is a group of community volunteers from Escambia, Santa Rosa, Okaloosa andWalton Counties that have agreed to work with DCF, its contractors, the judiciary, lawenforcement, other community agencies and stakeholders to steer initiatives to bettertheir communities for the benefit of children and families who reside in their servicearea. Voting membership is determined by legislative description and as determined bythe District One Community Alliance Bylaws. The meetings are held quarterly.

The coordinator for the Alliance in District One is FamiliesFirst Network’s CommunityLiaison who works closely with the executive committee and the members to ensure theAlliance is informed about child welfare issues and serves as a point of contact formembers and the community about the Alliance. This year the District One Alliance hasdefined its mission as “To build strong communities for children and families throughhuman services”, with the intention of broadening the Alliance’s scope.

Hillsborough County Community Alliance. The Community Alliance in HillsboroughCounty meets monthly, and HKI and the Suncoast Region report on specific outcomeand performance issues at each meeting. The Community Alliance Chair is an ex-officiomember on HKI’s Board of Directors.

Sarasota Community Alliance. While the majority of Alliances (e.g., the PinellasCounty Community Alliance, Children and Families Community Alliance of NortheastFlorida) discussed child welfare issues at every meeting, Sarasota’s Alliance presenteda variation on this theme. The Alliances were not mandated until two years after theoriginal CBC legislation, when Sarasota YMCA advisory group was already wellestablished. Naturally, the addition of a new advisory group raised issues of turf, power,and influence and a concern that the focus on CBC would be lost. There was also asense that the CBC advisory group represented a more community-driven planningapproach, in contrast to the Alliances that were mandated by state government.Therefore, the role of the existing CBC advisory group had to be reconciled and mergedwith the new Community Alliance. Sarasota County maintains both the CommunityAlliance and a separate stakeholders group run through the Sarasota YMCA leadagency. Thus the Community Alliance of Sarasota County was designed to examine allareas of health and human services and to work to improve the coordination of services.While the Sarasota Alliance does discuss child welfare issues, the conversationnormally deals with how such issues interface with other aspects of health and humanservices within Sarasota. In new CBC sites these issues should not be problematic, asthe Community Alliances should be involved in the planning of CBC and involved withthe lead agency from the beginning.

48

Ability of Alliances to Make Suggestions Regarding Lead Agencies

Some of the Alliances reported that while there was a preliminary communicationprocess in place to discuss issues with their respective lead agency(ies), the leadagency was too new to warrant much critical feedback. In these cases, Alliancesseemed pleased that the lead agency was routinely providing information on theirtransition process. In some communities, the Alliances have been able to assist newlead agencies with collective experience in areas such as foster care recruitment.

However, several Alliances reported that they are not able to make recommendations totheir lead agency because DCF Central Office has made it clear (via memo fromSecretary Regier) that the Alliance is “advisory only”, and therefore has no authority inthe opinion of many Alliance members. One Alliance Chair explained that while theymight occasionally make recommendations to their lead agency, the communicationprocess primarily involves listening to presentations and receiving updates, rather thanbeing solicited for recommendations.

Facilitators and Barriers to Communication Process with DCF and Lead Agencies

Several activities that promote the communication process were reported and areoutlined in Table 14, including regular attendance at Alliance meetings. As one Alliancechair explained, “The best benefit with our communications is the responsiveness of thelead agency. During their February report the Alliance requested a county-by-countybreakout of foster homes status. At the March meeting the lead agency provided acomplete breakdown that included the distribution of foster homes among the fourcounties, the number of new and closed cases by county, number of in-home and out-of-home care services provided by county, etc. This report gave an account to eachcounty as to what was happening in their communities relating to foster care.”

Another stakeholder explained how both sides benefit from a fluid dialogue: “TheAlliance benefits from being well informed so that information can be shared throughoutthe various agencies and communities represented by Alliance members. DCF benefitsfrom having Alliance member input on key issues, such as transition planning,readiness assessment, and resource development.” When Alliance members have theopportunity to discuss any items or concerns they might have with lead agency staff inthe context of an Alliance meeting, communication is not only facilitated, but a beneficialworking relationship is reinforced. Another Alliance stakeholder reported “They [the leadagency] have encouraged Alliance members to ask questions and visit them to see howthings are working.”

Several Alliance Chairs reported the concern that while the Florida Statute seems togive Community Alliances the authority to make recommendations and guide the leadagencies through community collaboration and consensus, the DCF Central Office hasemphasized that the Alliance is advisory only. The outcome of this statement has beenthat most negotiations between lead agencies and their district DCF office have beenstrictly between DCF and the lead agencies. The majority of Alliance members feel they

49

have been left “out of the loop” in choosing and approving lead agencies and thesystems of care developed by these agencies. The majority of Alliances expressed astrong desire to be involved in this process, but felt further clarification needed to comefrom the Legislature regarding the issue of whether or not they can have any authority.

One Alliance Chair explained the effect this lack of clarification has on other communitypartners: “I have to travel an hour each way for the meetings, and combined with thetime of the meeting, it ties up at least half of my day. When speakers/specialists do noteven show up, but rather send representatives because they sense that they don’t reallyneed to be there (e.g., sheriffs, school board members, judges, etc, sendrepresentatives in their place), this makes it seem questionable whether it’s critical forthe Alliance members to come together and meet.”

In sum, many Alliance Chairs are concerned that their members do not feel like they arecontributing stakeholders. Chairs suggested that problem resolution should be sharedby DCF, the lead agencies, and the Alliances. In addition, there is a need for moreongoing dialogue, as many Alliance members currently are only passively receivinginformation. Chairs suggest that lead agencies actively solicit feedback from Alliancemembers via open-ended questions regarding their system of care.

Other Alliance stakeholders pointed to the lack of communication with DCF CentralOffice on statewide issues such as budget or legislative matters impacting CBC andchild welfare. For example, one stakeholder suggested that DCF Central Office shouldhave notified Alliances about House Bill 723 and Senate Bill 1698 and provided briefanalyses of their impact so that Alliance members could make their points of viewknown to local constituents. Another specific concern was that only the larger(population wise) Alliances received attention from the State. Some smaller Alliancesfelt that they worked well with their community, but that the State did not care becauseof their smaller size and decreased political capital compared to other more populatedand politically connected areas of the State.

50

Table 14. Facilitations and Barriers to Communication

Facilitators• When DCF/lead agency is receptive to Alliance Chair’s request for

further engagement• High quality of professionalism and leadership of Lead Agency

administration• Excellent Communication between DA and lead agency CEO• When lead agency exceeds community’s initial expectations• When DA is member of the Alliance and any other pertinent

subcommittees• When lead agency is responsive to Alliance’s requests for

information• Email and/or newsletter updates

Barriers• Secretary and Central Office DCF have made it clear that Alliances

are “advisory only”• Alliances that cover several counties and multiple lead agencies may

lose local authority and momentum in shaping local systems of care• Alliances in less populated areas feel other Alliances are driving

decisions at the state level• No incentives to be an Alliance member, such as administrative

support, authority, or pay

Other Community Stakeholder Groups

Examples of community stakeholder groups beyond the Community Alliances can beseen in Table 15, as well as how their feedback is incorporated into lead agency designand activity.

For example, in addition to Heartland’s Board of Directors, Heartland receives guidancefrom Devereux’s (Heartland’s founding agency) national board. They have been anintegral part in providing financial assurances needed to proceed with CommunityBased Care. Heartland is also in the process of creating a Resource DevelopmentAdvisory Board so that local input can be sought and integrated into its businesspractices. This board may be involved in local resource development for prevention andcommunity involvement.

In addition, the Sarasota YMCA relies on feedback from two community advisorygroups whom they feel are closer to the ground on local child welfare issues than theirCommunity Alliance. The Manasota stakeholder group includes Manatee and SarasotaCounties, and tentatively Desoto County. It is comprised of community businessleaders, judges, attorneys, etc. There are approximately 40 members. They provideoversight of outcomes and financial records. The Sarasota YMCA provides this type of

51

information to them on a regular basis, and also prepares specific reports based on thegroups’ interests. This group has also taken over a form of the satisfaction survey sentto foster parents that the Department used to conduct. Reportedly, they have a higherresponse rate than DCF had, possibly due to the fact that they are an independententity. Currently, the Manatee County Commissioner is chair of the group. Similargroups have recently been formed in Pinellas and Pasco Counties as well.

The Sarasota YMCA also receives both resources and feedback from their CorporateCoalition, which spans Sarasota, Manatee, and Desoto Counties. Membership iscomprised of representatives from the local business community. The CorporateCoalition will adopt a specific program within the services the YMCA provides, and offerboth financial resources and volunteer work. Due to the extent of their involvement atthe programmatic level, they are able to offer many suggestions that the YMCAincorporates into programmatic design.

Table 15. Other Community Stakeholder Groups2

Lead Agency Stakeholder groups Incorporation of Feedback

CFC Faith-Based Advisory Board toassist with foster and adoptiveparent recruitment

Monthly meetings and agendas developed bythe advisory group.

FFN Promoting Safe & Stable FamiliesTask Force;Escambia County Chief JudgesChildren’s Council;Santa Rosa CEO Round Table;Emerald Coast Child AdvocacyCenter;FamiliesFirst Network ProviderForum;Foster Parent Support Groups;LCI/FFN and DCF SteeringCommittee

Input is recorded, compared between groups,trended over time, prioritized againstdemographics, performance measures andfiscal viability. Findings are incorporated intoNetwork goals, quality improvement plans andstrategic plans, with feedback to stakeholdersabout how their information was used.

HKI Hillsborough Children’sCommittee (all child servingagencies within the county) andthe HKI Collaborative Council(non-stakeholder agenciesinterested in direct participationwith HKI system of care andservice strategies).

Both committees meet monthly and HKI reportson specific outcomes and performancemeasures at each meeting. The collaborativecouncil has sub-committees tasked to addressspecific issues within the HKI system of care.The collaborative council has an ex-officio seaton the HKI Board of Directors and reportsmonthly at the Board meeting.

2 This reflects what was reported by lead agencies.

52

Lead Agency Stakeholder groups Incorporation of Feedback

Heartland Devereux’s national board;

Resource Development AdvisoryBoard

Heartland funnels the feedback from Devereux’sNational Board back through the Heartland forChildren Board of Directors. They haveprovided Heartland critical financial assurancesto support lead agency development.The goal for the newly created ResourceDevelopment Advisory Board is to gather inputto be integrated into lead agency businesspractices. The Board may also be involved inlocal resource development.

FSS Six neighborhood communityadvisory boards, representing allareas of the county

Teen Advisory Board

Meets monthly and is currently working ondistributing a prevention grant in their respectiveareas.

Assessments of services are conducted by thegroups.

UFF Leadership Council composed offoster parents, providers,substance abuse/mental healthstakeholders, local funders, andcommunity leaders

During start-up phase, Council met monthly andUFF gathered information useful to developingtheir system of care

ChildNet None as of now. Potentially aprofessional development board,but not currently.

Sarasota YMCA Manasota Stakeholder Group

Corporate Coalition

They provide oversight of outcomes andfinancial records. The Sarasota YMCA providesthis type of information to them on a regularbasis, and also prepares specific reports basedon the groups’ interests.The Corporate Coalition adopts a specificprogram within the services the YMCA provides,and offer both financial resources and volunteerwork. Due to the extent of their involvement atthe programmatic level, they are able to offermany suggestions that the YMCA incorporatesinto programmatic design.

Local Revenue Sources

The community can be valuable revenue contributors to the CBC initiative. As reportedin the CEO survey, several lead agencies are tapping into local community revenuesources. The following table (Table 16) lists local revenue sources as reported byseveral lead agencies.

53

Table 16. Local Revenue Sources

Lead Agency Local Revenue SourceCFC A local foundation and the local Substance

Abuse/Mental Health Office of DCFFFN For child welfare, local revenue sources consist of

matching efforts through the local provider network.HKI Grants from the Hillsborough County Children’s

Board, THINK Project, and various donations fromlocal businesses and individuals. Medicaid andADM are also funding sources that contribute anadditional $5+ million dollars (SIPP placements,Medicaid Therapeutic Group Homes, specializedtherapeutic foster care, mental health and substanceabuse services, BHOS, etc.)

Heartland A local Resource Development Advisory Council iscurrently under development. This Council will bemade up of Community Champions that will providesupport and guidance for community awareness andfundraising campaigns.

FSS Jacksonville Children’s Commission (at start-up)ChildNet Children’s Services Council; grants; Community

FoundationSarasota YMCA Corporate Coalition provides both financial

resources and volunteers.

Summary of Research Question 2

In conclusion, this research question sought to explore the effectiveness of CBC atinvolving the community as service partners and resource contributors. Lead agenciesoperating 18 months or more had a smaller average number and range in number ofproviders and contracts than agencies operating for a shorter periods. This reduction inproviders and contracts may be related to a trend towards stabilization and maturity ofthe service system, although this area will continue to be tracked over time. Leadagencies reported relatively little success in generating local revenues, especially afterthe start-up phase.

The Community Alliances represent a potential important community governancepartnership for lead agencies. Of the 7 Community Alliances who responded to thestudy’s survey, only 3 had representatives from mental health or the businesscommunity; only 2 included local foundations or foster parents, and only 1 CommunityAlliance had substance abuse or consumer representatives. The majority of CommunityAlliances reported that child welfare is a standing agenda item but several felt limited bytheir “advisory role” and expressed desire for more involvement in local system of caredevelopment. An unresolved issue for lead agencies is clarification of the governanceroles of boards of directors, Community Alliances, and other community stakeholdergroups.

54

RESEARCH QUESTION 3:How effective is Community-Based Care at meeting ASFA outcome requirementsfor child safety, permanency and well-being?

Introduction

The 1997 Adoption and Safe Families Act (ASFA) continues to offer the primary goals inchild welfare practice and monitoring these outcomes remains critical. Monitoring ofchild welfare is an important task because according to Adoption and Foster CareAnalysis and Reporting System (AFCARS) data, in fiscal year 2001, 295,000 childrenentered foster care nationwide and 542,000 were in care, with a median length of stayof 19 months (Administration for Children and Families (AFC); U.S. Department ofHealth and Human Services, 2003a).

ASFA required objectives (i.e., child safety, permanency, and well-being) are used inevaluating State outcome performance by the Child and Family Services Reviews(CFSR). Two measures assess attainment of child safety: the rate of recurrence ofmaltreatment and the rate of reentry into out-of-home care. According to theAdministration for Children and Families, the maltreatment recurrence rate within 6months of initial substantiated report nationwide was 8.9%, and 8.4% in Florida in 2001(ACF; U.S. Department of Health and Human Services, 2003b). The rate of reentry intofoster care varied by state. Among the states that provided sufficient data to calculatethe rate of reentry into foster care within 12 months of discharge from a prior foster careepisode, the percentages ranged from 0.9 to 27.7, with a median of 10.3%. Eighteenstates met the national standard of 8.6% for this measure (Child Welfare Outcomes2000: Annual Report; ACF, 2000; U.S. Department of Health and Human Services).According to data provided by DCF, 7.9% of children in Florida reentered out-of-homecare within 12 months after exiting by May, 2003.

ASFA also emphasizes timely permanency, regardless of the type of stable placement.Therefore, reunification, adoption, and placement with relatives are viewed as importantindicators of this outcome. Nationwide, 57% of children who exited foster care werereunified with their parents or primary caregivers, 10% were placed with relatives, and18% were adopted during 2001 (ACF, 2003; U.S. Department of Health and HumanServices, 2003).

Maintaining a focus on child safety, permanency, and child well-being is an ongoingchallenge for child welfare agencies nationwide. Case staffings become the mechanismby which this focus is maintained, as well as creating opportunities for supervisors, lineworkers, and other partners to discuss the needs of individual families while addressingthese issues. Focus on case staffings was introduced to the statewide evaluation ofCBC during the site visits in FY02-03. At that time, it was found that there was a primaryfocus on child safety, and that these meetings also allowed for ongoing supervision ofthe cases (Paulson, et. al., 2003). Variability existed across sites in how they wereimplementing statutorily or Department-mandated staffings.

55

The following evaluation questions, indicators, and data sources were used to evaluatethe effectiveness of CBC at meeting AFSA standards (Table 17):

Table 17. Research Question 3: Evaluation Questions, Indicator/Analysis, andSource

Evaluation Questions Indicator/Analysis Source

Are there mechanisms in place at thelead agency level to assure continuedfocus on child safety, permanencyand well-being?

Description of case staffingsat each lead agency

Lead AgencyCommunication

What proportion of children exit out-of-home care during first 12 monthsafter entry?

Proportion of children whoexited out-of-home care HomeSafenet

Data

Is there a significant differencebetween the proportion of childrenexiting out-of-home care for CBC bylead agency length of service contractcategory?*

Difference in proportion forchildren who have exited out-of-home care by category

HomeSafenetData

What proportion of children who re-enter out-of-home care within 12months after exiting their first episodeof out-of-home care?

Proportion of childrenreentering out-of-home care

HomeSafenetData

Is there a significant differencebetween the proportion of childrenreentering out-of-home care in CBCby service contract category?*

Difference in proportion forchildren who have reenteredout-of-home care by category

HomeSafenetData

How effectiveisCommunity-Based Careat meetingAFSAoutcomerequirementsfor childsafety,permanency,and well-being?

How effectiveisC it

What proportion of childrenexperience reoccurrence ofmaltreatment within 12 months oftheir first episode of maltreatment?

Proportion of children withreoccurrence of maltreatment

HomeSafenetData

56

Evaluation Questions Indicator/Analysis Source

Is there a significant differencebetween the proportion of childrenwith recurrence of maltreatment inCBC by lead agency length of servicecontract category?*

Difference in proportion forchildren with a recurrence ofmaltreatment by category

What is the proportion of children inout-of-home care who were reunifiedwith their parents?

Percentage of children whowere reunified

HomeSafenetData

Is there a significant differencebetween the proportion of childrenreunified with parents by site, time(FY01-02 and FY02-03) and site xtime?

Difference in proportion ofchildren reunified with parentsby site, time and Site x Time

What proportion of children areplaced with relatives for out-of-homecare?

Percentage of children placedwith relatives

HomeSafenetData

Is there a significant differencebetween the proportion of childrenplaced with relatives for out-of-homecare by site, time (FY01-02 andFY02-03) and site x time?

Difference in proportion ofchildren placed with relativesby site, time and Site x Time

What proportion of children have afinalized adoption?

Percentage of children withfinalized adoptions

HomeSafenet

Is there a significant differencebetween the proportion of childrenwith finalized adoptions by site, time(FY01-02 and FY02-03) and site xtime?

Difference in proportion ofchildren with finalizedadoptions

HomeSafenet

*Note: See page 63 for service contract categories.

57

Lead Agency Case Staffing Structures

Methods

During the first half of FY03-04, evaluators contacted the five lead agencies that wereoperating during FY02-03 (Family Continuity Programs, FamiliesFirst Network,Hillsborough Kids, Inc., Partners for Community-Based Care, Inc., and theSarasota YMCA). Each site was asked to complete the table in Appendix 6 as relatedto safety and permanency. It was found that each of the agencies had created anintricate schedule of staffings to address both routine and emergent issues. Theseschedules met statutory requirements, but also strived to match varying agency practiceapproaches regarding inclusion of parents, youth, and community providers.

During the latter half of FY03-04, an additional six lead agencies were contacted (Childand Family Connections [CFC], ChildNet, Family Support Services of NorthFlorida [FSS], United for Families [UFF], Heartland for Children [Heartland], andPartnership for Families [PFF]). The findings from these efforts are presented in Table18.

Table 18. Lead Agency Case Staffing Structures (Spring 2004)

Lead AgencyType of Staffing

CFC ChildNet FSS UFF Heartland

Authority

Transfer from CPI to the Lead AgencyESI staffing X X X X X CFOP 175-23; F.S.

Ch. 39ProtectiveInvestigations staffing

X Lead agency policy

Initial Family CaseConference

X Lead agency policy

Permanency StaffingsPermanency staffing X X X X X Administrative Rule

65C-13.020; s.39.012, F.S.

Family AssessmentReviews

X Florida’sPerformanceImprovement Plan

Three-monthPermanency Staffing

X X X Lead agency policy

IndependentLiving/Child aging outof system

X X X X X s. 409.1451, F.S.

Permanency/Reunification staffing

X X X X X Administrative Rule65C-13.020; s.39.012, F.S.

58

CFCChildNet

FSSUFF

Heartland

Case Review StaffingsCase Review staffing X X X X X Administrative Rule

65C-11, 12, 13, 15Case PlanConference

X X X X X Administrative Rule65C-11, 12, 13

Family TeamConference

X X Lead agency policy

Family ServicesTeams

To beimplement

ed

Lead agency policy

Adoption-related StaffingsAdoption staffing X X X X X Administrative Rule

65C-16Adoption ReviewCommittee

X Statutorilymandated onspecial concerncases

Adoption Action PlanReview

X X X X X Administrative Rule65C-16.003

Adoption MatchMeeting

X Lead agency policy

Separated Siblings X X X X X Statutorilymandated; Agencypolicy

Placement MaintenanceDisruption/Conflict Resolutionstaffing

X X Lead agency policy

Placement SupportStaffing

X X Lead agency policy

Unique Staffing PracticesCommunity Resources X Lead agency policyMultidisciplinary Team(MDT) Committee

X Lead agency Policy

New Baby Staffing X Lead agency policyTeen Runaway Staffing X Lead agency policyUtilization ManagementStaffing

X Lead agency policy

59

Early Services Intervention (ESI) staffing is required by statute to facilitate the transferof cases from the Investigating Agency to the CBC. These meetings are convened bythe Child Protective Investigator (CPI) unit (either DCF or Sheriff’s Office) andrepresentatives of the lead agency are invitees, not facilitators. As shown in Table 18,there is consistent occurrence of required staffings (e.g., ESI staffings, permanencystaffings, etc.). There is, however, some variability in the implementation. For example,all agencies hold permanency staffings, but several agencies have chosen to have astaffing at three months rather than waiting for the first Judicial Review at six months.ChildNet’s Family Assessment Review serves the same purpose as the three-monthPermanency Staffing,

Permanency Staffings are authorized by Administrative Rule 65C-13.020 and refer to awide range of staffing activities (e.g., three-month review, reunification, adoption, etc.).Therefore, some agencies call everything a Permanency Staffing while others haveadopted other names. As it currently exists, this Rule requires that the staffing beattended by the following individuals: foster care supervisor, foster care counselorresponsible for the case, adoption supervisor, district program specialist, child’s fosterparent or substitute caretaker, GAL, involved professionals, and district legal counsel.Adherence to this Rule does not preclude the inclusion of parents at these staffings.

As with the earlier survey, a high importance is placed on permanency as evidenced bythe frequency of Permanency Staffings. Heartland has chosen to conduct theirpermanency staffings at Months 3, 5, 8 and 11 from the date of removal. Specialpermanency staffings are held for families who exceed length-of-stay targets. The 3-month permanency staffing is an informal staffing between the case manager and theirsupervisor and usually does not involve other participants. The Permanency Staffingswere also found to serve various purposes. ChildNet chooses to use their PermanencyStaffings as their primary staffings. As needed, these staffings also serve the purpose ofIndependent Living staffing (6 months prior to a child’s 18th birthday), Reunificationstaffing (as requested by Child Advocate), Adoption staffing (quarterly), Adoption ActionPlan Review, Adoption Match Meeting (as needed), and Family Assessment Reviews. Anumber of unique staffing practices were identified that create and maintain a focus onchild safety, permanency, and child well-being. They also create opportunities forinclusion of family members, community providers, and individuals identified assignificant in the lives of the families served. These practices include:

• Initial Family Case Conference (Child & Family Connections)

• MDT (Multidisciplinary Team Committee) Staffing (Child & Family Connections)

• Family Service Teams (United for Families)

• Use of expert panel during ESI staffing (Heartland for Children)

• Review of sibling separation for reunification (Family Support Services)

• Visitation staffings for siblings separated in foster care (United for Families)

60

At the beginning of their involvement with a family, CFC implements an Initial FamilyCase Conference. These conferences are held within 72 hours of the transfer to casemanagement services. The meetings are convened by the Program Managers for casemanagement and typical invitees include parents, child protective investigator (CPI),CPI supervisor, lead agency supervisor/case management supervisors, and allsignificant people in the child or children’s lives. At the deeper end of out-of-homeplacement, CFC implements Multidisciplinary Team (MDT) Staffings. These staffingsare to review the progress and needs of children in residential treatment. TheCommittee is chaired by two CFC clinicians and is attended by District representativesfrom the Department of Juvenile Justice, Developmental Disabilities and MentalHealth/Substance Abuse within DCF, the SED Network (SEDNet) or EH/SED Program,the Agency for Health Care Administration (AHCA) and case managers and/orsupervisors or others as deemed clinically necessary. Reviews of cases occur every 90days for children in residential treatment (RTC) and therapeutic group care (TGC orSTGC). They occur every six months for therapeutic foster care (TFC or STFC)placements.

UFF plans to implement Family Service Teams to be held every 90 days. Thesestaffings will provide an opportunity for the family to meet with their supports andproviders as they work towards achieving case plan goals, similar to existing casereview staffings but more frequent. Invited attendees will include the parent and child (ifappropriate), the UFF case manager and supervisor, Guardian ad Litem, foster parent,child welfare legal services, therapists, school personnel and any other interestedparties.

Heartland participates in ESI staffings with an invitee list that has been expanded toinclude an Expert Panel with representation from DJJ, Mental Health/Substance Abuse,and Domestic Violence along with other providers deemed necessary to an individualfamily (e.g., Healthy Families, therapist, school guidance counselor, etc.). This panel oftreatment professionals uses their clinical expertise and knowledge of the child welfaresystem to review and assess the service intervention needs of the child and family andprovide expert opinion regarding the direction of the case including identification ofappropriate permanency options.

Sibling separation staffings are required following termination of parental rights (TPR)by Administrative Rule 65C. Some lead agencies, however, report additional siblingseparation staffings, recognizing the impact of sibling separation on the well-being ofchildren. These staffings are authorized by agency policies. FSS holds a staffingmonthly to review siblings who have been separated for reunification. These meetingsare attended by the caseworker, placement staff and program staff. United for Familiesholds additional meetings on a monthly basis to review siblings who have beenseparated in foster care and prior to TPR. These meetings are attended by the agencysupervisor, case manager, Guardian ad Litem, and foster parent support worker andsupervisor.

61

Programmatic Outcomes

DCF assesses the performance of CBC agencies in Florida according to ASFA requiredobjectives. The Department produces several management and performance reportsusing HomeSafenet (HSn) data to monitor outcomes for children (see Table 19)3. Datafor these indicators can be found on the DCF intranet website, and they are not includedin this report.

Table 19. FL DCF Reports Based on HSn

Report Description of the Indicator

Children in Out-of-Home Care by Timein Care

The number of all children who are in anactive removal, by length of stay (more orless than 12 months), with median lengths ofstay by type of placement.

Clients Active in Out-of-Home Care A listing of the children in a current removal.This report shows the length of time (inmonths) the client has been in the currentremoval and the number of placementsduring this period of removal.

Children Exiting Foster Care in 12Months

The percent of children exiting out-of-homecare within 12 months of those children whohave been removed from their home one-year prior to the report period.

Percentage of Children Entering FosterCare who Returned in Less Than 12Months of a Prior Reunification orRelease to Relatives

Percentage of children entering a removalepisode who returned to foster care less than12 months after a prior reunification. Theprior discharge reason must be “reunificationwith parents or removal home caregivers” or“long term custody to relative.” This reporthas two report periods: (1) the previousquarter of the current federal fiscal year; (2)the current federal fiscal year to date.

Children Adopted Within 24 Months A percentage of children from the precedingquarter that are adopted (adoption finalized)within 24 months of a removal.

AFCARS Adoption Compliance Rate The number of children with a dischargereason of “adoption finalized” at any timeduring the AFCARS reporting period.

3 Only reports that were relevant to the indicators selected for this evaluation are included in the table.

62

Indicators Calculated for Current Evaluation

For this evaluation six quantitative indicators were selected to measure child safety,permanency, and well-being. The rate of reentry into out-of-home care and the rate ofrecurrence of maltreatment were selected as indicators of CBC counties’ efforts toattain the goals of child safety and well-being. The proportion of children exiting out-of-home care was selected as an indicator to examine each CBC’s performance towardsachieving permanency. Finally, the proportion of reunification with parents or relativesand the proportion of children with finalized adoption were the indicators selected todescribe specific permanency outcomes. These quantitative child protection indicatorswere selected in consultation with DCF and were developed in collaboration with theOffice of Child Welfare and Community-Based Care. Although the quantitative childprotection indicators used in these analyses were modeled after the Performance-Based Program Budgeting (PB2) program4 and federal Adoption and Out-of-Home CareAnalysis and Reporting System (AFCARS) indicators, FMHI’s access to the original,disaggregated child-protection databases allowed us to construct indicators that weremore appropriate for detailed geographic and longitudinal statistical analysis. Forexample, analysis of these indicators developed by FMHI relied on entry or exit cohortsof children, as opposed to groups of children who were in care at a specific point-in-time, thus reducing selection bias. Moreover, the statistical methods used to analyze thedata were chosen to maximize the utility of the out-of-home care episode-level data thatwere made available by DCF. The above-mentioned indicators were calculated forFY01-02 and FY02-03.

Source of Data

The primary source of data for the quantitative child protection indicators used in thisreport was HomeSafenet (HSn), the State Automated Child Welfare Information System(SACWIS) for the State of Florida.

Methods

All counties that have transitioned or are currently transitioning to Community-BasedCare as of January 2004 were included in the analysis, and indicators were calculatedfor each lead agency. In addition, all lead agencies were categorized according to thelength of time they have had a service contract. The first category includes leadagencies that had a service contract for 18 months or longer (i.e., fully accountable), thesecond category includes lead agencies that had a service contract for 7-17 months(i.e., shared accountability), and the third category includes lead agencies that had aservice contract for 6 months or less (i.e., baseline). Lead agencies that did not beginimplementing CBC during FY02-03 but had a service contract for 6 months or less as ofJanuary 2004 were included in the 6 months or less category. Table 20 delineates theseservice contract categories.

4 See http://ewas1.dcf.state.fl.us/mspt/enc/encyclopedia.cfm (website available via DCF Intranet only).

63

Table 20. Lead Agencies (Counties Served) by Length of Service Contract

Service Contact held 18 months or more

Partners for Community-Based Care(Volusia, Flagler)

Sarasota YMCA(Sarasota, DeSoto, Manatee)

Service Contract held 7-17 months

Hillsborough Kids, Inc.(Hillsborough)

FamiliesFirst Network(Escambia, Walton, Okaloosa, Santa Rosa)

Family Continuity Programs(Pasco and Pinellas)

Service Contract held 6 months or less*

Child and Family Connections (Palm Beach)

ChildNet(Broward)

Family Support Services of North Florida(Duval)

Heartland for Children(Polk, Hardee, Highlands)

Partnership for Families(Holmes, Washington, Bay, Gulf, Calhoun,

Jackson)United for Families

(Okeechobee, St. Lucie, Indian River, Martin)*Note: These lead agencies did not begin the implementation of Community-Based Care during FY02-03but had service contracts for 6 months or less as of January 2004.

Statistical methods used for this analysis include:

• A survival analysis (a statistical procedure that allows analysis of datacollected over time as well as incorporation of information about cases wherethe event of interest did not occur during data collection [e.g., children whodid not exit out-of-home care during the 12-month period]) of time-to-eventindicators (i.e., percentage of children exiting out-of-home care, reentry intoout-of-home care, and recurrence of maltreatment).

64

• Logistic regression for point-in-time indicators (i.e., percentage of childrenreturned to parents, percentage of children placed with relatives, and percentageof children with finalized adoption).

The lead agency length of service contract categories were compared by site onpercentage of reentry into out-of-home care, recurrence of maltreatment and childrenexiting out-of-home care. The contract service categories were compared by site, bytime (i.e., FY01-02 was compared to FY02-03) and by site X time interaction on thepercentage of children returned to parents, the percentage of children placed withrelatives, and the percentage of children with finalized adoption.

These indicators were designed specifically for this evaluation and to most accuratelymeasure outcomes for children in the child welfare system as well as agencyeffectiveness. However, because specific algorithms were used to develop theindicators, the results based on these indicators and presented in this section may notbe directly comparable with national standards. Inclusion of the utilized indicators infuture analyses will be necessary to allow for a proper longitudinal comparison.

LimitationsA few limitations should be noted. First, because an FY01-02 entry cohort was used tocalculate the recurrence of maltreatment indicator, we needed to obtain the data fromThe Florida Abuse Hotline Information System (FAHIS) because FAHIS had nottransitioned to HSn during FY01-02. Subsequently, children who were not served byCBC counties could not be identified (i.e., could not be linked to HSn). Therefore, evenif CBC counties did not serve some children but recurrence of maltreatment occurred inthese counties, lead agencies were held accountable for the recurrence. Second, theservice contract categories were developed as of January 2004. However, the dataused for the calculation of the indicators included FY01-02 and FY02-03 entry and exitcohorts. Finally, longitudinal analyses were not performed for percentage of childrenexiting out-of-home care, reentry into out-of-home care, and recurrence of maltreatmentindicators because 12-month follow-up data for the FY02-03 entry cohort are not yetavailable.

Reentry into Out-of-Home Care

Description of the IndicatorThe calculation of this indicator, one of the two measures for child safety and well-being,was based on exit cohorts of children (i.e., children who exited their first out-of-homecare episode during FY01-02 or who had a discharge date during this fiscal year) andon unduplicated counts of children (i.e., only children who exited their first episode ofout-of-home care were counted). Survival analysis was used to calculate the proportionof children reentering out-of-home care within 12 months of FY01-02 after exiting theirfirst episode. Reentry into out-of-home care was identified either by a prior removal(indicated by a previous removal date) or by a new removal date after a discharge date

65

existed for the same child identified by a unique ID assigned by the HomeSafenetsystem.

ResultsGeneral results of the analysis of reentry into out-of-home care are shown in Table 21.The lead agencies were ranked in ascending order according to the proportion ofchildren reentering an out-of-home care episode.

Table 21. Percentage of Children who Exited Out-of-Home Care in FY01-02 andReentered Within 12 Months by Lead Agency

Lead Agency Total number of cases inexit cohorts

Estimatedprobability of

reentering out-of-home care (%)

Service Contract held 18 months or more

Partners for Community-Based Care

543 8.1

Family ContinuityPrograms

1,085 8.3

Sarasota YMCA 794 13.4Service Contract held 7-17 months

Hillsborough Kids, Inc. 1,442 10.5FamiliesFirst Network 1,100 16.8

Service Contract held 6 months or less*

Family Support Services 1,426 5.2ChildNet, Inc. 1,573 5.9

Child and FamilyConnections

784 6.9

Heartland for Children 1,433 10.1United for Families 728 12.1Partnership for Families 335 12.5

Median = 10.12% Mean = 9.99%

* Note: These lead agencies did not begin the implementation of Community-Based Care during FY02-03but had service contracts for 6 months or less as of January 2004.

When statistical comparison was done for service contract categories (see Figure 7),some statistically significant differences were found. The proportion of childrenreentering out-of-home care for lead agencies that had a service contract for 6 months

66

or less (8.3%) was significantly lower than the proportion of children reentering out-of-home care for lead agencies that had a service contract for 7 -17 months (13.3%) andsignificantly lower compared to lead agencies that had a service contract 18 months orlonger (9.9%). The proportion of children in lead agencies that had a service contract for7-17 months (13.3%) was significantly higher when compared to the proportion ofchildren reentering out-of-home care for lead agencies that had a service contract for 18months or longer (9.9%; see Appendix 7).

Figure 7. Percentage of Children who Exited Out-of-Home Care in FY01-02 and Reentered Within 12 Months After their Exit by Lead Agency Service Contract Category

9.9%

13.3%

8.3%

0

5

10

15

20

Service Contract 18 months or longer Service Contract between

7 and 17 monthsService Contract 6 months or less

n =2,422 n =2,542 n = 6,279

Proportion of Children with Recurrence of Maltreatment

Description of the IndicatorThe rates of recurrence of maltreatment, the second indicator for child safety and well-being, reported here are estimates produced by survival analysis. Data for this analysisconsist of unduplicated counts of maltreated children grouped by fiscal year entrycohort. Only children with “founded” maltreatment (i.e., when the protective investigationresulted in a verified finding of abuse, neglect, or threatened harm and when there wassome indication of any type of maltreatment) were included in the analysis. Recurrenceof maltreatment was defined as a second founded maltreatment episode within 12months of a child’s first founded episode. The first and second episodes of maltreatmentwere selected based on the received dates. The data extract used for this analysis

67

included all cases with reported maltreatment during FY01-02 and FY02-03. All childrenwho were found to be maltreated during FY01-02 were included in the analysis andwere followed for 12 months after their first episode.

ResultsGeneral results of the analysis of recurrence of maltreatment are shown in Table 22.The lead agencies were ranked in ascending order according to the proportion ofchildren who had a second episode of founded maltreatment.

Table 22. Percentage of Children who Experienced Recurrence of MaltreatmentWithin 12 Months After Their First Episode Based on FY01-02 Cohort by LeadAgency

Lead Agency Total number of cases inentry cohorts

Estimatedprobability ofmaltreatment

recurrence (%)Service Contract held for 18 months or more

Family ContinuityPrograms

10,749 9.9

Sarasota YMCA 4,844 10.9

Partners forCommunity-BasedCare

5,047 12.8

Service Contract held 7-17 months

FamiliesFirst Network 6,165 10.9Hillsborough Kids, Inc. 9,382 11.3

Service Contract held for 6 months or less*

ChildNet, Inc. 10,671 8.2

Family SupportServices

7,037 8.4

Child and FamilyConnections

8,613 9.9

United for Families 3,240 10.3Partnership forFamilies

3,684 11.7

Heartland for Children 7,878 14.2 Median = 10.86% Mean = 10.78%

*Note: These lead agencies did not begin the implementation of Community-Based Care during FY02-03but had service contracts for 6 months or less as of January 2004.

68

As shown in Table 22, FCP had the lowest (9.9%) proportion of children with recurrenceof maltreatment among CBC lead agencies with service contracts for 7 months orlonger. ChildNet, Inc. had the lowest (8.2%) proportion of children with recurrence ofmaltreatment compared to both CBC agencies that have held their service contracts for7 months or longer and agencies that have held their service contracts for 6 months orless.

The mean rate for this distribution was 10.78% and the median was 10.86%. Heartlandfor Children had the highest recurrence of maltreatment rate for the FY01-02 entrycohort (14.2%; see Appendix 7).

When statistical comparison was done between service contract categories (see Figure8), statistically significant differences were found. The proportion of children withrecurrence of maltreatment for lead agencies that had a service contract for 6 or lessmonths was significantly lower (10.2%) compared to the proportion of children withrecurrence of maltreatment for lead agencies that had a service contract for 7-17months (11.2%), and significantly lower compared to lead agencies that had theirservice contracts 18 months or longer (10.9%). There was no statistically significantdifference when the proportion of children with recurrence of maltreatment for leadagencies that had a service contract for 7-17 months was compared to the proportion ofchildren with recurrence of maltreatment for lead agencies that had a service contractfor 18 months or longer (see Figure 8 and Appendix 7).

69

Percentage of Children who Experienced Recurrence of Figure 8. Percentage of Children who Experienced Recurrence of Matreatment Within 12 Months after their First Episode Based on FY01-02 Cohort by Service Contract Category

10.9 11.210.2

0

5

10

15

20

Service Contract 18 months or longer

Service Contract between 7 and 17 months

Service Contract 6 months or less

20,640 15,54741,123

Proportion of Children Who Exited Out-of-Home Care Within 12 Months

Description of the IndicatorThe proportion of children who exited out-of-home care during the first 12 months afterentry in FY01-02 was an estimate of the percentage of children who had fewer than 12months length of stay in out-of-home care (one of the ASFA goals). This indicator was ameasure of permanency and was based on duplicated counts of children, that is,children could have multiple episodes of out-of-home care within a fiscal year. Hence,all children who entered out-of-home care during FY01-02 and began an out-of-homecare episode, including children who began an episode more than once during thesame year, were incorporated into the analysis.

This definition of an out-of-home care episode was developed in collaboration with theDCF Office of Child Welfare and Community-Based Care. An out-of-home care episodewas defined as a continuous period of time in out-of-home care that began on the datewhen the child was removed from parents or home caregivers (i.e., removal date) andended on the date when the child was discharged from an episode of out-of-home care

70

(i.e., discharge date). An out-of-home care episode may consist of multiple placements(e.g., family shelter home, residential treatment, pre-adoptive home, supervised practiceindependent living). The indicator was based on entry cohorts, that is, only children whoentered out-of-home care during this particular fiscal year were included in the analysis.The decision to use entry cohorts was made in order to more accurately assess CBCprogram effects. The proportion of children who exited out-of-home care was calculatedusing survival analysis. Therefore, the percentages represent estimates of probabilitiesthat a certain event (e.g., exit from out-of-home care) will occur. This indicator wascalculated for existing lead agencies and for three service contract categories.Statistical comparison was done for three categories of lead agencies.ResultsTable 23 shows the percentages of children exiting out-of-home care based on FY01-02entry cohort. The lead agencies were ranked in ascending order according to thepercentage of children exiting out-of-home care. Although some lead agencies (e.g.,CFC, ChildNet, Inc., etc.) had not begun implementing CBC in FY01-02, the decisionwas made to include them in order to obtain pre-CBC or baseline data for theseagencies.

71

Table 23. Percentage of Children who Entered Out-of-Home Care in FY01-02 andExited Within 12 Months by Lead Agency

Lead AgencyTotal number ofcases in entry

cohorts

Estimatedprobability ofexiting out-of-home care (%)

Service Contract held 18 months or more

Partners for Community-Based Care

623 42.3

Family ContinuityPrograms

1,425 45.0

Sarasota YMCA 679 59.2Service Contract held 7-17 months

Hillsborough Kids, Inc. 1,504 37.1

FamiliesFirst Network 1,074 63.3Service Contract held 6 months or less*

Child and FamilyConnections

850 32.8

ChildNet, Inc. 1,062 40.9Heartland for Children 1,268 42.4Family Support Services 1,440 52.9Partnership for Families 346 59.6United for Families 621 60.4 Median = 45.00% Mean = 48.73%

* Note: These lead agencies did not begin the implementation of Community-Based Care during FY02-03 but had service contracts for 6 months or less as of January 2004.

As shown in Table 23, FFN had the highest (63.3%) proportion of children exiting out-of-home care during 12 months after entry during FY01-02 compared to all examined CBCagencies. CFC had the lowest (32.8%) proportion of children exiting out-of-home carecompared to all examined CBC agencies. Two out of five CBC lead agencies (i.e.,Sarasota YMCA and FFN) had higher than average proportion of children exiting out-of-home care. This corresponds to shorter average lengths of out-of-home episodes forthese lead agencies. When statistical comparison was done between service contractcategories, no statistically significant differences were found (see Figure 9 and seeAppendix 7).

72

Figure 9. Percentage of Children who Entered Out-of-Home Care in FY01-02 and Exited Within 12 Months by Service Contract Category

48.5% 48.8% 47.4%

0

10

20

30

40

50

60

Service Contract 18 months or longer

Service Contract between 7 and 17 months

Service Contract 6 months or less

n = 2,727 n = 2,578 n = 5,587

Proportion of Children Returned to Parents or Legal Guardians andPlacement with Relatives After Exiting Out-of-Home Care

The following indicators (i.e., the proportion of children reunified with parents, theproportion of children placed with relatives, and the proportion of children with finalizedadoption) are not considered to be performance measures; they are utilized to describedifferent permanency outcomes for lead agencies with various lengths of servicecontracts.

Description of the Indicator

Analysis of the following two indicators was based on exit cohorts of children, definedhere as the duplicated number of children who exited out-of-home care during aparticular fiscal year. Children in out-of-home care whose case status was not closed(i.e., did not have a discharge date) and children who exited out-of-home care eitherbefore July 1, 2001 or after June 30, 2003 were excluded from the analysis.

73

Classification of children returned to “parents of removal home caregivers” was basedon “reasons for discharge” from the HSn.

The indicator “placement with relatives” for the FY01-02 cohort also was based on“reasons for discharge” from HSn and included children placed with relatives or non-relatives who obtained guardianship and children placed with relatives who were givenlong-term custody. For the FY02-03 cohort this indicator was based on “reasons fordischarge” from HSn and included children that were placed only with relatives whoobtained long-term custody or guardianship. All indicators were calculated aspercentages. The numerator for the percentage of children reunified with parents for theFY01-02 cohort is the number of children who exited out-of-home care and werereturned to parents or legal guardians during FY01-02. The denominator is the totalnumber of children exiting out-of-home care during FY01-02 (the denominator wasindicated as the number of cases in exit cohorts). Similarly, the numerator for thepercentage of children reunified with parents for the FY02-03 cohort is the number ofchildren who exited out-of-home care and were returned to parents or legal guardiansduring FY02-03.

The numerator for the percentage of children who were placed with relatives is thenumber of children who exited out-of-home care and were placed with relatives. Theserelatives were given either long-term custody or guardianship. The denominator is thesame as above. This indicator also was calculated for FY01-02 and FY02-03 cohorts.

Results: Children Returned to Parents After Exiting Out-of-Home Care

General results of the analyses based on FY02-03 regarding the proportion of childrenreunified with parents are shown in Figure 10. Findings indicated that during FY02-03,HKI and Heartland had the highest proportion of children who were returned to parents(61.2% and 59.2%, respectively). FSS had the lowest proportion (36.1%) of childrenwho were returned to parents. The median for this distribution was 51.1% and the meanwas 50.1%.

By comparison, during FY01-02 United for Families and Heartland for Children leadagencies had the highest proportions of children who were returned to parents (66.5%and 62.2%, respectively) and ChildNet, Inc. had the lowest proportion (45.4%) ofchildren who were returned to parents (see Appendix 7).

74

Figure 10. Percentage of Children Reunified with Parents, Placed with Relativesand with Finalized Adoptions Based on FY02-03 Cohorts by Lead Agency5

36.1%

38.5%

39.0%

47.1%

49.7%

51.1%

53.9%

57.3%

57.8%

59.2%

61.2%

38.4%

39.3%

29.6%

12.0%

20.4%

29.2%

26.6%

27.0%

21.6%

24.3%

20.8%

12.0%

9.1%

15.9%

24.4%

11.9%

12.3%

10.4%

8.8%

12.0%

9.9%

6.5%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Family Support Services*

Family Continuity

ChildNet, Inc.*

Child and FamilyConnections*

Partnership for Community-Based Care

Sarasota YMCA

Partnership for Families*

FamiliesFirst Network

United for Families*

Heartland for Children*

Hillsborough Kids, Inc.

Percentage of Chidren Reunified With Parents After Exiting Out-of-Home CarePercentage of Children Placed With Relatives After Exiting Out-of-Home CarePercentage of Children with Adoption Finalized After Exiting Out-of-Home Care

*Note: These lead agencies did not begin the implementation of Community-Based Care during FY02-03 but hadservice contracts for 6 months or less as of January 2004.

5 Percent exiting for each lead agency equals 100%; the remainder would consist of other reasons for exiting such asemancipation and children aging out of the system among others.

75

In order to examine the differences between proportions of children reunified with theirparents by service contract category, the following comparisons were made:

(a) by site (i.e., lead agencies that have been held their service contracts for 18months (n = 2,997) or longer, lead agencies that have been held their servicecontracts between 7 and 17 months (n = 3,016), and lead agencies that havebeen held their service contracts for 6 months or less (n = 7,231);

(b) by time (FY01-02 vs. FY02-03); and(c) by site X time (i.e., differences in the rate of change in indicator between sites

over time).Logistic regression was used to make a statistical comparison. Results of the logisticregression analysis for this indicator are shown in Appendix 7.

As shown in Figure 11, there was a statically significant relationship between time andthe percentage of children reunified with parents. Regardless of the length of time alead agency had a service contract, the percentage of children reunified with parentsdecreased over time between FY01-02 and FY02-03.

Figure 11. Percentage of Children Reunified With Parents After Exiting Out-of-Home Care by Site, by Time and Site by Time

44.5

54.4 59.559.8

46.754.3

05

101520253035404550556065

FY01-02 FY02-03

Fiscal Year

Perc

enta

ge

Service contractheld 18 months orlongerService contractheld 7-17 months

Service contractheld 6 months orless

76

Results: Children Placed with Relatives After Exiting Out-of-Home Care

The results of the analysis regarding children who were placed with relatives afterexiting out-of-home care are presented in Figure 10. The lowest percentage of childrenwho were placed with relatives for the FY02-03 cohort was observed for Child andFamily Connections (12.0%), and the highest percentage of children who were placedwith relatives for the FY02-03 cohort was found in FCP (39.3%). The averagepercentage of children placed with relatives during FY02-03 was 26.6%; the medianwas 26.3%.

By comparison, the lowest percentage of children who were placed with relatives for theFY01-02 cohort was observed for Child and Family Connections (14.7%). The highestpercentage of children who had custodial placement with relatives after exiting out-of-home care was observed in Family Support Services (35.7%; see Appendix 7).

When lead agencies that had a service contract for 18 months or longer (n = 2,997),lead agencies that had a service contract for 7 –1 7 months (n = 3,016), and leadagencies that had a service contract for 6 months or less (n=7,231) were compared onproportion of children placed with relatives, significant time, site, and site X time effectswere found. Results of comparison by time indicated that the proportion of childrenplaced with relatives increased over time, and the results of comparison by siteindicated that the proportion of children placed with relatives were significantly differentin the examined service contract categories. Finally, a significant site X time effectsuggested that the proportion of children placed with relatives increased over time at amuch higher rate in the lead agencies that had their services contracts for 18 months orlonger (i.e., increase from 26.1% to 32.3%) than in the lead agencies that had theirservice contracts of 7 months or less (Figure 12).

77

Figure 12. Percentage of Children Placed With Relatives After Exiting Out-of-Home Care by Site, by Time and Site by Time

78

79

32.3

26.1

23.625.4

2727

0

5

10

15

20

25

30

35

FY01-02 FY02-03

Fiscal Year

Perc

enta

ge

Service contractheld 18 months orlongerService contractheld 7-17 months

Service contractheld 6 months orless

Children with Adoption Finalized After Exiting Out-of-Home Care

Description of the IndicatorAnalysis of this indicator was based on exit cohorts of children, defined here as theduplicated number of children who exited from out-of-home care during a particularfiscal cohort. Classification of children with finalized adoption was based on “Reasonsfor Discharge” in HSn. The indicator was calculated as a percentage. The numerator forthis indicator is the percentage of children exiting out-of-home care whose adoption wasfinalized during FY01-02. The denominator is the total number of children exiting out-of-home care during FY01-02. Similarly, this indicator was calculated for the FY02-03cohort.

ResultsThe results of the analysis regarding children with finalized adoption after exiting out-of-home care are presented in Figure 10.Findings based on FY02-03 cohort indicated that Child and Family Connections had thehighest proportion of children whose adoptions were finalized (24.4%). The lead agencywith the lowest proportion of children with adoption finalized was HKI (6.5%). The

80

average percentage of children with adoptions finalized during FY02-03 was 12.1%, andthe median was 11.9%.By comparison, data based on FY01-02 exit cohort indicated that Child and FamilyConnections had the highest proportion of children whose adoptions were finalized(24.2%). In contrast, Family Support Services had the lowest proportion of children withfinalized adoptions (7.0%). The average percentage of children with adoptions finalizedduring FY01-02 for this distribution was 11.7, and the median was 10.6 (see Appendix7).

The results of logistic regression indicated that the proportion of children with finalizedadoption significantly decreased over time for the lead agencies that had a servicecontract for 7 -17 months and 18 months or longer. The proportion of children withfinalized adoption increased in the lead agencies that had their service contracts for 6months or less (from 11.1% to 14.1%), and this change is significantly different from thechange in the lead agencies that had their service contracts for over 6 months (seeFigure13).

Figure 13. Percentage of Children with Adoption Finalized After Exiting Out-of-Home Care by Site, by Time and Site by Time

10.611.4

7.57.6

14.1

11.1

0

5

10

15

FY01-02 FY02-03

Fiscal Year

Perc

enta

ge

Service contractheld 18 months orlongerService contractheld 7-17 months

Service contractheld 6 months orless

Summary of Research Question 3

81

Review of the various staffing models used by the lead agencies indicated the highimportance placed on permanency. Permanency staffings typically were held frequently.In addition, many examples illustrate unique staffing practices designed to create andmaintain a focus on safety, permanency, and child well-being.

Overall, the findings suggest that length of time with a service contract had the samebearing on lead agency performance measures. (Lead agencies did not differ whencompared on the proportion of children exiting out-of-home care.) The recurrence ofmaltreatment rates were not significantly different for lead agencies that had a servicecontract for 18 months or longer compared to lead agencies that had a service contractbetween 7 and 17 months. Lead agencies with a service contract for 18 months orlonger had a significantly smaller rate of reentry than lead agencies with a servicecontract of 7-17 months.

Similarly, lead agencies that had their service contracts longer appeared to perform aswell as lead agencies that had their service contracts for a shorter period of time onachieving child permanency. (When service contract categories were examined for theproportion of children reunified with parents, the proportion of children placed withrelatives, and the proportion of children with finalized adoption, a trend of decreasingpercentages of children reunified with parents was observed for all service contractcategories.) A similar trend can be observed for almost all service contract categoriesregarding the percentage of children with finalized adoption, with an exception for leadagencies with a service contract of 6 months or less (pre-CBC counties). In contrast, theproportion of children placed with relatives increased over time for almost all servicecontract categories, with a larger proportion of children placed with relatives in leadagencies that had a service contract for 18 months or longer.

RESEARCH QUESTION 4:

82

How effective is Community-Based Care at managing all resources and costsefficiently?

Introduction

Another important component in assessing the performance of community-based careis determining the extent to which the CBC initiative is managing resources and costsefficiently. Lead agencies are expected to use federal and state funds, as well aslocally generated revenue, to provide appropriate, evidence-based child protectiveservices to every child and family that enters their system of care. Because the twomajor federal funding sources (Medicaid IV-E and Temporary Assistance for NeedyFamilies [TANF]) for child protection are highly restricted, the lead agencies and DCFstruggle with developing budgets that allow maximum flexibility in services provided. Arelated challenge for DCF and lead agencies is managing the financial risk associatedwith providing services for a child welfare system with limited funding and anunpredictable need for care. This section of the report looks at average expendituresper child for direct services in the CBC sites compared with per-child expenditures in thenon-CBC counties, shows the differences by funding source between lead agencyappropriations and actual spending, and describes lead agency financial riskarrangements and the challenges related to financing child protective services.

Table 24: Research Question 4: Evaluation Questions, Indicator/Analysis, andSource

Evaluation Question Indicator/Analysis Source

How do average expenditures perchild for direct child protectiveservices in the CBC sites pre- andpost-CBC compare with expendituresin the non-CBC counties?

-Average expenditures perchild served-Average expenditures perchild day

-Average expenditures percapita

FLAIR, CIS,HSn, Floridapopulationdata

How do actual lead agencyexpenditures compare with leadagency budget amounts? Does thedifference between expenditures andbudget amount (i.e., variance) differby funding source?

-Variance (by funding source) FLAIR, leadagencycontracts

How effectiveisCommunity-Based Careat managingall resourcesand costsefficiently?

How do lead agencies handledfinancing issues and financial risk?

-Description of financial riskarrangements.

- Description of challengesrelated to financing.

CEO Survey

83

Analysis of CBC Expenditures for Direct Child Protective ServicesMethodsDirect services expenditures for child protective services were analyzed in every DCFdistrict. For this analysis, direct services expenditures are defined as expendituresincurred at the district level for out-of-home care (e.g., family out-of-home care,independent living support), in-home services (e.g., family preservation), servicesrelated to adoption, and other child protective services (e.g., family support services)that do not fit into the previous three categories. See Appendix 8 for a complete list ofservices by service category.

Several child protective services expenditures were excluded so that the CBC and non-CBC sites could be compared on the same basis. Our objective was to insure that theservices being compared would be comparable between CBC and non-CBC sites. Thisanalysis excludes expenditures for projects unique to a particular district, programs notmade available to CBC sites, and legislative member projects; see Appendix 9 for atable that aggregates these excluded expenditures by district and fiscal year.Expenditures for protective investigation and legal services were also excluded. Relatedexpenditures for childcare and adult protection services were excluded from thisanalysis because they are not provided by CBC’s or considered “child welfare relatedservices”(s. 409.167, F.S. as currently interpreted). Administrative expendituresassociated with child protective services were also omitted from this analysis but will bediscussed later in the report; an analysis of administrative expenditures can be found inthe June 2003 CBC evaluation report (see Paulson, et al., 2003).

DCF accounting data from the Florida Accounting Information Resource (FLAIR) wereused to calculate child protective services expenditures. The Office of FinancialManagement generated spreadsheets containing expenditures by District and OtherCost Accumulator (OCA) for Budget Entities (BEs) 60600700 Children, Youth andFamily Services, 60910303 Child Abuse Prevention/Intervention, and 60910304 ChildProtection Permanency. With guidance from Office of Revenue Management staff,appropriate BE-OCA combinations that were used for direct child protective servicesexpenditures were identified (see Appendix 8 for a complete list of current OCAs byservice category). This analysis covers the time period of FY 96-97 (the year prior to thefirst CBC implementation) through FY02-036, the most recent year for which data isavailable.

Two separate accounting system limitations prevented the evaluation from reportingexact expenditures for the Sarasota, Manatee, Pasco, and Pinellas CBC sites prior toFY00-01. The first limitation is that unique OCAs for CBC expenditures did not existduring FY96-97, FY97-98, FY98-99, and FY99-00. To overcome this limitation, Officeof Revenue Management staff helped identify the OCAs that were being used for CBCexpenditures during those years. While the CBC program has been rolled out at thecounty level, FLAIR cannot report expenditures any lower than the district level. This 6 Expenditures that were incurred during FY02-03 but certified forward and not paid until FY03-04 were included ifrecorded by November 30, 2003.

84

second limitation particularly affects the analysis of expenditures in Manatee andSarasota Counties prior to FY00-01, because Manatee and Sarasota did not comprisean entire district at that time in the same way as Pinellas and Pasco Counties (District5). An expert panel of current and former DCF district administrators was consulted tohelp accurately allocate district level expenditures to the counties within those districts.The panel advised to allocate 18% of District 6’s expenditures to Manatee and 27% ofDistrict 8’s expenditures to Sarasota during each of the above fiscal years. Both systemlimitations were alleviated in FY00-01 with the creation of unique OCAs for CBCexpenditure categories.

To adjust for inflation, expenditures from FY96-97, FY97-98, FY98-99, FY99-00, FY01-02, and FY02-03 were converted to FY00-01 dollars using the Consumer Price Index(CPI) for the South Region of the United States.

Ideally, total expenditures would be compared; however, counties vary by the number ofchildren they are obligated to serve and the number of children enrolled in the childwelfare system. Thus, average expenditures per child were calculated using 3 separatedenominators: (1) total number of children served, (2) total number of child-days, and(3) per capita. A “child served” is defined as any child receiving protective supervision,Intensive Crisis Counseling Program (ICCP), voluntary family services, other in-homeservices, out-of-home care, or adoptive home placement during the relevant fiscal year.The total number of children served is an unduplicated count (i.e., a child who exits andreenters the system in the same year is only counted once). A “child-day” is defined aseach day of service in which the child receives protective supervision, ICCP, voluntaryfamily services, other in-home services, out-of-home care, or adoptive home placementduring the relevant fiscal year. The “per capita” denominator reflects the number ofchildren under age 18 in the county population at the midpoint (January 1) of each fiscalyear. The advantage to presenting average expenditures per child using three separatemeasures of the number of children is that it is possible to analyze whether there weresimilar or dissimilar patterns that might aid interpretation.

Children served and child-day data were extracted from the Office of Family Safety’sClient Information System (for FY96-97 through FY00-01) and HomeSafenet (for FY01-02 and FY02-03). Due to the transition to HomeSafenet during FY00-01, data on thetotal number of children served and total child days for all 12 months were unavailable.Data available in CIS for that year ranged from 5-11 months, so the actual numbers ofchildren served and total child days were higher than the amounts used in this report.Another limitation associated with the change in data systems is that the fields forcomputing the number of children served and child-days have changed. Consequently,it cannot be determined whether the FY01-02 numbers are totally comparable withprevious years’ data. Population data came from the Florida Legislature’s Office ofEconomic and Demographic Research.

In order to compare average expenditures in CBC sites to average expenditures in non-CBC sites, each county was assigned a CBC status by fiscal year. For the purposes ofthis expenditure analysis, a county’s CBC status was determined by the first date that a

85

lead agency had a CBC service contract for that county. For counties in which theservice contract began during the middle of a fiscal year, their denominator data wereallocated to CBC and non-CBC based upon the proportion of the year each countyunder a CBC service contract. For example, HKI’s service contract for Hillsboroughbegan in May 2002, which means Hillsborough was considered a CBC site for 2months, or 16.7%, of FY01-02. The number of children served, child-days, and childrenin the population in Hillsborough County in FY01-02 were then allocated – 16.7% toCBC and 83.3% to non-CBC. This method is intended to more equitably attributeexpenditures to the responsible payer (the lead agency for the time period after theservice contract began, and the Department for the time period prior to the servicecontract). See Appendix 10 for a table that delineates the CBC/non-CBC allocationpercentages for each CBC site.

A Wilcoxon signed-rank test was used to compare the trend in average expenditures forCBC sites to the trend in average expenditures for non-CBC sites for each of the threeexpenditure measures. This nonparametric statistical test, which computes a Z-score,is appropriate when comparing groups with very few data points (Wilcoxon, 1945).

Findings

Average expenditures per child served have been consistently lower in the CBC sitesthan in the non-CBC sites during the last 3 years. As shown in Figure 14, non-CBC siteshad average expenditures per child served of $6,569 during the most recent year(FY02-03), which is 35% higher than the $4,876 average expenditures per child servedin CBC sites.

CBC sites have experienced a much more modest increase in average expendituresper child served than non-CBC sites over the study period. Average expenditures in theCBC sites were only 3% higher in FY02-03 than they were in the first year of CBCimplementation (FY96-97). In contrast, average expenditures per child served in non-CBC sites have increased steadily over the same time period, and were 78% higher inFy02-03 than in the first year of CBC implementation. However, average expendituresper child served in non-CBC sites did fall by 5% from FY01-02 to FY02-03. Overall, thetrend in average expenditures per child served in CBC sites was not statisticallydifferent from the trend in average expenditures per child served in non-CBC sites (Z = -0.34; p = .74).

86

Figure 14. Average Expenditures For Direct Child Protective Services Per ChildServed: CBC Sites vs. Non-CBC Sites

$4,086$4,768

$5,227

$4,296

$6,939

$3,687

$6,569$6,483

$5,304$4,876

$4,021

$4,929$4,738

$4,492

$0

$1,000

$2,000

$3,000

$4,000

$5,000

$6,000

$7,000

$8,000

$9,000

FY96-97 FY97-98 FY98-99 FY99-00 FY00-01 FY01-02 FY02-03

Non-CBC CBC

Average expenditures per child day also have been consistently lower in the CBC sitesthan in the non-CBC sites during the last 3 years, as shown in Figure 15. Non-CBCsites had average expenditures per child day of $32 during FY02-03, which is 45%higher than the $22 average expenditures per child served in CBC sites.

CBC and non-CBC sites have experienced opposite trends in average spending perchild day over the first 7 years of CBC implementation, although this difference is notstatistically significant (Z = 0.68; p = .50). Average expenditures per child day in CBCsites have been declining for the most part, while average spending per child day innon-CBC sites has steadily increased, rising by 68% since FY96-97.

87

Figure 15. Average Expenditures For Direct Child Protective Services Per ChildDay in System: CBC Sites vs. Non-CBC Sites

$26

$28$31 $32

$19$22 $23

$20

$36

$28

$32$35

$26$22

$0

$5

$10

$15

$20

$25

$30

$35

$40

FY96-97 FY97-98 FY98-99 FY99-00 FY00-01 FY01-02 FY02-03

Non-CBC CBC

Average expenditures per capita have been lower in the CBC sites than in the non-CBCsites during the last 3 years, but the gap is considerably smaller than for the previoustwo measures. As shown in Figure 16, non-CBC sites had average direct servicesexpenditures per capita of $134 during FY02-03, which is only 9% higher than the $123average expenditures per capita in CBC sites.

Both CBC and non-CBC sites have experienced consistent increases in per capitaspending over time. Average expenditures per capita in CBC sites rose by 62%, fromFY96-97 ($76) to FY02-03 ($123). Average expenditures per capita in non-CBC sitesgrew by 86% during the same time period. Overall, the trend in average expendituresper capita in the CBC sites was not statistically different from the trend in averageexpenditures per capita in non-CBC sites (Z = -1.35; p = .18).

88

Figure 16. Average Expenditures For Direct Child Protective Services Per Capita:CBC Sites vs. Non-CBC Sites

$123

$93$82

$109

$80

$86$76

$134$130

$115

$108$88

$80$72

$20

$40

$60

$80

$100

$120

$140

$160

FY96-97 FY97-98 FY98-99 FY99-00 FY00-01 FY01-02 FY02-03

CBC Non-CBC

While out-of-home care represents the largest portion of expenditures for direct childprotective services, there were small differences between CBC and non-CBC sites indirect child protective services by type of service (as shown in Figure 17). During FY02-03, non-CBC sites had a slightly higher proportion of direct services dollars spent onout-of-home care (59.7%) than non-CBC sites (53.9%). Conversely, CBC sites spent ahigher proportion of their service dollars on in-home services (27.7% vs. 20.1%). Non-CBC sites also had a higher proportion of expenditures for adoption-related servicesthan CBC sites (18.6% vs. 14.5%).

89

Figure 17. Direct Child Protective Services Expenditures by Type of Service(FY02-03)–CBC and Non-CBC Sites

53.9%

27.7%

14.5%

3.9%

20.1% 18.6%

1.6%

59.7%

0%

20%

40%

60%

80%

100%

Out of Home In-home Adoption Other

CBC Non-CBC

Conclusions & LimitationsOverall, average spending per child for child protective services has been lower for CBCsites than non-CBC sites during the last 3 years of CBC implementation. Although thereis some evidence that the annual increase in per child expenditures by non-CBC sites isslowing down, all 3 longitudinal cost measures demonstrated higher averageexpenditures by non-CBC sites.

Some of the difference in average expenditures per capita in FY02-03 may beattributable to an apparent difference in service mix between CBC sites and non-CBCsites. CBC sites devoted a smaller proportion (54%) of their total expenditures to out-of-home services than non-CBC sites (59%). Although determining the actual cost ofdelivering out-of-home services is beyond the scope of this analysis, it is likely that thecost of one month in out-of-home care far exceeds the cost of one month of in-homeservices or other types of direct protective services. Consequently, CBC sites may haveexperienced lower average expenditures per capita than non-CBC sites because theyprovided less out-of-home care than non-CBC sites during FY02-03.

90

The large difference in average CBC expenditures per capita compared with average,per capita expenditures for non-CBC sites during FY00-01 may have been caused by alarger increase in total expenditures for non-CBC sites than for CBC sites from FY99-00to FY00-01. One cause of higher expenditures in the non-CBC group in FY00-01occurred in District 10, where DCF spent additional dollars for child protective servicesin response to the Ward lawsuit. Excluding District 10’s expenditure and child data fromthe comparison group had a slight impact on narrowing the gap in average expendituresper capita between CBCs and non-CBCs for FY00-01 (data not shown), but does notappear to fully explain the wide gap between CBC and non-CBC sites that year.

Another important limitation to these findings is that there are unobserved county- ordistrict-level factors besides Community-Based Care that may affect expenditure levelsfor direct child protective services. For example, this analysis does not account for whatmany perceive as a historically inequitable distribution of child welfare system resourcesat the district level. Consequently, the differences in average expenditures per childreported here should be used for descriptive purposes only. It cannot be inferred fromthese findings that CBC sites had lower average expenditures per child than non-CBCsites because of the advent of CBC.

There are other limitations to these findings. FY00-01 was the first full year for system-wide changes to FLAIR. These changes included unique codes for CBC expenditures.While these updated OCAs are thought to be better suited for the type of analysiscontained in this report, there is a concern that DCF district staff and CBC staff may nothave used the new OCAs consistently initially. The inconsistent use of OCAs mayexplain some of the FY00-01 difference in average expenditures per child between CBCsites and non-CBC sites. The Office of Revenue Management has conducted a greatdeal of training on coding issues during the last 18 months, so the OCA level data wereexpectedly more reliable in FY01-02 and FY02-03.

This analysis is done from the perspective of DCF. With its focus on direct expenditures,the analysis fails to capture indirect expenditures such as lost productivity due to timelost from work/school and morbidity. The analysis also excludes the cost of informalsupports and child welfare services paid via other funding sources (e.g., dollars raisedby lead agencies in local communities). From a societal perspective, actual childprotective services expenditures are much higher. The effect of CBC on indirectexpenditures cannot be assessed from the data in this analysis, and conclusions cannotbe made about the overall cost of CBC without considering the impact of administrative,infrastructure, and indirect expenditures.

91

Analysis of Lead Agency Appropriations and Expenditures

Methods

Lead agency appropriations and expenditures for FY02-03 were analyzed for the 5 leadagencies that had a service contract during that fiscal year (FCP, FFN, HKI, PCBC, andSarasota YMCA). Appropriation amounts were pulled from the final version ofAttachment II from each lead agency’s FY02-03 CBC contract. FY02-03 expendituredata were extracted from FLAIR7. The overall difference between appropriation andexpenditures (i.e., the variance) for each lead agency was calculated; variance was alsocalculated by funding source for each lead agency. The variance %, which is equal tothe variance amount divided by the budget amount, was also calculated. Fundingsources include State general revenue, State matching, and other state funding sources(referred to here as State Funds); Title IV-E (referred to here as IV-E); TemporaryAssistance to Needy Families (TANF); and other federal funding sources (e.g., SocialSecurity Block Grant, Promoting Safe & Stable Families), which will be referred to hereas “Other”.

FindingsAll five lead agencies spent within 3% of their appropriation during FY02-03, as shownin Table 25. HKI, which had the highest total appropriation, spent every dollar that wasappropriated. PCBC, FFN, and the Sarasota YMCA, respectively, spent 1%, 3%, and3% less than appropriated, while FCP’s spending exceeded its budget by 2%. Totalspending by these five lead agencies was $119.8 million in FY02-03, which was $0.5million lower than what was appropriated.

Table 25. Lead Agency Budget Vs. Actuals (FY02-03)

Lead Agency Budget Actual Variance Variance %FCP $ 31,846,207 $ 32,597,335 $ 751,128 2%FFN $ 17,706,853 $ 17,150,464 $ (556,389) -3%HKI $ 37,736,873 $ 37,736,873 $ - 0%PCBC $ 13,642,332 $ 13,530,354 $ (111,978) -1%Sarasota YMCA $ 19,419,465 $ 18,816,185 $ (603,280) -3%

TOTAL $ 120,351,730 $ 119,831,211 $ (520,519) -0.4%

Analyzing variance by funding source shows that all five lead agencies exceeded theirIV-E appropriation. As shown in Tables 26-30, positive variance for IV-E funding rangedfrom 57% (PCBC) to 101% (FFN). Combining IV-E dollars for all five lead agencies,lead agencies spent $15.4 million more than their $22.4 million IV-E appropriation inFY02-03 (see Table 31).

7 Expenditures that were incurred during FY02-03 but certified forward and not paid until FY03-04 wereincluded if recorded by November 30, 2003 (FFN’s expenditures reflect certified forward adjustmentsmade in February 2004).

92

Table 26. Lead Agency Budget Vs. Actuals by Funding Source - FCP (FY02-03)

Budget Actual Variance Variance %State Funds $ 16,384,688 $ 11,394,571 $(4,990,117) -30%IV-E $ 6,499,550 $ 10,923,414 $ 4,423,864 68%TANF $ 3,701,764 $ 4,466,726 $ 764,962 21%Other $ 5,260,205 $ 5,812,623 $ 552,418 11%

TOTAL $ 31,846,207 $ 32,597,335 $ 751,128 2%

Similarly, four of the five lead agencies exceeded their TANF budget. The TANFvariance ranged from +6% for FFN to +21% for FCP. HKI, the only lead agency to notexceed their TANF appropriation in FY02-03, was 1% under budget for that fundingsource (see Table 28).

Table 27. Lead Agency Budget Vs. Actuals by Funding Source - FFN (FY02-03) Budget Actual Variance Variance %

State Funds $ 7,832,372 $ 4,021,358 $(3,811,014) -49%IV-E $ 1,830,985 $ 3,681,831 $ 1,850,846 101%TANF $ 5,262,870 $ 5,561,256 $ 298,386 6%Other $ 2,780,626 $ 3,886,019 $ 1,105,393 40%

TOTAL $ 17,706,853 $ 17,150,464 $ (556,389) -3%

Conversely, all five lead agencies spent less than their State Funds appropriation. TheState Funds variance ranged from –27% (HKI) to –52% (Sarasota YMCA). Statewide,the lead agencies spent only $37.0 million out of a possible $59.3 million, as shown inTable 31.

Table 28. Lead Agency Budget Vs. Actuals by Funding Source – HKI (FY02-03) Budget Actual Variance Variance %

State Funds $ 19,070,081 $ 13,846,603 $(5,223,478) -27%IV-E $ 7,682,365 $ 12,412,538 $ 4,730,173 62%TANF $ 5,615,107 $ 5,572,450 $ (42,657) -1%Other $ 5,369,320 $ 5,905,281 $ 535,961 10%

TOTAL $ 37,736,873 $ 37,736,873 $ - 0%

The variance associated with other funding sources was also positive for all five leadagencies. This variance ranged from +8% (Sarasota YMCA) to +98% (PCBC).

Table 29. Lead Agency Budget Vs. Actuals by Funding Source – PCBC (FY02-03)

Budget Actual Variance Variance %State Funds $ 7,178,971 $ 3,522,045 $(3,656,926) -51%

93

IV-E $ 2,159,858 $ 3,399,216 $ 1,239,358 57%TANF $ 2,208,027 $ 2,462,911 $ 254,884 12%Other $ 2,095,476 $ 4,146,181 $ 2,050,705 98%

TOTAL $ 13,642,332 $ 13,530,354 $ (111,978) -1%

Table 30. Lead Agency Budget Vs. Actuals by Funding Source – Sarasota YMCA(FY02-03)

Budget Actual Variance Variance %State Funds $ 8,846,655 $ 4,234,845 $(4,611,810) -52%IV-E $ 4,204,004 $ 7,403,721 $ 3,199,717 76%TANF $ 2,945,562 $ 3,466,404 $ 520,842 18%Other $ 3,423,244 $ 3,711,215 $ 287,971 8%

TOTAL $ 19,419,465 $ 18,816,185 $ (603,280) -3%

Table 31. Lead Agency Budget Vs. Actuals by Funding Source – All LeadAgencies Combined (FY02-03)

Budget Actual Variance Variance %State Funds $ 59,312,767 $ 37,019,422 $ (22,293,345) -38%IV-E $ 22,376,762 $ 37,820,720 $ 15,443,958 69%TANF $ 19,733,330 $ 21,529,747 $ 1,796,417 9%Other $ 18,928,871 $ 23,461,319 $ 4,532,448 24%

TOTAL $ 120,351,730 $ 119,831,208 $ (520,519) -0.4%

Conclusions & LimitationsEvery lead agency with a CBC service contract during FY02-03 spent within 3% of thetotal dollars appropriated. The variance ranged from -3% (FFN and Sarasota YMCA) to+2% (FCP).

There was considerable variation in budget versus actuals by funding source, butvariances across the five lead agencies were fairly similar for each funding source. Allfive lead agencies exceeded their Title IV-E budget, while four out of five spent moreTANF dollars than appropriated. Conversely, all five lead agencies spent fewer Statedollars than appropriated. The excess spending on Title IV-E services suggests therewas a higher than anticipated need for out-of-home care and/or miscoding ofexpenditures. It is also important to remember that restrictions on how Title IV-E andTANF funding can be used may hinder a lead agency’s ability to flexibly apply thesedollars to appropriate child protective services.

This analysis has a few limitations. The appropriation amounts and expendituresreported here are limited to those reported to DCF, and do not reflect lead agency

94

spending of non-DCF resources (e.g., state Medicaid funding not directly tied to childwelfare, locally generated revenue). Residential group care service dollars for childrenin FFN and PCBC’s service areas funded via the Comprehensive Residential Servicesand Programs legislation are excluded from this analysis because FFN and PCBC didnot receive those dollars in their FY02-03 budget. With both limitations, the dollaramounts involved are relatively small and should not impact the overall conclusionsdrawn here.

Financial Risk

In regard to financing mechanisms, in most privatization arrangements across thecountry, lead agencies have been financed through capitation or case rate paymentsthat reflect the actual number of people the agency is serving or likely to serve. Floridais the only state using a global budget transfer (see s. 409.1671, F.S.), that is, giving afixed amount of money to the lead agency and making it responsible for providing allservices needed to all children who enter the child welfare system (CWLA, 2000). Sincea lead agency cannot unilaterally control how many children it serves, but can impactthe number of children entering the system, it is at financial risk. This means that thefinancial stability and viability of the lead agencies and their provider networks must beevaluated as well. There have been reported instances where agency financial viabilitybecame problematic (Gibelman & Demone, 1998). DCF has recognized this possibilityand worked with the lead agencies around issues such as sudden increases inenrollment and has taken steps to mitigate the potential financial risks under thepurview of the Residential Group Care Bill. This bill directs DCF, collaboratively withlead agencies, to develop a proposal regarding the long-term use and structure of astatewide, shared earnings program that addresses the financial risk resulting fromunanticipated caseload growth or from significant changes in client mixes or serviceseligible for federal reimbursement.

As previously mentioned, it is a rare occurrence for child welfare agencies to receiveadequate funding. Funding inequities, which have most commonly resulted from specialappropriations, lawsuit settlements, and management decisions, are a common problemwithin Florida and other states. Existing lead agencies have worked with DCF todevelop an equity formula that is designed to make resource allotment fairer acrossgeographical areas, to provide incentives for best practice, and ideally, to allow forinvestment in prevention and diversion services. The FY03-04 proposal called for theformula to be based on 90% of a community’s child protection investigations and 10% ofan area’s child population. Total child population was determined to be the bestindicator of a community’s need for prevention resources, and the number ofinvestigations was found to be indicative of the demand for services.

Risk Sharing

DCF continues to struggle with the balance between the planned sharing of risk with itsnew partners and the negative implications of overwhelming fragile local systems ofcare with overly complex financial requirements. The adequacy of the current funding to

95

support both the existing local systems of care and fund, unanticipated changes tocurrent caseloads or eligibility criteria beyond the providers control is still in question.DCF has recently been given authority by the legislature to expand its limited risk poolby up to $10 million to deal with these issues of risk as well as incentives for innovativeservices and successful outcomes. Ideally, lead agencies would like to have a self-managing board composed of lead agency staff and community stakeholders, however,this may represent an extreme amount of local governance.

DCF is now attempting to stay “ahead of the curve” in anticipating the financial needs oflead agencies. In each new contract year, a provider is given one advance and twoestimated payments. Beginning with the fourth month of the contract year,reimbursement is based upon actual expenditures. However, some lead agencies viewDCF as hindering their financial flexibility via stated rules and policies. According to onelead agency stakeholder, “It has been tough for them to monitor rather than manage.They’ve done things for a certain way for so long that it’s hard for them to let those waysgo and to simply monitor what the agencies are doing. It’s been a learning situation forDCF as well.”

Table 32 presents a summary of fiscal issues faced by states across the nation duringprivatization of child welfare functions. This contextual information is provided as abackdrop for the funding challenges identified by Florida’s lead agency stakeholders(see Table 33). Many of the fiscal issues faced by Florida’s lead agencies, such asaligning fiscal incentives with performance outcomes and inequities in funding levels,are also being dealt with in other states as they implement privatization.

Table 32. Financing Issues Faced During National Privatization Initiatives

Issue Lessons Learned

Start-up expenses

• Expect start-up costs and plan for them inadvance

• Do not expect the old (long time in system) casestransferred over by the state to be adequately

Focus On: How Specific Agencies Deal with Risk

“Regarding risk, we have 5 “stakeholder” agencies that provided $75,000 in cash and$175,000 in irrevocable letters of credit each. These stakeholders must have beenaccepted to provide a core service of the system of care (care management,placement/adoption/foster home recruitment and licensure, prevention/diversionservices). Each stakeholder is an at-risk partner, wherein they have a cappedcontract amount and must provide services to all eligible clients referred to them.”

-Stakeholder respondent

96

covered by the new case rate established for theprivatized agency

• General underestimation of resources required toimplement such a major service system change

Deciding on reimbursementsystem

• Per month per child rather than case rate mayplace disincentive on expedient pathwaysthrough the child welfare system

• Those with case rates experience more financialrisk because they cannot control the number ofkids entering the system or pace of the courts

Deciding/determiningcapped or case rate

• Making sure the out-of-home case rate is nothigher than the adoption rate so that there is nota disincentive to keep children in care

• Base estimates on historical data and costs ofinflation (but this depends on reliable state data)

• Do not set case rate too low and graduallyincrease (quickly leads to deficits)—try setting itslightly high and then lower if needed

Eligibility• It is difficult to privatize a program that only takes

high-risk kids because historically, even withcase rates, agencies lose money—better tomaintain mixed case loads

Flex funding• Allows for more “concrete” services to families,

such as housing and post-reunification services,which facilitate better child welfare outcomes

Risk pool

• Align fiscal incentives with permanency, safetyand well-being goals

• Need to establish risk pool to control forunexpected shortfalls, more challenging cases orsudden entry rate increase

Historical inequities in levelsof funding

• Allocation method needs to factor in populationdifferences and geographic resource inequities

Table 33. Funding Challenges Faced by Lead Agencies

Lead Agency Funding Challenges8

8 Although DCF’s Central Office fiscal staff has been engaged in the CBC evaluation, they were notinterviewed regarding the issues in Table 28.

97

Managing the adoption subsidy and Road to Independence Scholarship lineitems, which are in a deficit in some areas.

Adequacy—one lead agency reported that they are currently funded at 63% ofthe State average per child in care. Additional lead agencies hope to bring theiraverage up to the average per child funding rate of $11,000 per active child.

Allocation mix of funding appropriate for actual case composition. The case mixbetween in home and out-of-home care in some lead agency systems does notalign with the funding that is allocated. Specifically, the case mix has a higherpercentage of in-home cases, but the funding only supports the Statewideaverage.

Disconnection between performance measures and fiscal support pose thegreatest challenges to fiscal stability.

Title IV-E funding: while evidence-based practice and performance outcomesseek to keep children at home with their families, IV-E funding only supportsservices to children who are in out-of-home care.

Demand for services exceeds revenue resources.

Independent living funding is a capped federal grant.

Frequent changes in the State’s fiscal system (e.g., changes in procedures leadagencies must follow).

Lack of a common vision between the District and Central Office DCF.

Overall challenge of trying to draw down funds from the State.

Time logs are burdensome and unnecessary; the job descriptions alone shouldbe adequate.

One lead agency stakeholder commented, “CBC is a cost reimbursement contract,which in a best case scenario offers no financial reward for good performance and in aworse case scenario may continue to foster poor management decisions. While CBCscertainly have more flexibility in spending than the State did, a system that isinadequately funded poses serious fiscal risks for agencies. The combination of costreimbursement and lack of funding has had an impact on the agency as a whole.”Another lead agency stakeholder explained that their agency continued to struggle toearn and account for the various federal funding sources. While their agency hassuccessfully navigated the system, it has not been without enormous effort.

98

On a positive note, one lead agency reported that CBC has had a positive fiscal impacton their organization via new sub-contracts to provide services. Yet another stakeholderindicated increased optimism that improvements would be seen with the transfer of therevenue maximization function from DCF to the lead agencies.

An additional theme, as noted in previous evaluation reports, is the challenge ofequitably allocating resources and managing service provision when the service area isprimarily rural. On the one hand, urban areas have a wider array of services andpartners, a much larger employee pool and fewer transportation issues. On the otherhand, densely populated communities have a much larger number of high-risk children.In rural areas, the distances people have to travel make it much more difficult to holdjoint meetings and to maintain good communication among the partners and betweenproviders and families.

Summary of Research Question 4

The evidence from this evaluation suggests that the five CBC lead agencies that hadservice contracts during FY02-03 have effectively managed the fiscal side of CBC.Average expenditures per child have been lower in CBC sites than in non-CBC sites,particularly during the last three years of CBC rollout. Total lead agency spending inFY02-03 was within 0.4% of the statewide budget, which is very reasonable. None ofthe five lead agencies (FFN) significantly overspent their budget. Lead agencies alsoappear to be managing risk adequately.

There are some caveats, however, to these findings. Fiscal problems were reportedlyone of the reasons why FCP was replaced recently as the Pinellas/Pasco lead agency,so past fiscal success does not guarantee future success at fiscal management.Successful risk management will require a plan and a significant investment by theLegislature in order to ensure continuity of care in the event of unforeseen demand forchild protective services. There is evidence that navigating the federal child welfarefunding requirements has been challenging for some lead agencies. Training of fiscalpersonnel and good communication about financial management and reporting betweenDCF and lead agency staff will be very important as new lead agencies begin servicecontracts during the coming year.

CONCLUSION: RESEARCH QUESTION 5:

How effective is Community-Based Care at identifying and meeting the needs ofthe families and children who have been maltreated?

Community-Based Care encompasses more than just privatization. It was intended thatCBC would: (1) create community ownership over child welfare issues; (2) improve the

99

safety and well-being of children; (3) shift the responsibility for child welfare services tonewly created lead agencies; (4) create a more integrated and comprehensive childprotective service system; and (5) achieve cost efficiencies and more flexiblemanagement of resources.

SUMMARY

Throughout our country, states and communities are struggling with the complexchallenges of ensuring the safety and well-being of children in the child welfare system,and many states, including Florida, are experimenting with various types of privatizationof child welfare services. This evaluation report examines the status of Community-Based Care (CBC) in Florida, with a special focus on quality, cost and child and familyoutcomes. The time period covered by the evaluation (FY03-04) was a time of rapidgrowth for CBC, resulting in 11 lead agencies in 28 counties serving over half of thechildren in Florida’s child welfare system, with additional counties in the start-up phase.

The evaluation is organized around a set of research questions regarding theeffectiveness of Community-Based Care. The final question --How effective isCommunity-Based Care at identifying and meeting the needs of children who have beenmaltreated and their families? --is the overarching question, and is answered in thissection of the report.

The first area of CBC studied was the effectiveness of lead agencies at designingand improving systems and services for child protection. This question wasaddressed through an analysis of the organizational structure of lead agencies, theirprovider networks, and boards of directors, and an inventory of promising approaches insupervision training and practice. An examination of organizational complexity revealedthat all lead agencies had developed 4-5 departments or divisions, indicating consensusregarding the types of specializations needed to administer Community-Based Care.The degree of vertical differentiation varied, with 4 lead agencies averaging 3 personsfrom the lowest level of the organization to the CEO, and 3 lead agencies averaging 4.6persons. Greater vertical differentiation indicates a need for more process andcommunication protocols and standards.

The analysis of provider networks revealed three models of organizational structures: astructure that includes parent or partner organizations, a more traditional providernetwork without parent/partner organizations, and a model that uses services centers.Many examples of re-design were evident in the early implementation stages at both thelead agency and provider network levels, indicating that the CBCs were willing to reviseinitial plans when there were indicators that systems and services were not optimal.Potential conflicts of interest through interlocking boards were indicated for several leadagencies by the inclusion of contracted providers on boards of directors. A secondpotential problem was the lack of diversity in board membership, with underrepresentation from community members, foster parents, and nonprofit organizationsbeyond contracted providers. The lack of diversity could result in a lack of community

100

involvement and ownership, and may prevent infusion of new local revenues and/orinnovative services.

Six of the 11 lead agencies submitted promising supervisory approaches, indicating thatseveral sites are trying to incorporate innovation in supervisory training and practice.There was also great variability in the types of approaches across sites, anotherindicator of individualized efforts to design effective systems and services.

The effectiveness of CBC at involving the community as service partners andresource contributors was the second area of study for the report. Lead agenciesoperating 18 months or more had a smaller average number and range in number ofproviders and contracts than agencies operating for shorter periods. This reduction inproviders and contracts may be related to a trend towards stabilization and maturity ofthe service system, although this area will continue to be tracked over time. Leadagencies reported relatively little success in generating local revenues, especially afterthe start-up phase.

In addition to the lead agency boards of directors, Community Alliances representanother potential community governance partnership for lead agencies. Of the 7Community Alliances who responded to the study’s survey, only 3 had representativesfrom mental health or the business community; only 2 included local foundations orfoster parents, and only 1 Community Alliance had substance abuse or consumerrepresentatives. The majority of Community Alliances reported that child welfare is astanding agenda item but several felt limited by their “advisory role” and expresseddesire for more involvement in local system of care development. An unresolved issuefor lead agencies is clarification of the governance roles of boards of directors,Community Alliances, and other community stakeholder groups.

The third area of CBC effectiveness reviewed was the ability to meet the outcomes ofchild safety, permanency, and well-being. A review of the various staffing modelsused by the lead agencies indicated the high importance placed on permanency.Permanency staffings typically were held frequently. In addition, many examplesillustrate unique staffing practices that allow lead agencies to deal with routine andemergent issues while maintaining a focus on safety, permanency, and child well-being.

Two indicators were examined for child safety and well-being: the rate of recurrence ofmaltreatment and re-entry into out-of-home care. The highest probability of re-entry intoout-of-home care was found in counties where CBC had been operating for 7-17months; this rate was significantly higher than CBC counties operating for 18 months ormore. Rates of recurrence of maltreatment were lowest for counties just prior to CBCimplementation; no differences were found between the categories of 7-17 months, and18 months or more of CBC operation.

Four indicators were used to examine the permanency outcome. Findings were mixedacross the 3 CBC length of time of service contract for the proportion of children exitingout-of-home care in less than 12 months, and no significant differences were found. For

101

the proportion of children returned to parents, there was a decrease for counties with 18months or more of operation, and a slight increase for the 7-17 months category. Forthe indicator of percent of children placed with relatives, the proportion increased overtime, with a significant difference in the increase for CBC counties with 18 months ormore of operation. The highest increase in the proportion of children with adoptionfinalized was found in the counties just prior to CBC implementation, indicating that thelead agencies have challenges with meeting the permanency goal of adoption.

The evaluation also examined the effectiveness of Community-Based Care atmanaging resources and costs efficiently. A comparison of CBC and non-CBCcounties found that the average spending per child for child protection services hasbeen lower for CBC counties than for non-CBC counties for the past three years. Onereason may be attributed to differences in the service mix. During FY02-03, CBCsdevoted a smaller proportion (54%) of resources to out-of-home care than non-CBCs(59%).

A comparison was made of actual lead agency expenditures with lead agencyappropriation amounts, and whether the differences varied by funding source. DuringFY02-03, most lead agencies spent within 3% of the total dollars appropriated. Therewas considerable variation by funding source, with all 5 lead agencies spending fewerstate dollars than appropriated, but spending more Title IV-E funds than appropriated.Several additional funding challenges were reported by lead agencies, includinginadequate statewide funds for adoption subsidies and the Road to Independencescholarships, the funding allocation formula versus actual case composition, and overallinadequate funding levels.

Additional Lessons Learned

Each lead agency was asked to rate themselves on several general outcomes (Table34). The results are presented in the following table, with the lowest score representingthe most progress and most positive outcome (scale of 1-5). For example, agenciesscored themselves highly on assured continuity of care for all children entering the childwelfare system, while they reported some difficulty controlling the costs of serviceswithout sacrificing quality of care.

Table 34. Self Reported Lead Agency Outcomes

Self-Reported Lead Agency Outcomes

Outcome Score

102

Improved assessment of local needs for child welfare services 2.3

Improved accountability for a local system of child welfare services tomeet performance outcomes and standards.

2.1

Controlled the costs of services without sacrificing quality of care. 2.5

Assured continuity of care for all children entering the child welfaresystem.

1.9

Improved the overall quality of child welfare services through the use ofbest practice models.

2.1

Reduced length of stay for children in the child welfare system. 2.3

POLICY RECOMMENDATIONS

• Several study findings indicate that lead agencies may need additional supportand technical assistance during the mid-implementation phase (7-17 months) sothat the lead agencies do not lose their start-up momentum and communitysupport as they struggle to solidify their organization’s infrastructure.

• Lead agencies should examine the composition of their boards of directors inrelationship to interlocking boards and lack of diversity (e.g., foster parents,children who have aged out of the foster care system, and other communitymembers).

• Community Alliances are potential community governance partners, butclarification is needed regarding their authority, and their role vis-à-vis the leadagency boards of directors and other community stakeholder groups.Legislatively appropriated incentives for Alliance membership and engagement inchild welfare issues should be considered.

• More direct communication is encouraged between DCF Central Office and theCommunity Alliances, for example, by legislative updates on child welfare relatedbills, so that the Alliance members can voice their opinion before items arepassed/vetoed.

• The lead agencies and DCF partnership should continue to address fundingissues, including adequacy of funding levels, and finding appropriate allocationmixes.

• Lead agencies need consultation and technical assistance regarding evidence-based and promising practices for the permanency goal of adoption.

103

• Lead agencies should be provided with support to advance their emergingpromising approaches to supervisory practice through technical assistance andthrough peer review of their processes.

• Lead agencies should be encouraged to continue developing individualizedpatterns of staffings so long as they meet statutory requirement. In addition tomaintaining a focus on child well-being, staffings also create opportunities forsupervision and inclusion of community stakeholders at the child and family level.

NEXT STEPS

In regard to examining the effectiveness of CBC in designing and improving systemsand utilizing the community as service partners and resource contributors, activities forthe Legislative report will center on clarifying responses and issues that arose duringprevious data collection for the organizational structure analysis, communitygovernance, and quality of services. While the June 2004 report provided an initialdescriptive analysis of organizational structure and community governance, theLegislative report will provide the next level of analysis. Specifically, the project teamwill:

1) Interview lead agency CEOs to refine visual representations and models of theirorganizational structure and interaction with community stakeholders.

2) Visit a sampling of lead agency sites and provider network agencies to examinethe spatial dispersion of their organization and contracting agencies (e.g., co-location, service centers, urban/rural issues).

3) Continue interviews to examine community stakeholder (e.g., CommunityAlliances, Florida Coalition, and other local stakeholder entities) input into theirrespective lead agency’s community-based system of care. In addition, wherepermitted, project team members will attend meetings and conduct observationsof discourse pertaining to child welfare and Community-Based Care issues.

4) Continue literature review on issues of community governance, financial risk, andorganizational structure, as well as document review of lead agency ITNs andprovider network contracts.

Quality of services will be focused on the topics of family involvement and child well-being. Throughout the evaluation activities, staff will work to identify national examplesof best practice models for the measurement and assessment of family involvement andchild well-being. Specific methodologies are yet to be determined, but the followingguiding principles have been established:

All CBC lead agencies will be included in the assessment. Efforts will maintain a strength-based focus as much as possible.

104

Methodologies will be developed in conjunction with the Quality Assurance Officeof the Department of Children & Families in Tallahassee and will build on theircurrent efforts.

Information gathered through application of the Child Welfare Integrated QualityAssurance (CWIQA) Tool developed for use by DCF will be incorporated into the FY04-05 evaluation. Once applicable data has been compiled, the evaluation team, inconjunction with DCF, will need to decide if additional data collection is needed toelaborate on findings (e.g., additional file reviews). One question that remains to beaddressed is whether findings will be further explored for all sites (e.g., small samplewith each lead agency) or only with high performers, in an effort to identify promisingapproaches to improving family involvement and child well-being.

To address the effectiveness of Community-Based Care at meeting the AFSArequirements of child safety, permanency and well-being, the calculation of the followingquantitative indicators will continue:

1. Proportion of children exiting out-of-home care (based on estimates provided bysurvival analysis). Entry cohorts from FY01-02 and FY02-03 will be used.

2. Proportion of children who reentered out-of-home care (based on estimatesprovided by survival analysis). Exit cohorts from FY01-02 and FY02-03 will beused.

3. Proportion of children with recurrence of maltreatment. This indicator will beredefined. Only cases where children were referred and served by CBC countieswill be included in the analyses.

4. Percentage of children who were reunified with parents. The numerator for thisindicator will be the number of children who exited out-of-home care and werereturned to parents or legal guardians. The denominator will be the total numberof children who exited out-of-home care.

5. Percentage of children who were placed with relatives for long-term care. Thenumerator for this indicator will be the number of children who exited out-of-homecare and were placed with relatives for long-term care. The denominator will bethe total number of children who exited out-of-home care.

6. Percentage of children with adoption finalized. The numerator for this indicatorwill be the number of children with adoption finalized. The denominator will be thetotal number of children who exited out-of-home care.

All indicators will be calculated for existing lead agencies and for three categories ofcounties. All counties that are transitioned or transitioning to Community-Based Carewill be included in the analysis.

105

The following comparisons will be done:

1. The county categories will be compared by site, by time (i.e., FY01-02 will becompared with FY02-03), and by site by time.

2. To examine the proportions of children reunified with their parents, placed withrelatives, and the proportion of children with adoption finalized, the followingcomparisons will be made: (a) CBC counties that are fully operational will becompared to counties that are in transition and in start-up phase (counties intransition and start-up will be merged in this comparison), and (b) CBC countiesthat are fully operational will be compared to counties that are in transition.

In addition, the sub-group analysis will examine children who stay in care for over 12months. Predictors of over 12 months stay in out-of-home care and predictors of reentryinto out-of-home care will be examined. The following predictors will be included inthese analyses: (a) child-level predictors (e.g., gender, race, age, physical disability,emotional or behavioral problems, and type of maltreatment), (b) county-level predictors(e.g., time county has been in CBC), and (c) system-level (e.g., legal status of thecase). Discharge reasons as a predictor for reentry into out-of-home care also will beexamined as well as placement stability.

To examine whether lead agencies are managing all resources and costs effectively,the primary focus of the CBC cost analysis for the FY04-05 evaluation is to continuepiloting and refining a methodology for analyzing child welfare administrativeexpenditures. The objective of this methodology is to have a more comprehensive toolfor exploring administrative and direct services expenditures at the lead agency level.Such a methodology should produce answers to the following questions:

1. What proportion of the dollars spent by a lead agency for child protectiveservices is used for administrative needs? What proportion of lead agencyexpenditures is used for direct services?

2. What proportion of the dollars spent by a lead agency for child protectiveservices is generated from sources other than the lead agency contract (e.g.,dollars contributed by the local community)?

3. How do lead agency budget amounts differ from reported expenditures?4. What are the relationships between the Department and lead agencies related to

administrative expenditures?

Answering Questions 1 and 2 is contingent upon being able to obtain the appropriatefinancial reports and documents from the lead agencies. Building on knowledge fromFY03-04 work with HKI, the evaluation team will work with other lead agencies (e.g.,FFN and the Sarasota YMCA) to understand their financial statements & documentspertaining to child protection. We also intend to seek guidance from the DCF-appointedorganization that is providing financial oversight to the District 10 lead agency. Data

106

collection and analytic efforts will be coordinated/triangulated with the organizationalanalysis component of this year’s evaluation.

Building on efforts from the last three years’ evaluations, the analysis of direct childprotective services expenditures will be extended by adding FY03-04 expenditure datafrom FLAIR. These data will enable the comparison of average expenditures per childby year using the same three measures of child “need” that were used in last year’sanalysis: number of children served, number of child days, and the number of childrenin each county‘s population. Last year’s graphical analysis of direct child protectiveservices expenditures by type of service (e.g., out of home, in-home, adoption) will alsobe extended to include FY03-04 expenditure data. All of the results mentioned in thisparagraph will be included in the January Legislative report.

107

REFERENCES

Adoption and Safe Families Act of 1997, P. L. No. 105-89 (1997).

Martone and Solek (2001). Retrived February 2, 2004 fromhttp://www.cwla.org/programs/Trieschman/2001fbwWilliamMartone, htm.

Barter, K. (2001). Building Community: A conceptual framework for child protection. ChildAbuse Review, 10, p.262-278.

Center for the Study of Social Policy. (1998). Creating a Community Agenda: Howgovernance partnerships can improve results for children, youth, and families.

Center for the Study of Social Policy, (2003). Community partnerships for protectingchildren. Safekeeping, 7(1), p. 1-7.

Child Welfare League of America. (2000). 2000 Management, Finance, and ContractingSurvey. Washington, DC: Author.

Child Welfare League of America. (2004) Effective Supervisory Practice Fact Sheet.Retrieved February 2, 2004 from http://www.cwla.org/programs/trieschman/espfact.htm.

Fitzgerald, S.P. (2002). Organizational models. Capstone Publishing ,Oxford,U.K[electronic resource] ExpressExec organizations ; 07.07http://www.netLibrary.com/urlapi.asp?action=summary&v=1&bookid=67240

Freundlich, M. & Gerstenzang. (eds.) (2003). An assessment of the privatization of childwelfare services. Washington, DC: Child Welfare League of America.

Geen, R. & Tumlin, K.C. (1999). State efforts to remake child welfare: Responses to newchallenges and increased scrutiny. Washington, DC: The Urban Institute.

Gibelman, M. & Demone, Jr. H. W. (eds.) (1998). The privatization of human services:Policy and practice issues. New York: Springer Publishing Company.

Hall, 1996. Organizations: Structures, Processes, and Outcomes. Prentice Hall, EnglewoodCliffs, NJ.

Malm, K., Bess, R., Leos-Urbel, Geen, R., & Markowitz, T. (2001). Running to keep inplace: The continuing evolution of our nation’s child welfare system. Washington DC:The Urban Institute.

Maximus and Florida State University (FSU) (2003). Best Practices Handbook, 4.

108

McCullough, C. & Schmitt, B. (2000). Child welfare league of America 2000: Management,finance, and contracting survey final report. Washington, DC: Child Welfare League ofAmerica, Managed Care Institute.

Paulson, R.I., Armstrong, M., Fitzpatrick, J., Jordan, N., Kershaw, M.A., Reyes, F., Vargo,A.C., Yampolskaya, S. (2003). Evaluation of the Florida Department of Children andFamilies Community-Based Care Initiative. Tampa, FL: University of South Florida,Department of Child and Family Studies.

Robbins, S.P. (1987). Organization Theory: Structure, Design, and Applications. PrenticeHall, Englewood Cliffs, NJ.

Schulzinger, R., McCarthy, J., Meyers, J., de la Cruz Irvine, M., & Vincent, P. (1999).Special Analysis: Child Welfare Managed Care Reform Initiatives. Research andTraining Center for Children’s Mental Health. Tampa, FL: Louis de la Parte FloridaMental Health Institute.

Sells, J. K. (2001). Child Welfare Privatization Reform Efforts in the States. Pages 1-20.Retrieved August 22, 2003, fromhttp://www.pacificresearch.org/pub/sab/health/childwelfare/index.html.

Snell, L. (2000). Child Welfare Reform and the Role of Privatization. Policy Study No.271: 1-26. Retrieved August 6, 2003 from http://www.rppi.org/ps271.html.

U.S. Department of Health and Human Services, Administration for Children and Families(2003a). Adoption and Foster Care Analysis and Reporting System (AFCARS).Washington, DC: Author

U.S. Department of Health and Human Services, Administration for Children and Families(2003b) Child Maltreatment 2001. Washington DC:Author

U.S. Department of Health and Human Services, Administration for Children and Families(2000). Child Welfare Outcomes 2000: Annual Report. Washington, DC: Author

U.S. General Accounting Office. (2000). Child welfare: New financing and servicestrategies hold promise but effects unknown (GAO/T-HEHS-00-158). Washington, DC:U.S. Government Printing Office.

Van Slyke, D.M. (2003). The mythology of privatization in contracting for social services.Public Administration Review, 63(3), 296-315.

Wilcoxon, F. (1945). Individual comparisons by ranking methods. Biometrics 1: 80-83.

109

APPENDIX 1. Review of Privatization Initiatives Across the United States

Kansas

Implementation. During the single year of 1996 Kansas privatized foster care, familypreservation, and adoption services statewide (Schulzinger, McCarthy, Meyers, de laCruz Irvine, & Vincent, 1999). Similar to Florida’s Community-Based Care, the initiativecame about in response to criticisms of the current system in Kansas, such as a crisis-oriented approach, an unorganized system of care, and lack of emphasis on child-leveloutcomes (Sells, 2001). The Kansas initiative was similar to managed care in that it heldlead agencies accountable for certain performance measures specified in eachagency’s contract and a case rate reimbursement system was initially utilized to coverall services needed by one family or one child. Another important principle was thatcost-efficiency and quality could be achieved through a competitive bidding process ofeach lead agency (Freundlich & Gerstenzang, 2003). Central goals of the initiativeincluded augmenting prevention efforts, increasing permanency, decreasing length ofstay, and the pooling of funds to allow for a more seamless system of care (McCullough& Schmitt, 2000).

Roles. No pilots were tried before statewide implementation, in contrast to Florida’spilot sites initiated the same fiscal year Kansas began (1996-1997) in Duval, Manatee,Lake, Sumter, and Sarasota counties. The Kansas Department of Rehabilitative andSocial Services (SRS) retained responsibility for investigations, intake and assessment,child abuse prevention efforts, and program monitoring, while the lead agencies wereresponsible for service delivery to any child referred by SRS with the exception ofchildren involved with juvenile justice (Freundlich & Gerstenzang, 2003; McCullough &Schmitt, 2000; Schulzinger, McCarthy, Meyers, de la Cruz Irvine, & Vincent, 1999). Aswas the case with Florida, the privatized foster care agencies in Kansas were expectedto identify and seek out additional funding sources (Freundlich & Gerstenzang, 2003).

Financing. The Kansas initiative is perhaps best known for it’s initial case ratereimbursement system across adoption, foster care and family preservation services(Schulzinger, McCarthy, Meyers, de la Cruz Irvine, & Vincent, 1999; Snell, 2000). Thismethod of repayment was seen as ground breaking and a step away from the permonth payments that may reward providers for keeping children in care longer ratherthan seeking expedient permanency. Despite initial hopes, both foster care andadoptions ran into extreme financial problems under the case rate system due to thecase rate not being adequate enough to cover both high end children and averageconsumers, and a lack of control over financial risk due to the fact that the courts andSRS often determined how quickly a child moved through the Kansas system. One ofthe Kansas lead agencies calculated that if a child stayed in their care longer than 6months (under the case rate reimbursement system), the agency would lose money. Infact, during the initial privatization year in Kansas, the lead agencies collectively spent25 million more dollars than were allotted to them. During 1998 the State gave them$24.7 million to make up for this, as well as an additional $51 million drawn fromwelfare-to-work dollars in 1999. It should also be noted that some of the excess costswere due to start-up expenses that were not expected (Snell, 2000).

110

Another interesting situation Kansas quickly learned from is that due to a lack of data onchildren before the privatization effort took place, there wasn’t much reliable informationby which to establish a baseline case rate for foster care. Problems ensued when theState transferred over all of its old cases under the same reimbursement structure asthe new intakes under privatization. The children previously under State care had oftenbeen there a long time, had increased service need, and were much more difficult forthe newly privatized agencies to place. The planning team behind the privatizationinitiative realized too late that there should have been a higher case rate for existingfoster care cases (Snell, 2000).

An initial corrective response was to increase the case rate and to allow agencies to setaside a specified number of high end cases that would be paid on the per-monthschedule, however, this was not enough to elicit fiscal survival. While familypreservation remains on the case rate system, both adoptions and foster care havebeen moved to a per-month per-child reimbursement (Freundlich & Gerstenzang, 2003).Critics of this more common reimbursement system lie in waiting for a potential return toincreased lengths of stay based on a “perverse incentive structure” (Snell, 2000: 7).

Outcomes. Outcomes measured included access to and availability of services,whether agencies accept all referrals that are directed to them, child safety, clientsatisfaction, adherences to professional and licensing standards, permanency planninggoals, cultural competency standards, and goals related to recidivism and reentry.These outcomes are measured via provider reports, cost and utilization data, andperformance data tracked through an MIS system (McCullough & Schmitt, 2000). TheState also contracts with an external review team in order to access whether or notcertain contract models are more successful (Schulzinger, McCarthy, Meyers, de laCruz Irvine, & Vincent, 1999). In fact, Kansas is heralded by some for their effort atmaking accountability through outcome measurement a focus of their privatization effort(Sells, 2001).

In general, positive outcomes have been achieved subsequent to privatization,particularly with respect to family preservation and child safety. Specifically, Kansas’sfamily preservation services exceeded many performance standards in the four yearsafter privatization. Engagement in program, safety during program participation, safetyafter 6 months of program participation, no placement during participation, noplacement within 6 months of case closure, and client satisfaction with service providedState performance standards were all surpassed (Snell, 2000).

However, permanency outcomes continue to be a challenge, as they were under publicchild welfare (Freundlich & Gerstenzang, 2003). Another study found that during the firstfour months of privatization adoptions increased from 30 to 50 children per month and inthe year subsequent to Kansas privatizing their child welfare system, adoptionsincreased by 26% (Sells, 2001). According to Sells, “Kansas increased both itsefficiency and adoption rate, while holding down costs” (ibid: 12). The State’sindependent evaluator has found that lead agencies are meeting or exceeding

111

standards set for sustainability of adoptive placements, placements finalized within 12months, children placed with siblings, children with fewer moves since referral, andchildren not experiencing abuse or neglect prior to finalization. However, Kansas isfalling short of measures of client satisfaction and timely placement. In addition, Kansashas been cited as being able to move children through the system more expediently andsafely, while reducing caseloads, despite significant challenges (Snell, 2000).

Lessons Learned. The overarching criticism of the Kansas initiative was thatprivatization was pushed forward in too short a time frame. Closely related wereaccusations from several sides, such as juvenile justice and foster parents that theywere not included in the planning process and there was a lack of communication aswell as a need for role clarity. There was much initial concern drummed up by the mediaand key stakeholders that privatization was leading to a drastic decline in foster homes(Demarre, 2001; Freundlich & Gerstenzang, 2003).

Another problem identified was that initially, privatization was heralded as improvingoutcomes via previously lacking competition among providers. However, the KansasLegislature did not realize that this competition would largely die off subsequent to thecompetitive procurement process. As it now stands, many of the monopoly issuespresent with a state run system are still present within the newly privatized jurisdictions.According to one Kansas judge, “Privatization, in my view, has lead not to the healthyfree competition one might expect, instead it has created enormous service monopoliesin this state” (Demarre, 2001: 647).

Closely related is the issue of accountability between private providers and stateinvestigators responsible for initial determination of need and determination of removalfrom a child’s home. Reportedly, when problems have come up with services provisionrelated to the initial recommendations, or based on a subsequently identified need by afoster parents, both public and private entities have been slow to react and quick topoint the finger back in the other direction, resulting in a missed opportunity for childadvocacy. Kansas courts have also had trouble dealing with the mix of public andprivate players central to a child’s case, particularly when trying to issue contempt ofcourt orders when the specified services are not delivered (Demarre, 2001).

According to a recent CWLA publication, however, the dust seems to have settled andboth operations and outcomes are reportedly showing much progress. At the beginningof FY00-01, contracts were rebid and altered to reflect lessons learned such asrequiring caseloads be 25 families or less and establishing service coordinators to fillservice gaps between adoptions and foster care (Freundlich & Gerstenzang, 2003). Inaddition, the new contracts included more kinship care arrangements and family fostercare homes. According to an SRS administrator, “We are working with the contractorsto find the best ways to keep children safe, move them through the system, and, whenpossible, get them back home” (Snell, 2000: 6).

Arizona: Family Builders

112

Implementation. Family Builders is a legislatively approved family preservationservice, originally started in Maricopa and Pima Counties, which make up the majority ofArizona’s child welfare population. The State contracts with providers to deliver serviceswith the goal of preventing child abuse and reabuse, in addition to taking the pressureoff of the public child protection workers and hopefully reducing their caseloads. Theprogram provides legislatively mandated case management, respite, transportation,parenting classes, assessment, and counseling services to low risk child protectioncases (McCullough & Schmitt, 2000). The program emphasizes a strength-based ratherthan investigative approach (Schulzinger, McCarthy, Meyers, de la Cruz Irvine, &Vincent, 1999).

Roles. Through a state-issued RFP process, non-profit lead agencies are selected tocontract with the state of Arizona. The State’s responsibilities include timely paymentsto the lead agencies, monitoring for quality and compliance with standards, as well ascontracting with an independent organization to conduct a program evaluation of thelead agencies. In turn, each lead agency must subcontract with local providers and workto create a community-based system of care, while accountability remains with the leadagency (Sells, 2001). The lead agencies are also responsible for accepting referrals,data collection, family assessment, and quality improvement (McCullough & Schmitt,2000). Within 48 hours of CPS referral, a Family Builders caseworker visit the familyhome unannounced, explains that an abuse report was made, and offers services to thefamily. If a child in the Family Builders program is ever determined to be at immediaterisk of maltreatment or has experienced sexual abuse, the child is referred to ChildProtective Services (Schulzinger, McCarthy, Meyers, de la Cruz Irvine, & Vincent, 1999;Snell, 2000).

Financing. Arizona operates within a case-rate reimbursement system. In 1998, therate per family was $3,500. Smaller payments of $100 for referral and $250 forassessment are paid initially, followed by the $3,500 upon completion of services(McCullough & Schmitt, 2000; (Schulzinger, McCarthy, Meyers, de la Cruz Irvine, &Vincent, 1999). Provider agencies must bear the financial risk of providing all of therequired services to families, however, since services are voluntary, not all families thatare eligible will seek services (Schulzinger, McCarthy, Meyers, de la Cruz Irvine, &Vincent, 1999). Arizona’s greatest achievement has been in aligning financial incentiveswith permanency goals and working to ensure that the foster care payment is not morethan the adoption payment, so that adoptions are encouraged over foster care (Sells,2001).

Outcomes. Lead agencies must collect satisfaction data from providers and families,keep pre- and post- data related to assessments, as well as enter cost, performance,and utilization data into their MIS system. Cumulatively, these measures are expectedto answer questions of program effectiveness, quality of care (e.g., access to services,client satisfaction), cost efficiency, and adherence to professional or licensing standards(e.g., caseload size) (McCullough & Schmitt, 2000). The Office of the Auditor General(OAG) evaluates the Family Builders program for client satisfaction, changes in financial

113

risk levels throughout the progression of a case plan, as well as demographicinformation (Schulzinger, McCarthy, Meyers, de la Cruz Irvine, & Vincent, 1999).

According to a recent national review, Arizona’s Family Builders program has “revisedArizona’s service delivery system, wrought with systemic problems, by streamlining theprocess and establishing clear performance measures to guarantee the desiredoutcome (Sells, 2001: 14). Subsequent to Family Builders implementation, the responserate to the child abuse hotline calls jumped from 84 to 100%. In 1999, only 1.12 percentof children referred to Family Builders had to be placed in out-of-home care, comparedto 16% of investigations conducted by CPS. Not only did Family Builder participantshave lower rates of reabuse, those families who declined services but initially met with aFamily Builders case worker , also had lower rates of recidivism. Findings are based oncomparison with a control group of children with similar level of risk whose cases wereinvestigated and handled by CPS. Overall, the State was much more likely to removechildren from their homes (Snell, 2000).

Lessons Learned. A clear lesson stands out as one examines the Arizona legislature’sinitial appropriation for the Family Builders initiative. In 1997 only $800,000.00 was setaside for the 1998 fiscal year. However, midway through, another $3,555,600.00 wasawarded. There may have been a severe underestimation of the resources required toimplement infrastructure for such an initiative (McCullough & Schmitt, 2000).

Arizona: Adoption Privatization Initiative

Implementation. This initiative was in the planning stage in 1998. The state of Arizonawanted to contract with lead agencies for comprehensive, outcome-driven childprotection services. Outcomes of significant importance to the State are length of stay,permanency, safety and well-being. All children in Arizona child welfare custody, as wellas families who had children in custody, would be eligible. The State is planning for aphased in implementation period, where in adoptions, out-of-home care, and in-homecare are privatized at different times, with the goal of all services eventually falling in theprivate sector with the possible exceptions of intake and investigations (McCullough &Schmitt, 2000).

Roles. The Arizona child welfare administration will enter into a risk-share contract withlead agencies in which the State is responsible for billing and timely payments,determining Medicaid and Title IV-E eligibility, and tracking outcomes via performanceindicators. The lead agencies are responsible for establishing community-basedprovider networks, case management, placement services, treatment and permanencyplanning, and data collection (McCullough & Schmitt, 2000).

114

Financing. Arizona planned to privatize adoptions first with a case-rate reimbursementmechanism. The overall initiative will be funded by a mixture of State and Federal funds,including Medicaid and Title IV-E. The State may ask for a Title IV-E waiver, but had notdone so as of 1998 (McCullough & Schmitt, 2000).

Outcomes. Arizona planned to conduct client satisfaction surveys as well asstakeholder and client focus groups. In addition, the lead agencies would be expectedto track data on assessments, cost, performance, and utilization through their MISsystems. These strategies were designed in order to answer questions of programeffectiveness, permanency outcomes, reentry rates, appropriateness of services,cultural competency, and adherence to professional or licensing standards (McCullough& Schmitt, 2000).

Colorado: El Paso’s Child Placement Agencies

Implementation. Although the state of Colorado is identified as having a managedcare approach to child welfare services, El Paso County has been described asincorporating aspects of service privatization. The El Paso County Department ofHuman Services has been in the role of a managed care organization since 1998, withinwhich, child placement agencies (CPAs) were privatized and given a more substantialrole in child welfare services, beyond the traditional case management functions thathad previously been performed by the State. This decision was made in hopes thatmore community driven and localized services would improve outcomes for children andtheir families, in addition to improving service quality. Many stakeholders were involvedin the planning of this decision, including clients, juvenile justice, the State legislature,local foundations, foster parents, and state, county, and private child welfare agencies(McCullough & Schmitt, 2000; Schulzinger, McCarthy, Meyers, de la Cruz Irvine, &Vincent, 1999).

Roles. With the switch from public to private, average CPA caseload size dropped from25 to 1 to 15 to 1, however a CPA’s responsibilities greatly increased. Formerly, CPAswere only responsible for working with children in care, as well as recruiting, training,and supervising foster parents. The state had been responsible for dealing with thecourts and communicating with biological parents. Under the new initiative, CPAs arenow responsible for children in care, foster parents, biological parents, and the courtsystem. Added requirements were also placed on each CPA for including families indevelopment of treatment plans, as well as ensuring culturally competent services. TheState then shifted into the role of monitoring CPAs (Schulzinger, McCarthy, Meyers, dela Cruz Irvine, & Vincent, 1999).

Financing. Per 1998 data, each CPA received a flat administrative rate of $600 permonth per child. Foster parents were compensated based on the level of care theyprovided to children, which in 1998 ranged from $420-$1,500 per month per child(Schulzinger, McCarthy, Meyers, de la Cruz Irvine, & Vincent, 1999).

115

Outcomes. The two most promising outcomes thus far for El Paso’s privatizationinitiative is that finalized adoptions more than doubled and children in foster caredecreased by 7% during the first year. Outcome measures that the CPAs are heldaccountable for include child safety, permanency, and family reunification. Contracts aredesigned with the enforcement of these outcomes in mind. Additional outcomes El Pasouses to determine system success include reduced caseloads, length-of-stay in out-of-home care, and reductions in adoption and out-of-home care budgets (Schulzinger,McCarthy, Meyers, de la Cruz Irvine, & Vincent, 1999). Information used to determinesuch outcomes includes client satisfaction reports, utilization and cost data, standardassessments, performance data, and pre- and post- data tracked through MIS systems(McCullough & Schmitt, 2000).

Michigan: Foster Care Permanency Initiative

Implementation. Michigan’s Family Independence Agency (FIA) entered intocontractual relationships with four privatized lead agencies in 1997, with the goal ofcollectively operating seven foster care initiatives. The was emphasis placed on theoutcome of each child’s placement, rather than solely focusing on number of childrenwith finalized adoptions, permanent placements, or family reunifications. However, themain goal of each program is achievement of each child’s specified permanency plangoals. All children in state custody, as well as their families, are eligible for services(McCullough & Schmitt, 2000).

Roles. While the State maintains responsibility for all management activities, each leadagency must accept all appropriate referrals, assess service need, design permanencyplans in conjunction with children and their caregivers, and provide a variety of servicesincluding parenting classes, substance abuse services, anger management, domesticviolence, general counseling, life skills (e.g., budgeting and homemaking) as well as jobpreparation, and behavior management techniques to use with children. Servicesspecifically geared toward children include self-esteem and communicationenhancement, academic assistance, conflict resolution strategies, and life skillsdevelopment (McCullough & Schmitt, 2000).

Financing. Lead agencies receive a set of payments per child in a combination of caserate and per diem distribution. Agencies are provided with an initial large payment of$1770 per child referred to the program) and then $14.94 per day. Additional paymentsof $1500 are made to the agency if a child is successfully placed within 315 days ortheir parental rights are terminated within 600 days (420 days for children out-of-state).Upon sustainable placement for a period greater than 6 months, the agency is awarded$730. Similar award is given if a child whose parental rights have been terminated isplaced with adoptive parents within 6 months of the termination. As a somewhatpunitive measure, children who return to the system from adoptive placements are onlyeligible for the initial $1700 and the per diem, but the agency forfeits all otherreimbursements (Freundlich & Gerstenzang, 2003; McCullough & Schmitt, 2000).

116

Outcomes. Performance indicators pertaining to cost, program effectiveness, childfunctioning, service utilization, permanency, and professional licensing requirements aretracked and reported back to the State. The agencies define successful placement asreturning a child to their biological family, placement with relatives or a legal guardian,or transition to independent living (McCullough & Schmitt, 2000). The flexible fundingsystem that includes both lump sum and per diem payments allowed agencies toprovide additional services to families such as housing assistance, which facilitatedfamily reunification efforts. Post reunification services have also increased under theflexible funding model, and families are visited on a weekly basis until the family issufficiently stable and connected to sustainable supports within their community.Preliminary data indicate that while the permanency initiative has not drasticallyimproved the number of reunifications and adoptions, it has made a statisticallysignificant difference in the expediency of these goals (Freundlich & Gerstenzang,2003).

Michigan: Adoption Contracting

Implementation. In 1992 the Michigan Family Independence Agency (FIA) begancontracting with licensed private adoption agencies (Sells, 2001). Each agency contractnow specifies requirements as to timeliness of placements. The two primary gorals ofthis initiative are increasing permanency rates, decreasing length of stay in pre-adoptive placements, and finding placements for children within their local communities.Child eligibility criteria requires that children must be in state custody with adoptionlisted on their permanency plan (McCullough & Schmitt, 2000).

Roles. While the state retains responsibility for determining Title IV-E and Medicaideligibility, the adoption providers are responsible for all service provision and datacollection for the purpose of Federal and State reports. Services include adoptionrecruitment, adoptive placement services, case management, working with the courts,and if needed, listing a child on the adoption exchange for the purpose of prospectiveplacement (McCullough & Schmitt, 2000).

Financing. Prior to 1992 providers were paid for each unit of service regardless of typeof care given, which led to a contradiction of greater funds for longer lengths of stay.Financial reimbursement is now contractually linked to achievement of permanency andpayments are increased based on the degree of difficulty in finding a child suitableplacement (Snell, 2000). Adoption agencies invoice the State within 6 weeks of placinga child in an adoptive home, and then again when the adoption is finalized. The FamilyIndependence Agency reimburses at each of the two junctures (McCullough & Schmitt,2000).

117

Outcomes. Adoption agencies are required to track cost, performance, and utilizationdata through MIS systems and report back to the State. Child and family functioning,program effectiveness, state licensing requirements, and achievement of permanencyplanning goals are specific outcomes of interest to the State (McCullough & Schmitt,2000). During the time period of 1991 (before privatized contracts started) to 1997,Michigan was able to double the number of children that were adopted per year from1,320 to 2,647 (Sells, 2001) and in 1999 Michigan’s Governor reported increased grantfunding due to a record number of adoptions involving special needs children in oneyear (Snell, 2000).

Missouri: The Interdepartmental Initiative for Children with Severe Needs and theirFamilies

Implementation. The Interdepartmental Initiative for Children with Severe Needs andtheir Families has been implemented across two states. Within Missouri, the MissouriAlliance for Children and Families, was created in 1997 as a for-profit conglomeration ofprovider agencies. This group came together to bid for the State contract, implementedthe initiative and then rebid for the contract in 2003. The initiative was viewed as a moveaway from residential treatment programs and toward a more community-based systemof care. In other words, there was a systemic shift from institution-based care to family-centered approaches to service delivery. Eight out of the nine founding provideragencies of the Alliance had been major players across the state in residentialprograms.

The Alliance and State’s focus on residential treatment has specific relevance to thecontext of Missouri’s state policy regarding the Department of Family Services andplacement in residential programs. Prior to 2002, it was impossible for families tovoluntarily place their SED children in a residential facility without relinquishing custodyto the State via substantiated reports of abuse or neglect. Thus, families without theproper resources were needlessly separated and more children were in State care thanneeded to be. This placed increased pressure on abuse investigators and Statecaseworkers, and unfortunately resulted in SED children remaining in highly restrictivesettings for extended amounts of time. The Department of Social Service’s Division ofFamily Services, the Juvenile Justice System’s Division of Youth Services, the State’sMedicaid agency’s Division of Medical Services, and the Department of Mental Health’sDivision of Comprehensive Psychiatric Services, Division of Alcohol and Drug Abuse,and Division of Mental Retardation and Developmental Disabilities all participated inplanning for the initial RFP (Freundlich & Gerstenzang, 2003; McCullough & Schmitt,2000).

The Alliance was the only organization to respond to the State’s RFP. Local providerswere interested but believed the case rate was too low to adequately serve an SEDpopulation. Thus, the Alliance was awarded the 40 month contract despite concernssurrounding the fact that they had no direct service staff or service infrastructure and

118

only 15% of contract funding could be used for administrative costs (Freundlich &Gerstenzang, 2003).

Roles. The Alliance was allowed four months of start-up time before being required toprovide direct services. Children 4 to 18 living in the Eastern region of the State, withserious behavioral health needs, at risk of long-term residential placement were eligiblefor the program. The Alliance was called upon to increase cultural competency,community empowerment, and to move toward an outcome-based system of care.Under the new contract, the Alliance was titled a Care Management Organization(CMO) and held responsible for accepting all eligible referrals, establishing acomprehensive provider network, recruitment and training of all staff, all aspects ofservice provision, data collection for the State and the establishment of a plan forcontinuous quality improvement (Schulzinger, McCarthy, Meyers, de la Cruz Irvine, &Vincent, 1999). The State remains responsible for determining eligibility, referringchildren, and monitoring outcomes of the initiative (Freundlich & Gerstenzang, 2003;McCullough & Schmitt, 2000).

Financing. Each of the founding providers of the Alliance contributed approximately$90,000 in start-up funds. In addition, the Alliance was also able to secure a $500,000credit line before planning even began to bid on the State contract. A fixed monthly caserate reimbursement system was set up based on a historical analysis of costs per childin 1996. The initial rate per child per month was $3,199, which was increased insubsequent years to $3,329 (the Alliance believes this rate is still below adequate levelsfor SED children). The RFP required the Alliance to be able to provide services toapproximately 400 SED children, with capacity to enroll up to 16 new cases eachmonth. Eighty-five percent of contract funds were required to be spent on directservices. In addition, there were incentive and penalty clauses to the contract funding. Ifa child required an out-of-home placement within 120 days of discharge from theprogram, the Alliance had to pay the first two months of that child’s care. On the otherhand, if a discharged child’s placement remained stable after 120 days, the Alliancereceives $1,600 (half of a monthly case rate), to which the Alliance gives $250 to rewardthe caseworker. Such a system has resulted in 90% of placements remaining stable. Inaddition, a State risk pool was established to help the Alliance in the event of extremecases of increased levels of care and cost (3.84% of the case rate and all savings mustbe reinvested into the system). In total, the Alliance receives $600,000 to $800,000monthly and the annual income from the initiative is approximately $10 million(Freundlich & Gerstenzang, 2003; McCullough & Schmitt, 2000).

Outcomes. The contracts specified outcomes pertaining to safety, community,education, family, and behavioral health care, but did not indicate how these measureswere to be calculated or what percentages should be the target goal. The contract didspecify that such outcomes were to be measured against a demographically matchedcontrol group, however, this was never carried out. The two central goals of theprivatization initiative were to decrease children entering foster care and residentialtreatment. However, due to publicized court cases against child welfare systems inother states, the initiative has not been able to affect the volume of intakes. The

119

initiative has achieved some success in keeping children out of residential care. Withinthe initiative’s catchment area, children entering residential care decreased by 11%,while it increased by 14% across the rest of the state between 1998 and 2001.

In contrast, the Alliance itself focuses on two additional goals: moving children inresidential care, or at risk of residential care, into community-based settings andsustaining them in community-based placements. 2001 data indicate that the Alliancehas been able to move a significant portion of children into less restrictive settings, andfor many back into their family’s home. In addition, 86% of children referred by DFS,97% by DMH, and 96% by DYS were still in the same community-based settings theyhad been upon discharge from the program (Freundlich & Gerstenzang, 2003).

Lessons Learned. Unfortunately, due to budget cuts at the state level, both DMH andDYS were forced to stop funding and participation in the privatization initiative. Now onlyDFS refers children to the Initiative and all DMH and DYS referred children weredischarged to the care of these two state divisions. While the Alliance did have theircontract extended until June 30, 2003, the future of the Initiative is uncertain. The Stateis currently examining contractual issues that may have hindered other organizationsfrom the initial bidding process.

It was observed that a key element of the Interdepartmental Initiative was that it washeaded by an independent organization, so that integration could truly occur, ratherthan one department trying to apply it’s service delivery philosophy and structure toother divisions of the State. However, with the creation of a new organization in such afield and political environment comes much financial risk. It was also noted that the factthat the Alliance was for-profit rather than nonprofit did not add to its success. Alliancemembers also suggested that the case rate initially be set higher than estimated needand then be adjusted down if appropriate, as well as transferring from a monthly caserate to a daily rate to facilitate budgeting and planning.

Missouri has learned, from the first two years of the privatization initiative, thatrelationships between state and privatized agencies will undoubtedly be strained.Missouri was fortunate in that there were already some pretty strong public-privaterelationships in place, but there was still the inevitable confusion around roles andcompetition of players for resources. Adequate start-up funding can work to ease someof this tension, however, change, particularly at the service provider level (where manyemployees have done things one way under the State), will take a long time. In addition,an initiative geared toward serving children with high levels of need (as was the case inMissouri) should not be expected to show improved outcomes in less than a year(Freundlich & Gerstenzang, 2003).

South Carolina: Privatized Adoption Services

Implementation. In 1998 South Carolina introduced an RFP that called for private,community-based, licensed agencies to provide adoptions service, including recruitment

120

and training of foster parents, as well as pre- and post- placement assessment. Goals ofthe initiative include decreasing length of stay in out-of-home care for children whosepermanency goal is adoption, in addition to ensuring the safety and stability of adoptiveplacements. Child eligibility for the program requires that the child has adoption as agoal listed in the permanency plan, the child is free for adoption, or the petition toterminate parental rights has been filed (McCullough & Schmitt, 2000).

Roles. The lead agency has the option of providing adoption services or contractingout for the services of assessment, training, recruitment, home studies, casecoordination and management, the provision of transportation, keeping up with courtreviews, and post adoption services. Custody of all children will remain with the Stateuntil adoptions are finalized. The State is also responsible for determining childeligibility, making referrals to the lead agencies, reimbursing and monitoring of theagencies, approving adoption placements, and mediating in situations of conflictingplacement options (McCullough & Schmitt, 2000).

Financing. South Carolina’s privatized adoption services are reimbursed at a caserate. In 1998, this rate was no more than $13,000 per child and $15,700 per siblinggroup. However, this amount is gradually paid to the lead agency and subcontractors asitems are accomplished such as placement in an adoptive or pre-adoptive home, amonthly fee in this placement for up to 12 months, time of final adoption decree, andpositive outcomes upon a 6 month post- adoption assessment. Other payments arepossible for home studies and the accomplishment of placing a high legal risk child withan adoptive family. In addition, the agencies must show proof of six months operatingreserve (McCullough & Schmitt, 2000).

Outcomes. Each lead agency must provide to the State quarterly reports on requiredface-to-face visits with children and adoptive families, the level of and appropriatenessof each service provided to each child, and progress made on permanency plans.Service plans must be completed within 30 days of intake and revised every 6 months.Each lead agency must also meet Federal and State requirements pertaining tocaseload size, staff and adoptive family training, and required credentials of providers(McCullough & Schmitt, 2000).

Hamilton County, Ohio: Creative Connections

Implementation. Creative Connections is a program designed to serve children in thechild welfare system and their families in Hamilton, County Ohio. Prior to 2003 the leadagency for this initiative was Beech Acres H however, the new contract has recentlybeen awarded to a behavioral health organization originally located in Indiana. HamiltonCounty started with a cluster approach to child welfare services in which several child-serving systems collaborated in planning and assessments. This evolved in the mid1990s to a managed care approach in which the newly created Family and ChildrenFirst Management (FCF) contracted with three care agencies (one of which was BeechAcres) to supervise 22 direct service agencies. Problems arose as the three care

121

agencies had very different philosophies of service provision and the public-privatepartnership devolved into a contract-vendor situation in the midst of funding deficits. In1998 the Department of Human Services contracted with Beech Acres as the lead,wanting to shift funding to a private entity in the hope that they would be able tocontribute additional resources as well as unify goals and approaches of the initiative(Freundlich & Gerstenzang, 2003).

Roles. Under Beech Acre’s four-year contract as lead agency, they were responsiblefor accepting all referrals, purchasing services from provider network agencies, workingwith the courts, administration, monitoring of service delivery, and quality improvement,evaluation, and information management. Beech Acres was a large agency andCreative Connections was actually only one of 14 programs that they ran. Interestingly,Creative Connections staff were drawn from Beech Acres, many from the former FCF,and several others from another organization that had been a care agency under FCF.Care Connections set as their major goals improvement of child level outcomes, costefficiency and containment, increase appropriateness of services, improved access toservices, and case coordination across systems (Freundlich & Gerstenzang, 2003;McCullough & Schmitt, 2000).

While eligibility for Creative Connections started off as including only children with ahigh level of need, those involved with juvenile justice, and those with SED, over timethe criteria became any child in need of residential care. This led to financial andprogram effectiveness challenges. Beech Acres itself did not provide many directservices, with the exception of therapeutic treatment and some residential care. BeechAcres contracted with 95 service providers as of 2000, in order to provide services suchas counseling, foster care, mentoring, residential care and partial hospitalization.However, subsequent to an actuarial analysis by Tony Broskowski in 2001, BeechAcres began to downsize the number of providers they worked with and consider othercommunity-based and nontraditional sources of support (Freundlich & Gerstenzang,2003).

Financing. Creative Connections was financed through Medicaid and county dollars,with the Department of Job and Family Services contributing the most significant level offunds. The system of reimbursement for Creative Connections was originally a caserate of $3,130.29 per child in FY98-99. A formula was then used to calculate the caserate for subsequent years, which involved the program costs of the proceeding year,adjustments for inflation and Medicaid contributions. During the third year of CreativeConnections, the case rate was $3,977.13, not including the Medicaid contribution.However, this still fell short of actual operating costs, particularly when administrativeexpenditures were factored into the equation. Due to a running deficit, by the end ofBeech Acres contract in 2002, the organization had to absorb 6 million dollars ofCreative Connection’s deficits, while the county dealt with the remaining 20%. Contractrenegotiations included different case-rates per level of services provided. Certain stop-loss provisions were included in the original contract that controlled for both loss andsurplus, however, additional strategies were proposed by the time of contractrenegotiation because Beech Acres had absorbed most of their financial loses

122

independently (Freundlich & Gerstenzang, 2003). Overall, Creative Connections onlyrepresents 18% of Hamilton County’s child welfare budget, due to the blending of fundswith juvenile justice, mental health, developmental disabilities, substance abuse andMedicaid (McCullough & Schmitt, 2000).

Outcomes. Creative Connections was contractually obligated to produce seven keyoutcomes pertaining to decreasing length of stay, ensuring appropriateness ofplacement and services, stability of services, quality of service, individually tailoredservice plans, and participation with State monitoring and data tracking. There werecontractually linked financial penalties for poor performance on these outcomes, suchthat administrative and case management funds could be withheld. Per contract in2002, up to $225, 000 could be withheld from the program (Freundlich & Gerstenzang,2003).

Initial program effectiveness was measured via CAFAS scores and it was determinedthat children enrolled in Creative Connections had lower scores. Also, as CreativeConnections matured, average CAFAS scores fell as well (e.g., in 2000 CAFASaverage was 90; in 2001 average was 80). Children sustained statistically different andimproved levels of interaction with others and overall functional improvement after sixmonths of service provision. Decreased delinquency and improved behavior both atschool and at home were also noted, though not statistically significant (Freundlich &Gerstenzang, 2003). An evaluation report conducted by the Auditor for the State of Ohio(2001) noted Creative Connections had improved both quality and coordination of care,as well as increased access to services and the development of innovative services.

Lessons Learned. Creative Connections is a good example of how changes inleadership can affect implementation efforts. The initial director of the program hadbeen part of the former FCFC management team, and it quickly became clear that shewas not capable of building the necessary program infrastructure. Unfortunately, overCreative Connection’s four-year contract, there were six different executive directors.This administrative instability was a constant challenge and did not help other internalproblems, such as dealing with MIS system issues.

The importance of a shared vision also became apparent, as well as clear delineation ofroles and responsibility between public and private entities. In addition, with risk-basedcontracts it is especially important to have agreement between the lead agency andpublic referring agencies as to child eligibility criteria as this determines the volume ofchildren that must be served through a mandated open-door policy. However, it wasalso noted that it is not recommended that a lead agency try to specialize in only SEDchildren due to the financial deficits this can lead to. A mix of cases of varying level ofneed was suggested. The accept all referrals policy also necessitates that a leadagency must have a great variety of expertise among it’s provider agency network,which in some communities can be very difficult to accomplish (Freundlich &Gerstenzang, 2003).

123

APPENDIX 2. Lead Agency CEO Protocol

IntroductionThe Louis de la Parte Florida Mental Health Institute of the University of South Florida isunder contract with the Florida Department of Children and Families to evaluateCommunity-Based Care (CBC), the privatization of child welfare in Florida. The purposeof this survey is to collect information about how CBC was planned and implementedand how the project has changed over time. Specifically, we need to gather informationabout community governance, organizational structure of lead agencies and theirprovider networks, and financial risk, as well as information on each lead agency’sBoard of Directors and Community Alliance.

1. Please describe your organizational structure, speaking to both strengths andweaknesses of that structure, especially pertaining to financial risk, your providernetwork, and community governance. In addition to a brief explanation, pleaseattach your organizational chart.

2. Were you provided with any guidance through the process of creating yourorganizational structure, provider network contracts, and community-basedsystem of care?

a. If yes, please explain what assistance was provided and by whom.

3. Please explain how the Community Alliance(s) in your area interact with yourlead agency.

4. Please list (or include in an e-mail attachment) all members of your Board ofDirectors and their professional and community roles. Please share examples ofpotential conflicts of interest or those you have encountered during the selectionprocess of Board members.

5. What other advisory groups does your lead agency have beyond the Board ofDirectors and Community Alliances?

a. How does your lead agency receive and integrate their input?

6. What aspects of your agency’s organizational structure facilitate and supportimplementation and operation of CBC? What aspects exist as barriers toimplementation?

7. What is the attitude of other cooperating agencies (e.g., schools, juvenile justice),key community stakeholders, and major provider agencies toward CBC?

8. How flexible has your agency been (and been able to be) in using its financialresources?

124

9. If it has not been sufficiently flexible, what has hindered financial flexibility?

10. What are your local revenue sources?

11. What challenges is your agency facing related to financial and fiscal systemissues?

12. If your agency existed prior to CBC implementation, how would you describe thefiscal impact that CBC implementation has had on your organization?

13. Is your current organizational structure different than what was submitted for theITN?

a. If yes, please describe any significant changes in your agency’sorganizational structure or administrative staff since CBC implementationand lessons learned.

14. How is your lead agency’s provider network structured? In addition to a briefexplanation, please attach your Network Management Plan.

15. How many contracts with providers does your lead agency currently maintain?

16. Please describe each contractual relationship your lead agency has with providerorganizations. Please describe the financial arrangements for or within thesecontracts (e.g., fee-for-service, capitation, case rates).

17. In what way, if any, do separate provider agencies interact with each other?Please provide a few examples of formal interaction (e.g., leadership meetings,co-location, information sharing).

18. Please describe the quality assurance and contract management feedback youprovide to your network providers. How often and by what method is it provided?Are there formal patterns for communication?

19. Please describe your methods, if any, of determining that your providers utilizethe feedback you provide.

20. Please describe any prior authorization processes or other utilizationmanagement processes you employ for managing services provided throughyour network. In addition to a brief explanation, please attach your UtilizationManagement Plan.

21. Regarding your service delivery structure, please describe your efforts tocoordinate or integrate case management and behavioral health services.

125

********************************************************************************************

VeryPlease use the following scale in Great Great Some Little To No Don’tAnswering these questions: Extent Extent Extent Extent Extent Know

1 2 3 4 5 DK

To what extent have the following been accomplished through CBC implementation inyour area?

22. Improved assessment of local needs for child welfare services.

23. Improved accountability for a local system of child welfare services to meetperformance outcomes and standards.

24. Provided for additional testing of creative and flexible strategies for financingchild welfare services to enhance individualized treatment and/or child-leveloutcomes.

25. Controlled the costs of services without sacrificing quality of care.

26. Assured continuity of care for all children entering the child welfare system.

27. Improved the overall quality of child welfare services through the use of bestpractice models.

28. Reduced length of stay for children in the child welfare system.

29. Demonstrated effort toward improved service integration between child welfareprograms and other system of care programs (e.g., behavioral health or juvenilejustice).

126

APPENDIX 3. Sample Letter to Lead Agency Soliciting Promising Approach toSupervisory Training/Practice

Dear ,

As you are likely aware, the Louis de la Parte Florida Mental Health Institute at theUniversity of South Florida is currently in its third year of the Evaluation of Florida’sCommunity-Based Care initiative. An issue that has repeatedly come up during ourevaluation is the importance of the role of supervisors in assuring the success of anychild protective system. Last year’s report made specific recommendations about: (1)the need to adapt training and supervision for case managers to meet the demands ofcommunity-based care and (2) for changes in training for case managementsupervisors to support the changes expected in frontline practice(http://www5.myflorida.com/cf_web/myflorida2/healthhuman/publications/pubs.html#cbc).

Across the country, researchers and practitioners alike are focusing their attentions onthe need to “train child welfare supervisors as child welfare supervisors” (Maximus andFSU, 2003). Poor supervisory relationships are pointed to as being one of the primaryfactors leading to employee dissatisfaction and turnover (Martone & Solek, 2001). Intheir review of “best practices Maximus and FSU identified community/universitypartnerships to improve training (Los Angeles County Inter-University Consortium forChild Welfare); specialized training for supervisors and interactive review of practice bysupervisors and case workers (Commonwealth of Pennsylvania); and the developmentof career and educational paths for supervisors.

The Center for the Study of Social Policy, in their implementation of CommunityPartnerships for Protecting Children (the forerunner of Florida’s NeighborhoodInitiatives), emphasizes the need for supervisors to have time available to work withtheir staff, being readily available to provide case consultation “spur of the moment” orto “share” difficult cases with their workers to improve teaching, learning and supportiveopportunities (“Community Partnerships for Protecting Children”, 2003). The WalterTrieschman Center (a division of the Child Welfare League of America) is devoted to thedevelopment and delivery of an “Effective Supervisory Practice” program that is basedon the premise that “effective state of the art treatment and service provision forchildren, youths and families requires highly competent, well-trained supervisors and acommitment from leadership to the development of competent supervisors.”

In short, our findings do not differ from what is being seen in the child welfare systemthroughout the country. This is where you come into the picture. As part of the FY03-04evaluation we are interested in identifying promising approaches to supervisorytraining/practices within our own CBC lead agencies. We are aware that some agenciesare early in their implementation and may just have a “plan” for how to address theseissues. We would love to hear those as well.

127

If you have a promising approach that you would like to share, please contact me at(813) 974-6419 or kershaw@fmhi.usf.edu. After gathering some initial information, I willschedule follow-up interviews with you and/or relevant staff in your community. Our goalis to include this under the “Quality Performance” heading in our report due to theDepartment of Children and Families in May 2004. As such, a response would be muchappreciated no later than March 1, 2004, allowing time for follow-up.

I look forward to hearing from you in the near future.

Sincerely,

Mary Ann Kershaw, CoordinatorQuality Performance ComponentCBC Statewide Evaluation

128

APPENDIX 4. Community Alliance Chair Protocol

IntroductionThe Louis de la Parte Florida Mental Health Institute of the University of South Florida isunder contract with the Florida Department of Children and Families to evaluateCommunity-Based Care (CBC), the privatization of child welfare in Florida. The purposeof this survey is to collect information about Community Alliances and their activities asthey relate to Community-Based Care. Specifically, we need to gather information aboutcommunity governance and what type of interaction occurs between CommunityAlliances and both DCF and CBC lead agencies.

1. Please list all of your Alliance’s voting members and their community andprofessional affiliations.

2. Please forward to the research team minutes of two recent Community Alliancemeetings.

3. Does your Alliance discuss issues of child welfare? If yes, how often?

4. Do you have any committees or workgroups to discuss child welfare? Pleaseexplain.

5. Is there a formal communication process in place with the DCF District/Regionaloffice regarding issues of child welfare? If yes, please describe.

6. Is there a formal communication process in place with the lead agency? If yes,please describe.

7. If the Alliance makes recommendations to the lead agency, do you receivefeedback? In what form?

8. What are some of the benefits and challenges to communicating with DCF onissues related to child welfare? Please list specific examples if possible.

What are some of the benefits and challenges to communicating with the leadagency? Please list specific examples if possible.

129

APPENDIX 5. Community Alliances: Responsibilities, Duties, Members, and FocusOutcome AreasResponsibilities Duties Initial Members Possible

Outcome Areas

1. To be informedabout theservices of theircommunity.

2. To discuss andreachagreement ondesiredoutcomes fortheircommunitychildren andfamilies.

3. To work toreduce theduplication andfill in the gapsin servicesacrossagencies.

4. To listen to theneeds andwants of theircommunitychildren andtheir families.

5. To coordinatefunding issuesacrossagencies.

1. Joint planning forresource utilizationin the community,including resourcesappropriated to thedepartment andany funds that localfunding sourcesshould choose toprovide.

2. Needs assessmentand establishmentof communitypriorities forservices delivery.

3. Determiningcommunityoutcome goals tosupplement state-required outcomes.

4. Serving as acatalyst forcommunityresourcesdevelopment.

5. Providing forcommunityeducation andadvocacy on issuesrelated to deliveryof services.

6. Promotingprevention andearly interventionservices.

s. 20.19 (6)(b) F.S.

1. The DistrictAdministrator

2. A representativefrom countygovernment

3. A representativefrom the schooldistrict

4. A representativefrom the countyUnited Way

5. A representativefrom the countysheriff’s office

6. A representativefrom the circuitcourtcorresponding tothe county

7. A representativefrom the countychildren’s board,if one exists

s. 20.19 (6) (d)F.S.

1. Child protectionoutcomes

2. Outcomesrelated toprevention andearlyintervention

3. Childbehavioralhealthoutcomes

4. Child physicalhealthoutcomes

5. Vocational andtransitionaloutcomes aschildrentransition toadulthood

6. Child educationoutcomes

7. Childdelinquencyoutcomes

8. Outcomes inother areasrelated tofamilies

Pg. iv, Community-Based CareCommunityAlliance ResourceHandbook Version1.0

Attachment A-2,Community-Based CareCommunity AllianceHandbook Version 1.0

Attachment A-3,Community-Based CareCommunity AllianceHandbook Version 1.0

Stage 2-2, Community-Based Care CommunityAlliance HandbookVersion 1.0

130

APPENDIX 6. Types of Child Protection Staffings

TYPE OFSTAFFING

MEETINGCONVENER

INVITEDATTENDEES

FREQUENCY AUTHORITY

EarlyIntervention

ProtectiveInvestigationsUnit

PI, PI supervisor,lead agencysupervisor

Once at time oftransfer

Statutorilymandated uponsheltering child

131

APPENDIX 7. Statistical Results

7.1: Results of Cox Regression of Exiting Out-of-Home Care

Comparison Categories Wald Chi-Square Number of Cases %Service contract 6 monthsor less 3.917 6101 47.40

Service contract between 7and 17 months 3.461 2849 48.81

Service contract 18months or longer 1.547 2964 48.49

Rest-of-State 7,869 56.17

7.2: Results of Cox Regression of Reentry in Out-of-Home Care

Comparison Categories Wald Chi-Square Number of Cases %Service contract 6 monthsor less 41.211* 6,279 8.25

Service contract between 7and 17 months 8.176* 2,542 13.26

Service contract 18months or longer 40.720* 2,422 9.91

Note: * p < 0.05

7.3: Results of Cox Regression of Recurrence of Maltreatment

Comparison Categories Wald Chi-Square Number ofCases %

Service contract 6 months orless 12.840* 41,123 10.22

Service contract between 7and 17 months 10.865* 15,547 11.16

Service contract 18 monthsor longer 5.677* 20,640 10.86

Note: * p < 0.05

132

7.4: Percentage of Children Reunified with Parents Based on FY01-02 Cohort byLead Agencies

Lead Agency Total number of cases in exitcohorts

Percentage

Sarasota YMCA 794 56.2Family Continuity Programs 1,085 54.4Partnership for Community-Based Care

543 51.7

FamiliesFirst Network 1,100 59.5Hillsborough Kids, Inc. 1442 60.1Child and FamilyConnections

784 51.5

ChildNet, Inc. 1,573 45.4Family Support Services 1,426 50.4United for Families 728 66.5Heartland for Children 1,433 62.2Partnerships for Families 335 58.8Median = 56.2 Mean = 56.06

7.5: Percentage of Children Reunified with Parents Based on FY01-02 by CountyCategories

County Category Total number of cases inexit cohorts

Percentage

Service contract 18 months orlonger

2422 54.4

Service contract is between 7 and17 months

2542 59.8

Service contract is 6 or lessmonths

6279 54.3

Rest-of-State 9254 55.2

7.6: Percentage of Children Reunified with Parents Based on FY02-03 by CountyCategories

County Category Total number of cases inexit cohorts

Percentage

Service contract 18 months orlonger

2,997 44.5

Service contract is between 7 and17 months

3,016 59.5

Service contract is 6 or lessmonths

7,231 46.7

Rest-of-State 9,658 52.6

133

7. 7: Results of Logistic Regression Analysis of Children Reunified with TheirParents

Wald Chi-Square

Comparison Site Time Site by Time

Service contracts havebeen held for 6 months orless(n = 30,376)Service contracts havebeen held between 7 and17 months (n = 13,845)Service contracts havebeen held for 18 months orlonger (n = 12,808)

0.17 7.38* 0.01

Note: * p < 0.05

7.8: Percentage of Children Placed with Relatives Based on FY01-02 Cohort byLead Agencies

Lead Agency Total number of cases in exitcohorts

Percentage

Sarasota YMCA 794 26.2Family Continuity Programs 1,085 29.7Partnership for Community-Based Care

543 18.8

FamiliesFirst Network 1,100 30.7Hillsborough Kids, Inc. 1,442 21.3Child and FamilyConnections

7,84 14.7

ChildNet, Inc. 1,573 34.9Family Support Services 1,573 35.7United for Families 728 17.9Heartland for Children 1,433 21.7Partnerships for Families 335 23.9Median = 23.90 Mean = 25.04

134

7.9: Percentage of Children Placed with Relatives Based on FY01-02 by CountyCategories

County Category Total number of cases inexit cohorts

Percentage

Service contract 18 months orlonger

2,422 26.1

Service contract is between 7 and17 months

2,542 25.4

Service contract is 6 or lessmonths

6,279 27.0

Rest-of-State 9,254 24.0

7.10: Percentage of Children Placed with Relatives Based on FY02-03 by CountyCategories

County Category Total number of cases inexit cohorts

Percentage

Service contract 18 months orlonger

2,997 32.3%

Service contract is between 7 and17 months

3,016 23.6%

Service contract is 6 or lessmonths

7,231 27.0%

Rest-of-State 9,658 22.3%

7.11: Results of Logistic Regression Analysis of Children Placed with Relatives

Wald Chi-Square

Comparison Site Time Site by Time

Service contracts havebeen held for 6 months orless(n = 30,376)Service contracts havebeen held between 7 and17 months (n = 13,845)Service contracts havebeen held for 18 months orlonger (n = 12,808)

10.35* 19.88* 16.08*

Note: * p < 0.05

135

7.12: Percentage of Children with Adoption Finalized Based on FY01-02 Cohort byLead Agencies

Lead Agency Total number of cases in exitcohorts

Percentage

Sarasota YMCA 794 12.5Family Continuity Programs 1,085 7.5Partnership for Community-Based Care

543 17.9

FamiliesFirst Network 1,100 8.1Hillsborough Kids, Inc. 1,442 7.2Child and FamilyConnections

784 24.2

ChildNet, Inc. 1,573 9.3Family Support Services 1,426 7.0United for Families 728 10.6Heartland for Children 1,433 11.2Partnerships for Families 335 12.8Median = 10.60 Mean = 11.66

7.13: Percentage of Children with Adoption Finalized Based on FY01-02 byCounty Categories

County Category Total number of cases inexit cohorts

Percentage

Service contract 18 months orlonger

2422 11.4

Service contract is between 7 and17 months

2542 7.6

Service contract is 6 or lessmonths

6279 11.4

Rest-of-State 9254 12.4

7.14: Percentage of Children with Adoption Finalized Based on FY02-03 byCounty Categories

County Category Total number of cases inexit cohorts

Percentage

Service contract 18 months orlonger

2,997 10.6

Service contract is between 7 and17 months

3,016 7.5

Service contract is 6 or lessmonths

7,231 14.1

Rest-of-State 9,656 13.1

136

7.15: Results of Logistic Regression Analysis of Children with Adoption Finalized

Wald Chi-Square

Comparison Site Time Site by Time

Service contracts havebeen held for 6 months orless(n = 30,376)Service contracts havebeen held between 7 and17 months (n = 13,845)Service contracts havebeen held for 18 months orlonger (n = 12,808)

3.51 7.81* 14.17*

Note: * p < 0.05

137

APPENDIX 8. Direct Services and OCAs by Service Category (FY02-03)Out-of-Home Services

OCA OCA_Title39EAS IV-A EMERGENCY ASSIST. FOR SUBSTITUTE CAREBX000 EMERGENCY SHELTER CARECHF0T CHAFEE FOSTER CARE INDEPENDENCE PROGRAM - OTHERCHFRB CHAFEE FC INDEPENDENCE PROGRAM ROOM AND BOARDCHKEM SECURITY CHECKS OF CAREGIVERSE1100 PSSF TIME-LIMITED FAMILY REUNIFICATION SERVICESE110R PSSF N’HOOD PARTNER. TIME-LIMITED FAMILY REUN.E1400 PSSF COMMUNITY FACILITATION - OHSE140R PSSF N’HOOD PARTNER. COMMUNITY FACILITATION-OHSFC00H FOSTER CARE ADMINISTRATION-OUT OF HOME CAREKR000 CHAFEE FC INDEPENDENCE PROGRAMKT000 FOSTER CARE RECRUITMENT AND RETENTIONNA000 NONPSYCHIATRIC RESIDENTIAL GROUP CAREPR005 MEDICAID ADMIN.- OHS - COMMUNITY-BASED CAREPR006 CHAFEE FC INDEPENDENCE PROGRAM - CBCPR020 IV-E FOSTER CARE CASE MANAGEMENT - CBCPR021 IV-E FOSTER CARE PREPLACEMENT/PLACEMENT - CBCPR022 IV-E FOSTER CARE ELIGIBILITY DETERMINATION - CBCPR023 IV-E FOSTER CARE - OTHER SERVICES - CBCPR024 SF CHILD WELFARE SVCS. OUT-OF-HOME ADMIN. - CBCPR025 CHILD WELFARE SVCS.- OUT OF HOME - TSTF - CBCPR026 IV-B CHILD WELFARE SVCS OUT-OF-HOME ADMIN. - CBCPR027 CHILD WELFARE SVCS. - OUT OF HOME - O&MTF - CBCPR050 IV-E FOSTER CARE MAINTENANCE PAYMENTS - CBCPR051 IV-E FOSTER CARE MAINTENANCE - TSTF - CBCPR052 IV-E FOSTER CARE MAINT.- EXCESS EARNINGS - CBCPR4A0 TITLE IV-A EMERGENCY ASST.-OUT-OF-HOME-CBCPR4A2 TITLE IV-A EMERGENCY ASSISTANCE - ADMIN. - CBCPRC0T CHAFEE FOSTER CARE INDEPENDENCE PRG. OTHER - CBCPRCRB CHAFEE FC INDEPENDENCE PROGRAM RM. & BOARD - CBCPRE11 PSSF TIME-LIMITED FAMILY REUNIFICATION - CBCPRE14 PSSF COMMUNITY FACILITATION - OUT-OF-HOME - CBCPRS11 SSBG FOSTER CARE SVCS. OHS ADMINISTRATION - CBCPRS29 SSBG OTHER CHILD WELFARE SERVICES - CBCPRSS2 SSBG TANF XFER CHILD WELFARE SVC OHS ADMIN - CBCPRT02 TITLE IV-A EMER. ASST. ELIGIBILITY DET. OH-CBCPRV01 IV-E FOSTER CARE ELIGIBILITY DET.- PRV CONTRACTSPRV02 STATE FUNDED OUT-OF-HOME ADMIN. - PRV. CONTRACTSPRV06 CHAFEE FC INDEPENDENCE PROGRAM - PRIVATIZATIONPRV12 TITLE IV-A EMERG. ASST. ADMIN.- OH-PRV. CONTRACTPRV13 IV-E FOSTER CARE PLACEMENT - PRIVATE CONTRACTPRV15 IV-E FOSTER CARE CASE MANAGEMENT-PRV. CONTRACTS

138

PRV17 IV-E FOSTER CARE OTHER SERVICES - PRV. CONTRACTSPRV18 IV-E FOSTER CARE LIC/RECRUIT/TRAIN-PRV. CONTRACTPRV72 TITLE IV-B FOSTER CARE OUT-OF-HOME SERVICESPRVB5 MEDICAID ADMIN.- FOSTER CARE - PRV. CONTRACTSPRVT0 IV-A EMERG. ASST. ELIGIBILITY DET.-PRV. CONTRACTRGC05 XIX MEDICAID ADMINISTRATION OUT-OF-HOME-RGCRGC20 IV-E FOSTER CARE CASE MANAGEMENT-RGCRGC21 IV-E FOSTER CARE - PLACEMENT OHS - RGCRGC22 IV-E FOSTER CARE - ELIG. DETERMINATION OH - RGCRGC23 IV-E FOSTER CARE OTHER-RGCRGC24 STATE FUNDED MAINTENANCE PAYMENTS OHC - RGCRGC50 IV-E FOSTER CARE MAINTENANCE - RGCRGC60 CHILD WELFARE SVCS. - OHC ADMINISTRATION - RGCSF00H STATE FUNDED OUT OF HOME CARESFBAP BEHAVIORAL ANALYST UNITSSFFPS FOSTER PARENT SUPPORTS CONTRACTED SERVICESWH000 FOSTER CARE PROGRAM ADMINISTRATIONWO004 CHILD WELFARE MAINTENANCE PAYMENTS - OHSWR000 IV-E FOSTER CARE MAINTENANCE PAYMENTS - OHS

In-Home Services

OCA OCA_Title2L000 FAMILY BUILDERS2LM0E FAMILY BUILDERS-TANF MOE39IHS TANF IN-HOME EMERGENCY SERVICES89K00 HOUSEKEEPER/HOMEMAKER SERVICES FOR CHILDREN89L00 SUPPORT SVCS FOR CHILDREN IN OUT-OF-HOME CAREE1300 PSSF COMMUNITY FACILITATION - IN-HOME SUPPORTE130R PSSF N’HOOD PARTNER. COMMUNITY FACILITATION-IHSE4000 PSSF FAMILY PRESERVATION SERVICESE400R PSSF N’HOOD PARTNER. FAMILY PRESERVATION SERVICEPR105 MEDICAID ADMINISTRATION - IN-HOME - CBCPR124 SF CHILD WELFARE SVCS. IHS ADMINISTRATION - CBCPR125 CHILD WELFARE SERVICES - IN-HOME - TSTF - CBCPR126 CHILD WELFARE SERVICES - IN-HOME - IV-B - CBCPR127 CHILD WELFARE SERVICES - IN-HOME - O&MTF - CBCPR2L0 FAMILY BUILDERS PROGRAM-COMMUNITY BASED CAREPR2LM FAMILY BUILDERS PROGRAM - TANF MOE - CBCPR4A1 TITLE IV-A EMERGENCY ASSIST. ADM. IN-HOME - CBCPRE04 PSSF FAMILY PRESERVATION SERVICES - CBCPRE13 PSSF COMMUNITY FACILITATION - IN-HOME - CBCPRS22 SSBG PROTECTIVE SERVICES - COMMUNITY-BASED CAREPRSS1 SSBG TANF TRANSFER - IN-HOME SVCS. - CBCPRT01 TANF ELIGIBILITY DETERMINATION - IH - CBCPRV05 MEDICAID ADMIN.- IN-HOME - PRV. CONTRACTS

139

PRV10 TANF PROTECTIVE SVCS. ADMIN.- PRV. CONTRACTSPRV21 STATE FUNDED PROTECTIVE SVCS.- PRV. CONTRACTSPRV34 SSBG-PREVENTION/INTERVENTION-IN HOME SUPPORTSPRV71 TITLE IV-B PROTECTIVE SUPERVISION-IH CASE PLANRGC15 XIX MEDICAID ADMINISTRATION IH - RGCRGC30 IV-E FOSTER CARE - CASE MANAGEMENT IH - RGCRGC34 STATE FUNDED CHILD WELFARE SERVICES IH -RGCW6000 INTENSIVE CRISIS COUNSELING PROGRAMWG000 PROTECTIVE SERVICES FOR CHILDREN

Adoption-Related Services

OCA OCA_Title39MAS TANF MAINTENANCE ADOPTION SUBSIDYAIA00 ADOPTION INCENTIVE AWARDE1200 PSSF ADOPTION PROMOTION & SUPPORT SERVICESLK000 PURCHASE OF ADOPTION SERVICESMCSA0 MEDICAL COSTS OF SUBSIDIZED ADOPTIONSMP000 NONRECURRING ADOPTION EXP. FOR SPEC. NEEDS CHILDPR003 IV-E ADOPTION ASSISTANCE ADMINISTRATION - CBCPR060 IV-E ADOPTION ASSISTANCE SUBSIDY PAYMENTS - CBCPR061 IV-E ADOPTION ASSISTANCE MAINT.- TSTF - CBCPR062 IV-E ADOPTION ASST. MAINT.- EXCESS EARNING - CBCPRA05 MEDICAID ADMINISTRATION - ADOPTION - CBCPRA24 SF CHILD WELFARE SVCS. ADOPTION ADMIN. - CBCPRA25 CHILD WELFARE SERVICES - ADOPTION - TSTF - CBCPRA25 CHILD WELFARE SERVICES - ADOPTION - TSTF - CBCPRA26 CHILD WELFARE SERVICES - ADOPTION - IV-B - CBCPRA27 CHILD WELFARE SERVICES - ADOPTION - O&MTF - CBCPRA70 STATE FUNDS CHILD WELFARE ADOPTION SUBSIDY - CBCPRA71 CHILD WELFARE ADOPTION SUBSIDIES - TSTF - CBCPRA80 MEDICAL COSTS OF SUBSIDIZED ADOPTIONS - GR - CBCPRA81 MEDICAL COSTS OF SUBSIDIZED ADOPTIONS-TSTF-CBCPRA90 MAINTENANCE ADOPTION SUBSIDY - TANF - CBCPRAIA IV-E ADOPTION INCENTIVE GRANT - CBCPRE12 PSSF ADOPTION PROMOTION & SUPPORT SERVICES - CBCPRS01 SSBG ADOPTION SERVICES ADMINISTRATON - CBCPRSSA SSBG TANF TRANSFER - ADOPTION - CBCPRT03 TANF ADOPTION ADMINISTRATION - CBCPRV03 IV-E ADOPTION ADMINISTRATION - PRV. CONTRACTSPRV04 STATE FUNDED ADOPTION ADMIN. - PRV. CONTRACTSPRV73 ADOPTION ASSISTANCE-NON IV-E ELIGIBLEPRVA5 MEDICAID ADMIN. ADOPTION - PRV. CONTRACTSPRVT2 TANF ADOPTION ASSISTANCE ADMIN. PRV. CONTRACTSWO006 CHILD WELFARE ADOPTION SUBSIDIESWR001 IV-E ADOPTION SUBSIDIES

140

WY000 ADOPTION PLACEMENT ADMINISTRATIONWYC00 ONE CHURCH ONE CHILD

Other Direct Child Protective Services

OCA OCA_TitleDR001 CAN BASIC GRANT PREVENTION & INTERVENTIONE3000 HOME VISITOR/HIGH RISK NEWBORNE6000 PSSF FAMILY SUPPORT SERVICES - IHSE600R PSSF N’HOOD PARTNER. FAMILY SUPPORT SERVICESPR008 CAPTA GRANT - COMMUNITY-BASED CAREPR010 TANF RELATED CHILD WELFARE SERVICES - CBCPRE06 PSSF FAMILY SUPPORT SERVICES - CBCPRS04 SSBG COUNSELING SERVICES - COMMUNITY-BASED CAREPRS20 SSBG PREVENTION/INTERVENTION - CBCPRS29 SSBG OTHER CHILD WELFARE SERVICES - CBCTCMCW STATE SHARE OF CHILD WELFARE TARGETED CASE MGT.

141

APPENDIX 9. Total Expenditures for District Specific Projects by District andFiscal Year

District FY96 FY97 FY98 FY99 FY00 FY01 FY02 FY03 Total1 141,225 127,718 129,854 152,860 26,105 241,482 0 0 819,2442 227,792 207,253 210,718 205,221 45,625 62,998 0 0 959,6083 148,590 149,784 152,288 152,181 158,838 385,576 225,604 233,858 1,606,7184 258,691 233,765 232,142 275,506 24,653 8,895 0 0 1,033,6515 147,788 131,542 154,341 232,547 56,618 0 0 0 722,8366 268,481 220,222 251,890 486,610 200,459 253,503 0 0 1,681,1647 258,164 238,929 244,204 321,587 72,436 0 0 0 1,135,3218 200,412 99,980 141,193 124,083 23,489 2,860 0 0 592,0179 68,687 62,103 63,232 114,597 12,638 0 0 0 321,25710 192,663 168,286 178,527 219,839 39,065 52,230 0 0 850,61011 191,067 172,670 163,514 341,743 184,173 723,028 456,614 392,538 2,625,34712 85,291 76,289 78,565 90,894 14,642 137,566 0 0 483,24713 123,421 107,950 145,480 138,355 1,631 2,418 0 0 519,25514 160,211 143,920 154,793 137,916 17,794 12,005 0 0 626,63915 113,103 110,111 119,138 120,399 32,354 250 0 0 495,35623 0 0 0 0 0 0 0 0 0

Total2,585,587 2,250,522 2,419,878 3,114,338 910,520 1,882,809 682,218 626,396 14,472,270

142

APPENDIX 10. Proportion of Each Study Year Data Were Allocated to CBC Status,by CBC Site

County ServiceContractDate

FY9596 FY9697 FY9798 FY9899 FY9900 FY0001 FY0102 FY0203

Sarasota 1/1/1997 0.0% 49.6% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%Manatee 5/1/1999 0.0% 0.0% 0.0% 16.7% 100.0% 100.0% 100.0% 100.0%Pinellas 6/1/2000 0.0% 0.0% 0.0% 0.0% 8.2% 100.0% 100.0% 100.0%Pasco 6/1/2000 0.0% 0.0% 0.0% 0.0% 8.2% 100.0% 100.0% 100.0%DeSoto 10/1/2001 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 74.8% 100.0%Hillsborough 5/1/2002 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 16.7% 100.0%Flagler 12/1/2001 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 58.1% 100.0%Volusia 12/1/2001 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 58.1% 100.0%Walton 12/21/2001 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 52.6% 100.0%Escambia 12/21/2001 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 52.6% 100.0%Santa Rosa 12/21/2001 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 52.6% 100.0%Okaloosa 12/21/2001 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 52.6% 100.0%