Analysis of the 2008-09 Budget Bill: Health and Social Services

Child Welfare Services (5180)

California’s state–supervised, county–administered Child Welfare Services (CWS) program provides services to abused and neglected children, children in foster care, and their families. The CWS program provides (1) immediate social worker response to allegations of child abuse and neglect; (2) ongoing services to children and their families who have been identified as victims, or potential victims, of abuse and neglect; and (3) services to children in foster care who have been temporarily or permanently removed from their family because of abuse or neglect.

In 2008–09, the Governor’s budget provides a separate CWS General Fund appropriation (Item 5180–153–0001) for the two counties (Los Angeles and Alameda) participating in the Title IV–E Child Welfare Waiver Demonstration Capped Allocation Project. The remaining 56 counties are budgeted in Item 5180–151–0001. Including the waiver counties, the Governor’s budget proposes $2.5 billion from all funds and $695 million from the General Fund for the child welfare system. This represents a decrease of 3.5 percent in total funds and a decrease of 7.4 percent in General Fund from the most recent estimates of current–year expenditures. This decrease in funding primarily results from the Governor’s budget–balancing reduction proposal to reduce CWS allocations (excluding automation, Adoptions, and Child Abuse Prevention) to counties by 11.4 percent.

Budget Proposes Reduction in CWS Allocations to Counties

The Governor’s budget proposes to reduce the total General Fund allocation to counties for Child Welfare Services (CWS) by $83.7 million. Counties will have the discretion to apportion their reduced allocation among various program components. We describe the potential impact of this proposed reduction on social worker caseloads and possible subsequent policy consequences resulting from fewer resources. We also provide three alternatives to the Governor’s proposal that more narrowly target reductions in CWS expenditures.

Background

There has been an ongoing effort in CWS to determine how many child welfare cases a social worker can carry and still effectively do his or her job. In 1984, the Department of Social Services (DSS) and the County Welfare Directors Association (CWDA) established an agreed–upon level of cases for each program component of CWS. These 1984 workload standards are still used by DSS to calculate the base level of funding for each county. In 2000, however, the Child Welfare Services Workload Study, which was required by Chapter 785, Statutes of 1998 (SB 2030, Costa), determined that the 1984 caseload standards were too high and that social workers had too many cases to effectively ensure the safety and well–being of the children for which they were responsible. The SB 2030 Study, as it is commonly called, proposed revised minimum and optimum caseload standards for social workers. Figure 1 compares the 1984 standards to the minimal and optimal standards developed in the SB 2030 Study.

Concerned about large social worker caseloads, over the years the Legislature has added additional funds known as the “augmentation” and the Outcome Improvement Project (OIP). The Governor’s workload budget proposes $152.7 million ($96.4 million General Fund) for these funding streams in 2008–09. These monies, in combination with the hold harmless budgeting methodology (which we discuss below), have enabled counties to hire more caseworkers and move toward standards established by the SB 2030 Study.

Governor’s Proposal

The Governor’s budget proposes to reduce CWS expenditures by $83.7 million General Fund. This represents a reduction of 11.4 percent to the total General Fund allocation for CWS, excluding funds for the Child Welfare Services Case Management System (CWS/CMS), the Adoptions Program, and the Child Abuse Prevention Program. Counties will have the flexibility to choose how to apportion the reduction to various CWS program expenditures. According to DSS, the department will work with CWDA to develop an allocation process for apportioning this proposed reduction. At the time this analysis was prepared, DSS could not provide further details on the implementation of the CWS reduction to county allocations and the potential program impacts.

Staffing Level Impacts of Proposed Reduction to CWS

The impact of the proposed reduction is difficult to measure because counties have multiple ways of responding to reduced funding. County options include reducing payments to service providers for preventive services, reducing transitional services for emancipated foster youth, reducing overhead expenses, and/or hiring fewer social workers. Nevertheless, because social workers and their support costs represent the majority of the CWS budget, counties are likely to substantially reduce the number of social workers.

Increase in Social Worker Caseloads. One potential program impact of the proposed reduction is an increase in county social worker caseloads because of a decrease in the number of funded full–time equivalent (FTE) social workers. The proposed reduction represents approximately 87 percent of the CWS augmentation and OIP monies. As a result, there may be a reversal of some of the progress made by counties in meeting or exceeding SB 2030 minimum standards.

In order to estimate existing staffing levels and the potential impact of the proposed reduction, we used the most recent caseload and budget data available from DSS and made a series of assumptions and adjustments related to non–case carrying social workers, the amount of OIP augmentation funds directed to hiring more social workers, and inflationary adjustments known as the cost–of–doing–business.

As Figure 2 shows, we estimate that in 2007–08, 20 counties, which have 9 percent of the total CWS caseload, are funded for enough FTE social workers to either exceed the SB 2030 minimum standards, or be within 10 percent of the standards. Additionally, 14 counties, which have approximately 43 percent of CWS cases, are between 80 and 89 percent of meeting the minimum standards.

As a result of the proposed reduction, we estimate an increase in the number of counties that are further away from meeting the minimum standards in the budget year. For example, we estimate that the number of counties that would be between 80 and 89 percent of meeting the minimum standards would decrease from 14 counties in 2007–08 to 6 counties in 2008–09. In addition, the number of counties below 80 percent of the standard would increase from 24 (representing 48 percent of the CWS caseload) to 38 (representing 90 percent of the CWS caseload) in the budget year.

From a statewide perspective, we estimate that the proposed reduction would result in an overall decrease of 522 FTE social workers. As a result, while the total number of funded FTE social workers in the state is at approximately 79 percent of meeting the minimum standards for 2007–08, for 2008–09 that figure would decline to 73 percent.

Potential Consequences of Fewer Resources

While counties will take different approaches to responding to reduced funding, there are several potential policy consequences from their actions:

• Counties that choose to reduce the number of social workers may decide to open fewer CWS cases or close cases earlier than they would otherwise because of limited resources. This could lead to leaving children in more marginally risky situations.
• Counties that choose to reduce spending on preventive services could see an increase in foster care cases. Rather than provide intensive and time–consuming family case–management services to cases in which the child remains in the home, counties with fewer social workers and limited resources may choose to change their policy to removing children from the home more frequently and placing them in foster care.
• Counties that choose to reduce spending in transitional services for emancipated foster youth could see an increase in unstable housing situations for this population.

Alternatives to the Governor’s Proposal

Below we present three alternatives to the Governor’s proposal which offer less budgetary savings, but are less likely to negatively impact services for children.

Suspend Hold Harmless. In preparing the budget for CWS, DSS adjusts proposed funding upwards when the caseload increases, but does not adjust funding downward when the caseload actually decreases. The practice of not adjusting the budget to reflect caseload decline is known as the “hold harmless” approach, though DSS technically refers to this as the “base funding adjustment.” Because of the way the hold harmless provision works, the number of social workers funded for the counties remains unchanged despite workload decreases. In other words, if an individual county’s caseload is declining, its number of caseworkers is held at the prior–year level. At the same time, if another county’s caseload is increasing, the state provides that county with funds to hire additional caseworkers. Therefore, on a statewide basis, despite an overall caseload decline, the funding for CWS continues to grow.

One alternative to the Governor’s proposal is to suspend the hold harmless budgeting methodology for 2008–09. For 2008–09, DSS reviewed estimated caseloads per CWS component and included $17.6 million ($6 million General Fund) in the budget for 29 counties with declining caseloads, pursuant to the hold harmless funding provision.

Under this option, the CWS case–management funding per child would remain at its 2007–08 level for these 29 counties. This would result in a General Fund savings of $6 million, while not reducing the level of care and service provided to the children and families in the child welfare system in the budget year. While the Governor’s proposed reduction would impact every county, suspending hold harmless would target CWS expenditure reductions to those counties with declining caseloads and would not reduce existing social worker caseload ratios.

Cap Social Worker Costs. Another option is to cap the total cost per social worker at $155,000, which would result in a General Fund savings of approximately $5.1 million. The average statewide “fully loaded” cost of a social worker, which is currently frozen at the level of funding provided in 2001–02, is $129,074. The fully loaded cost represents the social worker’s salary and benefits, in addition to the allocated cost of supervisors, data processing, departmental overhead, and other general expenses related to providing services.

The fully loaded social worker cost per county ranges from $72,788 to $176,930. This range in cost per county partially reflects cost–of–living differences, but there are also significant differences in costs between bordering counties. For example, while Sacramento County’s fully loaded social worker cost is $162,866, Yolo County’s cost is $101,468. Therefore, in some cases, the fully loaded funding for social workers in counties with similar cost–of–living rates are substantially different.

By capping the total cost per social worker at $155,000, which is the 2001–02 average statewide fully–loaded cost of a social worker adjusted for the California Consumer Price Index since that time, seven counties would experience a reduction in funding because their fully loaded social worker cost exceeds the proposed cap. Capping social worker costs is another alternative that targets a reduction in CWS expenditures to specific counties that have larger funding allocations per case, rather than an across–the–board reduction for all counties.

A Combined Approach. The Legislature could also choose a combination of a smaller across–the–board reduction to CWS county allocations, in conjunction with the hold harmless and social worker cost cap alternatives discussed above. For example, a 3 percent reduction to CWS county allocations, in combination with suspending the hold harmless provision and capping the fully loaded social worker cost at $155,000, results in an estimated General Fund savings of $33.1 million.

Conclusion

The Governor’s proposal to reduce CWS allocations to counties by 11.4 percent results in General Fund savings of $83.7 million. In deciding whether to adopt this proposal, the Legislature should weigh the budgetary savings against the potential for increased social worker caseloads as a result of fewer FTE social workers, as well as possible subsequent policy consequences resulting from fewer resources in CWS. Although the specific alternatives to reduce CWS expenditures that are outlined above save considerably less than the Governor’s proposal, these options set priorities and target the reductions which would lessen their statewide impact.

Rethinking the Future of CWS Automation

The Governor’s budget proposes to spend $247 million ($112 million General Fund) over the next seven years to continue with the development of a new Child Welfare computer system (referred to as the New System). Our review indicates that the current Child Welfare Services Case Management System (CWS/CMS) can be updated to meet federal and county functionality requirements. Accordingly, we recommend cancelling the New System project and updating the CWS/CMS, resulting in savings (all funds) of $184 million over the next seven years.

Current System

The CWS/CMS is a statewide computer system deployed in all 58 counties to support the administration of CWS. From 1992 until 1995, state and county staff participated with the vendor to develop system requirements and design. Statewide system implementation began in 1995, and by 1997 the CWS/CMS was in use in all 58 counties.

Federal Statewide Automated Child Welfare Information System (SACWIS)

Federal Funding. In 1993, the federal government offered “incentive funding” to states that would develop a SACWIS that met federal requirements. These systems would receive 75 percent federal funding for the first three years of system development and 50 percent thereafter. California received the 75 percent funding through 1997 when it implemented CWS/CMS and has received 50 percent federal funding since that time.

SACWIS Compliance. In 1999, a federal review raised concerns about the extent to which CWS/CMS complied with the requirements of SACWIS. In 2003, the federal government notified the state that CWS/CMS did not meet all SACWIS functional requirements. The missing functions included Adoptions case management, Foster Care eligibility, financial management, and automated interfaces to the Child Support and human services systems. In 2004, the state submitted a plan (referred to as the Go Forward Plan) to the Department of Finance (DOF) and the federal government for achieving SACWIS compliance and for meeting additional county business requirements. The counties had two business requirements beyond the SACWIS requirements: (1) a simplified data entry process and (2) the ability to access CWS/CMS from locations other than their office (remote access). The plan proposed to conduct a study to determine the technical viability of the current system to provide the additional functionality and a technical analysis of alternatives. The federal government approved the plan.

Technical Architecture Analysis Alternatives (TAAA)

In 2005, the state Office of Systems Integration (OSI) hired Eclipse Solutions and Gartner Group to conduct a technical analysis that would provide alternatives for meeting the following requirements:

• Achieve SACWIS compliance.
• Meet county requirements for simplified data entry and remote access.

In addition to these requirements, OSI instructed the consultants to propose solutions for making the system accessible from the web by abandoning the existing mainframe platform and moving it onto servers.

TAAA Report Did Not Consider All Possible Alternatives

State Instructions Constrained Analysis. The consultants conducted their analysis as they were instructed by OSI. The instruction that the system should be moved off the large, mainframe computer and onto servers represented a major constraint on the consultants’ analysis. It prevented them from considering all possible technical solutions for achieving SACWIS and county requirements.

Only Two Alternatives Were Examined. Because of the constraint placed on the consultants, only two alternatives were examined.

• The first alternative would move the current system, a piece at a time, off the mainframe and onto web servers. In the process of moving the system, software changes would be incorporated to meet the county requirements and the missing SACWIS components would also be added. This alternative would take eight years to accomplish.
• The second alternative was to develop a new system. This alternative would build in all the federal and county requirements. The new system would take three years to develop.

Third Alternative Was Not Considered. A third alternative was not considered by TAAA consultants because the state had specified that it wanted to eliminate use of the mainframe. This alternative would update the current system and leave it on the mainframe. In fact, a 2003 study also conducted by Gartner Group recommended this as a solution for making CWS/CMS accessible from the web in order to provide counties with a simplified data entry process and remote access.

Decision to Procure New System

Of the two alternatives provided in the 2005 TAAA Report, the state chose to develop a new system. A feasibility report was approved by DOF in April 2006. Since that time, DSS, OSI, and the counties have been working to document the detailed business requirements for a vendor bid to build a new Child Welfare system. The proposed technological solution is currently referred to as the “New System.”

Proposed New System Adds Risk and Cost. When replacement systems are built, the data from the old system must be moved to the new system. This is referred to as “data conversion.” In order to convert data, programmers must write software programs to locate and move the data from the old database to the new database. Data conversion efforts can be complex, time–consuming, expensive, and high risk. The high risk is attributable to the possibility that data can be accidentally altered or even lost during the conversion process. Both the alternatives considered by the TAAA require this costly and risky data conversion process. In order to avoid these cost and risk factors, many companies are choosing to retain their legacy database and modernize their systems by adding a software layer that allows the system to be accessed from the web. This software layer is referred to as an “enterprise service bus.” Adding an enterprise service bus enables application changes that can provide remote access and simplify data entry.

LAO Alternative

Update Current System. The CWS/CMS is built on software products currently under vendor support. That is, the vendors continue to maintain, upgrade, and market the software. Therefore, there is no reason to abandon CWS/CMS if it can play a role in meeting the additional SACWIS and county requirements. County requirements not met by the current system can be accommodated by making the system more modular and accessible from the web. This can be accomplished by adding an enterprise service bus as described above. This approach is increasingly being used by organizations to leverage their existing databases in order to minimize both the risk of data conversion and the cost of building a new system. Thus, the LAO alternative is to (1) update the current system and (2) add the missing SACWIS components. This will meet the federal and county business requirements.

Budget and Contract Availability. The CWS/CMS has been in use for more than ten years. There is $10 million in the baseline budget to keep the system current for changes in regulations and legislation. During the first five years that CWS/CMS was in operation, this baseline amount was being spent, most of it to adjust the system for changing business processes as social workers transitioned from a manual operation to an automated one. Over the past five years, approximately one–third has been spent of the $50 million budgeted. This reduced spending pattern is typical for new systems as they stabilize and attain user acceptance. The current vendor contract is effective through 2013 and allows up to $10 million annually for system changes. We estimate that $8 million could be made available each year from the existing baseline budget to update the system to make it accessible from the web and to add the missing SACWIS components. The remaining $2 million would be available to incorporate any regulatory and legislative changes.

Comparing New System to LAO Alternative

Figure 3 shows the total project cost for the New System and the LAO alternative. As the figure shows, the new system is estimated to cost $247 million (all funds), $184 more than the LAO alternative.

Cost of New System Was Understated. Over the past two years the state has spent $7 million for New System project planning. In November 2007, the administration estimated that it would take seven more years to procure a vendor and complete the system at a cost of $247 million. During the final three years of New System development, after the contract has been awarded, there will be a reduction in federal funding for the current system.

LAO Alternative Reduces Schedule, Cost, and Risk. As shown in Figure 3 above, the total cost of the LAO alternative is $63 million. The current contract provides adequate resources to perform the work necessary to update the current system to meet SACWIS and county requirements. Although there are separate costs for state and county staff to design and test the system, such costs are significantly less than they would be for the New System. This alternative also eliminates the risk and cost of data conversion, which is necessary under the other alternatives. In addition, federal funding levels for the current system will be retained if it is updated to meet SACWIS and county requirements.

Funding the LAO Alternative. The LAO alternative could be funded by applying $8 million of the existing CWS/CMS baseline budget to cover the system programming. In addition, the increased state and county staff needed to help design and test the system changes could be covered by redirecting funding from the New System for 2008–09 ($6.8 million) and 2009–10 ($8.2 million). Thus, through these redirections, there would be no net new cost under the LAO alternative for these years.

Analyst’s Recommendation

We recommend canceling the Child Welfare New System Project and updating the current system. This will result in reduced time, cost, and risk. This proposal is budget neutral in 2008–09 and 2009–10. Over the life of the project, total savings would be $184 million (all funds).