Introduction:

Congress and the Department have recognized in law and regulation that more information is essential for "better understanding of the foster care program and causes and other factors contributing to its expansion and other changes; and eventually, to make suggestions and proposals for change to improve the child welfare system." Congress requires that the Department collect this information from the States: the States ultimately rely on the caseworkers.

Caseworkers are currently hampered in their program efforts by unwieldy and burdensome systems and procedures. The process must be re-engineered to reduce administrative overhead, redesign inefficient paper-based processes, and eliminate manual interfaces. In place, caseworkers must redirect their attention to client support and to community outreach programs that will affect child welfare over the short and long term.

This systems effort is the beginning of the State's program to modernize its methods and leave legacy processes behind. Just as Congress and the Department recognize that fundamental reform is needed, so do we.

With this submission, the State requests approval and Federal participatory funding. As a summary of our justification, this systems project is projected to:

• Meet Federal mandates at the least cost alternative,
• Breakeven in nine months, and
• Achieve measurable benefits that reflect important program outcome improvements.

The State has evaluated the feasibility of and alternatives for modernizing the information technology and processing procedures supporting its child welfare programs. As detailed in the feasibility study, this Statewide Automated Child Welfare Information System (SACWIS) project has the following primary objectives as required by Federal regulations:

• Meet Adoption and Foster Care Analysis and Reporting System (AFCARS) reporting requirements,
• Interface with State child abuse and neglect information systems,
• Interface with State systems that determine IV-A eligibility and manage IV-D child support collections, and
• Support efficient, economical, and effective program administration.

This project also has program objectives to:

• Support families to prevent the unnecessary separation of children from parents by emphasizing prevention services,
• Speed the placement of at-risk children in foster care, and
• Reduce the time children spend in long-term care, by fostering return to parents or adoption.

During the alternatives analysis, the State selected (and justified the selection of) two alternatives for evaluation of costs and benefits in comparison to the status quo. Both alternatives are considered viable solutions, serving to distribute some degree of processing and to achieve the system objectives with equivalent quantitative benefits.

Although Alternative 2 is the more technologically complete solution, Alternative 1 is the State's selected approach for implementation, primarily because it is less costly and, consequently, will breakeven sooner. See the following breakeven chart and comparison of alternatives table.

We made this decision because our sensitivity analysis indicates that factors outside the control of the child welfare program could affect the ultimate realization of benefits. The less costly the solution, the more likely the system will prove cost-beneficial under post-implementation analysis. Even under the most negative assumptions, our projections indicate that this project will breakeven.

[The status quo (central data processing center and dumb terminals)is not a viable alternative, but is costed out as required by ACF instructions.]

Quantitative Benefits:

The status quo is not considered a viable alternative: no benefits Benefits: are evaluated.

Both alternatives are expected to generate the same specific dollar quantitative benefits:

• Improve the efficiency and effectiveness of caseworkers and reduce program costs,
• Use productivity gains to increase the foster home pool and decrease group and residential home placements,
• Reduce the duration of stay for children who can safely be returned home,
• Consolidate databases to reduce the interval until adoption,
• Increase child support collections, and
• Reduce AFDC overpayments.

Note that the first two benefits are based on cost changes from shifts in placement. Specifically, they project how productivity improvements will allow caseworkers to change the nature of their work, which should result in measurable program improvements and cost reductions. The next two benefits address changes in the duration of placement, first for children who can safely be returned home and second for children who can be adopted. The last two benefits are based on improvements resulting from systems interfaces with the child support and AFDC programs. Details on these benefits follow, beginning on page 2-15.

These benefits result in program cost avoidances and cost savings that offset the systems development cost, thereby achieving net benefits for the project. The project will breakeven. See page 2-14 for the cost/benefit profile of the selected alternative.

Related to these six benefits are several important measures. They are quantitative, but are measured by factors other than dollars. These outcomes are also part of the measurement plan. They are to:

• Reduce the amount of time caseworkers spend on data entry and administrative duties,
• Increase the amount of time caseworkers devote to prevention services, and
• Devote more time to prevention services, resulting in a shift in placement from more to less expensive care.

In addition, qualitative benefits are anticipated to accrue by:

• Providing more time for caseworkers' program evaluation, including cause and effect analysis and development of a correlation model with predictive risk factors,
• Reducing burden on and inconvenience to foster care providers, and
• Providing strategic support of agency program goals.

Although these qualitative benefits have value, at this time they cannot be measured in dollars for offsetting systems development costs.

The costs evaluated in this analysis are those that directly relate to the systems design, development, conversion, implementation, and operation. For the status quo, recurring costs include site and facility, equipment and software lease and maintenance, travel, training, supplies, security, and personnel salaries (including benefits) and support services directly supporting systems development and operation. The same categories are evaluated for the alternatives.

Nonrecurring costs for the status quo include a systems upgrade planned and budgeted for the third year of the systems life. Nonrecurring costs for the alternatives include costs for new site and facilities, equipment, system testing, conversion, studies, procurement, database preparation, and overhead. Details are provided in the requirements analysis and the cost/benefit analysis.

[As provided in ACF's cost/benefit guide, total project costs are analyzed regardless of funding source (State and Federal) and regardless of cost allowability for purposes of Federal Financial Participation, both of which are addressed by other documents.]

Costs. Actual costs will be measured against the selected alternative's projected costs by the finance office, subject to review and approval by the program office. Costs will be measured by category, but reported in the aggregate annually to ACF. Variances of over 10% will be explained by supporting documentation that addresses expenditures by category. The chart and table below depict the cumulative and annual baselines against which actual project costs will be measured.

[Editor's Note: As a reminder, the costs that States will measure against during implementation are the projected costs for the selected alternative from the cost/benefit analysis. Status quo costs are not used, present value discounted costs are not used, nor are measurement dollars discounted.]

Dollar-Quantifiable Benefits. The following chart and table depict the cumulative and annual baselines against which actual project benefits will be measured. Benefits will be measured in accordance with the measurement plan listed at the end of each narrative benefit description beginning on page 2-15.

Other Measurable Benefits. The State also plans to determine whether related measurable improvements are achieved. For example, the dollar-quantifiable benefits in Benefit 1 are based on reductions in caseworker administrative duties, an increase in caseworker prevention services, and a corresponding shift in the population from out-of-home to less expensive in-home care. We will use two measures to assess this outcome.

First, we plan to measure whether the projected decrease in the amount of time caseworkers must devote to data entry and administrative processing is realized. The current and projected measures (per caseworker) are in the performance table above. Second, we plan to measure the population distribution (in terms of percentages) to determine whether increasing caseworker prevention services had the projected effect on the population. The current and projected measures are shown in the population distribution table below.

Second, we plan to measure the population distribution (in terms of percentages) to determine whether increasing caseworker prevention services had the projected effect on the population. The current and projected measures are shown in the population distribution table below.

The dollar-quantified benefits in Benefit 2 are based on the development of a community outreach program that increases the foster home pool. We will separately evaluate whether the new program has the desired effect of adding fifty homes annually to the foster home pool. The value of this benefit is expressed by a shift in placement from the more expensive group and residential home care to foster home care. We have projected the effects over five years as indicated in the following chart.

Note that the analysis associated with these two benefits will not be a pure cause-effect analysis. We cannot isolate a portion of the foster care population to a laboratory environment and eliminate other external influences. These other factors, like

the availability of other support groups (church, extended family, school services) to the families and the size of the foster home pool, will affect the population and its distribution as well. Nevertheless, we will measure the actual population distribution and assess on a case-by-case basis the extent to which prevention services affected placement.

Note also that our sensitivity analysis (and reason) convince us that this benefit, although promising very high dollar returns, is the benefit most affected by factors outside of our control; it is, consequently, the most at risk of being unrealized. If drug usage escalates - especially drugs with side effects of violent behavior - no amount of family preservation services will protect the children.

Because of our uncertainty about achieving this benefit, we developed alternatives that would minimize expense and selected the least costly alternative to develop. Even if these benefits do not develop as projected, we still expect the system to breakeven.

In summary, this cost/benefit measurement plan provides that the State will measure system implementation against cost and benefit values - and against program performance goals. This information will serve as the baseline for reporting "actuals" in future APD Updates.

We thoroughly evaluated the performance of and described the systems life costs of the status quo in the feasibility study, alternatives analysis, and cost/benefit analysis.

During the alternatives analysis, we considered a broad range of alternatives. We addressed six alternatives, varying in terms of technology and source. Those alternatives included systems modification and transfer. The reasons for selection of the two alternatives for cost/benefit analysis are documented in the alternatives analysis.

We applied cost/benefit analysis to the status quo and two viable alternatives. We evaluated all on a systems life basis, using present value discounting at 7%. Constant dollars were used.

We consider the evaluation and documentation of costs and benefits to be thorough, detailed, and well documented. Back-up documentation and studies will be maintained in the State throughout the systems life of the project. The cost and benefit projections are well documented and provide a sound basis for cost/benefit measurement.

Net benefits (costs), benefit/cost ratios, and breakeven points were calculated for the two alternatives. We consider the selected alternative reasonable and fully capable of meeting our systems objectives.

We have set forth a clear set of projected costs and benefits against which actuals can be measured. We have also set forth qualitative measures, linked to program objectives, which can be measured.

A cost/benefit profile for the selected alternative and a narrative description of benefits (with benefit measurement plans) follow.

[Editor's Note: This section is based on the criteria set forth in ACF's "Feasibility, Alternatives, and Cost/Benefit Analysis Guide" on pages 1-5 and 1-6.]

The new system will dramatically change the mix in time spent on these functions. Specifically, it will reduce the need for these functions through capabilities such as centralized electronic files and automated notice and report generation. Automation of these functions will reduce caseworkers' data entry and administrative overhead 50%, to 15% of their time.

In place of these clerical duties, caseworkers will be able to devote more time to prevention services, reducing the percentage of children in out-of-home care (this benefit) and the duration of placement (Benefit 3).

Productivity Improvement. Caseworker workload distribution was documented using automated work measurement techniques and time and motion analysis conducted over two week intervals at four separate review periods during the last fiscal year. Management records and observation were used to verify that the performance of duties did not vary significantly from the norm during this time period.

Components of the administrative duties category include maintaining tickler files, performing work scheduling, manually tracking cases, and reporting to management. Current system inefficiencies require redundant data entry - a problem that will be corrected with the new system. The time distribution of data entry and administrative duties, by caseworker per week, is shown in the preceding table. Expected improvements are reflected in the column to the right.

(These productivity improvements will be monitored and measured under the State's Cost/Benefit Measurement Plan.)

Given the projected productivity improvement, management plans a redistribution of workload under the new system. With automated functions, caseworkers will be able to devote more time to client services (this benefit) and program analysis (a qualitative benefit). Once the new system is in place, caseworkers' performance plans will be revised to increase the standard for time devoted to family preservation services. The proposed work breakdown model follows.

The analysis and findings are documented in the State's study, Time Distribution of Caseworkers' Duties. A copy of this study will be retained in the State's files as an aid to future cost and benefit measurement.

Prevention Services. It costs less to support children in the home than in out-of-home care. The most recent statistics indicate that the 34% of the population in out-of-home care require over 60% of the costs.

The following table indicates the average daily distribution of children with costs per month each and total costs per month. The out-of-home care categories (foster homes, group homes, and residential homes) are all notably higher than in-home care (family preservation).

A review and reconstruction of case files indicates a positive correlation between the amount of time spent in direct contact with the family and the preservation of the child in the home. At each level of time invested, the ratio between in-home and out-of-home changes. See the chart on the next page.

To develop this data, we assembled a task force of caseworkers to review approximately 20% (chosen at random) of 1994's case records. Based on records, experience, and estimates, we correlated hours spent on cases with placement decisions. The data support what experience tells us; that is, positive support from caseworkers can affect outcomes.

Against this correlation model, we next analyzed the effect of investing additional time on each case - first one then two additional hours per case per month. The model projects a significant shift between the ratio of the in-home and out-of-home population. See the table below for a summary and the following page for details.

Recognizing that caseworker time is one of many factors that affect the ultimate placement of children, we used the more conservative model as a basis for projecting the effect (benefit) of increased caseworker time on the distribution of children under care.

* Percentage ratio maintained to current population in out-of-home care.
** Five years.

The detailed correlation table on the prior page shows the effect of shifts in the placement of children on costs. Given the State's commitment to preserve children in the home where safe and possible, increasing caseworkers' time in prevention activities creates a measurable goal.

Population and cost statistics and the results of the ad hoc case file study will be retained in the State's files as an aid to future cost and benefit measurement.

Recognizing that caseworker support is one of many factors that affect the ultimate placement of children, we have chosen to use the more conservative (one-hour) model.

As detailed on the previous pages, our correlation analysis predicts that investing one more hour per month per case in prevention activities such as family preservation should affect the placement ratio favorably. The estimated value of this benefit exceeds $35,000,000 (based on the current population) over the five year systems life.

The table below reflects current program costs based on the most recent complete year's data. It also shows projected new costs based on a change in ratio of placement as described previously. The difference is claimed as a benefit. Constant dollars are used throughout.

[Editor's Note: Please note especially that the value of this benefit is based on the value of decreased program costs of care, not caseworker time. Valuing the benefit based on caseworkers' time would be reasonable if payroll costs were reduced.]

Once the system is operational, the State will establish new standards for workload distribution for caseworkers, shifting the emphasis from administrative duties to family preservation. Detailed records will be maintained regarding hours invested and services delivered, so that we can measure the actual workload distribution achieved and verify the increase in time devoted to prevention activities. This is a measurable outcome and, as such, can be referred to as a quantitative benefit, measured in hours (rather than dollars).

Because factors (such as recessions, poor job markets, communicable disease, and teen pregnancy) outside of the control of the child welfare program affect the number of children under care, the State plans to measure this benefit on the basis of percentage and not absolute population numbers.

Our goal is to manage caseworkers' time and tasks so that we meet or exceed 73% of the population in safe, supported, at-home care. Using the status quo's 66% rate as a baseline, we will be able to evaluate the dynamic effects of the new system on program costs.

Note that these are management's confidential program evaluation goals. The State does not plan to require caseworkers to meet a quota nor will the State evaluate caseworkers on placement ratios. The effect of such a policy could place children at risk. Instead, this model serves as the basis for:

• Increasing the time caseworkers devote to family support, and
• Evaluating the effect and effectiveness of the change on the program.

There are two qualitative benefits related to this dollar-quantitative benefit.

The first is the qualitative benefit of the additional time that caseworkers will have for program evaluation. Congress and the Department have emphasized in law and regulation that more information is essential for "better understanding of the foster care program and the causes and other factors contributing to its expansion and other changes; and eventually, to make suggestions and proposals for change to improve the child welfare system." Congress requires that the Department collect this information from the States; the States ultimately rely on the caseworkers.

Under the new system, the State will commit additional resources to program evaluation. For example, we plan to refine and expand the correlation model by developing predictive risk factors. We also plan to analyze cause and effect in the child welfare system to support continuing program improvement. Although currently incalculable, such program evaluation could reap large benefits in the future.

The second qualitative benefit is the reduction of the burden on and inconvenience to clients. Workers will have client data more readily available and will therefore reduce the intake burden on clients.

"Time Distribution of Caseworkers' Duties"
(program office)

Correlation Analysis and Model on Time and Placement Data
(program office)

[Editor's Note: This is another example of developing a benefit resulting from productivity improvement by valuing the type of work done. In this case, caseworker productivity improvement is used to "free up" a caseworker for reassignment to a community outreach program. The projected effect is to increase the foster home pool, resulting in a shift in placement from more expensive means of care. Note that States could develop other benefits related to the foster home pool, such as improved provider retention supported by timely payments and issuance of medical assistance cards.]

Benefit 2:
Use Productivity Gains to Increase the Foster Home Pool and Decrease Group and Residential Home Placements

Currently, caseworkers spend 30% of their time interviewing clients and prospective foster care providers. (See graph below.)

These interviews are hampered by an unwieldy and inefficient paper-based system, adding unnecessary time to the interview process. By improving access to information, these interviews can be completed more expeditiously. The productivity improvement will enable reassignment of a caseworker to a new community outreach program with the goal of increasing the number of homes in the foster care program. With more homes available, the percentage of placements in group and residential homes will decrease.

Based on the requirements for the new system, we simulated the caseworker interview process to determine the effect of readily available on-line information.

Currently caseworkers spend an average of twelve hours a week conducting interviews, which average one hour and ten minutes each. With information readily available, our simulations indicate that interview time can be decreased to one hour (approximately a 15% reduction.)

(These productivity improvements will be monitored and measured as qualitative benefits under the State's Cost/Benefit Measurement Plan.)

In our headquarters office, we have 60 caseworkers, devoting 720 hours per week to interviews. With a 15% reduction, interview hours would drop 108 hours per week to 612 hours. This is more than enough time to select one caseworker for full-time assignment to community outreach services.

Our experience has been that for every two hundred homes contacted, we recruit one additional family as foster care providers. We estimate that a person assigned full-time to community outreach should be able to contact (primarily in groups) two hundred people per week. These would include groups such as churches, PTAs, and youth sports organizations. In addition, the caseworker could employ mass media (using television spots and news articles) to reach a larger audience. We conservatively estimate that fifty homes could be added to the foster home pool each year.

With more families in the foster care pool, we can reduce the percentage of the population in group and residential homes.

It costs less to support children in foster care than in group or residential homes. The table in Benefit 1 on page 2-18 indicates the average daily distribution of children with costs per month each and total costs per month. The costs for group homes and residential homes are notably higher than foster home care. If we can increase the pool of foster homes by fifty a year, we can reduce program costs.

To analyze the effect of a larger pool, we calculated system life costs, each year increasing the foster home pool while retaining the current ratio between group and residential home placement. Note several points. First, we increased the foster home pool beginning with year two to account for outreach start-up and placement. Second, for consistency throughout our analysis (and to avoid double-counting), we have based our projections on the correlation model (+one hour) developed in Benefit 1. Note also that we have held our population numbers constant for two reasons: growth of population is unpredictable and we plan to measure on the basis of percentage of population (as explained in Benefit 1).

The model projects that shifting the ratio of the out-of-home population will result in a sizable benefit. See the table on the next page for the effect on costs of shifts in the placement of children.

Population and cost statistics and the results of the ad hoc case file study will be retained in the State's files as an aid to future cost and benefit measurement. Other findings are documented in the State's white paper entitled Projected Effect of Increased Community Outreach Activities.

No major changes will take place in the duties assigned to caseworkers over the systems life.

As detailed in the previous pages, our analysis predicts that investing caseworker time in community outreach will affect the foster home pool, shifting the placement ratio in out-of-home care favorably. The estimated value of this benefit exceeds $14,000,000 over the five year systems life.

The table below reflects projected program costs based on the correlation model and shows projected decreases in costs based on a further change in ratio of placement as described previously. The difference is claimed as a benefit. Constant dollars are used throughout.

Once the system is operational, the State will measure caseworkers' interview time and monitor the effectiveness of the outreach program. If the program is successful, it will be expanded to multiple locations throughout the State, enabling further system benefits. Detailed records will be maintained regarding hours invested, groups, and increases in the foster home pool.

Because factors (such as recessions, poor job markets, communicable disease, and teen pregnancy) outside of the control of the child welfare program affect the number of children under care, the State plans to measure this benefit on the basis of percentage and not absolute population numbers.

Quantified Benefits Worksheet: Systems Life

[Editor's Note: The preceding two benefits were based on cost changes from shifts in placement. The next two benefits are based on cost changes from duration of placement.]

Benefit 3:
Reduce the Duration of Stay for Children
Who Can Safely be Returned Home

As indicated in Benefit 1, caseworkers currently spend 30% of their time entering data and performing routine administrative functions. The new system will dramatically reduce the need for these functions through capabilities such as centralized electronic files and automated notice and report generation. Automation of these functions will reduce caseworkers' data entry and administrative overhead 50%, to 15% of their time.

In place of these clerical duties, caseworkers will be able to devote more time to client services, reducing the percentage of children in out-of-home care (Benefit 1) and the duration of placement (this benefit).

If children can safely be returned home, costs will decrease because it costs less to support children in the home than in foster care. The most recent statistics indicate that family preservation services cost $451 per child, compared to foster care costs of $546 per child. Multiplied by the population, these costs run in the millions. See below.

There are many reasons that children are in foster care, the foremost of these being safety. However, for a small percentage of the foster care population (approximately 15%), placement is temporary, caused not by concerns of safety, but by conditions such as temporary lack of housing, lack of day care, or parental illness. Although such children should be returned to their homes quickly, delays caused by administrative processing backlogs and lack of caseworker time are all too common.

With the new system, caseworkers will have more time for client services and administrative processing will be more highly automated. By also changing intake procedures to track the temporary-needs children, the average duration of foster care placement for these children should drop from 120 days to 45 days, without additional risk to the children.

No major changes will take place in the duties assigned to caseworkers over the systems life. The number of temporary-needs children will remain relatively constant.

Typically, about 300 children in the foster home population are children in temporary care. Daily costs for supporting these children is about $18 in foster care and $15 in family preservation, a $3 difference per child. If the average duration of stay can be reduced from 120 days to 45 days, meaningful savings will accrue. [120 days less 45 days equals 75 days times $3 less cost per day equals $225 less cost per child: $225 times 300 children equals $67,500 (in constant dollars) per year times five years for a systems life cost difference of $337,500.]

Quantified Benefits Worksheet: Systems Life

[The following benefit is based on the comparative costs between children in foster care and adopted children. If the time until adoption can be reduced from ten months to two months by the system's consolidated database, measurable benefits will result, because it costs approximately 50% less in continued care costs for adopted children.]

Benefit 4:
Consolidate Databases to Reduce the Interval Until Adoption

Under the current system, there is no centralized database from which children in foster care who are eligible for adoption can be matched with adoptive parents. This would be corrected with the new system, enabling matching of children and parents, not just in the community, but within the State and among other States.

Last year, the State Department of Social Services completed a laborious study that involved a statewide sampling of records from the numerous listings (automated and non-automated) that are the State's records of children and potential adoptive parents. The findings were grim. Of the 2,000 children in foster care, approximately 10% (200) were eligible for adoption. Of the 200, approximately half (100) could be matched with potential adoptive parents. Despite the fact that these children could be placed with screened adoptive parents, the average time until placement was approximately ten months. This delay is attributable to the lack of centralized information and inefficient procedures and scheduling information. The study, "From Foster Care to Adoption: The Long Road," estimated that the ten month interval could be reduced to two months with centralized information.

nbsp;

Quantified Benefits Worksheet: Systems Life

[Editor's Note: This benefit is based on savings derived from automated interfaces. Note that savings from automated interfaces can only be claimed by one system. For example, savings in child support collections cannot be claimed for child welfare systems, if they have already been claimed for child support systems.]

Benefit 5:
Increase Child Support Collections

Based on a sampling of the 1994 foster care population of 3,110 (in foster homes, group homes, and residential homes), approximately 75% (2,333) are under court order for child support payments. However, under the current system, only 10% of the cases in the foster care population are referred to IV-D for follow-up and collection of child support payments.

The new system will feature an automated interface with the IV-D child support system. Through this mechanism, foster care cases will be automatically referred to IV-D for follow-up and collection of child support payments. These collections will offset foster care payments, reducing program costs considerably.

We estimate that 75% of the 1994 foster care population had a parent under court order for child support payments. All will be automatically referred under the new system. Using the current collection rate (15%) and average monthly collection ($250) achieved by the IV-A program's automated interface, we should collect annual offsetting costs of well over a half million dollars. The table on the following page details estimated collections for the current and projected populations (see Benefit 1). The difference is claimed as a benefit for the new system. Constant dollars are used.

Quantified Benefits Worksheet: Systems Life

[Editor's Note: This benefit is based on savings derived from automated interfaces.]

Benefit 6:
Reduce AFDC Overpayments

Based on a State inspector general review, 400 children who entered foster care in 1994 had come from homes that had received Aid to Families with Dependent Children (AFDC). Yet when the children entered foster care, AFDC payments to their families continued on average about five months. These payments were rarely recouped.

The new system will feature an automated interface with the State AFDC system. Through this mechanism, foster care admissions records will automatically update AFDC eligibility records, cancelling AFDC overpayments.

The automated interface will reduce AFDC overpayments to families whose children have entered foster care.

The number of children entering foster care whose families receive AFDC will remain relatively constant.

With an automated interface, AFDC overpayments to families whose children have entered foster care will be reduced from an average of five months to one month. Based on an average monthly payment of $146, over $200,000 in overpayments will be saved each year. [$146 times four months fewer payments equals $584 per child, times 400 children equals $233,600 a year in constant dollars]

Quantified Benefits Worksheet: Systems Life