Companion Guide 2 Cost/Benefit Analysis Illustrated for Child Welfare Systems
Chapter 1: Introduction

Introduction and Purpose

The Office of State Systems (OSS) of the Administration for Children and Families (ACF) has a continuing interest in helping States timprove the quality and usefulness of the plans and studies that support their development of public benefit information systems. Tthis end, OSS has published the Feasibility, Alternatives, and Cost/Benefit Analysis Guide and the Companion Guide: Cost/Benefit Analysis Illustrated (for generic public benefit systems); has sponsored State systems planning working groups; has developed a set of model spreadsheet templates for cost/benefit analysis; and has prepared cost/benefit training materials.

Taugment these efforts, and especially trespond trequests from State personnel, OSS has developed this Companion Guide 2: Cost/Benefit Analysis Illustrated for Child Welfare Systems. This optional guidance responds tthe States' requests for more program-specific guidance and provides (as does the original Companion Guide):

• Examples of sound cost/benefit studies,
• Clarification (by example) of what is required by law and regulation tbe submitted tACF, and
• Examples on reporting actuals against a baseline.

This Companion Guide 2 is, in a sense, three documents in one.

Chapter 1: Introduction and Purpose. This introductory chapter provides general information tsupplement the information presented in the Feasibility, Alternatives, and Cost/Benefit Analysis Guide and the Companion Guide: Cost/Benefit Analysis Illustrated.

Chapter 2: APD Documentation. This chapter provides an example of the part of the Implementation APD which addresses cost/benefit analysis. This example illustrates the summary or key information that ACF considers important. Among the most important factors are:

• Detailed descriptions of benefits, and
• Clear establishment of a baseline for later cost/benefit measurement and reporting.

This section in nway implies a standard, approach, or format that States must use. It does illustrate a sufficient level of detail for ACF's purposes since this section (and the other chapters) underwent review in ACF's program offices.

Chapter 3: APDU Documentation. This chapter is an example of a cost/benefit measurement report. It is written as though reporting in the second year of the project described in Chapter 2. This clarifies the relationship between the planning stage studies and the post-implementation measurement and reporting phase.

This Companion Guide 2 is a supplement, not a replacement, for the prior guides. The Feasibility, Alternatives, and Cost/Benefit Analysis Guide remains the definitive ACF reference on the subject of cost/benefit analysis tsupport State public benefit information systems advanced planning. The Companion Guide: Cost/Benefit Analysis Illustrated, which provides a generic example of a cost/benefit analysis, alsaddresses material not duplicated in this guide, such as definition and clarification of terms, the importance of consistency, use of more sophisticated techniques, level of effort, compilation of data, development of benefits in general, and sensitivity analysis. The Companion Guide alsillustrates the difference between State cost/benefit documentation and that submitted tACF.

Child Welfare Benefits in Perspective

Congress requires by law that the Federal government and States together support children and families in need. Government, therefore, assumes the obligation of financial and program support for those in need. Unfortunately, those needs have risen due tfactors such as the economic recession, teen pregnancy, substance abuse, and the AIDS epidemic.

Some benefits typically used in cost/benefit analyses are simply not reasonable for child welfare systems. What if an effort were made treduce costs by trying tkeep children out of foster care and in homes? What if an effort were made treduce casework by decreasing the time spent with those in need? What if an effort were made treduce cases by failing tidentify or keeping children or families out of support? The unfortunate result of all these "benefits" would be an increase in injuries tor deaths of children - incidences our society considers intolerable. Sthe benefits that an effective, economical, and efficient child welfare system would deliver might be a system that:

• Controls costs by supporting children in their homes, where possible - and where not, by placing them as soon as possible in permanent care,
• Reduces administrative casework, allowing caseworkers more time for contact with clients, analysis, and decision-making, and
• Reduces cases in the long-term by identifying solutions that work.

Congress and the Department recognized in law and regulation that improvement is needed. The Department suggested that comprehensive child welfare information systems will provide the "data needed by policy makers and, more importantly . . . the tools [needed] tassist caseworkers and managers in making decisions and providing comprehensive support tfamilies in need." In summary, the Department added that these information systems "will result in more efficient and effective practices in administering child welfare programs which in turn will ultimately result in improved service delivery." The emphasis is not solely on cost reduction but on program improvement.

This concept is critical: Cost/benefit analysis must prove that the projected benefits are sufficient twarrant the expenditure for the system project. This implies that the justification will be based on measurable benefits and that the outlay for those benefits is reasonable. It is a two-step process that answers twquestions:

• What am I buying in terms of outcomes? 1
• Is the cost of achieving those outcomes reasonable?

The most common and straightforward approach tjustifying an acquisition is tproject that the dollar value of the benefits for the proposed acquisition will exceed the costs. In other words, the system will breakeven. Such justifications can normally be approved at face value, slong as the stated benefits and costs appear reasonable.

What if, however, projected benefits dnot exceed costs? This situation is much more difficult tjustify and approve, at both the Federal and State levels.

When a system will breakeven, the measurable, dollar-quantified benefits are projected texceed the costs. This is an easy approval. However, if the system will not breakeven, a State must prove that the measurable outcomes are worth the cost of investment - and that those costs are reasonable. The onus is on the States.

Even though the standard of review is the same - cost/benefit analysis must prove that the projected benefits are sufficient twarrant the expenditure for the system project - the justification and the decision are much more difficult.

Consider, for example, the following hypothetical situation. A State submits an Implementation APD that will not breakeven, will cost nine million dollars, projects nsignificant offsetting dollar-quantified benefits, will support a foster care population of 200, and will achieve (as its most significant outcome) report processing in twdays rather than five days. Clearly, this submission fails on twcounts. Its measurable outcome is not meaningful and its costs are not reasonable.

In contrast, consider the following hypothetical situation. A State submits an Implementation APD that:

• Meets, but does not exceed, mandatory Federal requirements for child
• welfare systems,
• Will not breakeven but will recover ninety percent of costs through
• measurable benefits,
• Will cost one million dollars ($100,000 in unrecovered costs),
• Is established as the State's least costly alternative,
• Will reduce time tintervention from five days tone day, and
• Based on prior year statistics, will reduce the incidences of death or injury tchildren during the elapsed time from notice tintervention by 200 cases.

In this case, $100,000 buys a system that meets Federal mandates at the lowest cost and prevents the death or injury of 200 children annually. Does this submission clearly indicate what is being bought and does it meet the requirements of meaningful, reasonable, and measurable outcomes? Yes.

What if projected benefits will not exceed costs? It is clearly preferable that systems be justified as cost effective. However, ACF will not automatically disapprove Implementation APDs with cost/benefit analyses that fail tbreakeven. ACF will consider other justifications, especially for these types of systems in which child welfare program improvements are both clear and well supported in the State's analysis.

The onus is on the States tpresent a compelling case that establishes that the cost of the investment is worth the project outcomes. In addition, because States must report actual benefits, the onus is alson States timplement systems that achieve the projected outcomes.

Developing Benefits, in General

Cost/benefit analysis in the private sector is normally concerned with determining whether expenditures will result in increased income. The effect on the bottom line is the primary concern. What net profit will result?

Only a handful of federal programs can conduct cost/benefit analyses as "profit" decisions - notably, the IRS and the Child Support Enforcement Program. These organizations generate collections (income) that offset (in a sense) the costs of the government's programs tcollect or provide welfare support. In these systems, improved and integrated information systems can increase collections, resulting in a net gain for the government.

However, most public sector cost/benefit analysis is concerned with net program effect. The government does not charge for its services: public services or benefits are required by law tbe provided. Sthe government's obligation is not tmaximize profit, but tmake cost-effective expenditures and tdeliver maximum benefits within the budget.

Therefore, most public sector cost/benefit analysis does not seek tincrease program funds - but tchange the distribution of costs enough tsupport system development within the overall budgetary limitations of the organization. By this means, public agencies prove projects tbe cost-effective.

ACF views cost/benefit analysis as serving four fundamental and equally important needs - to:

• Evaluate alternative mixes of financial, human, and information resources,
• Support wise economic decisions on proposed information system investments,
• Establish a performance baseline against which tmeasure the success of the systems project, and
• Provide fundamental management tools tmaximize benefits and minimize costs.

• Can enough be "saved" in other categories t"pay for" the costs of developing the new system, and
• Will the system project result in measurable improvements over current operations?

Public sector cost/benefit analysis is not an accounting process. When benefits equal costs, the analysis has not proven the system will cost nothing. It has proven that the organization will remain within the overall, projected program budget - and that the projected benefits are sufficient twarrant the expenditure for the system project.

Using a Structured Approach tIdentify Benefits: Several Examples

When first faced with the task of identifying benefits, many are overwhelmed. It may seem impossible tidentify and quantify the millions of dollars of benefits needed toffset the costs of developing and acquiring a new information system. What is needed is a structured approach or a framework within which tanalyze the effect of the systems project.

The Feasibility, Alternatives, and Cost/Benefit Analysis Guide provides a number of examples of quantitative and qualitative benefits, categorized as cost/resource, functional/programmatic, technical (system), legislative, and socio-political. (See page 1-8 in this Guide.) This is one framework an analyst might use. Following such an approach, States have developed benefits for child welfare systems, such as, to:

• Increase amount of child support collected (toffset child welfare program costs),
• Reduce provider payment errors,
• Reduce AFDC overpayments tfamilies with children in foster care,
• Replace current payment system,
• Reduce pre-printed forms costs,
• Eliminate existing county-based management systems,
• Eliminate data center costs, and
• Increase staff retention (and reduce staff training).

Another framework was discussed at the July 1995 SACWIS Conference. That framework focused primarily on program and system benefits, such as might result from the development of a child welfare system. Specifically, the analytical approach suggested was texamine the benefits of a proposed information system from the perspective of five functional program areas:

• Intake and assessment,
• Family services,
• Custody cases,
• Resource development and management, and
• Management and administration.

Intake and assessment describe the process of evaluating each report of a child in need of care tdetermine whether the situation represents an emergency, tassess the degree of potential risk tthe child, and testablish appropriate agency responses.

Family services are provided tfamilies that are at risk of losing children tstate custody, but that currently dnot have children in custody.

EXAMPLES OF BENEFITS BY TYPE

Custody cases refer tfamilies that have at least one child in custody (as a child in need of care, a juvenile offender, or both). Custody generally refers tfoster homes, group homes, or institutions.

Resource development and management encompass the activities necessary testablish and maintain relationships with providers of the facilities and services that the child welfare agency utilizes in the performance of its mission.

Management and administration include the general functions of office administration, communications with internal and external organizations, and reporting.

Using this framework, the analyst would assess these functional areas, their domains (affected populations), and the effects or outcomes of the project.

Examples of effects or outcomes (benefits) identified using such a programmatic framework are shown in the table on page I-10 and described below:

• Improved client services, through the provision of rapid (electronic) access tcomprehensive and up-to-date information.
• Improved staff utilization, by reducing or eliminating redundant and labor-intensive processes.
• Reduced costs, for duplication, mailing, and file storage space.
• Fewer manual functions, specifically related ttracking review dates, court dates, and legislatively-mandated schedules or timeframes.
• Less data redundancy, by eliminating duplicate data entry, and by features tautomate completion of key sections of forms and documents.
• Less processing time, by automatic generation of basic assessment information and automation of comprehensive records-searching.
• Improved access tinformation, such as by electronic data transfer or electronic mail.
• Increased automation of administrative functions, such as payment processing for services.

EXAMPLES OF OUTCOMES OR BENEFITS BY PROGRAM FUNCTION

• Fiscal improvements, such as lower error rates and increased collections of overpayments.
• Increased and better-used resources, by electronic access tcurrent information on placement and service providers, improved matching of resources tneeds, and better information for planning.
• Improved management information through automation, particularly in areas of budget planning, management reporting, and workload scheduling.

By considering the effects or outcomes of the project, benefits can be identified, both quantitative and qualitative.

For example, an analyst might consider the effect of the system on the intake and assessment function, concluding that electronic access tcomprehensive and current information would result in improved client services. (The domain is the client population and the effect is improved services.) The benefit tbe derived from this effect or outcome of the system project is a saving in staff time that could be used treduce current staffing levels or be reinvested intdirect services tchildren and families. Depending on the ultimate result, the benefit would be quantified based on the reduction in staff salary or the value of reinvesting the hours intdirect services.

Notice that the analysis requires a number of steps within the framework:

1. 1. Identify functional areas.
2. 2. Analyze effects or outcomes of the system project on functional areas.
3. 3. Analyze effects or outcomes of the system project on affected populations.
4. 4. Determine the benefit of the effects or outcomes.
5. 5. Decide whether the benefit is qualitative or quantitative (by dollars or other measures).
6. 6. Decide how tvalue or measure the benefit.

Using this analytical framework, a State analyst might identify anticipated benefits (such as those on the following page) for further evaluation.

EXAMPLES OF BENEFITS DEVELOPED BY ASSESSING EFFECT OF SYSTEM ON FUNCTIONAL AREAS

Although the initial list developed by the analyst might be quite lengthy, the State would select only the most critical, in terms of program or dollar impact, tdevelop in the cost/benefit analysis.

EXAMPLES OF BENEFITS BY OUTCOME DOMAINS

That Guide alsmakes clear that qualitative benefits can have cost implications, but may be difficult or impossible tquantify. Examples might include enhanced compatibility between State benefits systems, improved public assistance, improved management information, and improved security.

Quantitative benefits, however, are at the heart of the cost/benefit analysis. Quantitative benefits may be quantified on the basis of dollars or by other measures, such as time, percentages, number of foster homes, caseloads, service delivery, placements, adoption rates, and sforth.

Although dollar-quantified benefits are the most typical and are essential tcost/benefit analysis, benefits quantified by other measures can be extremely important, especially tprograms such as child welfare. Consider, for example, the twbenefits listed under safety in the table on the preceding page: reductions in response time from notice taction after receiving a child abuse or neglect report. This benefit can be easily quantified and measured by time. The State should know its current performance and could set a goal for improvement under the new system. This benefit can not be easily quantified and measured by dollars. How can the cost of injury or death ta child be measured? There may be a long-term loss tsociety, but one less child could mean less expense tthe child welfare system. Should States even try tplace a value on a child's life? No.

As indicated previously, Congress and the Department place an emphasis, not on cost reduction in child welfare programs, but on program improvement. In addition, current ACF policy requires developing benefits that accrue tthe Federal government, not society as a whole.

Furthermore, the importance of the cost/benefit analysis is not only tprove that a course of action is cost-beneficial, but alstestablish a baseline for performance measurement. What could be more important than measures linked tthe safety of children? Accordingly, ACF recommends that States develop cost/benefit analyses that include both qualitative and quantitative measures - and that quantitative measures not be limited tthose valued in dollars.

Applying Values or Measures tBenefits

Once benefits have been identified, the State faces the difficult task of assigning values or measures tthe benefits. Perhaps the easiest way tdetermine where savings or improvements can be achieved is ttake a close look at the budget and management reports - and tvisit the accounting department.

Generally speaking, benefits may be derived from both the systems or program area.3 Examples of systems-related quantitative benefits include future cost savings by avoiding such expenses as scheduled equipment upgrades, charge-back expenses for central data processing staffs, contractor support fees, and telecommunications fees. Examples of benefits derived from more current technology might be avoidance of courier fees, long distance tolls, postage, printing and large square-footage fees for housing systems and staff. Examples follow. 4

Reductions in system-related building overhead. Even though computer systems have expanded in capabilities and price/performance, their environmental (overhead) requirements have decreased. For example, processor and storage capacity that recently required thousands of square feet of reinforced, raised floor, water cooling, and special air-conditioning can now be located in a much smaller area, in a normal office environment.

The savings in lease costs, utilities, and special environmental systems are quantifiable. The current annual costs for building and utilities overhead should be available from the operations support or budget staff. Using this and information available from the marketplace (for space and energy costs for new technology), the power and environmental expenses can be compared.

Reductions in telephone, postage, and printing costs. If the new system will reduce the number of telephone calls made or the number of letters, memoranda, or other documents printed and mailed by caseworkers, then a dollar value for this benefit can be developed. The dollar value can be estimated by assessing the effects of automation in other offices, then projecting a percentage reduction in current costs for these services.

For example, a program is currently paying $1,000,000 per year for telephone, printing, postage, and delivery costs. The agency has information from a pilot study and from contact with a recently automated office that access telectronic communication will reduce communication costs 15% in the first year and 25% per year after that as the system is implemented statewide. A five-year benefit of $1,150,000 is projected. This benefit can be monitored through implementation by reporting the actual expenditures in these categories.

Even more significant than system benefits are benefits derived from program-related productivity improvements, because large staffs and expenditures are involved. However, a common approach is tclaim the productivity improvement as a direct cost savings. It is not, unless staffing will be reduced an equivalent percentage. If staffing is not reduced, analysts need tdetermine the secondary effect. How will staff use the time saved by automation? Can a value be placed of the results of their new efforts?

In Chapter 2, productivity improvements are the basis for redirecting the efforts of staff from administrative duties tprevention services. The secondary effect is a realization of program benefits resulting from shifts in placement and duration of placement.

Other program-related examples follow. 5

Reduced staff turnover. Frequently, high rates of staff turnover are directly related tcauses such as obsolete equipment, limited technological support, and excessive administrative overhead - causes the project may be designed teliminate. The human resources office should have information regarding historic levels of support staff turnover, and may have conducted exit interviews tidentify the reasons that staff have left. There may be evidence tsuggest that more effective technological resources will reduce this turnover.

If this is the case, the human resources office and program management should be able tprovide reasonable estimates of the cost of replacing an employee. The costs would be derived from the expenses of recruiting, management time dedicated tinterviewing and reference checking, training, and lost productivity. The total of these costs, for the percent of staff whleft for reasons related tthe obsolete system, is reasonably a benefit of a new support system. Staff turnover can be monitored during implementation and operation of the new system, tdetermine the actual value of this projected benefit.

Improved ability trespond tprogram (legislative) changes. Child welfare programs tend tbe highly dynamic, as evidenced by welfare reform programs currently planned or underway at the State and Federal level. Changes in procedure, forms, or reporting may be mandated at short notice by legislative changes or executive order. The costs of making such changes can be substantial; they typically involve system staff tmodify or enhance the system, as well as program staff timplement the changes. Significant changes may require extensive retraining and may involve the production of new forms and instructions.

The historic costs of accommodating such changes should be available (or estimable) as hours of effort by various staff categories. Hours can be turned intdollar costs by applying average loaded hourly rates. Note that in order tproject a benefit in this area, it will be necessary tshow that specific features of the design and implementation of the new system will result in improved flexibility or ability trespond tnecessary changes or enhancements.

The examples used in Chapter 2 of this guide are specific tthe child welfare program. Other examples of child welfare program benefits were cited in this chapter. In addition, the Companion Guide: Cost/Benefit Analysis Illustrated cites other generic types of program and system benefits. States should view these examples of benefits as representative, not comprehensive. States should alskeep in mind that the studies cited in the examples in this guide are fictional.

In Summary

Several key points were made in this chapter. When developing cost\benefit analyses for child welfare systems, States should:

• Emphasize not only cost reductions, but alsprogram improvements;
• Develop values or measures for program improvements wherever possible;
• Identify a broad list of potential benefits, but develop values or measures for only the most critical, in terms of program or dollar impact;
• Evaluate how time savings will be reinvested tbenefit the program;
• Develop both qualitative and quantitative benefits;
• Use dollar and other quantitative measures for benefits testablish the performance baseline and goals; and
• Justify acquisitions on the basis of dollar-quantifiable benefits where possible.

¹ Note that the answer is not simply "a computer system;" it is what the system will achieve.
^2
3 See the Companion Guide, Chapter 1, page I-9 for further information on developing benefits.
⁴ The figures in these examples are fictional and should not be cited in States' cost/benefit analyses.
⁵ The figures in these examples are fictional and should not be cited in States' cost/benefit analyses.

Back to Table of Contents