This chapter provides guidelines for assessing the economic and financial costs of HIV/AIDS interventions. In turn, policymakers will use this information to assure that fully informed decisions are made regarding the cost-effective use of limited resources for HIV/AIDS care and prevention. These guidelines are intended to allow the reader to understand the basic concepts of cost analysis and to actually perform an assessment. The guidelines will walk the reader through the process of:

• defining the question that needs to be addressed;
• collecting cost information;
• categorizing costs in a way that will focus the analysis on the relevant questions; and
• writing a technical report that lays out the cost of an HIV/AIDS intervention, projects future costs, and identifies possible means of cost recovery.

These guidelines are intended to provide a rapid overview of the issues and approaches that need to be considered in performing a cost analysis. More detailed guidelines have been developed for assessing the costs of health1, family planning2 and HIV/AIDS3 programs.

Defining The Question

Accurate and full cost information is crucial for evaluating the current level of resource use and assessing opportunities for future expenditures on the control of the AIDS pandemic. However, cost analyses are not designed to answer all questions with a single number. In fact, the approach selected for performing a cost analysis is highly dependent on the type of questions being asked. For example, an analysis designed to evaluate all resources required to carry on an intervention will likely be performed very differently from an economic evaluation designed to identify the incremental resources required to add HIV/AIDS services to an existing clinic.

Most cost analyses are designed to address at least one of the following types of questions:

• Planning and budgeting–What are the current and future financial requirements of each project?
• Sustainability–What is required to sustain each project?
• Economic evaluation–What is the cost per unit of output or outcome produced from each project?

Planning and Budgeting
One possible objective in performing a cost analysis is to plan and budget for current and future resource requirements. A well-performed cost analysis will permit decision makers to perform more accurate planning and develop new budgets. It will also allow project planners to project the financial requirements necessary to continue, expand, or replicate an ongoing project.

Sustainability
Program planners are frequently asked to ensure that their interventions can be sustained beyond the current funding cycle. One key to sustaining interventions is to estimate and plan for the costs of programs over time1. A sustainability plan can include data about the cost of services and can identify feasible opportunities for recovering costs. For example, a cost analysis of a sexually transmitted infections (STI) clinic can help program planners to identify the proportion of project costs that can be recovered by charging various prices for services.

Financial sustainability does not necessarily require that clients pay for the full cost of the services they are receiving. A sustainability plan might include the development of financial resources from any combination of sources, including clients, the government, the private sector, and international donors. For example, a country that wants to expand its blood banking services to ensure that the entire blood supply is screened for HIV, could perform a cost analysis to accurately ensure the start-up and recurrent costs that would be required to sustain such an endeavor. In addition to financial sustainability, interventions need political commitment and institutional capacity to sustain their operations. From the perspective of the consumer (for example, the user of a socially-marketed brand of condom), there also needs to be a sustainability of consumption to assure that uptake of services is maintained.

Economic Evaluation
A cost analysis is a critical tool for performing various types of economic evaluations of HIV/AIDS interventions, such as cost-effectiveness, cost-utility, and cost-benefit analyses. The analysis collects estimates as to the annual cost of the interventions in terms of economic requirements. The economic assessment also provides information so that changes in project costs over time can be projected.

In addition to evaluating economic inputs, it is also necessary to collect outputs that correspond with specific activities. Outputs can include items such as the number of condoms distributed, the number of patients treated for STIs, or the number of pamphlets distributed.

Cost-effectiveness analysis (CEA), covered in greater detail in the next chapter, is one technique to evaluate interventions by combining project costs and outcomes. When used for health care planning, CEA can help policymakers and program managers examine alternative ways of achieving a given objective and enable them to select the method that uses available resources most effectively and efficiently.

CEA is used primarily as a decision making tool to help policymakers and program managers select a future course of action. Although it often is used to evaluate completed programs, the ultimate use of this research method should be to help decision-makers decide what to do in the future. The next chapter will provide a detailed methodology for conducting cost-effectiveness analysis of HIV intervention programs.

Cost analyses are also necessary for performing a cost-benefit analysis. Cost-benefit analysis take cost-effectiveness analysis one step further by putting both the numerator (cost) and the denominator (benefits) in monetary terms.

Collecting Data On Costs

This section provides practical guidelines that are intended to help project planners systematically collect and analyze project costs. Methods for estimating and reporting project costs are presented throughout the rest of this section. Once these project costs are collected, a technical report can be written that incorporates the findings and conclusions derived from the cost analysis. The later section in this chapter, "Preparing Technical Reports," provides an overview of technical report writing using cost data.

In the longer term, economists should make collecting local costs a routine matter by incorporating quarterly cost reports into the requirements of project agreements. By doing this, the process will become more efficient and sustainable.

Macroeconomic Indicators
The first step in performing a cost analysis involves collecting and presenting macroeconomic data. This information will be used in the cost analysis to project future costs and to annualize capital costs. These macroeconomic indicators should be explicitly presented in the technical report.

Inflation
The current consumer price index should be presented in the technical report, along with available predictions concerning the future inflation rate. If future inflation projections are not available, the average rate over the previous five years should be used as the best estimate of the future inflation rate.

Exchange Rates
Official and black market exchange rates should be included in the technical report, as well as the dates and source of the data.

Official exchange rates may be artificial (not reflecting the true value of foreign exchange) and significantly different from black market exchange rates. Therefore, black market exchange rates are occasionally used in performing an economic cost analysis. Their use will provide better cost estimates when converting costs for analysis and comparison purposes.

Discount Rates
The nominal and real interest rates (the nominal interest rate minus the inflation rate) should be provided in the technical report. The real interest rate for a long-term treasury note should be used to annualize capital costs because it will reflect the current annual return at which money can be invested in a secure financial note.

Possible sources for a discount rate include the country's Ministry of Health, Ministry of Finance, local banks, and the World Bank. A range of interest rates should be collected for the purpose of performing a sensitivity analysis of the results.

Property Values
Information should also be presented in the cost analysis that summarizes the average cost of land and building space. This will only be necessary if the intervention is located within a building for which there is no estimated replacement value.

General Principles
In order to ensure consistency, reliability, and comparability of all data, it is necessary to collect cost data based on several general principles.

Collecting Full Cost Data
It is necessary that all resources, including resources provided in-kind, that are directly or indirectly used by the project be recorded and assigned a value by the economist. There are a number of reasons for assigning a value to in-kind resources. First, measuring the full cost permits the economist and project planners to identify the full resource requirements of the effort and provides the information needed to plan for long-range financing. For example, condoms and other in-kind inputs represent limited resources that may not be provided for free in the future. Finally, for evaluation purposes, it is necessary to assess all resources used in carrying out activities. An economic appraisal of two different intervention programs or two different activities within a program will be meaningless if based only on partial cost information.

The monetary value of all in-kind resources (material or human) should be estimated based on the opportunity costs of those resources. An opportunity cost represents the amount that one would have to sacrifice in order to obtain the particular good or service, regardless of whether or not it is actually paid for.

Partially or fully subsidized rental space, administrative services, materials, or equipment should be valued at the price at which they could be purchased on the open market. For example, condoms that are received at no cost from USAID have been valued at $0.045 per piece (not including shipping and storage). Resources such as volunteer labor should also be valued in the cost analysis. The value of volunteer labor should be based on the current wage rate for similar activities.

Including and Excluding Taxes
Taxes, duties, and other transfers from the project to the government are financial costs to the project and therefore should be included in the cost finding forms. However, it is important to separate the purchase price of items from the tax for that asset on the cost finding sheets. In performing an economic cost analysis, taxes should be excluded because they represent a transfer of funds rather than a true economic cost.

Allocating Joint Costs
Joint costs represent resources that are used to carry out more than one activity. Most interventions involve the use of some resources that are shared across multiple activities or services. The following are examples of joint costs that have to be allocated across activities or services:

• The salary of a receptionist who greets patients at a health center and directs patients to an STI clinic or various other health services.
• The cost of electricity at an STI clinic
• that performs both STI treatment and outreach activities.
• The equipment in a blood bank that is used to run the equipment for identifying both STIs and HIV.

In the first example, it would be inappropriate to fully assign the receptionist's salary to the STI clinic. Similarly, it would be inappropriate to fully exclude the receptionist's salary from the cost of providing STI services. Therefore it is necessary to allocate a portion of the receptionist's salary to the STI clinic through some allocation technique.

Determining how to allocate joint costs across projects or activities can sometimes be difficult. Below are some possible methods of allocating joint costs. The method used for allocating costs should be carefully described and justified within the technical report. In most cases, there is no definitive rule concerning methodologies for allocating joint costs, and it is therefore necessary to use the most logical approach available.

• One method to allocate labor is to observe resource utilization and to allocate costs accordingly. For example, the analyst can observe the accountant in a hospital to determine how much time is actually spent on the accounts of the STI clinic that is being costed. While this method is the most accurate way of allocating joint costs, it is also one of the most difficult and time-consuming.

• Indirect costs may be allocated based on the amount of space used (area of rooms used for the intervention) by the project being analyzed. For example, in order to allocate the use of building space, it may be possible to divide the building into areas that are predominantly used for the project being surveyed and to allocate costs based on the amount of space used. An STI clinic that comprises 10 percent of a hospital might therefore be assigned 10 percent of the costs of utilities and rent.

• Indirect costs can also be allocated based on the number of clients seen. For example, consider a voluntary counseling and testing (VCT) clinic that has 100 clients per month that is located within a health center, which has a total of 500 clients per month. In this case, 20 percent of the receptionist's salary could be allocated to VCT services, because 20 percent of the clients are attending the health center for VCT services.

• Indirect costs may also be allocated in the same proportion as the direct costs. Thus, it would be necessary to add up the costs of all direct costs and determine what proportion has been assigned to the project being studied. This percentage would then be used to allocate indirect costs. For example, if the budget of an STI clinic comprised 30 percent of a hospital's operating budget, it may be possible to allocate 30 percent of the hospital's indirect costs to the STI clinic.

• Finally, it is possible to simply divide costs evenly among ongoing projects. This is the least accurate method of allocating costs and should only be used when no other alternative is feasible.

Documenting All Information
All sources of information for cost estimates and specific calculations should be well documented in the technical report.

Recurrent Costs
Recurrent costs value resources that are expected to be consumed (or replaced) within a year. The following instructions indicate how individual recurrent costs should be collected and recorded in the cost finding sheets.

Salaries and Benefits
First, identify all personnel who are directly or indirectly connected in any way with the program. Include clerical and maintenance staff, volunteers, and consultants.

Second, calculate the amount of time each person spends on the program during the study period. Because most employees work on more than one project, it will be necessary to allocate time to the various projects based on interviews and through observation. If any employee (such as maintenance staff, administrators, or accountants) works indirectly on the project, their labor time should be allocated based on one of the methods described under "Allocating Joint Costs" in this section.

Third, determine the compensation paid to each worker. The figures may be listed as annual rates; however, in some cases, they will be noted as hourly, daily, weekly, or monthly rates.

Fourth, multiply the amount of time each person spends on the program by the compensation figures to compute the cost for that person.

Fifth, add the cost of fringe benefits if these are not included in personnel costs. Employer contributions to medical plans, insurance plans, social security, and other benefits are sometimes computed as a percentage of total salary. Some fringe benefits have to be computed separately. Consultants are usually paid an hourly or daily rate, or sometimes a flat honorarium. They do not usually receive fringe benefits.

Sixth, identify the value of any income taxes paid. This should be included in the financial analysis but may be excluded from the evaluation of economic costs.

Finally, sum all personnel costs to obtain the total personnel costs.

Materials and Supplies
First, list all of the relevant materials and supplies directly or indirectly used in the program during the study period. Second, determine the amount of materials and supplies used during this period. One way to do this is to conduct an inventory of all relevant materials and supplies at the beginning of the time period and at the end and subtract the ending from the beginning inventories to determine the amount used. Third, determine the cost of each item used. Finally, multiply the quantity of each item used by its cost and sum the total.

Travel and Associated Expenses
This category includes all transportation and per diem costs attributable to the intervention, whether for transporting people or supplies. Travel and transportation often make up a significant proportion of the costs of outreach services, supervision, and training. They may also account for a significant portion of the costs of imported supplies, materials, and equipment.

Operation and Maintenance
Buildings–The ongoing costs of running and maintaining a facility will include all building maintenance functions (such as repairs, housekeeping) and all utility payments (such as telephone, electricity, water, gas).

If operation and maintenance costs are not available, these costs can be calculated as a percent of the building's value. This percentage will vary, however, depending on the quality of construction materials, the age of the building, and the nature of the services that are delivered within the facility. If the facility is heavily used, then maintenance and operation costs will be higher. In general, operation and maintenance costs range from 2.5 to 4 percent of the building's investment cost, although figures of 8 to 9 percent may be applicable when facilities are in poor condition, heavily used, or both.

Equipment–The estimated percentage of total costs for required operation and maintenance of equipment increases with the equipment's complexity. When possible, the cost of operating and maintaining equipment should be itemized. In general, office furniture and general office maintenance has been estimated to be approximately 5 percent of investment cost per year; office equipment and audiovisual aids are approximately 10 percent; and laboratory and technical equipment vary between 10 percent and 20 percent of investment costs. These estimates, based on World Health Organization (WHO) and World Bank projects, will vary depending on circumstances and intensity of use.

Vehicles–To determine the cost of vehicle operation and maintenance, it is necessary to identify the actual expenditures for gasoline, oil, lubrication services, and other expenses. If the vehicle is being used for more than one project, the costs should be allocated based on estimates of the number of kilometers traveled for each project. Studies by the World Bank have recommended the following calculations for determining operating and maintenance costs:

• Gas = # km traveled/month X 12 months X mileage (litres/100 km) X unit cost/liter of gasoline
• Oil = 15% times annual expenditures on gasoline
• Spare parts and labor = 24% times purchase price of vehicle
• Insurance = Varies greatly by country; check with ministry of health or USAID

Studies suggest that the operating and maintenance costs of motorcycles are 40 percent of the purchase price and mopeds are 20-30 percent of the purchase price.

Capital Costs
The distinction between recurrent and capital costs is life expectancy. Items that are consumed or replaced within a year are categorized as recurrent costs, and items with a life expectancy of more than a year are categorized as capital costs. The cost of a capital item is spread over its expected life. The analyst must calculate the portion of the cost that should be annually attributed to the project, as follows:

• Identify all capital cost items that are expected to be used during the study period.
• Determine the current market value of each asset, including the cost of transporting the asset to the site. This information can be obtained by contacting manufacturers or reviewing catalogs.
• Compute the annualized cost of each item (using the estimated life expectancy of the resource, the prevailing interest rates, and annualization tables).
• Determine the proportion of the item to be charged to the program.
• Sum the total.

Buildings
The cost of space is frequently the largest capital cost and should be carefully calculated to be certain that a standard approach is being used. This calculation will involve making cost estimates of the building. In calculating building costs, it should first be determined whether the space is rented or owned.

Rented space–The rental cost, including any rental fees that are incurred, should be obtained. Because the rental cost can be used as a proxy for the annualized cost of a building that is owned, it is not necessary to annualize the rental cost. Rent should be classified as "rented/leased capital."

Owned space–For a building that is owned, the replacement value of the building should be estimated. If a cost cannot be obtained for the specific building, it should be possible to develop an estimate based on the size of the building and the cost per square meter of comparable buildings that are for sale in the community. Estimates for number of square meters and for the unit cost per square meter should be made in consultation with architects in the Ministry of Health or the Ministry of Public Works. Unit cost per square meter may also be verified with local construction firms and with bilateral and multilateral organizations that have recently completed or are in the process of financing projects with similar civil works components. It is important to note that the analysis is concerned with the economic cost of the building, so even if an occupant neither pays rent nor owns the building, a cost still needs to be assigned.

Equipment
The current replacement value of equipment should be calculated. This can be obtained with information about the type of equipment, the manufacturer, and the model.

Furniture
Furniture costs should be itemized wherever possible. However, if no other source of cost information is available, the capital cost of furniture may be broadly estimated at 8 percent of the construction cost of the building.

Vehicles
The cost of vehicles should be estimated on the basis of whether or not the items will be purchased locally or imported. Quotes (including import duties) should be solicited from potential suppliers and verified by the Ministry of Health. All import duties and taxes should be included in the analysis under a separate line item.

Annualizing Capital Costs
Value of Assets–The value of assets will be estimated by determining the current market value. The use of current market value will facilitate the data collection process because it is not necessary to determine historical prices.

Life Expectancy–The analyst should identify the life expectancy of an existing asset by determining the remaining life of the existing item. For highly technical items, such as medical equipment, the life expectancy is frequently limited by technological obsolescence. This should be carefully considered in determining an appropriate life expectancy.

There is no clear-cut rule for establishing the useful life of assets. In general, it is recommended that Table 16-1 be used as a guide.

Annualization Factor–To compute the annualized cost of a capital item, an annualization table is necessary. As an example, assume a vehicle has recently been purchased for the exclusive use of a project. The replacement value of the vehicle is $10,000, with a 7-year life expectancy and a 5 percent real interest rate. The annual cost can be found in the table by finding the interest rate column (r = 5%), locating the annualization figure for the expected remaining lifetime of 7 years (this is 0.173), then multiplying the current replacement value of the vehicle by this annualization factor: $10,000 x 0.173 = $1,730.

Technical Assistance Costs
In order to place the cost of technical assistance in a comparable format with the capital and recurrent costs incurred in the field, the technical assistance cost should be annualized over the expected length of the intervention. Thus, if a donor's in-house costs were $100,000 over 3 years and the real annual interest rate were 15 percent, the annual cost would be $100,000 x 0.438 = $43,800.

Categorizing Costs

This section defines the categories of costs that will be applied in the cost finding forms. The first step involves deciding if economic or financial costs will be calculated. The second step involves actually categorizing costs.

Economic versus Financial Costs
Before beginning data analysis, it is necessary to determine if the cost analysis is going to focus on economic or financial costs. Financial costs are those narrowly defined as having been paid for by the project. This could include such things as salaries and materials, but not donated items. These costs are particularly relevant when trying to determine the payments required to keep an intervention operational. Implicit in financial costs is the assumption that items that are donated (by the government, international donors, volunteers, or others) will continue to be donated into the indefinite future.

An economic cost analysis involves a broader evaluation of resources used, regardless of who actually pays for those resources. Economic costs include financial costs, but also values resources that are obtained from other sources. An analysis of economic costs would include items such as donated condoms or volunteer labor. This information is particularly useful when evaluating the opportunity cost of an intervention or when comparing interventions.

One pitfall in performing a cost analysis occurs when two interventions are compared using a financial cost analysis. Imagine an inefficient and expensive intervention that is highly subsidized by either the government or an international donor. In contrast, a second intervention produces significant "value for money," but does not receive any outside support. An economic analysis would correctly demonstrate that the latter intervention was the less costly intervention. However, a financial analysis might conclude that the first intervention was less costly, simply because the costs of carrying out the intervention were being incurred by the government or the donor.

Defining Costs
As already indicated, it is necessary to determine whether the analysis is trying to address issues of budgeting, planning, sustainability, or economic evaluation. This step of defining a question is critical because it will assist the evaluator in defining what costs should be included in the analysis.

Each resource used to carry out an intervention should be identified and categorized. Even though resources may be subsequently removed from the main focus of the cost analysis, it is still important to identify each of these resources.

Once a resource has been identified, it is necessary to ask the following questions about each item.

• What is the functional classification of this resource?
• Is this a start-up cost or a recurrent cost?
• Is the cost fixed or variable?
• Is it a direct or an indirect cost?
• Who paid for this resource?

Functional Classification of Costs
A prerequisite to the standard collection of cost data is establishing proper functional categories for each project. A critical step in creating a useful costing system is to define clear program categories that allow project evaluation by project components or activities. Cost and output data collection by activities will allow cost-effectiveness analysis and comparisons between alternative programs and activities within a program. This will also facilitate planning, budgeting, and monitoring tasks by meaningful program components.

All costs should be assigned to a specific programmatic activity (such as lab support, condom distribution, data collection, evaluation, training, production and distribution of material, STI diagnosis and treatment, or health education). When categorized in this manner, the activity will have its own measure of costs and its own measure of output.

Start-up versus Recurrent Costs
Costs can be categorized as either start-up or recurrent costs. Start-up costs involve the purchase of goods or services during the initial stage of an intervention. This can include capital items (resources with a life expectancy of more than one year), research, consultancies, and the design of information, education, and communication (IEC) materials. Recurrent costs are those required for the normal operation of the intervention.

The distinction between start-up and recurrent costs is important because:

• Some donors limit their contribution to start-up costs and expect the host country to be responsible for the recurrent costs.
• In many countries, there is one budgeting and accounting process for start-up costs and another for capital costs.

Fixed versus Variable Costs
Costs that remain unchanged despite small fluctuations in output are fixed costs. Thus, a blood bank that takes one additional blood sample would probably not incur additional staff or equipment costs. Items such as a building, salaries of permanent staff and laboratory equipment are generally considered to be fixed costs.

Variable costs, in contrast, fluctuate directly with changes in the related total activity or volume. Thus, an STI clinic would probably incur variable costs for such items as swabs and medications because the need for these items increases directly with the number of patients treated. Variable costs might include items such as medications or materials.

This distinction between fixed and variable costs is important because the cost of expanding a project in the short run will depend on the variable costs but not the fixed costs. The distinction, however, is also somewhat subjective. For a clinic that is currently working at 100 percent capacity, the addition of one additional patient might require that another doctor be hired, or that another item of equipment be purchased. To sufficiently address the costs of expanding a program, the analyst should also look at other relevant factors, such as space constraints, trained labor availability, and sufficiency of demand.

Direct versus Indirect Costs
In performing a cost analysis, it is a common mistake to omit costs that are only indirectly attributable to an intervention. However, it is critical that both direct and indirect costs be included in the cost analysis. Direct costs represent resources that are directly attributable to the service being provided. For example, the direct costs of an intervention with sex workers may consist of personnel directly involved in the intervention (including volunteer time), condoms, and transportation directly attributable to the intervention.

Indirect costs represent resources that are only indirectly attributable to the intervention. For example, the salary of a security guard at a hospital that provides STI services is an indirect cost that must be included in an analysis of the cost of providing STI services. Similarly, the water and electricity bill of that same hospital needs to be considered as a component to the cost of providing STI services.

Indirect labor costs may include the salaries for accountants, administrators, secretaries, and maintenance employees. Indirect capital costs may include buildings, office equipment, and furniture. The approaches for allocating a portion of indirect costs to a particular intervention have already been discussed in this chapter under "Allocating Joint Costs."

Funding Source
Project activities may be funded by international donors or through other local sources, such as local institutions or the private sector. Resources provided in-kind will be identified under "donated goods and services."

In-house costs incurred by donors for providing technical assistance to the project need to be collected separately. These costs should

be collected and disaggregated into two time periods: 1) from project approval to the initiation of the intervention, and 2) from project initiation until the conclusion of the technical assistance. In this way it is possible for donors to separate their initial investment from the resources that are required for actually maintaining the project over time.

Items that are currently provided to the field at no cost to either the donor or the in-country counterparts (for example, voluntary labor and free rent) should be categorized as donated resources. Because an analysis will collect the full cost of all resources, the monetary value of these "free" resources will be estimated (an explanation on how to value donated resources was provided in the section on "Collecting Full Cost Data").

Preparing Technical Reports

A technical report generally includes at least three parts. The first, the cost analysis, concerns the actual assessment of costs. This includes an initial assessment of all resources used to carry out the intervention, along with a more narrow assessment of specific costs that are relevant for addressing the particular questions that need to be addressed.

The second part involves developing cost projections. This part will describe the resources required to assure that an intervention is sustained into the future. It will take into consideration changes in the price of resources as well as possible contingencies (such as salary increases).

The final part should identify possible opportunities for cost recovery. This might include an assessment of how much clients might be willing to pay for services. However, such an assessment should also evaluate opportunities for sustaining interventions through resources from the public and private sector and donors. Furthermore, this part of the technical report might be used to evaluate ways in which the cost of services might be reduced.

Cost Analysis
The first part of the report will include a clear statement of the questions that are being addressed in the analysis. This part should include justification for the approach taken, as well as an explanation of all assumptions. The report should ideally provide the cost finding forms that were used to collect data.

The analysis should begin with an estimate of the cost of all resources used in implementing the project. This should be followed by a description of costs that were excluded in addressing specific issues. For example, an analysis of the costs to add voluntary counselling and testing to existing health centers will exclude the cost of indirect costs that would have been paid for regardless of the introduction of the intervention.

Cost Projections
The technical report should include estimates of changes in costs five years into the future. While cost projections are likely to have a potential for error, such estimates will provide some indication as to the potential future growth and financial needs of the project.

In order to make cost projections, it is necessary to make assumptions about the future.

The assumptions should be clearly and carefully stated in the technical report. The assumptions concerning cost projections will depend on the expected future growth of the project.

The technical report should identify alternative future scenarios and the expected cost under each scenario. In addition, it will be necessary to project future outputs under the various assumptions. Recommendations for possible scenarios under targeted sexual interventions are presented below. Cost projections should be made for all scenarios and the most likely growth path should be identified in the technical report.

One scenario should assume that the existing project continues to reach the current target audience with essentially no growth. Cost projections should include expected price changes for project inputs (beyond the national inflation rate), the usual recurrent costs, and the replacement of expired capital items.

The second scenario will involve an increase in the target population without an expansion in the regions covered. Cost projections will increase due to an increase in the size of the target audience (for example, reaching more commercial sex workers in a city), but not an increase in the size of the region involved. Thus, for a condom distribution program, such a scenario would involve increased spending for variable costs, such as condoms and materials, but only moderate or no growth for fixed costs, such as permanent staff.

Finally, the third scenario would involve an increase in the target audience within the existing region and an expansion into other regions. To obtain data for this scenario, it is necessary that project personnel realistically determine into which regions the project could reasonably expand. This scenario will probably involve significant increases in both fixed and variable costs.

Price Changes
All cost projections should be presented in constant currency. If information is available indicating that certain cost items will increase in value at rates different from the national inflation rate, the costs should be adjusted appropriately to reflect the expected real price increase.

For example, assume that an intervention is being assessed that used 1,000 pesos worth of fuel in the current year. In the next year, it is expected that the national inflation rate will rise at a rate of 10 percent but fuel is expected to increase by 35 percent. Fuel prices should then be adjusted by the expected price increase less the inflation rate (35 percent minus 10 percent). In this case, fuel costs should then be adjusted as follows:

a) Current project cost of fuel (1999): 1,000 Pesos*
b) Expected fuel price increase (2000): 35%
c) Expected national inflation rate (2000): 10%
d) Adjusted fuel cost (2000): 1,250 Pesos*
* Prices in 1999 Dominican pesos

Contingencies
When estimating resource needs for planning and budgeting purposes, an allowance should be made for physical contingencies. Physical contingencies represent the possible increase in base costs as a result of changes in quantity and/or design of the project. For example, operating costs may vary considerably depending upon the level of use in the project. Therefore, cost estimates should reflect anticipated increases in use and project scale over time. Similarly, construction costs will significantly vary according to geography, terrain, and labor availability factors. Acceptable ranges for physical contingencies will vary among and within projects. They are most appropriately expressed as a percentage of base costs. They should be calculated, presented separately, and then aggregated to provide an estimate of the total cost of project components.

Cost Recovery
The technical report should also include an analysis of opportunities for cost recovery. This may include an analysis of the possible cost recovery that could be feasible from clients of the service. For example, a condom distribution program should evaluate opportunities for charging for condoms, and include an assessment of what is affordable for consumers. For an STI clinic, the technical report should include an evaluation of what the patients are willing and able to pay for STI services.

This section of the report should also include an evaluation of additional options for sustaining the program. For example, the report should include an overview of resources that could be reasonably expected from the government, donors, and the private sector over the following five years.

Conclusion

A cost analysis is a useful and necessary tool for evaluating interventions. It is particularly useful for budgeting, sustainability planning, and economic evaluations. Regardless of whether or not the project is being implemented by a governmental or a non-governmental organization, it remains vital to track costs and to project the value of resources that will need to be purchased in the future. Cost analysis is also a necessary input when performing a cost-effectiveness, cost-utility, or cost-benefit analysis.

The most critical step in performing a cost analysis involves clearly defining the questions to be addressed. This will assist both in terms of identifying the costs that should be included and determining the ways in which costs should be categorized. As illustrated in these guidelines, there are a variety of ways in which costs can be categorized.

Throughout the process of identifying costs, it is important to include all resources that are used for carrying out an intervention. This should include resources that are only indirectly used, as well as the use of resources that are provided in-kind, such as volunteer time or donated condoms. The technical report outlining the findings of the cost analysis should clearly address the original economic questions that were being asked and provide feasible solutions.

References

1. Creese A, Parker D, editors. Cost analysis in primary health care: a training manual for programme managers. Geneva: World Health Organization; 1994.
2. Janowitz B, Bratt JH. Methods for costing family planning services. New York: United Nations Population Fund and Family Health International; 1994.
3. Kumaranayake L, Pepperall J, Goodman H, Mills A. Costing guidelines for HIV/AIDS prevention strategies. London: London School of Hygiene and Tropical Medicine; 1998.