Cover Image
close this bookCost Analysis in Primary Health Care: A Training Manual for Programme Managers (WHO - OMS, 1994, 158 p.)
close this folderPart C. Using cost data in planning
close this folderModule 10. Future costs
View the document(introduction...)
View the documentHow to estimate future costs
View the documentHow to use future cost estimates for decision-making

How to estimate future costs

The method of estimating the likely future costs of a health programme has many features in common with the method of measuring existing costs covered in Modules 4 and 7. For one thing, the same framework is appropriate: resources (and costs) can be classified according to type of input, the activity they are used for, who is expected to contribute to them, and what kind of currency is involved. Use the checklists in Module 1 to guide you in adapting the framework to budgets.

Also, as with existing programmes, you need to identify clearly the health services whose future costs you wish to estimate. In particular, this means you must:

· Describe the proposed programme in as much detail as possible.

· Examine the resources that already exist and determine whether they have the capacity to take on additional tasks (i.e. is there any spare capacity?).

· Determine the additional (incremental or marginal) resources required.

Cost estimation for a future programme has other features in common with that for existing programmes. It is sometimes appropriate to estimate in detail a sample of the activities of the planned future programme and extrapolate from these to the full programme. It is also important to distinguish between financial costs (money outlay) and economic costs. (You will often want to estimate both types of costs.)

However, there are a number of areas where differences in emphasis or procedures emerge. For example, in the choice of the future time period, you have considerably greater flexibility than is possible when measuring the actual costs of an existing programme. Actual costs rely on information on use and expenditure: there are often delays in gaining access to this information, and its usability deteriorates over time. Information can become obsolete quite quickly, and you should always be careful when using historic data for projections. Future costs do not face these problems. It is usually appropriate to estimate future costs for a period of 5 - 10 years. Naturally, the time period selected will depend on your specific needs.

There are three general approaches to estimating future costs (both economic and financial). Sometimes it is possible to use them in combination. These approaches are explained below.

The "ingredients" approach

The "ingredients" approach involves translating the general description of whatever programme is to be costed into specific resource requirements. To do this, you should list in detail the number and types of staff, the number and types of vehicles, the quantity and types of drugs, and so forth. Calculate the cost for each item by multiplying the quantity by the price, and then add up all the individual costs.

For the whole programme (or for each identified activity of the programme) you would prepare a detailed list of all the resources required, the quantities of each resource, the unit price, and (for capital items) the expected life of the item. For each year of the proposed programme (at least five years for a long-term programme), you would fill in the details of resource requirements as indicated in the table below, to which details for all the "ingredients" of each input category must be added.

Consider the "ingredients" approach further by doing exercise 10A, page 136.

The "ingredients" approach requires a thorough understanding of the resources required for a new project. Specifying the details of planned project implementation can be a very useful exercise for the good design and management of programmes. For this reason, it is recommended that you start by using the "ingredients" approach when estimating future costs.

This approach is particularly appropriate to use when:

· the programme to be costed is new or rapidly changing, and existing programme costs may not provide a reliable guide;

· the available cost data are incomplete or unreliable, and it is feasible to document the resources required in some detail.

Cost estimation for long-term budget preparation ___ year___


Description (type and unit of measure)

Quantity (number of units)

Price per unit

Useful life

Annualization factor

Annualized cost




Buildings, space

Training, non-recurrent

Social mobilization, non-recurrent

Subtotal, capital




Vehicles, operation & maintenance

Buildings, operation & maintenance

Training, recurrent

Social mobilization, recurrent

Other operating inputs

Subtotal, recurrent


* Annualization factor = 1.0 for all recurrent costs,

Unfortunately, for many health projects detailed resource requirements are difficult to predict. Losses of supplies will occur to a greater or lesser extent, affecting the full cost of delivering the planned services. The combination or mix of personnel employed, even at the basic delivery unit level, may change over time in unexpected ways. (The substitution of a community health worker for a physician or nurse for certain tasks is a case in point.) The ratio of labour to equipment or other resources may be equally unpredictable.

Perhaps not surprisingly, estimates of future costs that rely on itemizing resources tend to focus on the items of large cost (mostly capital items, which are discussed in earlier modules). Much less attention is likely to be paid to estimating recurrent costs. This is unfortunate for several reasons:

· The items classified under recurrent costs, although often of low individual value, are frequently purchased in large amounts and thus can contribute significantly to overall costs.

· Inadequate funding for recurrent costs is likely to jeopardize the effectiveness of the whole programme. Recurrent items are often complementary to capital inputs. Many capital items, such as vehicles and equipment, will not function adequately without, for example, fuel or power or maintenance.

· Recurrent costs are usually borne by recipient governments in developing countries, even when outside donors provide capital items such as equipment or vehicles, and the government is likely to have limited flexibility to provide more resources if the initial budget is incorrect.

Because detailed documentation of resources is tedious (particularly for recurrent costs) and there is a danger that some costs will be left out, a second approach, described below, has some advantages.

The "adaptation" approach

This second approach to prediction involves using your experience of what a certain kind of programme actually costs to estimate what a similar programme planned for the future might cost. This has the advantage of accommodating not only obvious recurrent costs but also wastage and other costs which might easily be overlooked in the "ingredients" approach. This second method shares with the first the difficulty of allowing for changes in the proportions of resource inputs employed, but it has some advantages.

The "adaptation" approach involves the following steps:

· Select a programme that (a) is as similar as possible to the planned one; (b) operates as rationally and efficiently as the planned one is likely to; (c) has accessible cost data.

· Estimate the costs of that programme, using Modules 4 and 7 as guides or making use of available estimates.

· Adjust your estimates to take into account any known differences between the programme costed and the planned programme.

One of the major challenges is to know how to modify appropriately the cost data you have. One important difference that is likely to exist between the costed programme and the one you are planning is size. Total costs are likely to be misleading in this case. As a first approximation, you will need to know unit costs, such as total cost per contact, per facility and per district. Which of these is most relevant will depend on the aspects of the costed (current or past) programme you are going to reproduce. For example, if you are establishing similar health centres in another area, you might use cost per facility (and multiply this by the number of planned new facilities to get the total future cost estimate). If you are expanding the catchment areas of existing facilities, cost per contact would be more appropriate. If you are establishing a similar programme in a new district, cost per district may be the most useful guide, assuming a reasonable similarity between the two populations served in terms of size and demographic structure.

Often, you will have to make adjustments to take into account the existing infrastructure. For example, your planned programme may not need additional buildings, but the cost data available might include building costs. In that case, if possible, you would eliminate building costs from the estimates. In other words, for budgetary purposes you do not want the average total cost of the existing programme, but the average additional or incremental costs of the new one.

Even if it is difficult to find programmes that are very similar to the planned one, you may be able to make use of some parts of existing cost estimates. In particular, you should use available cost data to help estimate the likely recurrent costs associated with capital investments. Usually, it is relatively easy to work out what capital inputs your new programme requires (using the "ingredients" approach), for example the number of vehicles and the number of refrigerators. You could then use available cost data on the relationship between capital inputs and their recurrent costs (the "adaptation" approach). You could look at the recurrent cost for each capital item. For example, expenditure data for existing programmes could give you an estimate of the average annual cost of running a vehicle (i.e. the operation and maintenance cost per vehicle). This could save you the difficult (and probably unreliable) exercise of attempting to document all the different running costs a vehicle is likely to have (fuel, oil, spare parts, registration, insurance). Your existing cost data may show you that, on average, it costs about $ 800 a year to operate a vehicle, and you could apply this estimate to your future programme.

Even with such information, you should not assume that it is appropriate to apply costs from one situation directly to another. For example, the running costs for a vehicle in a city are likely to be very different from those for the same vehicle used on poorly maintained rural roads. Or maintenance might have been inadequately funded. Fortunately, it is usually easier to understand the factors that influence the value of a specific category of costs than it is to understand those that influence total costs, so that you can often make appropriate adjustments.

In general, use the "adaptation" approach when:

· the programmes for which cost data are available are broadly representative of the planned programmes;

· the programmes costed have been adequately funded and have functioned efficiently (check, for example, that supplies of fuel and drugs have been adequate);

· costs for existing programmes can be reasonably readily obtained (i.e. more easily than documenting resource requirements in detail as in the "ingredients" approach).

The "mark-up" approach

For programmes that are already under way, a third, and very common, way of estimating future costs is simply to add a certain percentage to cost estimates from the recent past - a kind of "mark-up" of past values. This might be an arbitrary percentage, perhaps just taking into account expected inflation.

The danger with this approach is that it assumes that the programme has been adequate in the past, and discourages consideration of new initiatives or modifications. It also depends on reasonably accurate past cost estimates for the programme. However, a very detailed estimate of future costs will not be needed every year. Once a programme has been initiated, it is often necessary only to:

· determine current and recent costs of the programme (i.e. take stock);

· modify these estimates to reflect general cost trends, e.g. by using official price indexes, and allow for demographic changes, health service developments and any special problems expected to arise;

· check annually that expenditure is in line with the budgets, and adjust the budget if necessary;

· prepare a detailed budget to incorporate any new items into the programme.