Using source financial data from accounting systems (data from organizational financial transactions), all costs associated with an output are collected, summed (total cost) and then divided by the number of units of output produced (data from management information systems). The result is the cost per unit of that output, or unit cost. This is identical to the model presented in Figure 1. To properly understand unit cost in DoD, we must understand what is meant by "output" and what and how costs are included in "total cost." The following two sections explore these areas.

In contrast to the more simple examples cited above, in general, DoD support activities produce more than one output, often complex in nature. That is, they are not producers of a homogenous output or single product. The training school trains navigators, bomber pilots, attack pilots, ground crews, etc., while the finance center pays bills, maintains account balances, produces internal and external reports, etc. Arguably, each of these is a different output calling for different mixes and types of labor skills, materials and capital (equipment, etc.), resulting in different total cost and unit costs.

Consider a distribution depot that handles millions of different items. Clearly large items require different resources to receive, store and issue than small items. Does this mean there are millions of different outputs at distribution depots? Some argue there are, but as a practical matter it would be difficult to manage effectively so many discrete unit costs. Thus, DoD unit cost methodology uses output measures that are aggregated. This means that the various outputs of an organization are grouped together to form a single (or few) measure(s) of output which generally describe the output of the organization. This aggregation is considered only when those outputs share similar characteristics and costs. These output measures provide a quantified measure of the workload an organization has produced. Output measures differ for each support area and are designed uniquely for each area included in unit cost.

In a few instances, proxy output measures are used to represent outputs that may be impracticable to count individually. For example, the principle function of an inventory control point (ICP) is to provide supply items to customers upon receipt of a customer order. However, this supply function includes such supply management activities as procurement, cataloging, standardization, storage and distribution. Since millions of types of supply items are managed by each ICP and distributed to ICP customers, managing unit cost information for each item is not presently practical. Therefore, the dollar value of sales of all supply items managed by an ICP is used as a proxy measure of ICP output. This proxy measure consolidates the cost of a vast variety of individual inventory items and supply management functions into one single measure.

An output is characterized as a measure of work produced, supported by a data system capable of tracking it. An output must be quantifiable, auditable and produced in response to a customer requirement. That is, to qualify as an output, an activity cannot control the demand for an output. Defining an output requires identification and inclusion of all tasks related to producing the output. These characteristics enable us to use automated tracking systems, foster definitional consistency and eliminate ambiguity in reporting. Knowing the way outputs are defined for a business area is a critical element in understanding how to manage with unit cost information. It is essential that the output measure represents the inclusion of all tasks performed.

Unit cost is a move to more businesslike accounting. Businesses rely on linking or tracing costs to outputs, and managerial or cost accounting provides managers this information. Management accountants strive to establish causal relationships between costs and cost objects to determine why costs were incurred. The process of tracing costs to cost objects forges a necessary link so that ultimately, we can relate costs to outputs, even if at an aggregated level. To enable treating costs in this manner, in DoD unit cost, we categorize costs as direct costs, indirect costs, and general and administrative costs (G&A). The sum of these categories yields total cost as shown in the model in Figure 2. These categories are briefly described below.

Direct Costs: Direct costs are those costs that can be traced exclusively to one output, such as hands on labor or material consumed directly in the production of an output. Direct costs tend to (but not always) change proportionally with the quantity of output.

Indirect Costs: Indirect costs are those costs which benefit two or more outputs but not all outputs. Indirect costs are often relatively insensitive to changes in quantity of output. An example of an indirect cost may be a second line supervisor who oversees some specific, but not all, production processes. Typically these costs are allocated among the various outputs which they benefit. It is important to note that while cost accounting attempts to establish causality, ultimately there is a certain amount of arbitrariness in any allocation scheme used. This is to say that no allocation scheme will be completely accurate, but a reasoned allocation scheme for these costs helps better identify the "true" total cost.

General and Administrative Costs: General and administrative (G&A) costs are those costs that cannot be reasonably associated with any particular product or service produced. Commonly referred to as overhead, these costs are allocated over all outputs produced. Again, any allocation scheme will be somewhat arbitrary in nature, but will help better reveal the "true" total cost of an output. However, every attempt is made to choose an allocation methodology that best suits each functional area. As with indirect costs, G&A costs tend to be relatively insensitive to small changes in output. Examples of G&A costs include functions such as local comptroller, security, facilities engineering, fire protection, custodial services, snow removal and similar types of base support functions.

As with the cost accounting view, cost behavior is generally viewed as either fixed or variable. In this view of cost, total cost is the sum of all fixed costs and all variable costs. Fixed costs are those costs that, over some specific time period, do not vary with quantity of output. These are costs that we must pay no matter how much we produce. For example, an annual lease that requires monthly rent payments whether or not we produce anything is a fixed cost.

Variable costs are those costs that vary directly with quantity of output. If we produce nothing, variable costs are zero. Consider the cost of materials used to repair radios. Each radio repaired requires one unit of material at some cost. The more radios repaired, the greater the total material cost. If we repair no radios, we use no materials and incur no cost for materials.

There is also a cost category called mixed costs. Mixed costs behave as fixed costs up to some minimum quantity of output, then, like variable costs, increase as the number of units of output increases. Only fixed and variable costs will be discussed in this section on unit cost.

It is important for managers to recognize how costs behave in order to plan and manage properly. In the short run, changes can normally be effected only by making decisions about variable costs. For example, reducing the amount of packing material used for shipments could reduce the unit cost of issues at a distribution point. Changes to fixed cost tend to be long run oriented. The buy versus rent decision is one example.

Managers should not use the notion that all (or most) of their costs are fixed costs and therefore management is unable to control costs. In today's environment of increasingly scarce resources, creative, cost conscious managers should want to drive production cost down. To do so enables them to "do more with less." This challenge to managers means reengineering the process in which work (output) is produced. In this pursuit, it is necessary to recognize which costs can be affected quickly by management action and which will take more planning and management attention to influence. More will be said about these challenges later.

Within DoD financial management, there is no clear consensus or generally accepted definitions as to which costs are fixed and which costs are variable. But regardless of cost behavior, fixed or variable, all the costs of a business must be addressed. Managers are responsible for their decisions, and using unit cost helps focus attention on and communicate information about the total cost of production. The unit cost process reinforces responsibility and accountability for decisions.

Historically, there have been some production inputs that were treated as "free goods" because the performing activity did not actually pay for them. But DoD has always paid the full cost. Unit cost simply identifies costs with the outputs produced. For example, military personnel, paid from a separate military pay appropriation account, were considered "free" assets. If military labor is required in a production process, the cost of military labor should be captured as part of total labor to know the total cost associated with that production process. Using unit cost facilitates cost visibility by making clearer connections between costs and outputs.

Effectively implementing unit cost principles and improving the financial management of support activities requires managers to understand the types and behavior of the costs incurred in the production process. Increasing the level of cost awareness should be encouraged by organizations. This can be done by actively researching production processes, investigating how costs behave and seeking means to reduce costs and improve value to customers.