Evaluation o f Inputs and Outputs

After inputs and outputs for each activity have been identified, they need to be evaluated. The chosen time frame in which to evaluate the costs and benefits of the activities is termed the base year for PAM analysis. The base year may be the current year or any past year. Research objectives and practical considerations determine the choice of base year. If current policy issues are the focus of research, the base year will be as near the present as possible. But current-year data, especially price data, will often be incomplete, so the data must come from one or two years in the past. Because policy-makers may be wary of dated results, relevance to current issues requires the use of a base year as close as possible to the present year. Alternatively, if historical issues are of particular interest, the base year or years could go far into the past.

Both quantity and unit price information for the estimation of costs and returns are desirable to facilitate social valuation. The most common procedure used in the estimation of social input cost or social output value is to apply social prices for inputs or outputs to the relevant quantity measure. For some inputs, quantity and unit price data cannot be isolated. In this circumstance, social values are approximated by proportional adjustment of the private value of a particular input or output. Sometimes information about divergences can be used to generate estimates of social values. For example, if the farm wheat budget has only a total cost for pesticide input, without any indication of the quantities used, information on the percentage distortion of import prices can allow the researcher to impute a social value. If tariffs on imports of pesticides are 50 percent, the private value of pesticides is 50 percent higher than the social value. Division of the total private value by 1.5 gives an estimate of the social value. Such procedures entail the assumption that quantities are unaffected by the price change.

When proportional adjustments to private values are impossible, equality between private and social values is often presumed. If the input or output accounts for a small proportion of total input costs or output revenue of the activity, little harm is done to the results. Even if the assumption of equal private and social values is incorrect, incor-poration of the "true" social value will have an insignificant effect on the magnitudes of total social costs and social revenues. But if the item in question is a large component of costs or revenues, the assumption of convenience can prove a grave error in practice. At this point, further analysis must be postponed until a more comprehensive set of data can be assembled.

Explicit recognition of the time frame of analysis provides another justification for the collection of separate price and quantity estimates for the major inputs and outputs of the system. From the policy-maker's perspective, the long-run profitability of the system is often most germane to the policy formation process. Because many policies are not changed with great frequency, the policy-system interaction over a long time period must be understood. In the portrayal of the longer-run interactions of policy and profitability, expected prices replace prices observed at a particular time as the correct measures for calculation of input costs and output revenues.

Disaggregating Input Costs into Domestic Factor and Tradable-Input Components

After all private and social input costs have been standardized to an annual basis, they are allocated to their domestic factor and tradable-input components. This disaggregation is necessary to permit identification of tradable-input and domestic factor divergences. Figure 8.4 illus-trates the complete organizational format for the activity budgets. Both total private and total social costs are decomposed into their domestic factor and tradable-input components. In principle, many classes of domestic factors could be recognized. But for most purposes, four categories of domestic factors-unskilled labor, skilled labor, land, and capital-are sufficient. Because the commodity-in-process category is used only as an accounting device in the construction of the commodity system model, only the first three categories of input costs are disaggre-gated.

The decomposition exercise could be applied to every input listed in the fixed input, direct labor, and intermediate input categories. For example, the cost of fixed inputs reflects some marketing margin in addition to the basic cost of the machine. This margin incorporates the payments to factors and tradable inputs needed to operate the retail shop. Payments to hired labor could implicitly include payments for transportation to the activity site. Like the marketing margin, transportation costs reflect payments to a range of domestic factors and tradable inputs.

Decomposing all input costs into their exact domestic factor and tradable-input components is a formidable task that can absorb substantial resources. Moreover, adjustment often will have only a trivial effect on the results. The noncapital cost components of fixed inputs and the nonlabor cost components of direct labor inputs are usually a very small proportion of total costs. Unless information about decomposition is readily available, fixed input costs are usually classified entirely in the capital cost category and direct labor inputs are classified entirely in the categories of unskilled labor and skilled labor.

In practice, the exercise is generally limited to the intermediate inputs. Again, on the basis of available information and resources for the research effort, many intermediate inputs can be classified into a single domestic factor or tradable category. Seeds, fertilizer, and pesticides are examples of intermediate inputs whose costs reflect marketing margins in addition to the pure tradable cost. But if these margins are judged to be relatively small, costs of intermediate inputs can be allocated exclusively to the tradable category without causing major errors in the results.

With other intermediate inputs, including electricity, transportation, and most services, no particular cost category appears to dominate total costs. Such inputs are denoted as nontradable inputs, because they are not available on international markets. The decomposition of these inputs implies the construction of an activity budget for production of the intermediate inputs that is as complicated as the one in Figure 8.4. Electricity production could be analyzed as an activity, for example, with a budget that identifies the fixed inputs, direct labor, and intermediate inr ats necessary to produce electricity. In the process of decomposing the input costs for electricity production, more nontradable inputs-for example, machinery service and repairs-would be encountered. A budget could be constructed for the service and repairs in order to determine the proper allocation of these costs among the domestic factor and tradable categories.

Such calculations can take the analyst away from the original purpose-turning all policy analyses into studies of the nontradable industries in the economy. If these inputs are relatively unimportant elements of the commodity system costs, substantial research resources would be expended with little effect on the results. A rule of thumb is that unless the nontradable input represents more than S percent of total production costs of the system, separate budgeting exercises should be avoided.

More rapid approximations of the decomposition of nontradable inputs can be obtained in two ways. The most common technique utilizes an input-output matrix of the national accounts. These aggregate portraits of the economy allow calculation of the shares of labor and capital in each sector of the economy. Land costs typically are ignored, because they are a small component of nontradable-goods production costs. When the nontradable input of interest is associated with a particular sector, the capital and labor cost shares can be approximated. The remainder is allocated to tradables. This exercise provides a decomposition of the private costs of the nontradable input. Social costs of the nontradable input are then estimated by adjustment of the labor, capital, and tradable components to reflect the impacts of divergences. The sum of the social values of the domestic factor and tradable components gives the total social value of the nontradable input.

A second alternative for the treatment of nontradable inputs relies more on the analyst's judgment. When input-output matrices of the economy are unavailable, the distribution of costs among domestic factor and tradable categories must be estimated. In the absence of any information, an operational rule for distributive shares is the assumption that nontradable inputs contain one-third labor, one-third capital, and one-third tradables. Each private cost component is then adjusted to its social value, and the social values of the labor, capital, and tradable components are summed to generate an estimate of total social cost. Such estimation exercises are not much different from pure guess-work. If these arbitrary calculations are commonplace in the system evaluation, more data collection is essential before the system analysis should proceed. Box 8.3 illustrates the decomposition of nontradable inputs for a wheat system in Portugal.