TOCPREVNEXTINDEX

 


ECONOMIC ANALYSIS OF AGRICULTURAL PROJECTS

SECTION I. IDENTIFYING PROJECT COSTS AND BENEFITS

We undertake economic analyses of agricultural projects to compare costs with benefits and determine which among alternative projects have an acceptable return. The costs and benefits of a proposed project therefore must be identified. Furthermore, once costs and benefits are known, they must be priced, and their economic values determined. All of this is obvious enough, but frequently it is tricky business.

What costs and benefits in agricultural projects are, and how we can define them in a consistent manner, are the topics of this chapter. In chapter 3 we will examine how we can obtain market prices. After the financial analyses are discussed in chapters 4-6, the economic analysis is addressed in chapter 7 with a discussion of how to adjust market prices to reflect the real resource flows.

Objectives, Costs, and Benefits

In project analysis, the objectives of the analysis provide the standard against which costs and benefits are defined. Simply put, a cost is anything that reduces an objective, and a benefit is anything that contributes to an objective.

The problem with such simplicity, however, is that each participant in a project has many objectives. For a farmer, a major objective of participating is to maximize the amount his family has to live on. But this is only one of the farmer's interests. He may also want his children to be educated; as a result, they may not be available to work full time in the fields. He may also value his time away from the fields: a farmer will not adopt a cropping pattern, however remunerative, that requires him to work ten hours a day 365 days a year. Taste preference may lead a farmer to continue to grow a traditional variety of rice for home consumption even though a new, high-yielding variety might increase his family income more. A farmer may wish to avoid risk, and so may plan his cropping pattern to limit the risk of crop failure to an acceptable level or to reduce the risk of his depending solely on the market for the food grains his family will consume. As a result, although he may be able to increase his income over time if he grows cotton instead of wheat or maize, he would rather continue growing food grains to forestall the possibility that in any one year the cotton crop might fail or that food grains might be available for purchase in the market only at a very high price. All these considerations affect a farmer's choice of cropping pattern and thus the income-generating capacity of the project. Yet all are sensible decisions in the farmer's view. In the analytical system presented here, we will try to identify the cropping pattern that we think the farmer will most probably select, and then we will judge the effects of that pattern on his incremental income and, thus, on the new income generated by the project.

For private business firms or government corporations, a major objective is to maximize net income, yet both have significant objectives other than simply making the highest profit possible. Both will want to diversify their activities to reduce risk. The private store owner may have a preference for leisure, which leads him to hire a manager to help operate his store, especially during late hours. This reduces the income-since the manager must be paid a salary-but it is a sensible choice. For policy reasons, a public bus corporation may decide to maintain services even in less densely populated areas or at off-peak hours and thereby reduce its net income. In the analytical system here, we first identify the operating pattern that the firms in the project will most likely follow and then build the accounts to assess the effects of that pattern on the income-generating capacity of the project.

A society as a whole will have as a major objective increased national income, but it clearly will have many significant, additional objectives. One of the most important of these is income distribution. Another is simply to increase the number of productive job opportunities so that unemployment may be reduced-which may be different from the objective of income distribution itself. Yet another objective may be to increase the proportion of savings in any given period so there will be more to invest, faster growth, and, hence, more income in the future. Or, there may be issues to address broader than narrow economic considerations-such as the desire to increase regional integration, to upgrade the general level of education, to improve rural health, or to safeguard national security. Any of these objectives might lead to the choice of a project (or a form of a project) that is not the alternative that would contribute most to national income narrowly defined.

No formal analytical system for project analysis could possibly take into account all the various objectives of every participant in a project. Some selection will have to be made. In the analytical system here, we will take as formal criteria very straightforward objectives of income maximization and accommodate other objectives at other points in the process of project selection. The justification for this is that in most developing countries increased income is probably the single most important objective of individual economic effort, and increased national income is probably the most important objective of national economic policy.

For farms, we will take as the objective maximizing the incremental net benefit-the increased amount the farm family has to live on as a result of participating in the project-derived as outlined in chapter 4. For a private business firm or corporation in the public sector, we will take as the objective maximizing the incremental net income, to which we will return in chapter 5. And for the economic analysis conducted from the standpoint of the society as a whole, we will take as the objective maximizing the contribution the project makes to the national income-the value of all final goods and services produced during a particular period, generally a year. This is virtually the same objective, except for minor formal variations in definition, as maximizing gross domestic product (GDP). It is important to emphasize that taking the income a project will contribute to a society as the formal analytical criterion in economic, analysis does not downgrade other objectives or preclude our considering them. Rather, we will simply treat consideration of other objectives as separate decisions. Using our analytical system, we can judge which among alternative projects or alternative forms of a particular project will make an acceptable contribution to national income. This will enable us to recommend to those who must make the investment decision a project that has a high income-generating potential and also will make a significant contribution to other social objectives. For example, from among those projects that make generally the same contribution to increased income, we can choose the one that has the most favorable effects on income distribution, or the one that creates the most jobs, or the one that is the most attractive among those in a disadvantaged region.

Thus, in the system of economic analysis discussed here, anything that reduces national income is a cost and anything that increases national income is a benefit. Since our objective is to increase the sum of all final goods and services, anything that directly reduces the total final goods and services is obviously a cost, and anything that directly increases them is clearly a benefit. But recall, also, the intricate workings of the economic system. When the project analyzed uses some intermediate good or service-something that is used to produce something else-by a chain of events it eventually reduces the total final goods and services available elsewhere in the economy. On the one hand, if we divert an orange that can be used for direct consumption-and thus is a final good-to the production of orange juice, also a final good, we are reducing the total available final goods and services, or national income, by the value of the orange and increasing it by the value of the orange juice. On the other hand, if we use cement to line an irrigation canal, we are not directly reducing the final goods and services available; instead, we are simply reducing the availability of an intermediate good. But the consequence of using the cement in the irrigation project is to shift the cement away from some other use in the economy. This, in turn, reduces production of some other good, and so on through the chain of events until, finally, the production of final goods and services, the national income, is reduced. Thus, using cement in the project is a cost to the economy. How much the national income will be reduced by using the cement for the project is part of what we must estimate when we turn, in chapter 7, to deriving economic values. On the benefit side, we have a similar pattern. Lining a canal increases available water that, in turn, may increase wheat production, and so on through a chain of events until in the end the total amount of bread is increased.

By this mechanism, the project leads to an increase in the total amount of final goods and services, which is to say it increases the national income. Again, part of the analyst's task in the economic analysis is to estimate the amount of this increase in national income available to the society; that is, to determine whether, and by how much, the benefits exceed the costs in terms of national income. If this rather simple definition of economic costs and benefits is kept in mind, possible confusion will be avoided when shadow prices are used to value resource flows, a matter taken up in chapter 7.

Note that, by defining our objective for economic analysis in terms of change in national income, we are defining it in real terms. (Real terms, as opposed to money terms, refer to the physical, tangible characteristics of goods and services.) To an important degree, economic analysis, in contrast to financial analysis, consists in tracing the real resource flows induced by an investment rather than the investment's monetary effects.

With these objectives defined, we may then say that in financial analysis our numeraire-the common measurement used as the unit of account-is a unit of currency, generally domestic currency, whereas in economic analysis our numeraire is a unit of national income, generally also expressed in domestic currency. We will return to this topic in our discussion of determining economic values in chapter 7.

In the economic analysis we will assume that all financing for a project comes from domestic sources and that all returns from the project go to domestic residents. [This is one reason why we identify our social objective with the gross domestic product (GDP) instead of the more familiar gross national product (GNP).] This convention-almost universally accepted by project analysts-separates the decision of how good a project is in its income-generating potential from the decision of how to finance it. The actual terms of financing available for a particular project will not influence the evaluation. Instead, we will assume that the proposed project is the best investment possible and that financing will then be sought for it at the best terms obtainable. This convention serves well whenever financing can be used for a range of projects or even versions of roughly the same project. The only case in which it does not hold well is the rather extreme case in which foreign financing is very narrowly tied to a particular project and will be lost if the project is not implemented. Then the analyst may be faced with the decision of implementing a lower-yielding project with foreign financing or choosing a higher-yielding alternative but losing the foreign loan.

"With" and "Without" Comparisons

Project analysis tries to identify and value the costs and benefits that will arise with the proposed project and to compare them with the situation as it would be without the project. The difference is the incremental net benefit arising from the project investment. This approach is not the same as comparing the situation "before" and "after" the project. The before-and-after comparison fails to account for changes in production that would occur without the project and thus leads to an erroneous statement of the benefit attributable to the project investment.

A change in output without the project can take place in two kinds of situations. The most common is when production in the area is already growing, if only slowly, and will probably continue to grow during the life of the project. The objective of the project is to increase growth by intensifying production. In Syria at the time the First Livestock Development Project was appraised, for example, production in the national sheep flock was projected to grow at about 1 percent a year without the project. The project was to increase and stabilize sheep production and the incomes of seminomadic flock owners and sheep fatteners by stabilizing the availability of feed and improving veterinary services. With the project, national flock production was projected to grow at the rate of 3 percent a year. In this case, if the project analyst had simply compared the output before and after the project, he would have erroneously attributed the total increase in sheep production to the project investment. Actually, what can be attributed to the project investment is only the 2 percent incremental increase in production in excess of the 1 percent that would have occurred anyway (see figure 2-1).

A change in output can also occur without the project if production would actually fall in the absence of new investment. In Guyana, on the north coast of South America, rice and sugarcane are produced on a strip of clay and silt soil edging the sea. The coast was subject to erosion from wave action. Under the Sea Defense Project, the government of Guyana has built seawalls to prevent the erosion. The benefit from this project, then, is not increased production but avoiding the loss of agricultural output and sites for housing. A simple before-and-after comparison would fail to identify this benefit (figure 2-2).

In some cases, an investment to avoid a loss might also lead to an increase in production, so that the total benefit would arise partly from the loss avoided and partly from increased production. In Pakistan, many areas are subject to progressive salinization as a result of heavy irrigation and the waterlogging that is in part attributable to seepage from irrigation canals. Capillary action brings the water to the surface where evaporation occurs, leaving the salt on the soil. If nothing is done to halt the process, crop production will fall. A project is proposed to line some of the canals, thus to reduce the seepage and permit better drainage between irrigations. The proposed project is expected to arrest salinization, to save for profitable use the irrigation water otherwise lost to seepage, and to help farmers increase their use of modern inputs. The combination of measures would not only avoid a loss but also lead to an increase in production. Again, a simple before-and-after comparison would fail to identify the benefit realized by avoiding the loss (figure 2-3).

Of course, if no change in output is expected in the project area without the project, then the distinction between the before-and-after comparison and the with-and-without comparison is less crucial. In some projects the prospects for increasing production without new investment are minimal. In the Kemubu Irrigation Project in northeastern Malaysia, a pump irrigation scheme was built that permitted farmers to produce a second rice crop during the dry season. Without the project, most of the area was used for grazing, and with the help of residual moisture or small pumps some was used to produce tobacco and other cash crops. Production was not likely to increase because of the limited amount of water available. With the project now in operation, rice is grown in the dry season. Of course, the value of the second rice crop could not be taken as the total benefit from the project. From this value must be deducted the value forgone from the grazing and the production of cash crops. Only the incremental value could be attributed to the new investment in pumps and canals (figure 2-4).

Another instance where there may be no change in output without the project is the obvious one found in some settlement projects. Without the project there may be no economic use of the area at all. In the Alto Turi Land Settlement Project in northeastern Brazil, settlers established their holdings by clearing the forest, planting upland rice, and then establishing pasture for production of beef cattle. At the time the settlers took up their holdings the forest had not been economically exploited-nor was it likely to be, at least for many years, in the absence of the project. In this case, the output without the project would be the same as the output before the project (figure 2-5).

Direct Transfer Payments

Some entries in financial accounts really represent shifts in claims to goods and services from one entity in the society to another and do not reflect changes in national income. These are the so-called direct transfer payments, which are much easier to identify if our definition of costs and benefits is kept in mind. In agricultural project analysis four kinds of direct transfer payments are common: taxes, subsidies, loans, and debt service (the payment of interest and repayment of principal).

Take taxes, for example. In financial analysis a tax payment is clearly a cost. When a farmer pays a tax, his net benefit is reduced. But the farmer's payment of tax does not reduce the national income. Rather, it transfers income from the farmer to the government so that this income can be used for social purposes presumed to be more important to the society than the increased individual consumption (or investment) had the farmer retained the amount of the tax. Because payment of tax does not reduce national income, it is not a cost from the standpoint of the society as a whole. Thus, in economic analysis we would not treat the payment of taxes as a cost in project accounts. Taxes remain a part of the overall benefit stream of the project that contributes to the increase in national income.

Of course, no matter what form a tax takes, it is still a transfer payment-whether a direct tax on income or an indirect tax such as a sales tax, an excise tax, or a tariff or duty on an imported input for production. But some caution is advisable here. Taxes that are treated as a direct transfer payment are those representing a diversion of net benefit to the society. Quite often, however, government charges for goods supplied or services rendered may be called taxes. Water rates, for example, may be considered a tax by the farmer, but from the standpoint of the society as a whole they are a payment by the farmer to the irrigation authority in exchange for water supplied. Since building the irrigation system reduces national income, the farmer's payment for the water is part of the cost of producing the crop, the same as any other payment for a production input. Other payments called taxes may also be payments for goods and services rendered rather than transfers to the government. A stevedoring charge at the port is not a tax but a payment for services and so would not be treated as a duty would be. Whether a tax should be treated as a transfer payment or as a payment for goods and services depends on whether the payment is a compensation for goods and services needed to carry out the project or merely a transfer, to be used for general social purposes, of some part of the benefit from the project to the society as a whole.

Subsidies are simply direct transfer payments that flow in the opposite direction from taxes. If a farmer is able to purchase fertilizer at a subsidized price, that will reduce his costs and thereby increase his net benefit, but the cost of the fertilizer in the use of the society's real resources remains the same. The resources needed to produce the fertilizer (or import it from abroad) reduce the national income available to the society. Hence, for economic analysis of a project we must enter the full cost of the fertilizer.

Again, it makes no difference what form the subsidy takes. One form is that which lowers the selling price of inputs below what otherwise would be their market price. But a subsidy can also operate to increase the amount the farmer receives for what he sells in the market, as in the case of a direct subsidy paid by the government that is added to what the farmer receives in the market. A more common means to achieve the same result does not involve direct subsidy. The market price may be maintained at a level higher than it otherwise would be by, say, levying an import duty on competing imports or forbidding competing imports altogether. Although it is not a direct subsidy, the difference between the higher controlled price set by such measures and the lower price for competing imports that would prevail without such measures does represent an indirect transfer from the consumer to the farmer.

Credit transactions are the other major form of direct transfer payment in agricultural projects. From the standpoint of the farmer, receipt of a loan increases the production resources he has available; payment of interest and repayment of principal reduce them. But from the stand-point of the economy, things look different. Does the loan reduce the national income available? No, it merely transfers the control over resources from the lender to the borrower. Perhaps one farmer makes the loan to his neighbor. The lending farmer cannot use the money he lends to buy fertilizer, but the borrowing farmer can. The use of the fertilizer, of course, is a cost to the society because it uses up resources and thus reduces the national income. But the loan transaction does not itself reduce the national income; it is, rather, a direct transfer payment. In reverse, the same thing happens when the farmer repays his loan. The farmer who borrowed cannot buy fertilizer with the money he uses to repay the loan his neighbor made, but his neighbor can. Thus, the repayment is also a direct transfer payment.

Some people find the concept of transfer payments easier to understand if it is stated in terms of real resource flows. Taking this approach in economic analysis, we see that a tax does not represent a real resource flow; it represents only the transfer of a claim to real resource flows. The same holds true for a direct subsidy that represents the transfer of a claim to real resources from, say, an urban consumer to a farmer. This line of reasoning also applies to credit transactions. A loan represents the transfer of a claim to real resources from the lender to the borrower. When the borrower pays interest or repays the principal, he is transferring the claim to the real resources back to the lender-but neither the loan nor the repayment represents, in itself, use of the resources.

Costs of Agricultural Projects

In almost all project analyses, costs are easier to identify (and value) than benefits. In every instance of examining costs, we will be asking ourselves if the item reduces the net benefit of a farm or the net income of a firm (our objectives in financial analysis), or the national income (our objective in economic analysis).

Physical goods

Rarely will physical goods used in an agricultural project be difficult to identify. For such goods as concrete for irrigation canals, fertilizer and pesticides for increasing production, or materials for the construction of homes in land settlement projects, it is not the identification that is difficult but the technical problems in planning and design associated with finding out how much will be needed and when.

Labor

Neither will the labor component of agricultural projects be difficult to identify. From the highly skilled project manager to the farmer maintaining his orchard while it is coming into production, the labor inputs raise less a question of what than of how much and when. Labor may, however, raise special valuation problems that call for the use of a shadow price. Confusion may also arise on occasion in valuing family labor. Valuing family labor will be discussed with farm budgets in chapter 4, and the overall question of valuing unskilled labor will be taken up in chapter 7.

Land

By the same reasoning, the land to be used for an agricultural project will not be difficult to identify. It generally is not difficult to determine where the land necessary for the project will be located and how much will be used. Yet problems may arise in valuing land because of the very special kind of market conditions that exist when land is transferred from one owner to another. These valuation problems will also be considered with farm budgets in chapter 4 and with determining economic values in chapter 7.

Contingency allowances

In projects that involve a significant initial investment in civil works, the construction costs are generally estimated on the initial assumption that there will be no modifications in design that would necessitate changes in the physical work; no exceptional conditions such as unanticipated geological formations; and no adverse phenomena such as floods, landslides, or unusually bad weather. In general, project cost estimates also assume that there will be no relative changes in domestic or international prices and no inflation during the investment period. It would clearly be unrealistic to rest project cost estimates only on these assumptions of perfect knowledge and complete price stability. Sound project planning requires that provision be made in advance for possible adverse changes in physical conditions or prices that would add to the baseline costs. Contingency allowances are thus included as a regular part of the project cost estimates.

Contingency allowances may be divided into those that provide for physical contingencies and those for price contingencies. In turn, price contingency allowances comprise two categories, those for relative changes in price and those for general inflation. Physical contingencies and price contingencies that provide for increases in relative costs underlie our expectation that physical changes and relative price changes are likely to occur, even though we cannot forecast with confidence just how their influence will be felt. The increase in the use of real goods and services represented by the physical contingency allowance is a real cost and will reduce the final goods and services available for other purposes; that is, it will reduce the national income and, hence, is a cost to the society. Similarly, a rise in the relative cost of an item implies that its productivity elsewhere in the society has increased; that is, its potential contribution to national income has risen. A greater value is forgone by using the item for our project; hence, there is a larger reduction in national income. Physical contingency allowances and price contingency allowances for relative changes in price, then, are expected-if unallocated-project costs, and they properly form part of the cost base when measures of project worth are calculated.

General inflation, however, poses a different problem. As we will note in chapter 3 in discussing future prices, in project analysis the most common means of dealing with inflation is to work in constant prices, on the assumption that all prices will be affected equally by any rise in the general price level. This permits valid comparisons among alternative projects. If inflation is expected to be significant, however, provision for its effects on project costs needs to be made in the project financing plan so that an adequate budget is obtained. Contingency allowances for inflation would not, however, be included among the costs in project accounts other than the financing plan.

Taxes

Recall that the payment of taxes, including duties and tariffs, is customarily treated as a cost in financial analysis but as a transfer payment in economic analysis (since such payment does not reduce the national income). The amount that would be deducted for taxes in the financial accounts remains in the economic accounts as part of the incremental net benefit and, thus, part of the new income generated by the project.

Debt service

The same approach applies to debt service-the payment of interest and the repayment of capital. Both are treated as an outflow in financial analysis. In economic analysis, however, they are considered transfer payments and are omitted from the economic accounts.

Treatment of interest during construction can give rise to confusion. Lending institutions sometimes add the value of interest during construction to the principal of the loan and do not require any interest payment until the project begins to operate and its revenues are flowing. This process is known as "capitalizing" interest. The amount added to the principal as a result of capitalizing interest during construction is similar to an additional loan. Capitalizing interest defers interest cost, but when the interest payments are actually due, they will, of course, be larger because the amount of the loan has been increased. From the standpoint of economic analysis, the treatment of interest during construction is clear. It is a direct transfer payment the same as any other interest payment, and it should be omitted from the economic accounts. Often interest during construction is simply added to the capital cost of the project. To obtain the economic value of the capital cost, the amount of the interest during construction must be subtracted from the capital cost and omitted from the economic account.

In economic analysis, debt service is treated as a transfer within the economy even if the project will actually be financed by a foreign loan and debt service will be paid abroad. This is because of the convention of assuming that all financing for a project will come from domestic sources and all returns from the project will go to domestic residents. This convention, as noted earlier, separates the decision of how good a project is from the decision of how to finance it. Hence, even if it were expected that a project would be financed, say, by a World Bank loan, the debt service on that loan would not appear as a cost in the economic accounts of the project analysis.

Sunk costs

Sunk costs are those costs incurred in the past upon which a proposed new investment will be based. Such costs cannot be avoided, however poorly advised they may have been. When we analyze a proposed investment, we consider only future returns to future costs; expenditures in the past, or sunk costs, do not appear in our accounts.

In practice, if a considerable amount has already been spent on a project, the future returns to the future costs of completing the project would probably be quite attractive even if it is clear in retrospect that the project should never have been begun. The ridiculous extreme is when only one dollar is needed to complete a project, even a rather poor one, and when no benefit can be realized until the project is completed. The "return" to that last dollar may well be extremely high, and it would be clearly worthwhile to spend it. But the argument that because much has already been spent on a project it therefore must be continued is not a valid criterion for decision. There are cases in which it would be preferable simply to stop a project midway or to draw it to an early conclusion so that future resources might be freed for higher-yielding alternatives.

For evaluating past investment decisions, it is often desirable to do an economic and financial analysis of a completed project. Here, of course, the analyst would compare the return from all expenditures over the past life of the project with all returns. But this kind of analysis is useful only for determining the yield of past projects in the hope that judgments about future projects may be better informed. It does not help us decide what to do in the present. Money spent in the past is already gone; we do not have as one of our alternatives not to implement a completed project.

Tangible Benefits of Agricultural Projects

Tangible benefits of agricultural projects can arise either from an increased value of production or from reduced costs. The specific forms in which tangible benefits appear, however, are not always obvious, and valuing them may be quite difficult.

Increased production

Increased physical production is the most common benefit of agricultural projects. An irrigation project permits better water control so that farmers can obtain higher yields. Young trees are planted on cleared jungle land to increase the area devoted to growing oil palm. A credit project makes resources available for farmers to increase both their operating expenditures for current production-for fertilizers, seeds, or pesticides-and their investment-for a tubewell or a power thresher. The benefit is the increased production from the farm.

In a large proportion of agricultural projects the increased production will be marketed through commercial channels. In that case identifying the benefit and finding a market price will probably not prove too difficult, although there may be a problem in determining the correct value to use in the economic analysis.

In many agricultural projects, however, the benefits may well include increased production consumed by the farm family itself. Such is the case in irrigation rehabilitation projects along the north coast of Java. The home-consumed production from the projects increased the farm families' net benefit and the national income just as much as if it had been sold in the market. Indeed, we could think of the hypothetical case of a farmer selling his output and then buying it back. Since home-consumed production contributes to project objectives in the same way as marketed production, it is clearly part of the project benefits in both financial and economic analysis. Omitting home-consumed production will tend to make projects that produce commercial crops seem relatively high-yielding, and it could lead to a poor choice among alternative projects. Failure to include home-consumed production will also mean underestimating the return to agricultural investments relative to investments in other sectors of the economy.

When home-consumed crops will figure prominently in a project, the importance of careful financial analysis is increased. In this case, it is necessary to estimate not only the incremental net benefit-including the value of home-consumed production and money from off-farm sales-but also the cash available to the farmer. From the analysis of cash income and costs, one can determine if farmers will have the cash in hand to purchase modern inputs or to pay their credit obligations. It is possible to have a project in which home-consumed output increases enough for the return to the economy as a whole to be quite attractive, but in which so little of the increased production is sold that farmers will not have the cash to repay their loans.

Quality improvement

In some instances, the benefit from an agricultural project may take the form of an improvement in the quality of the product. For example, the analysis for the Livestock Development Project in Ecuador, which was to extend loans to producers of beef cattle, assumed that ranchers would be able not only to increase their cattle production but also to improve the quality of their animals so that the average live price of steers per kilogram would rise from S/5.20 to S/6.40 in constant value terms over the twelve-year development period. (The symbol for Ecuadorian sucres is S/.) Loans to small dairy farmers in the Rajasthan Smallholder Dairy Improvement Project in India are intended to enable farmers not only to increase output but also to improve the quality of their product. Instead of selling their milk to make ghee (cooking oil from clarified butter), farmers will be able to sell it for a higher price in the Jaipur fluid milk market. As in these examples, both increased production and quality improvement are most often expected in agricultural projects, although both may not always be expected. One word of warning: both the rate and the extent of the benefit from quality improvement can easily be overestimated.

Change in time of sale

In some agricultural projects, benefits will arise from improved marketing facilities that allow the product to be sold at a time when prices are more favorable. A grain storage project may make it possible to hold grain from the harvest period, when the price is at its seasonal low, until later in the year when the price has risen. The benefit of the storage investment arises out of this change in "temporal value."

Change in location of sale

Other projects may include investment in trucks and other transport equipment to carry products from the local area where prices are low to distant markets where prices are higher. For example, the Fruit and Vegetable Export Project in Turkey included provision for trucks and ferries to transport fresh produce from southeastern Turkey to outlets in the European Common Market. The benefits of such projects arise from the change in "locational value."

In most cases the increased value arising from marketing projects will be split between farmers and marketing firms as the forces of supply and demand increase the price at which the farmer can sell in the harvest season and reduce the monopolistic power of the marketing firm or agency. Many projects are structured to ensure that farmers receive a larger part of the benefit by making it possible for them to build storage

facilities on their farms or to band together into cooperatives, but an agricultural project could also involve a private marketing firm or a government agency, in which case much of the benefit could accrue to someone other than farmers.

Changes in product form (grading and processing)

Projects involving agricultural processing industries expect benefits to arise from a change in the form of the agricultural product. Farmers sell paddy rice to millers who, in turn, sell polished rice. The benefit to the millers arises from the change in form. Canners preserve fruit, changing its form and making it possible at a lower cost to change its time or location of sale. Even a simple processing facility such as a grading shed gives rise to a benefit through changing the form of the product from run-of-the-orchard to sorted fruit. In the Himachal Pradesh Apple Marketing Project in northern India, the value of the apples farmers produce is increased by sorting; the best fruit is sold for fresh consumption while fruit of poorer quality is used to make a soft drink concentrate. In the process, the total value of the apples is increased.

Cost reduction through mechanization

The classic example of a benefit arising from cost reduction in agricultural projects is that gained by investment in agricultural machinery to reduce labor costs. Examples are tubewells substituting for hand-drawn or animal-drawn water, pedal threshers replacing hand threshing, or (that favorite example) tractors replacing draft animals. Total production may not increase, but a benefit arises because the costs have been trimmed (provided, of course, that the gain is not offset by displaced labor that cannot be productively employed elsewhere).

Reduced transport costs

Cost reduction is a common source of benefit wherever transport is a factor. Better feeder roads or highways may reduce the cost of moving produce from the farm to the consumer. The benefit realized may be distributed among farmers, truckers, and consumers.

Losses avoided

In discussing with-and-without comparisons in project analyses earlier in this chapter, we noted that in some projects the benefit may arise not from increased production but from a loss avoided. This kind of benefit stream is not always obvious, but it is one that the with-and-without test tends to point out clearly. In Jamaica, lethal yellowing is attacking the Jamaica Tall variety of coconut. The government has undertaken a large investment to enable farmers to plant Malayan Dwarf coconuts, which are resistant to the disease. Total production will change very little as a result of the investment, yet both the farmers and the economy will realize a real benefit because the new investment prevents loss of income. The Lower Egypt Drainage Project involves the largest single tile drainage system in the world. The benefit will arise not from increasing production in the already highly productive Nile delta, but from avoiding losses due to the waterlogging caused by year-round irrigation from the Aswan High Dam.

Sometimes a project increases output through avoiding loss-a kind of double classification, but one that in practice causes no problem. Proposals to eradicate foot-and-mouth disease in Latin America envision projects by which the poor physical condition or outright death of animals will be avoided. At the same time, of course, beef production would be increased.

Other kinds of tangible benefits

Although we have touched on the most common kinds of benefits from agricultural projects, those concerned with agricultural development will find other kinds of tangible, direct benefits most often in sectors other than agriculture. Transport projects are often very important for agricultural development. Benefits may arise not only from cost reduction, as noted earlier, but also from time savings, accident reduction, or development activities in areas newly accessible to markets. If new housing for farmers has been included among the costs of a project, as is often the case in land settlement and irrigation projects, then among the benefits will be an allowance for the rental value of the housing. Since this is an imputed value, there are valuation problems that will be noted later.

Secondary Costs and Benefits

Projects can lead to benefits created or costs incurred outside the project itself. Economic analysis must take account of these external, or secondary, costs and benefits so they can be properly attributed to the project investment. (Of course, this applies only in economic analysis; the problem does not arise in financial analysis.)

When market prices are used in economic analysis, as has been the custom in the United States for water resource and other public works projects, it is necessary to estimate the secondary costs and benefits and then add them to the direct costs and benefits. This is a theoretically difficult process, and one easily subject to abuse. There is an extensive and complex literature on secondary costs and benefits that specifically addresses this analytical approach. For those who would like to review this literature, a good place to begin is the article by Prest and Turvey (1966), which outlines the historical development of the discussion. A highly technical review of the arguments can be found in Mishan (1971).

Instead of adding on secondary costs and benefits, one can either adjust the values used in economic analysis or incorporate the secondary costs and benefits in the analysis, thereby in effect converting them to direct costs and benefits. This is the approach taken in most project analyses carried out by international agencies, in the systems based on shadow prices proposed in more recent literature on project analysis, and in the analytical system presented here.

Incorporating secondary costs or benefits in project analysis can be viewed as an analytical device to account for the value added that arises outside the project but is a result of the project investment. In the analytical system here, as will be explained in more detail in chapter 7, every item is valued either at its opportunity cost or at a value determined by a consumer's willingness to pay for the item. The effect is to eliminate all transfers-both the direct transfers discussed earlier in this chapter and the indirect transfers that arise because prices differ from opportunity costs. By this means we attribute to the project investment all the value added that arises from it anywhere in the society. Hence, it is not necessary to add on the secondary costs and benefits separately; to do so would constitute double counting.

One qualification must be made. If a project has a substantial effect on the quantity other producers are able to sell in imperfect markets-and most markets are imperfect-there may be gains or losses not accurately accounted for. Squire and van der Tak (1975, p. 23) cite the example of an improved road that diverts traffic from a railway that charges rates below marginal cost. This diversion entails a social gain from reduced rail traffic (in avoiding the social losses previously incurred on this traffic) in addition to the benefits to the road users measured directly. In agricultural projects, this is a rather infrequent case because prices generally are more flexible than in other sectors of the economy. In any event, in the practice of contemporary project analysis the size of these gains or losses is generally assumed to be insignificant, and no provision is made for them in the analysis.

Although using shadow prices based on opportunity costs or willingness to pay greatly reduces the difficulty of dealing with secondary costs and benefits, there still remain many valuation problems related to goods and services not commonly traded in competitive markets. One way to avoid some of these problems is to treat a group of closely related investments as a single project. For example, it is common to consider the output of irrigation projects as the increased farm production, since valuing irrigation water is difficult. Another example is found in development roads built into inaccessible areas. It is argued that the production arising from the induced investment activities of otherwise unemployed new settlers should be considered a secondary benefit of the road investment. One way of avoiding the problem is to view this case as a land settlement project in which the road is a component. New production is then properly included among the direct benefits of the project and can be included in the project accounts at market or shadow prices, and no attempt need be made to allocate the benefits between road investment and the other kinds of investment that must be made by settlers and government if settlement is to succeed.

Another group of secondary costs and benefits has been called "technological spillover" or "technological externalities." Adverse ecological effects are a common example, and the side effects of irrigation development are often cited as an illustration. A dam may reduce river flow and lead to increased costs for dredging downstream. New tubewell development may have adverse effects on the flow of existing wells. Irrigation development may reduce the catch of fish or may lead to the spread of schistosomiasis. When these technological externalities are significant and can be identified and valued, they should be treated as a direct cost of the project (as might be the case for reduced fish catches), or the cost of avoiding them should be included among the project costs (as would be the case for increased dredging or for investment to avoid pollution).

It is sometimes suggested that project investments may give rise to secondary benefits through a "multiplier effect." The concept of the multiplier is generally thought of in connection with economies having excess capacity. If excess capacity exists, an initial investment might cause additional increases in income as successive rounds of spending reduce excess capacity. In developing countries, however, it is shortage of capacity that is characteristic. Thus, there is little likelihood of excess capacity giving rise to additional benefits through the multiplier. In any event, most of the multiplier effect is accounted for if we shadow-price at opportunity cost. Since the opportunity cost of using excess capacity is only the cost of the raw materials and labor involved, only variable costs will enter the project accounts until existing excess capacity is used up.

It is also sometimes suggested that there is a "consumption multiplier effect" as project benefits are received by consumers. Consumption multipliers are very difficult to identify and value. In any case, they presumably would be much the same for alternative investments, so omitting them from a project analysis would not affect the relative ranking of projects.

Intangible Costs and Benefits

Almost every agricultural project has costs and benefits that are intangible. These may include creation of new job opportunities, better health and reduced infant mortality as a result of more rural clinics, better nutrition, reduced incidence of waterborne disease as a result of improved rural water supplies, national integration, or even national defense. Such intangible benefits are real and reflect true values. They do not, however, lend themselves to valuation. How does one derive a figure for the long-term value of a child's life saved, or for the increased comfort of a population spared preventable, debilitating disease? Benefits of this kind may require a modification of the normal benefit-cost analysis to a least-cost type of analysis, a topic we will take up when we discuss valuation. Because intangible benefits are a factor in project selection, it is important that they be carefully identified and, where at all possible, quantified, even though valuation is impossible. For example, how many children will enroll in new schools? How many homes will benefit from a better system of water supply? How many infants will be saved because of more rural clinics?

In most cases of intangible benefits arising from an agricultural project, the costs are tangible enough: construction costs for schools, salaries for nurses in a public health system, pipes for rural water supplies, and the like. Intangible costs, however, do exist in projects. Such costs might be incurred if new projects disrupt traditional patterns of family life, if development leads to increased pollution, if the ecological balance is upset, or if scenic values are lost. Again, although valuation is impossible, intangible costs should be carefully identified and if possible quantified. In the end, every project decision will have to take intangible factors into account through a subjective evaluation because intangible costs can be significant and because intangible benefits can make an important contribution to many of the objectives of rural development.

SECTION II. PRICING PROJECT COSTS AND BENEFITS

Once costs and benefits have been identified, if they are to be compared they must be valued. Since the only practical way to compare differing goods and services directly is to give each a money value, we must find the proper prices for the costs and benefits in our analysis.

Prices Reflect Value

Underlying all financial and economic analysis is an assumption that prices reflect value-or can be adjusted to do so. In this chapter we will discuss how to find these prices. Before proceeding, however, it is necessary to define two economic concepts crucial to project analysis: marginal value product and opportunity cost.

Consider a Filipino farmer who applies nitrogenous fertilizer to his rice. In the 1979-80 season this fertilizer cost him P3.98 per kilogram of elemental nitrogen (N), and he received P1.050 for every kilogram of paddy rice he sold. (The symbol for Philippine pesos is P.) Table 3-1 shows the responsiveness of his rice to fertilizer. At low levels of application, fertilizer has a great effect on rice yield. Increasing the application from no fertilizer to 10 kilograms of elemental nitrogen increased the farmer's

Table 3-1. Crop Response to Nitrogen Fertilizer in the Philippines
 
Paddy rice
Shelled maize
Nitrogen (kgs/ha)
Yield (kgs/ha)
Value
MVP
Yield (kgs/ha)
Value
MVP
0
3,442
3,614
 
2,600
2,688
 
10
3,723
3,909
29.50
2,830
2,926
23.80
20
3,971
4,170
26.10
3,040
3,143
21.70
30
4,187
4,396
22.60
3,230
3,340
19.70
40
4,370
4,588
19.20
3,400
3,516
17.60
50
4,520
4,746
15.80
3,550
3,671
15.50
60
4,637
4,869
12.30
3,680
3,805
13.40
70
4,721
4,957
8.80
3,790
3,919
11.40
80
4,772
5,011
5.40
3,880
4,012
9.30
90
4,791
5,031
2.00
3,950
4,084
7.20
100
4,777
5,016
-1.50
4,000
4,136
5.20
110
 
 
 
4,030
4,167
3.10
120
 
 
 
4,040
4,177
1.00
130
 
 
 
4,030
4,167
-1.00

Personal communication from Pedro R. Sandoval, University of the Philippines at Los Banos, September 1980. Rice responses are based on Changes in Rice Farming in Selected Areas of Asia (Manila: International Rice Research Institute, 1978), p. 61. Maize responses are based on University of the Philippines at Los Banos Experiment Station records. Prices are from the Bureau of Agricultural Economics, Ministry of Agriculture, Republic of the Philippines.

a. The farm-gate price of elemental nitrogen (N) in 1979-80 was F3.98 per kilogram
b. The farmgate price of paddy rice in 1979-80 was F1.050 per kilogram.
c. The marginal value product is the extra revenue that comes from increasing the quantity of an input used by one unit, all other quantities remaining constant. In this instance, the marginal value product is the increased value of paddy rice or shelled maize from using 1 additional kilogram of elemental nitrogen. Note that in this table the interval between levels of elemental nitrogen is 10 kilograms. Thus, the marginal value product of elemental nitrogen applied to rice between the 60- and 70-kilogram levels of application is the difference in value of output between the two levels divided by 10, or P8.80 [(4,957 - 4,869) - 10 = 8.80].
d. The farm-gate price of shelled yellow maize in 1979-80 was P1.034 per kilogram.
e. Beyond application of 100 kilograms of elemental nitrogen, all marginal value products for paddy rice are negative; therefore, figures for these applications of nitrogen to rice are not reported.

yield from 3,442 kilograms to 3,723 kilograms per hectare and increased the value of his output by P295, from P3,614 to P3,909. Thus, for every additional kilogram of elemental nitrogen the farmer applied at this level, he received P29.50 in return [(3,909 - 3,614) _ 10 = 29.50]. The extra revenue from increasing the quantity of an input used, all other quantities remaining constant, is the marginal value product of the input. In this case, then, the marginal value product of a kilogram of fertilizer is P29.50.

If the farmer could buy fertilizer for P3.98 a kilogram and use it to increase output by PP9.50, it obviously would have paid him to apply more. But as the intensity of application increases, each additional kilogram of fertilizer has less and less effect on production. If the farmer had increased his application from 80 to 90 kilograms per hectare, he would have increased the value of his production by only P20, from P5,011 to P5,031, and the marginal value product of a kilogram of fertilizer would have fallen to only P2.00 [(5,031 - 5,011) - 10 = 2.00]. Since he would have had to pay P3.98 per kilogram, it clearly would not have been worthwhile to apply fertilizer at this rate. In fact, it would only have paid the farmer to apply fertilizer up to the rate at which the marginal value product just equaled the price. For this Filipino farmer, it would have paid him to apply approximately 80 kilograms of nitrogen: between 70 and 80 kilograms the marginal value product of each additional kilogram was some P5.40, whereas between 80 and 90 kilograms it fell to P2.00. Thus, the farmer would have expanded his fertilizer use until he reduced the marginal value product of the fertilizer to its market price, and the market price, therefore, is an estimate of the marginal value product of the fertilizer.

The optimal amount of fertilizer to use will change, of course, when the price of fertilizer changes relative to the price of rice. If the relative price of fertilizer were to rise, the farmer would respond by reducing the amount of fertilizer he applies, increasing the marginal value product of the fertilizer (but reducing the total amount and value of production) until the marginal value product of the fertilizer again just equals its price. Suppose fertilizer were to double in price to P8.00 per kilogram of elemental nitrogen, and rice prices remained unchanged. Then, table 3-1 indicates the farmer should reduce the amount of fertilizer applied to a hectare from 80 kilograms to 70 kilograms, since between 60 and 70 kilograms the marginal value product was some P8.80 but between 70 and 80 kilograms it was only some P5.40.

In practice, because of risk and limited resources, the farmer would probably not have applied the amounts indicated here. We may consider that the farmer reduces his expected return by some "risk discount." Even so, the principle we are illustrating remains the same: the farmer equates the expected marginal value product less some risk discount to the price of fertilizer.

If this farmer also grew maize, for which in 1979-80 he would have received P1.034 per kilogram of shelled grain, table 3-1 indicates it would have paid him (in the absence of risk) to apply some 100 kilograms of elemental nitrogen to each hectare, because between 90 and 100 kilograms the marginal value product of a kilogram of nitrogen applied to maize was P5.20, whereas between 100 and 110 kilograms the marginal value product fell to P3.10, below the price of fertilizer.

Now, suppose the farmer had limited resources and could not obtain sufficient credit to increase his fertilizer application on both rice and maize to where the marginal value product equaled the price. Suppose the farmer had only 2 hectares, 1 planted in rice and 1 in maize, and resources sufficient to purchase just 80 kilograms of nitrogen. How should he have used it? Should he have put it all on rice and none on maize? If he did, he would have applied fertilizer to his rice at the level where the marginal value product was just about equal to its market price. But suppose he had shifted some fertilizer, instead, to maize. If he had shifted 10 kilograms, he would have reduced the value of his rice production by P54- from P5,011 to P4,957, or by P5.40 for each kilogram shifted-but he could have obtained some P238 for the 10 kilograms applied to maize, since the marginal value product between 0 and 10 kilograms was some P23.80 per kilogram. In other words, at these levels each kilogram of nitrogen shifted would reduce the rice value by P5.40 but increase the value of maize output by some P23.80. In the language of economics, the opportunity cost of fertilizer shifted from rice to maize was P5.40. Opportunity cost, thus, is the benefit forgone by using a scarce resource for one purpose-in this case applying fertilizer to maize-instead of for its best alternative use-in this case using the fertilizer to produce rice. Said another way, the opportunity cost is the return a resource can bring in its next best alternative use. What would be the opportunity cost if the farmer were to move a kilogram of fertilizer in the other direction, back from maize to rice? He would have given up P23.80 to gain only P5.40-not a very attractive proposition-and the opportunity cost, obviously, would be some P23.80.

Given his limited resources, it would pay the farmer to shift fertilizer from rice to maize until the marginal value product of fertilizer applied to both crops is the same. In the case of the Filipino farmer who could buy only 80 kilograms of fertilizer, if on the one hand he were to move 40 kilograms to maize, reducing his application on rice from 80 kilograms to 40 kilograms, he would have increased the marginal value product of the fertilizer on his rice to some P15. On the other hand, the 40 kilograms shifted away from rice and put on maize would have decreased the marginal value product of nitrogen applied to maize also to about P15. At these levels, there would be no advantage in shifting fertilizer between the two crops-the opportunity cost of shifting more fertilizer from rice to maize would be about P15, but the gain would also be only about P15-and the farmer would have reached the optimal level of application to both crops.

Note, however, that if the farmer could somehow have bought as much fertilizer as he wanted at the market price of P3.98 per kilogram-perhaps through a credit program-then the market price of fertilizer would have become its opportunity cost, and (in the absence of a risk discount) he should have increased his application to 80 kilograms on rice and 100 kilograms on maize.

From a single farmer to the economy as a whole, the same principles apply. In a "perfect" market-one that is highly competitive, with many buyers and sellers, all of whom have perfect knowledge about the market-every economic commodity would be priced at its marginal value product, since every farmer will have expanded his fertilizer use to where its marginal value product equals its price, and the same will have happened for every other item in the economy. That is, the price of every good and service would exactly equal the value that the last unit utilized contributes to production, or the value in use of the item for consumption would exactly balance the value it could contribute to additional production. If a unit of goods or services could produce more or bring greater satisfaction in some activity other than its present use, someone would have been willing to bid up its price, and it would have been attracted to the new use. When this price system is in "equilibrium," the marginal value product, the opportunity cost, and the price will all be equal. Resources will then have been allocated through the price mechanism so that the last unit of every good and service in the economy is in its most productive use or best consumption use. No transfer of resources could result in greater output or more satisfaction.

Without moving further into price theory, we can consider some direct implications for agricultural projects of the assumption that prices reflect value.

First, as everyone knows, markets are not perfect and are never in complete equilibrium. Hence, prices may reflect values only imperfectly. Even so, there is a great deal of truth in this price theory based on the model of perfect markets. In general, the best approximation of the "true value" of a good or service that is fairly widely bought and sold is its market price. Somebody in the economy is willing to pay this price. One can presume that this buyer will use the item to increase output by at least as much as its price, or that he is willing to exchange something of value equal to the price to gain the satisfaction of consuming the item. Hence, the market price of an item is normally the best estimate of its marginal value product and of its opportunity cost, and most often it will be the best price to use in valuing either a cost or a benefit. In financial analysis, as we have noted, the market price is always used. But in economic analysis some other price-a "shadow price"-may be a better indicator of the value of a good or service; that is, a better estimate of its true opportunity cost to the economy. When prices other than market prices are used in economic analysis, however, the burden of proof is on the analyst.

Finding Market Prices

Project analyses characteristically are built first by identifying the technical inputs and outputs for a proposed investment, then by valuing the inputs and outputs at market prices to construct the financial accounts, and finally by adjusting the financial prices so they better reflect economic values. Thus, the first step in valuing costs and benefits is finding the market prices for the inputs and outputs, often a difficult task for the economist.

To find prices, the analyst must go into the market. He must inquire about actual prices in recent transactions and consult many sources-farmers, small merchants, importers and exporters, extension officers, technical service personnel, government market specialists and statisticians, and published or privately held statistics about prices for both national and international markets. From these sources the analyst must come up with a figure that adequately reflects the going price for each input or output in the project.

Point of first sale and farm-gate price

In project analysis, a good rule for determining a market price for agricultural commodities produced in the project is to seek the price at the "point of first sale." If the point of first sale is in a relatively competitive market, then the price at which the commodity is sold in this market is probably a relatively good estimate of its value in economic as well as financial terms. If the market is not reasonably competitive, in economic analysis the financial price may have to be adjusted better to reflect the opportunity cost or value in use of the commodity.

For many agricultural projects in which the objective is increased production of a commodity, the best point of first sale to use is generally the boundary of the farm. We are after what the farmer receives when he sells his product-the "farm-gate" price. The increased value added of the product as it is processed and delivered to a market arises as a payment for marketing services. This value added is not properly attributed to the investment to produce the commodity. Rather, it arises from the labor and capital engaged in the marketing service. Usually the price at point of first sale can be accepted as the farm-gate price; even if this point is in a nearby village market, the farmer sells his output there and thus earns for himself any fee that might be involved in transporting the commodity from the farm to the point of first sale. But if any new equipment is necessary to enable the farmer to do this-say, a new bullock cart or a new truck-then that new equipment must be shown as a cost incurred to realize the marketing benefit in the project.

In projects producing commodities for well-organized markets, the farm-gate price may not be too difficult to determine. This would be true for most food grains traded domestically in substantial quantities. One may think of wheat in most countries of the Middle East and South Asia, of rice in South and Southeast Asia, and of maize in much of Latin America. It would also be true of farm products for which the processor is generally the first buyer (such as fresh fruit bunches for palm oil in Malaysia or milk in Jamaica), where the price quoted to the farmer is the price on his farm, and the firm responsible for the marketing comes to the farm to pick up the product.

In many cases, however, the prices in a reasonably competitive market or in the price records kept by the government statistical service will include services not properly attributable to the investment in the project itself. This may happen, for instance, when the only price series available for a product records the prices at which it has been sold in a central market-such as the price for eggs in Madras, for melons in Tehran, or for vegetables in Bogota. In that case, the project analyst will have to dig deeper to find out how to value the marketing services. Then he can adjust the central market price to reduce it to the farm-gate price.

The farm-gate price is generally the best price at which to value home-consumed production. In some cases it may be extremely difficult to determine just what a realistic farm-gate price is for a crop produced primarily for home consumption because so little of the crop appears on markets. This is the case, for example, for manioc and cocoyam in Africa. On the one hand, some argue that the true value of the crop is overstated if the market price is used as a basis for valuation because such a small proportion of the product is actually sold. On the other hand, the same crop in different situations may not be so difficult to value. Manioc is sold extensively in Nigeria to make gari flour, and it is commonly traded in local markets in tropical Latin America and the Caribbean.

The farm-gate price may be a poor indicator of the true opportunity cost we want to use in economic analysis. In Ghana the Marketing Board takes some proportion of the cocoa price as a tax for development purposes. In Thailand, a rice "premium"-that is, a tax on rice exports-effectively keeps the domestic price well below what the international market would pay. In these cases, when the commodity is traded its economic value would have to be considered higher than the actual farm-gate price, and this price distortion will have to be corrected in the economic analysis. In other cases, just the opposite happens. In Mexico the price of maize is maintained at a high level to transfer income to ejidatarios, the small farmers. In Malaysia, the price of rice is supported above world market levels to encourage local production and to reduce imports. In these cases, part of the price does not really reflect the economic value of the product-its cost if it could be imported-but rather an indirect income transfer to small farmers. Again, this price distortion will have to be corrected in the economic analysis.

Pricing intermediate goods

By emphasizing the point of first sale as a starting point for valuing the output of our projects, we are also implying that imputed prices should be avoided for intermediate goods in our analysis. An intermediate good is an item produced primarily as an input in the production of another good. If an intermediate good is not freely traded in a competitive market, we cannot expect to obtain a price established by a range of competitive transactions. Fodder produced on a farm and then fed to the dairy animals on the farm is an example of such an intermediate product. If increased fodder production is an element in the proposed agricultural project, the analyst would avoid valuing it. Instead, the analyst would treat the whole farm as a unit and value the milk produced at its point of first sale or value the calves sold as feeder cattle. Treatment of intermediate products will vary from project to project depending on the particular marketing structures. In some countries it would hardly make sense in an egg production project to value the pullets produced in a pullet production enterprise and then "sell" these pullets to the egg production enterprise on the same farm. But in other countries there might be an active market in pullets, which would mean that we could expect to find a reasonably competitive price to use in the economic analysis. To avoid most of the problems that might be introduced by trying to impute values for intermediate products, the financial accounts in agricultural projects are based on budgets for the whole farm instead of on budgets for individual activities on the farm; that is, on the budget for the egg farm as a whole rather than on the budget for a pullet production activity.

A frequently encountered intermediate good in agricultural projects is irrigation water. The "product" of an irrigation system-water-is, of course, really intended to produce agricultural commodities. The price farmers are charged for the water is generally determined administratively, not by any play of competitive market forces. If the analyst were to try to separate the irrigation system from the production it makes possible, he would be faced with a nearly impossible task of determining the value of irrigation water. Hence, it is not surprising that the economic analyses of most irrigation projects take as the basis for the benefit stream the value of the agricultural products that are offered in a relatively free market at the point of first sale.

Other problems in finding market prices

Considerable confusion often arises in determining the values for two important inputs in agricultural projects, land and labor. This happens primarily when the analysis moves from the financial project accounts to the economic analysis (to which we will turn in chapter 7). In the accounts prepared for the financial analysis, the treatment of prices for land and labor is quite straightforward: the price used is the price actually paid. Thus, if the farmers in a settlement project are expected to pay the project authority a price for the land they acquire, perhaps through a series of installments, then the actual price in the year it is paid is entered in the project accounts. In the financial analysis, we do not question whether this is a "good" price in economic terms. Similarly, if land must be bought for the right-of-way for canals in an irrigation project, the actual price to be paid is entered in the project accounts in the financial analysis. Or, if the project includes tenant farmers who will receive help in increasing wheat production, then in the financial accounts for these tenant farmers the analyst will enter the rent paid each year at the amount actually paid, or at the farm-gate value of the wheat delivered to the landowner if the tenants pay rent in kind.

If farm accounts are laid out on a with-and-without basis following the format suggested in chapter 4, in those instances where the project involves only changing the cropping pattern (say, a shift from pasture to irrigated sorghum), the cost of the land (in this instance an opportunity cost) need not be separately entered because of the form of the account. When the net benefit without the project is subtracted from the net benefit with the project, the contribution of the land to the old cropping pattern is also subtracted and only the incremental value remains.

In valuing labor for the financial analysis accounts, again, the problems arise when the financial accounts are adjusted to reflect economic values. For financial analysis, the analyst enters the amounts actually paid to hired labor, either in wages or in kind, in the farm budgets or project accounts. Family labor is treated differently. It is not entered as a cost; instead, the "wages" for the family become a part of the net benefit. Thus, if our project increases the net benefit, it also in effect increases the family's income or "wages" for its labor. Again, if we follow the format suggested in chapter 4, the account will automatically value the family labor at its opportunity cost, and the incremental net benefit will reflect any increased return the family may receive for its labor.

Prices for agricultural commodities generally are subject to substantial seasonal fluctuation. If this is the case, some decision must be made about the point in the seasonal cycle at which to choose the price to be used for the analysis. A good starting point is the farm-gate price at the peak of the harvest season. This is probably close to the lowest price in the cycle. The line of reasoning here is that as prices rise during the cycle at least some part of that rise is a result not of the production activities of the farmer but of the marketing services embodied in storing the crop until consumers want it. But, markets being what they are, there may be an element of imperfection in the harvest price level. Market channels may become so glutted that merchants try actively to discourage farmers from immediately bringing their crop to the market by offering a price that even the merchants themselves would admit is too low. Even so, the need to sell immediately to meet debt obligations may force farmers to offer their crops despite these artificially low, penalty prices. In some cases, therefore, a price higher than the farm-gate price in the harvest season may be selected. But there is an obligation here to justify the price chosen as more valid than the lowest seasonal price. One way to resolve this problem may be to include an element of credit in the project design. This would permit farmers to withhold their product from the market until prices have had a chance to rise from their seasonal lows but at the same time to have enough money to meet their cash obligations and family living expenses. The credit element may also include credit for building on-farm storage so that farmers will have a safe place to store their production until they decide to market it at a better price.

Prices vary among grades of product, of course, and picking the proper price for project analysis may involve making some decisions about quality of the product. In general, it can be assumed that farmers will produce in the future much the same quality as they have in the past and will market their product ungraded. In many agricultural projects, however, one objective is to upgrade the quality of production as well as to increase the total output. Small dairy farmers, for instance, may be able with the help of the project investment to meet the sanitation standards of the fluid milk market and to command a higher price; or reduced time for delivery may hold down sucrose inversion in sugarcane; or better pruning will increase the average size of the oranges Moroccan farmers can offer European buyers. In such cases, the proper price to select is the average price expected for the quality to be produced.

A special problem occurs in pricing housing. If project investment includes housing construction, as would be the case for a settlement project, then one benefit arising from the investment is the rental value of the house. Since the rental value will usually be an imputed value rather than a real market price, care must be exercised in determining it. No more should be allowed for the rental value than would normally be paid by a prospective tenant family. Nor should more rental value be allowed than the family would be expected to pay for a comparable house in the vicinity or in a similar area elsewhere (if the new settlement is in a distant locale). In particular, the temptation should be avoided to take as a rental value some arbitrary proportion of the housing cost. Otherwise, overly elaborated housing construction might be justified simply by assigning it an unrealistically high imputed value.

Project boundary price

Prices used in analyzing agricultural projects are not necessarily farm-gate prices. The concept of a farm-gate price may be expanded to a "project boundary" price if a project has a marketing component or if it is a purely marketing project. Many projects have a marketing component, perhaps because there is no competitive channel reaching down to the farm-gate level for the unprocessed product. Of concern in these projects are both the farm-gate price (on which to base the estimates of the net benefit to the farmer) and the price at which the processed product is sold in the market (after being handled in the facilities financed by the project). Such a case is found in the Rahad project in the Sudan. There the Roseires dam on the Blue Nile will provide irrigation water for the production of cotton, which will be ginned in new facilities financed by the project. The analyst, of course, is interested in the price of cotton paid to the farmers so that their incomes can be estimated. But, since this price is set administratively, it could not be used directly in the economic analysis of the project. The analyst is also interested in the price of ginned cotton because that is the first product the project will actually sell in a reasonably competitive market. In this case, the point of first sale is f.o.b. (free on board) Port Sudan, and the price there becomes the basis for the benefit stream.

Predicting Future Prices

Since project analysis is about judging future returns from future investment, as analysts we are immediately involved in judging just what future prices may be. This is a matter of judgment, not mechanics. No esoteric mathematical model exists to come to the aid of the project analyst; like everyone else he must take into consideration all the facts he can find, seek judgments from those he respects, and then come to a conclusion himself. It tends to be a rather unsettling process. The only consolation is that careful, considered judgment about the course of future prices is better than giving the matter no thought at all and wasting scarce resources on incompletely planned projects.

We have been discussing how to find market prices, and it is from these current prices that we begin. The best initial guess about future prices is that they will retain the present relationships, or perhaps the average relationship they have borne to each other over the past few years. We must consider, however, whether these average relationships will change in the future and how we will deal with a general increase in the level of prices owing to inflation.

Changes in relative prices

We may first raise the question of whether relative prices will change. Will some inputs become more expensive over time in relation to other commodities? Will some prices fall relatively as supplies become more plentiful? Not easy questions to deal with, but some approaches to answers can be made. In financial analysis, of course, a change in a relative price means a change in the market price structure that producers face either for inputs or for outputs. A change in a relative price, then, is reflected directly in the project's financial accounts. A rise in the relative price of fertilizer reduces the incremental net benefit-the amount the farm family has to live on. It is thus clearly a cost in the farm account. The same line of reasoning can be applied in the financial analysis for any other group participating in the project.

A change in the relative price of an item implies a change in its marginal productivity-that is, a change in its marginal value product-or a change in the satisfaction it contributes when it is consumed. In economic analysis, where maximizing national income is the objective, a change in the relative price of an input implies a change in the amount that must be forgone by using the item in the project instead of elsewhere in the economy; it is therefore a change in the contribution the output of the project makes to the national income. Thus, changes in relative prices have a real effect on the project objective and must be reflected in project accounts in the years when such changes are expected.

There are several kinds of commodities subject to future changes in relative prices. Most agricultural project analysts would probably agree that the relative price of energy-intensive agricultural inputs is likely to continue to rise over the next several years, just as it has done over the past few years. Thus, on the input side the project accounts might show an annual increase, at least for the first decade or so, in the cost of fuel for tractors, for transporting the harvested crop, for drying grain, and for such petroleum-based inputs as fertilizers and chemical pesticides. On the output side, there may be some commodities that will probably continue to be in short supply and whose prices will rise as incomes increase-one might think of mutton from fat-tailed sheep in Iran, or, for that matter, of most meat products worldwide. How much will prices increase relative to those of other products? Certainly a difficult question, but one the project analyst must confront. For a range of products-from industrial crops such as fibers or oilseeds to food grains and vegetables-judgments will have to be made on the best possible basis.

In some countries, relative wages of rural labor may rise as economic development proceeds during the life of a project. This will have implications not only for the prices assumed for hired labor, but also for the incentive effect exerted by a given change in net benefit and for the technology assumed as a basis for projections in the farm budgets and project accounts.

Inflation

In the past few years, virtually every country has experienced inflation, and the only realistic assessment is that this will continue. No project analyst can escape deciding how to deal with inflation in his analysis.

The approach most often taken is to work the project analysis in constant prices. That is, the analyst assumes that the current price level (or some future price level-say, for the first year of project implementation) will continue to apply. It is assumed that inflation will affect most prices to the same extent so that prices retain their same general relations. The analyst then need only adjust future price estimates for anticipated relative changes, not for any change in the general price level. By comparing these estimates of costs and benefits with the constant prices, he is able to judge the effects of the project on the incomes of participants and its income-generating potential for the society as a whole. Although the absolute (or money) values of the costs and benefits in both the financial and the economic analyses will be incorrect, the general relations will remain valid, and so the measures of project worth discussed in chapter 9 may be applied directly. Working in constant prices is simpler and involves less calculation than working in current prices; for the latter, every entry has to be adjusted for anticipated changes in the general price level.

It is quite possible, however, to work the whole project analysis in current prices. This has the advantage that all costs and benefits shown would be estimates of what the real prices will be in each year of the project. Furthermore, estimates of investment costs will be in current terms for the year in which they are expected to occur, so that the finance ministry can more easily anticipate these needs and budget the amounts necessary to finance the project on schedule. The problem in this approach is that it involves predicting inflation rates. For items to be imported, some help is available in the World Bank report on Price Prospects for Major Primary Commodities (1982a), which is published biennially and updated in six-month intervals and includes an estimate of inflation in developed countries. For domestic inflation rates in developing countries, other sources will have to be consulted, but obtaining an estimate in which one can place even minimal confidence will be difficult, to say the least. Even casting the project analysis in current terms may raise problems for the project analyst. Many governments have policy goals that call for greatly reduced inflation, and they cannot permit the circulation of official documents that assume rapid inflation will continue.

The mere mechanics of using current prices presents no analytical problem in project analysis, although it does complicate the computations. When we consider measures of project worth, some means of deflating future prices must be adopted for comparing future cost and benefit streams in terms that are free from the effects of general price increases. We will illustrate the methodology in chapter 10 in the section "Calculating Measures of Project Worth Using Current Prices."

Even when constant prices are used in the more conventional approach to project analysis, a table estimating the budgetary effects of the project in current terms that will prevail at least during the investment phase should be included either in the analysis or as a separate memorandum. It would list in current prices domestic currency needs, foreign exchange requirements, and subsidies. The finance ministry would then have better estimates to work with, and delays because of budgetary shortfalls could more easily be avoided.

Prices for Internationally Traded Commodities

For commodities that enter significantly in international trade, whether inputs or outputs, project analysts usually obtain price information from various groups of specialists who follow price trends and make projections about relative prices in the future. In many countries where agricultural exports are important, there are groups in the agriculture ministry or the finance ministry whose help may be sought.

There are also several international organizations and trade groups to which the analyst may turn. The World Bank, for instance, publishes its projections under the title Price Prospects forMajorPrimary Commodities. The Food and Agriculture Organization (FAO) sponsors intergovernmental groups that publish price information on rice; grains (other than rice); citrus; hard fibers; fibers (other than hard fibers); oilseeds, oils and fats; bananas; wine and wine products; tea; meat; and cocoa. Information may be obtained from the secretary of the relevant intergovernmental group at the FAo headquarters in Rome or from the FAo representative in individual countries.

Several international commodity organizations keep detailed price information for the products of their interest. These include the International Tea Committee, the International Cocoa Organization, the International Wool Secretariat, the International Coffee Organization, the International Association of Seed Crushers, the International Rubber Study Group, and the International Sugar Organization, all with headquarters in London; the International Olive Oil Council in Madrid; and the International Cotton Advisory Committee in Washington.

Some individual nations systematically collect production and price information for crops and livestock products of interest to them, and they often are willing to share this information with analysts in other countries without charge or restriction. The United States Department of Agriculture-probably the most important of these-publishes detailed studies about most major crops traded in international markets. Information may be obtained from agricultural attaches in American embassies, or directly from the department's Foreign Agriculture Service. The Commonwealth Secretariat in London publishes information about price trends for commodities of interest to its member nations. A detailed list of "Sources of Information on World Prices" is available from the World Bank (Woo 1982).

Financial Export and Import Parity Prices

In projects that produce a commodity significant in international trade, the price estimates are often based on projections of prices at some distant foreign point. The analyst must then calculate the appropriate price to use in the project accounts, either at the farm gate or at the project boundary.

If the farm-gate or project boundary prices for the internationally traded commodities in the project are already known, and the prices in the particular country tend to follow world market prices, the farm-gate prices may be adjusted by the same relative amount as indicated, say, by the medium trend projected in the future relative prices supplied by one or another international organization. Also, in financial analysis, if the farm-gate price is set administratively and is not allowed to adjust freely to world prices, the relevant price to use is the administratively set price.

Simply adjusting domestic prices by the same relative amount as foreign prices often arrives at figures too rough for project analysis. The approach ignores the fact that marketing margins in commodity trade tend to be less flexible than the commodity prices themselves. There are also many instances in estimating the economic value of a traded commodity that involve deriving a shadow price based on international prices. In such instances it is necessary to calculate export or import parity prices. (See chapter 7, the subsection "Economic export and import parity values.") These are the estimated prices at the farm gate or project boundary, which are derived by adjusting the c.i.f. (cost, insurance, and freight) or f.o.b. prices by all the relevant charges between the farm gate and the project boundary and the point where the c.i.f. or f.o.b. price is quoted. The elements commonly included in c.i.f. and f.o.b. are given in table 3-2.

Table 3-2. Elements of C.i.f. (Cost, Insurance, Freight) and F.o.b. (Free on Board)
Item
Element
C.i.f.
Includes:
 
F.o.b. cost at point of export
Freight charges to point of import
Insurance charges
Unloading from ship to pier at port
 
Excludes:
 
Import duties and subsidies
 
Port charges at port of entry for taxes, handling, storage, agents’ fees and the like
F.o.b.
Includes:
 
All costs to get goods on board-but still in harbor of exporting country:
Local marketing and transport costs
Local port charges including taxes, storage, loading, fumigation, agents’ fees and the like
Export taxes and subsidies
Project boundary price
Farm-gate price

Source: William A. Ward, “Calculating Import and Export Parity Prices,” training materials of the Economic Development Institute, CN-3 (Washington, D.C.: World Bank, 1977), p.8.

One common case for which an export parity price has to be calculated is that of a commodity produced for a foreign market. Table 3-3 gives an example based on the Rahad project in the Sudan. It shows the generalized elements for calculating export parity prices so that the same methodology can be applied in other cases. As noted earlier, the Rahad project included cotton gins. Since the gins produce lint and cottonseed for export and scarto, a by-product of very short fibers not suitable for export and sold locally, the analyst needed three prices. For the lint and seed estimates, he began with forecasts of the 1980 c.i.f. prices in current terms at Liverpool, which were available from World Bank publications. From these c.i.f. prices, he then deoucted insurance, ocean freight, export duties, port handling costs, and rail freight from the cotton gin at the project site to Port Sudan, thus obtaining the export parity prices at the project boundary: LSd 178.650 for lint and LSd18.097 for seed. (The symbol for Sudanese pounds is LSd.) The price for scarto, which was not exported, was based on the prevailing domestic price.

To illustrate, we may continue to calculate the export parity price at the farm gate, although in the Rahad example, where the farm-gate price was set administratively, this calculation was not made. The computations are laid out in the part of table 3-3 that continues from the entry for "Equals export parity price at project boundary." Here a new issue arises. The three products that the gin produces-lint, seed, and scarto-must be converted into their seed cotton equivalents, since it is seed cotton that the farmer sells. Similar conversions have to be made in many other instances-for example,

,

Table 3-3. Financial Export Parity Price for Cotton, Rahad
Irrigation Project, Sudan (1980 forecast prices)
Step in the calculation
Relevant step in the Sudanese example
Value per ton
 
 
Lint
Seed
Scarto
C.i.f. at point of import
C.i.f. Liverpool (taken as estimate for all European ports)
US$639.33
US$103.39
 
Deduct unloading at point of import
Deduct freight to point of import
Deduct insurance
Freight and insurance
 
 
 
F.o.b. Port Sudan
39.63
24.73
 
Equals f.o.b. at point of export
 
F.o.b. Port Sudan
US 599.70
US$ 78.66
 
Convert foreign currency to domestic currency at official exchange rate
Converted at official exchange rate of LSd 1.000=US$2.872
LSd 208.809
LSd 27.389
 
Deduct tariffs
Export duties
17.813
1.000
 
Add subsidies
(None)
 
 
 
Deduct local port charges
Port handling cost
Lint: LSd 5.564 per ton
Seed: LSd 1.510 per ton
5.564
1.510
 
Deduct local transport and marketing costs from project to point of export (if not part of the project cost)
Freight to Port Sudan at LSd 6.782 per ton
6.782
6.782
 
Equals export parity price at project boundary
Export parity price at gin at project site
LSd 178.650
LSd 18.097
 

Conversion allowance (if necessary)

Convert ot seed cotton (LSd 178.650 x 0.4 + LSd 18.097 x 0.59 + LSd 110.200 x 0.01)
71.460
10.677
1.102
Deduct local storage, transport, and marketing costs (if not part of project cost)
Ginning, baling, and storage (LSd 15.229 per ton)
 
15.229
 
 
Collection and internal transfer (LSd 1.064 per ton)
 
1.064
 
Equals export parity price at farm gate
Export parity price at the farm gate
LSd 66.946

LSd Sudanese pounds. US$ U.S. dollars
Source: Adapted from World Bank, “Appraisal of the Rahad Irrigation Project,” PA-139b (Washington, D.C., 1973; restricted circulaton) annex 16, table 6. The format of the table is adapted from Ward, “Calculating Import and Export Parity Prices,” p.9
a. Scarto is a by-product of very short, soiled fibers not suitable for export and is sold locally at a price of LSd 110.200 per ton.
b. Seedcotton is converted into lint, seed, and scarto assuming 1 ton of seed cotton yields 400 kilograms seed, and 10 kilograms scarto.

rice milling or groundnut decortication. For the Rahad project, a weighted price of LSd83.239 forthe seed cotton was calculated using a ginning outturn of 40 percent lint, 59 percent seed, and 1 percent scarto. From this weighted price were deducted the ginning, baling, and storage charges and the costs of collection and transport from the farm gate to the gin, thus arriving at the farm-gate export parity price of LSd66.946.

A parallel computation leads to the import parity price. Here the issue is the price at which an import substitute can be sold domestically if it must compete with imports. Table 3-4 illustrates this issue with the example of maize production in Nigeria. The same example is presented diagrammatically in figure 3-1. Nigeria is a net maize importer, and the project is to produce maize for domestic consumption to replace imported maize.

Table 3-4. Financial Import Parity Price of Early-crop Maize, Central
Agricultural Development Projects, Nigeria
Steps in the calculation
Relevant steps in the Nigerian example
Value
per ton
F.o.b. at point of export
F.o.b. U.S. Gulf ports, No. 2 U.S. yellow corn in bulk
US$116
 
Freight and insurance
31
Add freight to point of import
Add unloading at point of import
Add insurance
Equals c.i.f. at point of import
C.i.f. Lagos or Apapa
US$147
Convert foreign currency to domestic currency at official exchange rate
Add tariffs
Deduct subsidies
Add local port charges
Converted at official exchange rate of N1 = US$1.62
none
none
Landing and port charges (including the cost of bags)
N91
 
 
 
22
Add local transport and marketing costs to relevant markets
Transport (based on a 350-kilometer average)
18
Equals price at market
Wholesale price
N 131
Conversion allowance if necessary
(Not necessary)
 
Deduct transport and marketing costs to relevant market
Primary marketing (includes assembly, cost of bags, and intermediary margins)
-14
 
Transport (based on a 350-kilometer average)
- 18
Deduct local storage, transport, and marketing costs (if not part of the project costs)
Storage loss (10 percent of harvested weight
- 9
Equals import parity price at farm gate
Imported parity price at farm gate
N 90

N = Nigerian naira
Source: Adapted from World Bank, “Supplementary Annexes to Central Agricultural Development Projects,” 1370-UNI (Washington, D.C., 1976; restricted circulsation), supplement 11, appendix 2, table 4. The format of the table is adpated from Ward, “Calculating Import and Export Parity Prices,” p. 10.
a. Forecast from World Bank, Price Prospects for Major Primary commodities, 814/76 (Washington, D.C., 1976), annex 1, p. 12.

We begin with the f.o.b. price at the point of export-in this case U.S. ports on the Gulf of Mexico-derived from World Bank commodity estimates. To this we add freight and insurance to obtain the c.i.f. price at either Lagos or Apapa, the two Nigerian ports concerned. Then we would add any tariffs and subsidies (in this case there are none); add local port charges for harbor dues, fumigation, handling, and the like; and add local transport to the relevant inland market. The result is the wholesale price of imported maize. It is this wholesale price of maize in the inland market that is the focal point of our calculation. The alternative to project production is not to import the maize and transport it to the project area. Rather, the alternative is to import it and market it directly on the inland market. Thus the price the farmer can expect to receive in the absence of tariffs, subsidies, or an import ban is the wholesale price less the cost of moving his maize to the market. If the project had included processing facilities, then the relevant project boundary price would have been this wholesale price less handling costs from the processing facility to the wholesale market. In the Nigerian project, no processing facilities were included, so the relevant import parity price is the farm-gate price. As we move back from the wholesale market to the farm gate, we would have to provide for any conversion allowance. In this case none is necessary, since it is assumed that the farmer will sell shelled maize. From the wholesale price, then, we deduct local marketing costs including assembly, bags, and intermediary margins, transport from the farm to the market, and storage losses, thus obtaining the import parity price at the farm gate of N90. (The symbol for Nigerian naira is N.) This is the maximum price the farmer could expect to receive, again in the absence of tariffs, subsidies, or an import ban.

SECTION III. DETERMINING ECONOMIC VALUES

Once financial prices or costs and benefits have been determined and entered in the project accounts, the analyst estimates the economic value of a proposed project to the nation as a whole. The financial prices are the starting point for the economic analysis; they are adjusted as needed to reflect the value to the society as a whole of both the inputs and outputs of the project.

When the market price of any good or service is changed to make it more closely represent the opportunity cost (the value of a good or service in its next best alternative use) to the society, the new value assigned becomes the "shadow price" (sometimes referred to as an "accounting price"). In the strictest sense, a shadow price is any price that is not a market price, but the term usually also carries the connotation that it is an estimate of the economic value of the good or service in question, perhaps weighted to reflect income distribution and savings objectives.

In chapter 2, for purposes of project analysis, we took the objective of a farm to be to maximize the farm family's incremental net benefit, the objective of the firm to maximize its incremental net income, and the objective of the society to maximize the contribution a project makes to the national income-the value of all final goods and services produced in the country during a particular period. These objectives, and the analysis to test their realization, were seen in financial terms for farms and firms. But economic analysis of a project moves beyond financial accounting. Strictly speaking, we may say that in financial analysis our numeraire-the common yardstick of account-is the real income change of the entity being analyzed valued in domestic market prices and in general expressed in domestic currency. But in economic analysis, since market prices do not always reflect scarcity values, our numeraire becomes the real, net national income change valued in opportunity cost. As we will note below, one methodology expresses these economic values in domestic currency and uses a shadow price of foreign exchange; the shadow price increases the value of traded goods to allow for the premium on foreign exchange arising from distortions caused by trade policies. Another method in use expresses the opportunity cost value of real national income change in domestic currency converted from foreign exchange at the official exchange rate and applies a conversion factor to the opportunity cost or value in use of nontraded goods expressed in domestic currency; the conversion factor reduces the value of nontraded goods relative to traded goods to allow for the foreign exchange premium.

Before a detailed discussion of adjusting financial accounts to reflect economic values commences, an important practical consideration must be emphasized. Many of the adjustments to the financial accounts can become quite complex. Not every point made in this chapter will apply to every agricultural project, nor will all points have the same importance in those projects where they do apply. The complexity of some calculations and the relative importance of some adjustments recall the reason for undertaking an economic analysis of a project: to improve the investment decision. Some adjustments will make a considerable difference to the economic attractiveness of a proposed project; others will be of minor importance, and no reasonable adjustment would change the investment decision. What we need to do here is to adopt an accounting practice-the doctrine of materiality. The analyst must focus his attention on those adjustments to the financial accounts that are likely to make a difference in the project investment decision. He should use rough approximations or ignore trivial adjustments that will not make any difference in the decision. There is an important balance to be struck between analytical elegance and getting on with the job.

In this chapter we will adjust the financial prices of tangible items to reflect economic values in three successive steps: (1) adjustment for direct transfer payments, (2) adjustment for price distortions in traded items, and (3) adjustment for price distortions in nontraded items. Before embarking on this series of adjustments, we will examine the problem of determining the appropriate premium for foreign exchange. After completing the adjustments, we will summarize the main points in a "decision tree" for determining economic values.

The series of successive adjustments to the financial accounts will lead to a set of economic accounts in which all values are stated in "efficiency prices," that is, in prices that reflect real resource use or consumption satisfaction and that are adjusted to eliminate direct and indirect transfers. These values will be market prices when market prices are good estimates of economic value or they will be shadow prices when market prices have had to be adjusted for distortions. When we adjust financial prices to reflect economic values better, in the vast majority of cases we will use the opportunity cost of the good or service as the criterion. We will use opportunity costs to value all inputs and outputs that are intermediate products used in the production of some other good or service. For some final goods and services, however, the concept of opportunity cost is not applicable because it is consumption value that sets the economic value, not value in some alternative use. In these instances, we will adopt the criterion of "willingness to pay" (also called "value in use"). We need to do this, however, only when the good or service in question is nontraded (perhaps as a result of government regulation) during some part of the life of the project-a point to which we will return later in our discussion. Because the ultimate objective of all economic activity is to satisfy consumption wants, all opportunity costs are derived from consumption values, and thus from willingness to pay.

An example may clarify our use of willingness to pay and opportunity cost. Suppose a country that is a rather inefficient producer of sugar has a policy to forbid sugar imports to protect its local industry. The price of sugar may then rise well above what it would be if sugar were imported. Even at these higher prices, most consumers will still buy some sugar for direct consumption-say, in coffee or tea-even though they may use less sugar than if the price were lower. The domestic price of sugar will be above the world market price and will represent the value of the sugar by the criterion of willingness to pay. If we were now to consider the economic value of sugar from the standpoint of its use in making fruit preserves, its value would become the opportunity cost of diverting the sugar from direct consumption, where willingness to pay is the criterion and has set the economic value.

Economic analysis, then, will state the cost and benefit to the society of the proposed project investment either in opportunity cost or in values determined by the willingness to pay. The costs or values will be determined in part by both the resource constraints and the policy constraints faced by the project. The difference between the benefit and the cost-the incremental net benefit stream-will be an accurate reflection of the project's income-generating capacity-that is, its net contribution to real national income.

The system outlined here will make no adjustment for the income distribution effects of a proposed project nor for its effect on the amount of the benefit generated that will be invested to accelerate future growth. Rather, the economic project analysis, stated in efficiency prices, will judge the capacity of the project to generate national income. The analyst can then choose from those alternative projects (or alternative formulations of roughly the same project) the high-yielding alternative that in his subjective judgment also makes the most effective contribution to objectives other than maximizing national income-objectives

such as income distribution, savings generated, number of jobs produced, regional development, national security, or whatever. The choice about the kind of project will of course be made rather early in the project cycle. Thus, it may be determined early on that for reasons of social policy a project will be preferred that encourages smallholder agriculture rather than plantations. Then, the choices will likely be several projects or variants of projects that encourage smallholders; the analytical technique presented here can determine from among the projects that will further the desired social objective the ones that are more economically efficient.

Although the system outlined here makes no adjustment for income distribution effects or for saving versus consumption, it is compatible with other analytical systems that do. In particular, Squire and van der Tak (1975) recommend evaluating proposed projects first by using essentially the same efficiency prices that will be estimated here and then by further adjusting these prices to weight them for income distribution effects and for potential effects on further investment of the benefits generated. The systems in Little and Mirrlees (1974) and the uxmo Guidelines for Project Evaluation (I 972a), with minor departures, also propose evaluating the project by first establishing its economic accounts in efficiency prices and then by adjusting these accounts to weight them for income distribution and savings effects. Making allowances for income distribution and savings effects involves somewhat more complex adjustments than those necessary to estimate efficiency prices; it also unavoidably incorporates some element of subjective judgment. Although these systems have attracted widespread interest among economists, their application has been only partial or on a limited scale. The system of economic analysis using efficiency prices that is outlined here is essentially the one currently used for all but a few World Bank projects and also the one used for most analyses of projects funded by other international organizations.

The economic analysis follows on the financial analysis presented in the preceding chapters; it will be based on projected farm budgets similar to those in chapter 4, on projected accounts for commercial firms such as those in chapter 5, and on projected government cash flows such as those in chapter 6. Since these accounts are projected for the life of the project, there will be no separate allowance for depreciation. Instead, as noted earlier, the costs will have been entered in the years they are incurred and the returns in the year they are realized.

In the economic analysis, we will want to work with accounts cast on a constant basis; thus we will want to be sure that any inflation contingency allowances have been taken out. As noted in chapter 2, however, physical contingency allowances and contingency allowances intended to allow for relative price changes are properly incorporated in the economic accounts, even when the accounts are in constant prices. Of course, any of the items included among the contingencies may be revalued, if necessary, to adjust them from their market prices to economic values. The projected financial accounts will usually not have any entry for cash. Instead, they will show separately the cash position of the farmer or note a cumulative cash surplus or deficit. It is possible, however, that some accounts may have a cash balance included in an entry for working capital or the like. If such an entry exists, it must be removed from the economic analysis; since we will be working on a real basis in the economic accounts, we will show real costs when they occur and real benefits when they are realized.

Determining the Premium on Foreign Exchange

Adjusting the financial accounts of a project to reflect economic values involves determining the proper premium to attach to foreign exchange. That determination quickly involves issues of obtaining proper values and of economic theory. Fortunately for most agricultural project analysts, the answer to the question about how to determine the foreign exchange premium is simple (and simplistic): ask the central planning agency. The point is that if various alternative investment opportunities open to a nation are to be compared, the same foreign exchange premium must be used in the economic analysis of each alternative. Otherwise we will be mixing apples and oranges and cannot use our analysis reliably to choose among alternatives. Sometimes, however, the analyst will be forced to make his own estimate of the foreign exchange premium. A practical approach, along with some of the theoretical and applied problems of the computation, is given by Ward (1976). Little and Mirrlees (1974), Squire and van der Tak (1975), and the UNIDO Guidelines (1972a) also outline in considerable detail how to make the conversion between foreign exchange and domestic currency when their analytical systems are used.

The need to determine the foreign exchange premium arises because in many countries, as a result of national trade policies (including tariffs on imported goods and subsidies on exports), people pay a premium on traded goods over what they pay for nontraded goods. This premium is not adequately reflected when the prices of traded goods are converted to the domestic currency equivalent at the official exchange rate. The premium represents the additional amount that users of traded goods, on an average and throughout the economy, are willing to pay to obtain one more unit of traded goods. Since all costs and benefits in economic analysis are valued on the basis of opportunity cost or willingness to pay, it is the relation between willingness to pay for traded as opposed to nontraded goods that establishes their relative value.

The premium people are willing to pay for traded goods, then, represents the amount that, on the average, traded goods are mispriced in relation to nontraded items when the official exchange rate is used to convert foreign exchange prices into domestic values. By applying the premium to traded goods, we are able to compare the values of traded and nontraded goods by the criterion of opportunity cost or willingness to pay. Although this premium is commonly referred to as the foreign exchange premium, it should be recognized that the premium is actually a premium for traded goods; foreign exchange itself has no intrinsic value. The premium for traded goods is a premium on the particular "basket" of traded goods that the present and projected trade pattern implies. Of course, future patterns of trade could change the exact composition of the basket, and thus the premium would change; to estimate these changes involves a knowledge of elasticities-the way demand and supply of goods and services vary when prices change-that is generally not available. Where such elasticities are known, it is possible for a well-trained economist to provide the project analyst with a more accurate estimate of the expected premium on foreign exchange.

If traded items were to be taken into the project analysis at an economic value obtained by simply multiplying the border price by the official exchange rate without adjusting for the foreign exchange premium, imported items would appear too cheap and domestic items too dear. This would encourage overinvestment in projects that use imports. For example, if combine harvesters look cheap because no allowance is made for the premium on traded goods, then imported combines might displace local harvest labor, even though the local labor might have no other opportunities for employment.

There are two equivalent ways of incorporating the premium on foreign exchange in our economic analysis. The first is to multiply the official exchange rate by the foreign exchange premium, which yields a shadow foreign exchange rate. [Note that this derivation of the shadow exchange rate is appropriate for efficiency analysis of projects and thus has a discrete definition. Other definitions of the shadow exchange rate are appropriate depending on the uses to which the rate will be put. Bacha and Taylor (1972) discuss some of these alternatives.] The shadow exchange rate is then used to convert the foreign exchange price of traded items into domestic currency. The effect of using the shadow exchange rate is to make traded items relatively more expensive in domestic currency by the amount of the foreign exchange premium. (An alternative arithmetic formulation is to convert the foreign exchange price into domestic currency at the official exchange rate and then multiply by 1 plus the foreign exchange premium stated in decimal terms.) The shadow exchange rate approach has been used in the past in most World Bank projects when adjustments have been made to allow for the foreign exchange premium on traded goods, and it is also used in the UNIDO Guidelines (1972a).

An alternative way to allow for the foreign exchange premium on traded items that is increasingly coming into use is to reduce the domestic currency values for nontraded items by an amount sufficient to reflect the premium. This is sometimes called the "conversion factor" approach. In its simplest form, based on straightforward efficiency prices, a single conversion factor-the "standard conversion factor" of Squire and van der Tak-is derived by taking the ratio of the value of all exports and imports at border prices to their value at domestic prices (Squire and van der Tak 1975, p. 93). In this form, the standard conversion factor bears a close relation to our shadow exchange rate; indeed, the standard conversion factor may be determined by dividing the official exchange rate by the shadow exchange rate or by taking the reciprocal of 1 plus the foreign exchange premium stated in decimal terms. Market prices or shadow prices of nontraded items are then multiplied by this standard conversion factor, and this reduces them to their appropriate economic value. Little and Mirrlees and Squire and van der Tak both adopt the conversion factor approach. In addition, both pairs of authors recommend deriving specific conversion factors for particular groups of products that will allow for any difference between market prices and opportunity costs and for the foreign exchange premium on traded items. As a result, their specific conversion factors may always be applied directly to domestic market prices. These authors also recommend that their conversion factors be calculated in social prices by including distribution weights.

In the valuation system followed here, all items are valued at efficiency prices without allowance for distribution weights (the issue of selecting projects to achieve distributional objectives is treated as a subsequent decision). This being the case, consideration of the distribution-weighted conversion factors proposed by Little and Mirrlees and Squire and van der Tak may be left aside, and we may focus our discussion on the Squire and van der Tak standard conversion factor as it relates to efficiency prices.

The relation between the official exchange rate (in the equations below, OER), the foreign exchange premium (Fx premium), the shadow exchange rate (SER), and the standard conversion factor (SCF) is perhaps easier to understand in equation form:

and

so that, as Squire and van der Tak note (1975, p. 93),

and

We may illustrate these relations by an example taken from the Agricultural Minimum Package Project in Ethiopia. At the time the project was appraised, the analyst knew that the official exchange rate of Eth$2.07 = US$1 failed to account for a foreign exchange premium of at least 10 percent. (The symbol for the Ethiopian dollar is Eth$; since this project was appraised the name of the currency unit has been changed to birr.) Thus, the analyst multiplied the official exchange rate by 1 plus a 10 percent foreign exchange premium to obtain a shadow exchange rate of Eth$2.28 = US$1 (2.07 x 1.1 = 2.28) that he rounded up to Eth$2.30 = US$ 1. The shadow exchange rate was then applied to all'traded items in the financial accounts, thereby increasing their relative value.

If the domestic currency is worth more per unit than the foreign exchange, the arithmetic is somewhat different. At the time the Nucleus Estate/Smallholder Oil Palm Project in Rivers State, Nigeria, was appraised, the official exchange rate was N 1 = US$1.54. (The symbol for Nigerian naira is N.) The project analysts were given a shadow exchange rate of N1 = US$1.27 to use in their economic evaluation. If, however, they had simply been informed that the foreign exchange premium was 21 percent, they could have determined the shadow exchange rate by dividing the dollar value by 1 plus the premium stated in decimal terms (1.54 - 1.21 = 1.27).

Of course, the effect of applying the shadow exchange rate to the traded items in the Ethiopian project was to make all nontraded items 10 percent less expensive in relation to the traded items in the economic accounts as opposed to the financial accounts. Now, instead of increasing the relative value of traded items, we could reduce the value of all nontraded items appearing in the financial accounts so that in the economic account they are relatively 10 percent less expensive. To do this we calculate the standard conversion factor, which is 1 divided by 1 plus the amount of the foreign exchange premium stated in decimal terms. In this case, the result is a factor of 0.909 (1 - 1.1 = 0.909). To obtain the economic values, we would then multiply all financial prices for non-traded items by this factor if these market prices have been judged good estimates of opportunity cost or good estimates of economic value on grounds of willingness to pay. For nontraded items such as wage rates for unskilled labor for which it is felt that the market price has overstated the economic values, we would first determine a good estimate of the economic value in domestic currency and then multiply that by the standard conversion factor. Financial prices for traded items, whether imports or exports, would be left unchanged in the economic accounts except that any transfer payment included in these prices would be taken out. To get all values into the same currency, we would convert all foreign currency prices to domestic currency values using the official exchange rate.

When we turn to determining measures of project worth in chapter 9, we will find that the absolute value of the net present worth differs depending on which approach we use, shadow exchange rate or conversion factor, but that the relative net present worths of different projects analyzed by the same approach will not change. Whichever approach is used, the internal rate of return, the benefit-cost ratio, and the net benefit-investment ratio do not change. (Using a number of disaggregated conversion factors, rather than a standard conversion factor, can give different values for the measures of project worth. Hence, for projects at the margin of acceptability, using specific conversion factors rather than a standard conversion factor or a shadow exchange rate may result in a different decision on whether to accept or reject, but such cases are infrequent.)

Adjusting Financial Prices to Economic Values

Let us now proceed with the adjustments necessary to convert financial prices to economic values. We will divide these into three steps: (1)

adjustment for direct transfer payments, (2) adjustment for price distortions in traded items, and (3) adjustment for price distortions in non-traded items. We will then note that, for what are termed "indirectly traded" items (locally produced items that use a high proportion of traded inputs, such as locally assembled tractors, or construction that uses imported materials), steps 2 and 3 must be done at the same time.

Step 1. Adjustment for direct transfer payments

The first step in adjusting financial prices to economic values is to eliminate direct transfer payments.

Direct transfer payments (see chapter 2) are payments that represent not the use of real resources but only the transfer of claims to real resources from one person in the society to another. In agricultural projects, the most common transfer payments are taxes, direct subsidies, and credit transactions that include loans, receipts, repayment of principal, and interest payments. Two credit transactions that might escape notice are accounts payable and accounts receivable. All these entries should be taken out before the financial accounts are adjusted to reflect economic values.

Many important subsidies in agriculture operate not by means of direct payments but through mechanisms that change market prices. These subsidies are not direct subsidies treated as direct transfer payments but rather are indirect subsidies. The financial price of an item for which the price has been changed because of an indirect subsidy is converted to an economic value according to the procedures outlined below for traded items in step 2 and, as appropriate, for nontraded items in step 3.

Step 2. Adjustment for price distortions in traded items

The second step in adjusting financial prices to economic values is the adjustment for distortions in market prices of traded items.

Traded items are those for which, if exports,f.o.b. price > domestic cost of production, or the items may be exported through government intervention by use of export subsidies and the like, and, if imports,domestic cost of production > c.i.f. price.

Conceptually-and usually in practice, too-prices for traded items in project analysis are more easily dealt with than those for nontraded items. We begin the valuation by determining the "border price." For imports, this normally will be the c.i.f. price and, for exports, normally the f.o.b. price. The border price is then adjusted to allow for domestic transport and marketing costs between the point of import or export and the project site; the result is the efficiency price to be used in the project account (see the subsection on "Economic export and import parity values," below).

If the proposed project produces something that can be used in place of imported goods-that is, if it produces an "import substitute"-the value to the society is the foreign exchange saved by using the domestic product valued at the border price, in this case the c.i.f. price. But if the project uses items that might otherwise have been exported-that is, if it uses "diverted exports"-then the opportunity cost to the society of these items is the foreign exchange lost on the exports forgone valued at the border price, this time the f.o.b. price.

If we are using conversion factors to allow for the foreign exchange premium, the economic value of a traded item would be obtained by converting the foreign exchange price to its domestic currency equivalent using the official exchange rate.

If we are using the shadow exchange rate to allow for the foreign exchange premium, the economic value of a traded item would be obtained by converting the foreign exchange price to its domestic currency equivalent using the shadow exchange rate.

To illustrate how these computations are made, we may take as an example an imported item such as a combine harvester for which the c.i.f. price is US$45,000. In the financial accounts, we will convert this price to domestic currency using the official exchange rate of, say, Rs10 = US$1, obtaining a c.i.f. price in domestic currency of Rs450,000 (45,000 x 10 = 450,000). To this would be added any import duty, say 10 percent, or Rs45,000 (450,000 x 0.10 = 45,000); the price of the combine in our financial accounts would therefore be Rs495,000 (450,000 + 45,000 = 495,000). (The costs of moving the harvester to the project site would also be added; see the subsection on "Economic export and import parity values," below.) If we are using the conversion factor approach to allow for the foreign exchange premium in our economic accounts, we would enter the combine in the accounts at the c.i.f. price expressed in domestic currency converted at the official exchange rate, or Rs450,000 (45,000 x 10 = 450,000). There would be no allowance for the duty because that is a transfer payment. If we are using the shadow exchange rate approach to allow for the foreign exchange premium, however, we would increase the price of the imported items to reflect the premium. Suppose we assume the foreign exchange premium to be 20 percent; our shadow exchange rate thus becomes Rs12 = US$1 (10 x 1.2 = 12). Now the Rs495,000 item in our financial accounts becomes Rs540,000 in our economic account (45,000 x 12 = 540,000). We could have accomplished the same thing, of course, by multiplying our domestic financial price (net of transfer payments) by 1 plus the foreign exchange premium (450,000 x 1.2 = 540,000). The effect of our computation, obviously, is to make imported items more expensive in our economic analysis.

The same logic works in reverse for exports. The ton of wheat that is worth $176 a ton f.o.b. at the port of export will be entered in the financial accounts by converting the foreign exchange price to its domestic currency equivalent using the official exchange rate. This gives a value of Rsl,760 (176 x 10 = 1,760), assuming that there is no export subsidy. The same rupee value would be entered in the economic accounts if we are using the conversion factor approach to allow for the foreign exchange premium. If we are using the shadow exchange rate approach to allow for the foreign exchange premium, we multiply the foreign exchange border price of the wheat by the shadow exchange rate instead of the official exchange rate to calculate the economic value expressed in domestic currency. This increases the relative value of the wheat, which now will be valued at Rs2,112 (176 x 12 = 2,112). We could have accomplished the same thing, of course, by multiplying our financial domestic price by 1 plus the foreign exchange premium stated in decimal terms (1,760 x 1.2 = 2,112). Now the ton of wheat, like other exported goods, is valued at its opportunity cost and is seen to be relatively much more valuable.

Diverted exports and import substitutes are valued by the same line of reasoning, except that for a diverted export we would take the f.o.b. price as the basis for valuation and for import substitutes we would take the c.i.f. price. In the examples of the previous paragraphs, if the country exported combines but diverted them to a domestic project, the opportunity cost would be based on the f.o.b. price instead of the c.i.f. price we assumed for imported combines. Similarly, if the wheat produced were to substitute for imports, we would base its value on the c.i.f. price of wheat rather than on the f.o.b. price we assumed for the case of exports.

In practice, values for most traded items are determined by taking the border price as we have been using it and then either subtracting or adding the domestic handling costs to obtain an economic value at the farm gate or project boundary-the economic export or import parity value (see the subsection on "Economic export and import parity values," below). Also, many items that are locally produced incorporate a significant proportion of imported components and may be considered indirectly imported items (see the section on "Indirectly Traded Items," below). To determine either parity values or values for indirectly traded items involves valuing separately not only the traded component but the nontraded component as well, so we will defer detailed discussion of these values until we have discussed valuing nontraded items.

Step 3. Adjustment for price distortions in nontraded items

The third step in adjusting financial prices to economic values is the adjustment for distortions in market prices of nontraded items. Nontraded items are those for which c.i.f. price > domestic cost of production > f.o.b. price, or the items are nontraded because of government intervention by means of import bans, quotas, and the like.

Often, nontraded items will be bulky goods such as straw or bricks, which by their very nature tend to be cheaper to produce domestically than to import but for which the export price is lower than the domestic cost of production. In other instances, nontraded items are highly perishable goods such as fresh vegetables or fluid milk for direct consumption.

In general, these are produced under relatively competitive conditions-they are produced either by many small farmers or by a few industrial producers for whom entry into the market is relatively easy; thus prices cannot rise too far out of line before new competition appears.

If we are using the shadow exchange rate approach to allow for the foreign exchange premium, and if the market price of a nontraded item is a good estimate of the opportunity cost, or willingness to pay is the criterion, we will accept the market price directly as our economic value. Otherwise, we will adjust the market price to eliminate distortions by the methods outlined in this section and then use the estimate of the opportunity cost we obtain as the shadow price to be entered in the economic accounts.

If we are using the conversion factor approach to allow for the foreign exchange premium, an additional step is necessary. All prices for non-traded items are reduced by multiplying them by the appropriate conversion factor. When willingness to pay is the criterion or when the market price is considered to be a good estimate of opportunity cost, the market price is accepted as the basis for valuation and then reduced by multiplying it by the conversion factor to obtain the economic value. But if we are using the standard conversion factor and the market price must be adjusted to obtain a better estimate of the opportunity cost, then the opportunity cost must, in turn, be multiplied by the standard conversion factor. (If specific conversion factors have been developed, as Little and Mirrlees and Squire and van der Tak suggest in their systems, then these factors incorporate the adjustments for nontraded goods distortions, opportunity costs, and distribution weights; the market price need only be multiplied by the specific conversion factor to reach the economic value.) Whether we use a shadow exchange rate or a standard conversion factor to allow for the foreign exchange premium, the adjustments we make to allow for distortions in market prices of nontraded items are essentially the same; only the step of multiplying the market price or the opportunity cost by the standard conversion factor differs.

As we said earlier in the chapter, prices for traded items are more easily adjusted to economic values than are prices for nontraded items. The following subsections treat some of the difficulties encountered in determining economic values for various nontraded items.

Market prices as estimates of economic value.

In a perfectly competitive market, the opportunity cost of an item would be its price, and this price would also be equal to the marginal value product of the item (see chapter 3). If a nontraded item is bought and sold in a relatively competitive market, the market price is the measure of the willingness to pay and is generally the best estimate of an opportunity cost. Most agricultural projects are expected to meet a growing demand for food or fiber and are small relative to the total agricultural production of the nation. If that is the case, in general we can accept the market price directly as our estimate of the economic value of a nontraded item. Also, if we are valuing a domestically produced project input that is produced by a supply industry operating near full capacity, we can generally accept the market price of the input as its economic value.

In some instances more common in industrial and transport projects than in agricultural, the output of the project is large relative to the market. The output from the project may therefore cause the price to fall. But the economic value of the new production, despite the fall in price, is not lower to the old users of the product; to them, it is still worth what the price was without the project. Yet to new users, the project output is not worth what the old price was; otherwise, the price would not have fallen. Under these circumstances, the economic value of the new output is neither the old price nor the new; rather, it is estimated by some weighted average of the old and new values. In technical economic terms, the total value of the new output is measured by the additional area under the demand curve as project output is increased, and the marginal value in use for each new buyer is measured by the demand curve at the point the buyer enters the market. The problem is that the precise shape of the demand curve is rarely known. As a result most project economists, when dealing with a project whose output is large relative to the market, adopt a simplifying rule of thumb-they assume that the demand curve is linear and downward sloping at 45 degrees. They then take the new estimate of the average value in use or opportunity cost-hence, of economic value-to be the average of the price without the project and the lower price with the project.

Sometimes a project will be proposed that does not meet new demand but replaces other goods or services in the market. Again, this is more common in industrial and transport projects than it is in agricultural. In such situations, if the project accounts are cast on a with-and-without basis, the economic value of the incremental net benefit stream would reflect only the saving from the new project compared with the old. This is because one of the costs of the new project would be the benefit forgone from the old production no longer realized and because one of the benefits would be the cost avoided for the old production. Such a case might arise, for instance, if an inefficient food processing plant were to be replaced by a more modern and efficient one, or if a high-cost railway branch line were to be replaced by bus and truck transport along an existing highway. Occasionally, however, a project will be proposed for a new plant that will replace existing output, and the analyst fails to recognize the with-and-without situation. Instead, he values the output from the new plant as if it were meeting new demand and forgets to charge as a cost to the project the benefit forgone from the production of the old plant that is to be displaced. If the project is not to be cast on a with-and-without basis, then the analyst must take as his gross benefit only the economic value of the resources saved by replacing the old plant, not the economic value of the output from the new plant.

Note that some nontraded items may involve using significant amounts of imported raw materials. These will be considered below, in the discussion of indirectly traded items. Such items might include machinery assembled domestically from imported components or electricity that is generally nontraded but that may require imported generating equipment and traded fuels for production.

One nontraded item that can sometimes lead to confusion is insurance. At first glance, insurance might look like a transfer payment and thus would not be included in the economic accounts of the project. We may, however, look upon insurance as a kind of sharing of the risk of real economic loss. This would be the case for fire insurance if project buildings were to be pooled with many other buildings in the society. In the event of a fire, there is a real economic cost. The resources used to replace a burned building, or the output forgone because a building no longer is available, reduce the amount of final goods and services available to the society and thus create a real reduction of the national income. Therefore, to the extent an insurance cost represents sharing of risk, it represents a proportionate sharing of real economic cost and should be included in the economic accounts. The insurance rate is usually based on the probability of a real loss and the value of the item insured.

Although the market price can frequently be accepted as a good estimate of the economic value of a nontraded item, for institutional reasons of one kind or another the market price can vary significantly from the opportunity cost of the item to the society. Two such nontraded items are important in most agricultural projects: land and labor.

Valuing Land

The opportunity cost of land is the net value of production forgone when the use of the land is changed from its without-project use to its with-project use.

The simplest case to value is one in which land changes use but not management control, either because an owner-operator is farming the land or because the same tenant continues to farm it. This is a common case in agricultural projects in which farmers are simply encouraged to adopt a more productive technology. If the analyst has laid out the financial accounts to show the situations with and without the project for farm budgets as suggested in chapter 4, then the incremental net benefit (that is, the incremental cash flow) of the project, when financial prices have been converted to economic values and the accounts aggregated as suggested in chapter 8, will include an allowance for the net value of production forgone by changing the land use. Take, for example, the Kemubu Irrigation Project in Malaysia in which new irrigation water permitted changing the land use in the dry season from rather unproductive pasture to second-crop paddy rice production. The contribution of the land to the value of the pasture-hence, its opportunity cost-would be properly accounted for when the value of the weight gain of the livestock pastured on the land without the project is subtracted from the value of the paddy rice produced on the land with the project. Converting project financial prices to economic values-say, changing the market price of the weight gain of the animals on the pasture and the market price of paddy rice to their economic equivalents if these are seen to be different from the market prices-automatically revalues the opportunity cost of the change in land use from financial to economic terms.

In other instances, however, the financial accounts must show a cost for purchasing land or the right to use it. Here problems arise because in many countries agricultural land is hardly sold at all, and, when it is, considerations of investment security and prestige may push its price well above what the land could reasonably be expected to contribute to agricultural production. In these instances, we will not want to accept the market purchase price as a good estimate of the economic opportunity cost of the land and must search for an alternative. Many times that alternative will be to take the rental value of the land. In a number of countries, although land is infrequently sold, there is a fairly widespread and competitive rental market. This may be true if there is considerable tenancy in the country, of course, but it may also hold true if the dominant form of land tenure is the owner-occupied farm. Older farmers may not wish to cultivate all of their holdings themselves and will be willing to rent a field to a younger neighboring farmer; widows may not wish to operate their holdings themselves; or a farmer suffering from an illness may wish to rent part of his farm for a season while he recovers. When such a rental market exists, it probably provides a fairly good indication of the net value of production of the land and, hence, of the opportunity cost if the land use is changed. A renter is not likely to pay any premium for prestige or investment security and thus will not pay a rent higher than the contribution the land can make to the crop he proposes to grow. That rental value may then be entered in the project's financial account year by year as a cost. Alternatively, it may be capitalized by dividing the rent by an appropriate rate of interest stated in decimal terms; the capitalized value is then entered in the first year of the project's financial accounts. The appropriate rate of interest actually would be the economic rate of return (see chapter 9), but this may well involve repetitive computations. Some analysts prefer to use the opportunity cost of capital (also discussed in chapter 9). If this rate were, say, 12 percent and the going rental rate were Rs525 a hectare, then the capital value of a hectare would be Rs4,375 (525 - 0.12 = 4,375). If we were using the conversion factor approach to allow for the foreign exchange premium, this capitalized value would be, in turn, multiplied by a conversion factor. If the standard conversion factor were 0.909, for instance, the land would then have an economic value of Rs3,977 (4,375 x 0.909 = 3,977). At the end of the project, the same value of the land could be credited to the project as a residual value.

Inevitably, however, there will be instances in which neither the purchase price nor the rental value is a good estimate; we then will have to make a direct estimate of the productive capability of the land. Such a direct estimate is not difficult if idle land is to be used for a settlement project. In the projects financed by the World Bank in the Amazon basin at Alto Bene in Brazil and in the Caqueta region of Colombia, the land without the project would in effect have produced no economically valuable output at all. Hence, the net value of production forgone was clearly zero, and no value for the land was entered in the project economic accounts. If settlers were required to pay the government a purchase price, either all at once or in installments, the farm budgets at market prices in the financial analysis would have to show those payments as a cost. When these financial farm budgets were converted to economic values, however, there would be no cost entered for the land because there was no reduction in national income as a result of shifting its use from jungle to farmland. (Of course, the cost of clearing jungle land should be reflected somewhere in the project costs.)

In other cases it will not be so simple. The analyst will have to make a direct estimate of the net value of production forgone for bringing the land into the project. A straightforward approach is to take the gross value of the land's output at market prices and deduct from that all the costs of production-including allowances for hired and family labor and for the interest on the capital engaged, again all at market prices. The analyst can assign the residual as the contribution of the land to the production of the output and take that as the opportunity cost of the land in financial terms. This set of computations can then be converted to economic terms by using economic values for each of the input and output entries. For those familiar with the technique, estimating a production function would provide a much more accurate estimate of the contribution of the land to the value of the output than the direct method described here and thus is a preferable approach.

Valuing Labor

Wage rates for labor in many developing countries may not accurately reflect the opportunity cost of shifting labor from its without-project occupation to its with-project use.

The price of labor in a perfectly competitive market, like other prices in that impossible place, would be determined by its marginal value product. That is, the wage would be equal to the value of the additional product that one additional laborer could produce. It would pay a farmer to hire an additional laborer-for harvesting, for example-so long as that extra worker increased total output by a value more than the wage the farmer had to pay him.

Even in labor-abundant societies, there are probably peak seasons at planting and harvesting when most rural workers can find employment. At those seasons, the market wage paid rural labor is probably a pretty good estimate of its opportunity cost and its marginal value product; therefore, we could accept the market wage as the economic value of the rural labor.

The problem of course is that, except for the peak seasons, in many crowded countries the addition of one more laborer may add very little to the total production-in an extreme instance, nothing at all. That is, if there is a surplus of agricultural workers, there may be very little or virtually no productive outlet for their energies in the off-season. In technical language, we may say that the marginal value product of such labor-the amount such labor adds to the national income-is very close to zero. Because the marginal value product of labor is also the opportunity cost of labor in the economic accounts, we may make another statement: if we take a laborer away from a farm community where he is producing very little or nothing and put him to work productively in an agricultural project that produces something of value, we do not have to forgo very much to use this labor to realize new production. This being the case, we can consider the cost of the laborer to be very low-some economists would say even zero. By this line of reasoning, the proper value to enter in the economic (not financial) account as the cost of labor would be very small, perhaps only a fraction of the going market wage. If the opportunity cost of labor in an agricultural project is properly priced at a very small amount, then it is likely that the rate of return on the project will look very favorable in comparison, say, with a capital-intensive alternative project that uses labor-saving tractors or expensive imported harvesting machinery.

Note that the validity of this reasoning is not changed by the fact that agricultural labor is, in fact, paid a wage well above its opportunity cost. A common example of a "wage" paid, even though little productive work is available on the margin, is found in the case of family labor. Older children and the farmer's wife will be entitled to a share of the family income even if the family farm is too small to give them an opportunity to be productive. In this instance, if an older son were to find productive employment elsewhere, the total production on the farm might be reduced by very little or none at all. Yet, because the older son is entitled to a share of the total family income, he would accept new employment far away from his home only if he were offered a wage in excess of his share-and that might be well above what his marginal value product would be and the reduction in farm output that would occur if he were to leave.

Rural wages may be above the marginal value product because of a traditional concept of a "proper" wage or because of social pressure on the more prosperous farmers in a community to share their wealth with their less fortunate neighbors. In parts of Java, for example, social custom prevents even quite small farmers from harvesting their own rice. Instead, they permit landless laborers to do the work, even though the farmer himself may well have the time to do it. This is explicitly seen by the community as a means of providing at least something for the poorest agricultural laborers. Unfortunately, increasing economic pressures on small farmers and continued population growth are leading to a break-down of this system.

Virtually all economists now agree that the marginal value product of agricultural labor on an annual basis worldwide is more than zero, so that in every instance our opportunity cost of labor, at least in some season or another, will be positive-even though it may still be very low. [A more detailed discussion of the marginal value product of agricultural labor can be found in McDiarmid (1977) and in Barnum and Squire (1979).]

To begin our discussion of how actually to determine an economic value for labor, we can take the easiest case. In most instances, skilled labor in developing countries is considered to be in rather short supply and would most likely be fully employed even without the project being considered. Hence, the wages paid workers such as mechanics, foremen, or project managers are in general assumed to represent the true marginal value product of these workers, and the wages are entered at their market values in the economic accounts. The rationale here is that, if those skills are in such scarce supply that they would be worth more than the going wage, then someone in the society would be prepared to pay more, and the skilled worker would then move to where he could earn that higher wage, thus establishing a new equilibrium. This convention of accepting market wages as good estimates of economic value may substantially undervalue skilled labor or the management skills of such top civil servants as extension specialists and project managers-or project analysts!

Note too that, as we consider the opportunity cost of labor and how to estimate it, if we set the financial accounts so they correctly show the situations with and without the project, then the opportunity cost of family labor will be appropriately priced in financial terms. Suppose that, in the dry season without the project, a farmer along the north coast of Java could find essentially no gainful employment. With the advent of the Jatiluhur Irrigation Project he now is able to produce a second crop of rice, and his net benefit rises accordingly. When we subtract his without-project net benefit (which would be essentially only what the family could earn for a rainy-season rice crop) from the with-project net benefit (which will include earnings from two crops), the incremental net benefit will correctly show the labor return the family had to give up during the dry season (essentially nothing) to participate in the project and produce a second crop of rice. Shifting the financial prices in the farm budget to economic values also automatically converts the opportunity cost of family labor to economic values.

To make our farm budgets work this way, we must remember to include any off-farm earnings in the accounts. Suppose we assume that the farmer from the north coast of Java goes to Jakarta and finds employment in the construction industry during the dry season, as many such farmers do. The without-project net benefit will thus be increased by the amount of the farmer's off-farm earnings. If he wishes to use Jatiluhur irrigation water to produce a second crop of rice, he must now give up the construction wages he could otherwise have earned in the dry season. In turn, when we subtract the without-project net benefit from the with-project net benefit, which includes the returns from two crops of rice, the incremental net benefit will be smaller by the amount of the opportunity cost of labor at the market wage, that is, by the amount of construction earnings the farmer must forgo. We may proceed to convert these financial accounts to economic terms by revaluing the appropriate entries at their shadow prices. In doing so, however, we must remember that one shadow price will be the shadow wage rate for the construction earnings the farmer had to forgo. It is to estimating this shadow wage rate that we now may turn.

In most discussions of the marginal value product of labor-hence, of its economic opportunity cost-the standard is the productivity of the marginal agricultural laborer. This is true not only for agricultural projects but also for projects in other sectors, since it is assumed that additional manufacturing employment, for example, will tend to reduce the number of unemployed agricultural laborers. This would be true even if it is urban workers drawn from some other urban occupation who actually take the new factory jobs, since it is assumed the jobs they vacate will, in turn, be filled by workers drawn from agriculture.

Cast in this form, our estimate of the shadow wage rate must now focus on how to estimate the marginal value product of agricultural labor without the project. We can begin by noting that in most agricultural communities there is usually a season when virtually everyone who wants work can find it. Even unemployed urban laborers may return to their home villages in these peak seasons to help their families or to work as hired laborers. This happens at harvest time in Java, and may happen at the peak planting time in other areas where transplanted rice is grown. Thus, we may reasonably assume that this peak season labor market is a relatively competitive one, that labor is in relatively short supply at this period, and that the daily wage at this period is a good indicator of the daily marginal value product of the labor engaged.

With this accepted, a good estimate of the annual shadow wage for agricultural labor is the number of days in the year when most rural labor can expect to find employment, multiplied by the daily wage rate at such times, and reduced by a conversion factor if appropriate. If an agricultural worker's daily wage at harvest were Rs7.50, and during harvest and other peak seasons most people in the rural work force could find employment for 90 days, then his annual shadow wage might be Rs675 if we are using the shadow exchange rate approach to allow for the foreign exchange premium (7.50 x 90 = 675), or Rs614 if we are using the conversion factor approach and the factor is 0.909 (7.50 x 90 x 0.909 = 614). Now if we wanted to hire an agricultural laborer to work in our project for 250 days a year, all the society would give up in production-the opportunity cost-would be Rs675 if we are using the shadow exchange rate approach, or Rs614 if we are using the conversion factor approach. This opportunity cost is the economic value of the annual earnings of the laborer without the project. Note that we surely would have to expect to pay a wage much greater than this amount, and thus our financial accounts at market prices would have quite a different cost for this same agricultural laborer. It is possible, for instance, that the hired laborer would expect a wage of Rs7.50 a day for all 250 days he worked during the year, or an annual wage of Rs1,875 (7.50 x 250 = 1,875). More probably, he would be willing to work for rather less a day outside the harvest season-say, Rs5.00 a day. Thus, his annual wage might be something more on the order of Rs675 for 90 days and Rs5.00 a day for the remaining 160 days, or an annual total wage of Rs 1,475 [(7.50 x 90) + (5.00 x 160) = 1,475]. The project analyst would clearly have to form a judgment of the shadow wage of hired labor on the best basis he could, just as he must for every other price estimate he makes.

Of course, in many agricultural projects labor is not engaged on a year-round basis. Rather, the work is quite seasonal, and we must consider in which particular season hired labor would be engaged. If our new cropping pattern calls for work to be done during the peak season, then we will have to consider that the peak season market wage is probably a good estimate of the marginal value product, and we could not justify using a lower wage as the basis for our shadow wage rate, even though there might be considerable unemployment in the off-season. In Egypt, for example, a common rotation calls for both rice and cotton to be harvested in October. If we were to propose a project incorporating these crops-or another crop requiring hired labor at this period-then the going wage (in 1975 about E^0.30 a day; the symbol for Egyptian pounds is E^) would be paid. Since even in a country as populous as Egypt most rural labor can find employment at this peak season, the use of a shadow wage rate derived from a basis less than the market wage would be unjustified. But suppose our project called for growing maize, which is planted in May when there is little other agricultural work available and harvested in August before the peak harvest season for rice and cotton. Then we might find that, on the margin, many agricultural laborers were either unemployed or not very productively engaged at that season and that to draw them into maize planting might entail an opportunity cost considerably less than the going wage, although it would perhaps not be zero. Thus, we might estimate that at this season the combination of being able to work only two or three days a week on the average, and then at jobs of rather low productivity, would justify taking a shadow wage rate based on half the going market rate. This would mean the equivalent of E^0.15 in 1975 if we are using the shadow foreign exchange rate approach (0.30 - 2 = 0.15), or E^0.14 if we are using the conversion factor approach and the conversion factor is 0.909 (0.30 - 2 x 0.909 = 0.14), even though our farm budget at market prices would continue to show a wage for hired labor of E^0.30.

All of these considerations will have to be adapted to fit the circumstances of any given project. For example, in India nationwide we might expect a shadow wage rate for agricultural labor rather less than the going wage rate. But using a nationwide shadow wage rate in particular projects might underestimate the true opportunity cost of the labor actually engaged in a project. The peak season in the Punjab, for instance, finds virtually all agricultural labor fully engaged, but in the neighboring state of Haryana the marginal labor in agriculture is not fully engaged. While many laborers from Haryana do migrate in search of peak season employment in the Punjab, not enough do so to meet the demand for labor completely. Using a very low shadow wage rate for a project in the Punjab might be unjustified because at the peak season the project would have to bid labor away from harvesting. Thus, although the shadow wage rate might not be as high as the harvest wage (but it might), neither would it be as low as conditions in neighboring Haryana might otherwise indicate.

This discussion of how to value labor applies whether labor is to be paid a money wage or is to be compensated in kind. The discussion so far has emphasized that it is the opportunity cost that determines the value of labor in the system of economic analysis we have adopted. The value of the payment actually made to labor-whether in money or in kind-is not the issue. If we shadow-price labor, we already are acknowledging that the wage the labor receives is different from the benefit forgone by using that labor in the project instead of in its next best alternative use without the project. It is the opportunity cost of the labor, not the form of payment, that sets the economic value of labor. Hence, it is irrelevant in a determination of the economic value of labor whether labor is paid a money wage or is compensated in kind-for example, in food grain, even though the food grain may be a tradable commodity and even though the food grain itself might need to be shadow-priced if it is to be valued.

Excess Capacity

In some projects, a domestically produced input may come from a plant that is not operating at its full capacity. If that is the case, then the opportunity cost of using the input in a new project is only the marginal variable cost of producing the input, and no allowance need be made for the fixed capital cost of the plant itself. If the national cement industry is operating at less than its full capacity and it is proposed to line irrigation canals with cement, then the cost of the cement for the canals would be only the marginal variable cost of producing the cement. This would be less than the average cost of cement production, which would include some allowance for fixed costs of production.

Situations such as these are more common in industrial projects than in agricultural projects. When they do occur, however, they may influence the timing of projects. A canal-lining project might be quite attractive if it is begun soon, while there is excess cement-manufacturing capacity, but much less attractive later, when demand has caught up with the cement industry's capacity. To supply cement for canal lining later, after demand has picked up, would entail constructing an additional cement plant. At that time, new fixed as well as variable costs would be incurred, and the analyst would include all costs, both fixed and variable, plus an estimate of the "normal" profit in calculating the cost of cement.

TRADABLE BUT NONTRADED ITEMS. In the system of project analysis presented here, we lay out the economic accounts as best we can to reflect the real resource costs and benefits of the proposed project. The project will be carried out within a framework of economic policies set by the government. The project analyst must make the best judgment about what those policies are and will be, not just what they ought to be, and work the economic analysis accordingly. This can lead to difficult choices when the analyst must evaluate the real effects on resources of a project that involves items that could be traded but probably will not be because of government regulation. These items, which are "tradable but non-traded" across national boundaries, are valued as nontraded.

Such items would usually be imported were it not for an import quota .00 or an outright ban that is enforced against them. Their domestic price may well rise high above the prevailing price on the world market. The import restriction might be enforced to protect domestic industries, even though the imported item may be preferred by consumers. Import of foreign engines for tubewells, for example, may be forbidden so that domestic manufacture might be encouraged. Yet, the domestic equivalent may not be as efficient or as durable as the imported engine and may cost more to produce. The domestic engine clearly could not compete on the world market, and it would therefore be a nontraded item. For those few imported engines allowed to enter the country, the price may rise quite high. This indicates that to some buyer the imported item is worth more than its domestic equivalent. If our project will use one of these engines, the economic value is not a price based on the world market as if the engines could be relatively freely traded. Rather, it is the higher domestic market price of the imported engine, which indicates its high opportunity cost. Upon reexamination, of course, we might consider changing the project design to use the domestic engine-for example, we might do so if we find the domestic engine to be less costly when valued at shadow prices.

For the domestic equivalent of an imported item, the market price usually will closely approximate the real resource use that went into producing it. But if there is a shortage and the price is bid up, in the absence of additional imports the market price will rise above the cost of production. In this case, the opportunity cost of the item will not be determined by the resources used to produce it but by its marginal value product in its best alternative use. If the price is higher than is justified by the resources used to produce the item, it may well be because to someone that high price for the domestic engine is worth it-for this buyer's purposes, the marginal value product of the scarce engine at least equals the market price. If we wish to bid that engine away for use in our project, we are denying its use to the other potential buyer. If we use the engine in our project, the economy must forgo the productive contribution of the engine in the alternative use the other potential buyer had in mind-our standard concept of opportunity cost. Again, in this instance the opportunity cost is most likely well estimated by the market price; if it were not, other buyers would not have bid the price up so high for the limited number of engines available.

If there is an import ban on an imported final good or service, then we will base the economic valuation on the criterion of willingness to pay and accept the market price as a good indicator of the economic value of the product-provided that we expect the trade ban to remain in force throughout the life of the project. Earlier we cited the example of a ban on sugar imports that would force the domestic price of sugar above its border price. If the ban on imports will continue, then the higher price of sugar indicates a willingness to pay that, in turn, is an indicator of the economic value set on sugar by the consumers. In the project analysis, we would accept this market price as the economic value, not a border price as if the sugar were being traded.

For both kinds of import substitutes we have cited, the analyst may want to prepare an analysis that will indicate the effect on the proposed project of lifting the import ban. We will discuss this topic further below, and value each separately. Take locally assembled tractors, for example. We may be told that the market price of Rs65,000 includes a 30 percent local component (in other words, 30 percent of the market price represents domestic value added) and that 70 percent of the market price represents the imported component, which includes a 15 percent tariff. Thus, the local component will amount to Rs19,500 (65,000 x 0.3 = 19,500), and the imported component including the tariff will amount to Rs45,500 (65,000 x 0.7 = 45,500). The domestic value added will most likely arise from sources such as wages paid domestic skilled labor and domestically manufactured items that use mainly domestic raw materials. If so, we probably can accept the market price as a good indicator of the opportunity cost to the economy of these items.

To determine the economic value of the imported component of the tractor, the tariff must first be eliminated. This may be done by dividing the value of the imported component including the tariff by 1 plus the percentage of the tariff stated in decimal terms; this calculation gives a value for the imported component without the tariff of Rs39,565 (45,500 1.15 = 39,565). This is, of course, the c.i.f. price converted to its domestic equivalent at the official exchange rate.

Now, if we are using the shadow exchange rate to allow for the foreign exchange premium, we will want to revalue the imported component of the indirect import (after the tariff has been eliminated) to reflect the distortion in the prices of traded goods. To do this, we can take the c.i.f. price converted at the official exchange rate and multiply it by 1 plus the foreign exchange premium stated in decimal terms. If the official exchange rate is Rs10 = US$1 and the foreign exchange premium is 20 percent, then for the imported component of the tractor we derive a value of Rs47,478 (39,565 x 1.2 = 47,478). (We could, of course, have taken the c.i.f. price in foreign exchange and converted it to its domestic equivalent by the shadow exchange rate; this would have given the identical result.) The shadow price of the tractor is now the market price of the domestic component, which we calculated to be Rs19,500, plus the shadow-priced value of the imported component of Rs47,478-or a total economic value of Rs66,978 (19,500 + 47,478 = 66,978).

If we are using the conversion factor to allow for the foreign exchange premium, the economic value of the imported component will be the c.i.f. price converted to the domestic currency equivalent at the official exchange rate after eliminating the tariff, or Rs39,565. To obtain the economic value of the domestic component we will need to multiply it by the conversion factor. For efficiency prices, we would use the standard conversion factor of 1 divided by 1 plus the foreign exchange premium stated in decimal terms. In this instance, the foreign exchange premium is 20 percent, so the standard conversion factor becomes 0.833 (1 - 1.2 = 0.833). Applying this to the domestic component of the tractor, estimated to be Rs19,500 at market prices, gives us an economic value of Rs16,244 (19,500 x 0.833 = 16,244). The shadow price of the tractor now becomes the sum of the imported component valued at c.i.f. converted at the official exchange rate and the shadow price for the domestic component, or Rs55,809 (39,565 + 16,244) = 55,809).

In some agricultural projects, electricity is an important cost that may raise valuation problems. Electricity is usually thought of as a nontraded commodity. In reality, part of the value of electricity in most developing countries arises from the imported generating and transmission equipment and, perhaps, from imported fuel. Thus, in our system of project analysis, electricity might be an indirectly traded item. The first difficulty is that the price charged for electricity is not competitively set, since there is no competition in electricity. Rather, electricity rates are administered prices, and electricity prices thus may bear little relation to marginal value product or to opportunity cost. No easy means exists to resolve this problem. Some average rate, or perhaps some weighted average rate, will probably have to suffice as an estimate of opportunity cost at market prices. Once a rate is accepted, an estimate will have to be made of the domestic and imported components, and the components revalued using the shadow exchange rate or a conversion factor as appropriate, just as for any other indirectly imported item (and as we illustrated earlier by the example of tractors assembled from imported components). These calculations would usually not be undertaken by agricultural project analysts. The planning office should estimate a shadow price for electricity and other utilities to be used in all project analyses.

For some agricultural projects, new generating facilities will be required. In the simplest case, we might think of a project remote from the electric grid, such as a settlement project, in which a diesel generating unit might be included as a cost of the project. In that instance, there would be no particular problem of valuation. When new generating facilities would be needed to meet the demand on the power grid arising from an irrigation project, however, the problem would not be so simple. Here, the best approach would probably be to ask the electricity authority for an estimate of the additional cost the authority would incur for this particular project, and then to treat that cost-properly shadow-priced to allow for the imported component-as the opportunity cost. The cost of the additional facilities needed for the project will probably have to be reduced to a kilowatt-hour basis (using, perhaps, the capital recovery factor to estimate the annual charge for the new facilities).

We have contrasted use of a shadow exchange rate and a conversion factor to correct for price distortions caused by import and export tariffs and subsidies, and we have noted that the same correction can be realized whichever approach is used. This is illustrated in table 7-1, in which an economic account for a hypothetical project is drawn up using both a shadow exchange rate and a standard conversion factor.

When indirectly traded items will be used repeatedly in projects, it may be convenient to have specific conversion factors that, once they are derived, can be directly applied to the same class of indirectly traded items. This is the approach both Little and Mirrlees (1974) and Squire

Table 7-1. Use of Shadow Exchange Rate and Standard Conversion Factor Compared

 
Economic Value (Rs)
Remarks
 
Financial values
Using shadow exchange rate`
Using standard conversion factor
Item
Rs
US$
Inflow
 
 
 
 
 
Gross value of wheat produced
1,750
175
2,100
1,750
Traded item
Total
1,750
175
2,100
1,750
 
Outflow
 
 
 
 
 
Unskilled labor (shadow wage rate= 50% market wage)
600
60
300
250

Nontraded item

Imported fertilizer
200
20
240
200

 

Tractor services
 
 
 
 

 

75% imported component
90
9
108
90

Traded item

 
25% domestic component
30
3
30
25

Indirectly traded item

 
Total
920
92
678
565
 
Net benefit
830
83
1,422
1,185
 
Ratio of inflow to outflow
1.90
1.90
3.10
3.10
 

Rs= Indian rupees; US$ =U.S. dollars.

a. The official exchange rate is assumed to be Rs10 = US$1. Financial prices are converted by this official exchange rate.

b. The foreign exchange premium is assumed to be 20 percent. As in note a, the official exchange rate is assumed to be Rs10 = US$1.

c. The shadow exchange rate is the official exchange rate of Rs 10 multiplied by 1 plus the percentage of the foreign exchange premium stated in decimal terms, or Rs12 (10 x 1.2 = 12), so that Rs12 = US$ 1. Foreign exchange prices

are converted into domestic currency values by multiplying the foreign currency price by Rs12.

d. The standard conversion factor is the reciprocal of 1 plus the foreign exchange premium stated in decimal terms, or 0.833 (1 - 1.2 = 0.833). Foreign currency prices are converted into decimal currency values at the official exchange rate. Domestic currency prices are multiplied by the standard conversion factor of 0.833.

and van der Tak (1975) suggest, and both sets of authors recommend that some central agency prepare specific conversion factors for project analysts to use. It is possible in a parallel manner to derive "specific shadow exchange rates" that may then be applied repeatedly, although in practice this has rarely been done. Instead, when the shadow exchange rate approach is followed, nontraded items are decomposed into their traded and nontraded elements and each is valued separately. Use of a specific conversion factor can be illustrated by referring to table 7-1. Suppose we planned a number of projects in which tractor services would be important and we wanted a specific conversion factor for tractor services. Once we had the conversion factor in hand, we could multiply the domestic market price of items in each project by the same specific conversion factor to obtain the various economic values. In table 7-1, in the column illustrating use of the standard conversion factor, we have a value for the imported component of the tractor services of Rs90, which was converted at the official exchange rate. The domestic component was multiplied by the standard conversion factor to obtain an economic value of Rs25. If we accept this as a good estimate of the value of the domestic component, then by adding the two we reach an economic value for the tractor services of Rsl 15. If we divide this economic value by the domestic price, we obtain a specific conversion factor of 0.958 (115 - 120 = 0.958). In the future, we can simply multiply the market price of tractor services by the specific conversion factor to obtain the economic value directly.

Economic export and import parity values

The economic value of a traded item-either an export or an import-at the farm gate or project boundary is its export or import parity value. These values are derived by adjusting the c.i.f. (cost, insurance, and freight) or f.o.b. (free-on-board) prices (converted to economic values) by all the relevant charges (again converted to economic values) between the farm gate or project boundary and the point where the c.i.f. or f.o.b. price is quoted. The general method of calculating export and import parity prices was discussed in the last section of chapter 3. When these financial prices are adjusted to derive their economic equivalent, both traded and nontraded elements must be valued simultaneously.

The methods for deriving import and export parity values are parallel. Thus, it is unnecessary to discuss the method for both; instead, we will discuss only derivation of the import parity price as an example because import parity values tend to be a bit more complicated to derive.

We may return to the example of the imported combine harvester used earlier in the chapter to illustrate economic valuation of a traded item. In our financial accounts, the c.i.f. price of US$45,000 was converted to its domestic currency equivalent at the official exchange rate of Rs 10 = US$1, to which we would add, say, a 10 percent duty, Rs 1,500 in domestic handling and marketing charges, and Rs2,250 in internal transport costs to the project site-for an import parity price at the farmgate of Rs498,750 [(45,000 x 10) + (45,000 x 10 x 0.10) + 1,500 + 2,250 = 498,750].

To obtain the economic import parity value at the farm gate or project boundary when using the shadow exchange rate to allow for the foreign exchange premium, we would make the same computations except that we would use the shadow exchange rate and omit the tariff, which is a transfer payment. In the illustration of valuing traded items, we assumed that the foreign exchange premium on the imported combine was 20 percent, and so we assumed a shadow exchange rate of Rs12 = US$1 (10 x 1.2 = 12). Now, to obtain the import parity value of the harvester, we would convert the c.i.f. price to its domestic equivalent using the shadow exchange rate, omit the tariff, and then add the value of the nontraded domestic items. To simplify matters, we will assume that all costs of moving the combine to the project site reflect only nontraded items-although that might not be acceptable if, say, the transport costs included significant amounts of petroleum fuel. We now reach an economic import parity value of Rs543,750 [(45,000 x 12) + 1,500 + 2,250 = 543,7501.

If we are using the conversion factor to allow for the foreign exchange premium, the foreign exchange would be converted to its domestic currency equivalent in the economic accounts by using the official exchange rate, and every nontraded item would be reduced by the conversion factor. Recalling that the standard conversion factor is 1 divided by 1 plus the foreign exchange premium stated in decimal terms, we obtain a standard conversion factor of 0.833 (1 - 1.2 = 0.833). Now, to obtain the economic import parity value of the harvester at the farm gate or project boundary, we convert all foreign exchange costs to domestic currency at the official exchange rate and reduce all prices of nontraded items by applying the standard conversion factor. Again, we will assume that the transport costs are predominantly made up of nontraded items. As before, we will omit the tariff because it is a transfer payment. The economic import parity price thus becomes Rs453,124 [(45,000 x 10) + (1,500 x 0.833) + (2,250 x 0.833) = 453,124].

In certain instances, the value in local currency of an imported item at the project site will be known, as will the rate of tariff and local transport charges from the point of import to the project site. If this is the case, to determine the economic value it is necessary to determine the c.i.f. price, take out the tariff, and allow for the cost of domestic transport. Using our previous values, we may know, for example, that a combine harvester delivered to the project site costs Rs498,750, that the tariff on imported harvesters is 10 percent, and that local transport and domestic handling from the point of import to the project site costs Rs3,750. We know that the official exchange rate is Rs10 = US$1 and that the foreign exchange premium is 20 percent, so the shadow exchange rate would be Rs12 = US$1 (10 x 1.2 = 12) and the standard conversion factor 0.833 (1 - 1.2 = 0.833). We deduct the cost of local transport to obtain a financial value of Rs495,000 at the point of entry, which includes the c.i.f. price plus the duty (498,750 - 3,750 = 495,000). To take out the duty, we divide by 1 plus the percentage of the duty stated in decimal terms to obtain

Rs450,000 (495,000 - 1.1 = 450,000). This is the c.i.f. value at the official exchange rate. We can then divide by the official exchange rate to obtain the c.i.f. value in foreign exchange of US$45,000 (450,000 - 10 = 45,000). If we are using the shadow exchange rate to allow for the foreign exchange premium, we can obtain our c.i.f. economic value by multiplying by the shadow exchange rate of Rs12 = US$1 to obtain a value of Rs540,000 (45,000 x 12 = 540,000). Then, to obtain the economic value at the project site, we would add the cost of transport from the point of entry to the project site; this yields an economic import parity value for the harvester at the farm gate or project boundary of Rs543,750 (540,000 + 3,750 = 543,750). If we are using the conversion factor to allow for the foreign exchange premium, the economic value of the combine at the port will be the c.i.f. foreign exchange price converted at the official exchange rate, or Rs450,000 (45,000 x 10 = 450,000). To obtain the economic import parity value at the farm gate or project boundary, we would add to this c.i.f. value the cost of domestic transport and domestic handling, reduced by the standard conversion factor, to obtain an economic import parity value of Rs453,124 [450,000 + (3,750 x 0.833) _ 453,124].

It is clear that to derive the import and export parity values in the economic analysis we must omit transfer payments, allow for the foreign exchange premium, and use shadow prices for those domestic goods and services for which prices are inaccurate indicators of opportunity cost. The same examples from the Sudanese and Nigerian projects used to illustrate the discussion of import and export parity prices in chapter 3 (tables 3-3 and 3-4) are used again in tables 7-2 and 7-3 to show economic parity values using both the shadow exchange rate and the conversion factor to allow for the foreign exchange premium.

Trade Policy Signals from Project Analysis

Up to this point, we have been discussing an analytical system that estimates the contribution of a proposed project to national income within a policy framework that the project analyst considers will exist during the life of the project. We have assumed that the project analyst has very little influence on trade policies, for this is true in the agriculture sector in most countries. Questions often arise, however, about the effects on a proposed project if trade policies were to change, and about whether changes in trade policies should be recommended. Unfortunately, when assessing the effects on a project of policies that would lift or impose a ban on trade, the analytical issues become very complex, and the analysis of a single project is of limited usefulness. The limitations of project analysis in influencing policy arise from the partial nature of project analysis and from the assumption that the project investment does not significantly change price relations in the economy as a whole.

Two important cases involving trade policy often arise that cause soul-searching among project analysts. The first is when a quota or prohibitive tariff prevents entry of a crucial input-perhaps fertilizer-

Table 7-2. Economic Export Parity Value of Cotton, Rahad Irrigation Project, Sudan (1980 forecast prices)

 
 
Value per ton
Steps in the calculation
Relevant steps inthe Sudanese example
Lint
 
Seed
Scarto

 

Using shadow exchange

 

 

 

C.i.f. at point of entry
 

C.i.f. Liverpool taken as estimate for all European ports)

US$639.33

US$103.39

 

Deduct unloading atpoint of import

 

 
 
 
Deduct freight to point of import

Freight and insurance

 

- 39.63

- 24.73

 

Deduct insurance
 
 
 
 
Equals f.o.b. at point
of export

F.o.b. Port Sudan

US$599.70

US$78.66

 

 

Converted at shadow

exchange rate of

 
 
 
Convert foreign currency
to domestic currency
at shadow exchange rate

LSd1.000 = US$2.61lb

LSd229.682

LSd30.126

 

 

Port handling cost

 
 
 
Deduct local port
charges

Lint: LSd5.564 per ton

- 5.564

 
 
 

Seed: ^Sd1.510 per ton

 

- 1.510

 

Deduct local transport

Freight to Port Sudan

 
 
 
and marketing

at ^Sd6.782 per ton

- 6.782

- 6.782

 

costs from project to
 
 
 
 
point of export (if not
 
 
 
 
part of project cost)
 
 
 
 
Equals export parity

Export parity value

 
 
 
value at project

at gin at project

 
 
 
boundary

site

^Sd217.336

^Sd21.834

 

Scartoa

Conversion allowance

Convert to seed cotton

86.934

12.882

1.102

if necessary

(^Sd217.336 x 0.4

 
 
 
 

+ ^Sd21.834 x 0.59

 

^Sd100.918

 
 

+ ^Sd110.200 x 0.01)

 
 
 

Deduct local storage,

Ginning, baling, and

 
 
 

transport, and

storage (^Sd15.229

 
 
 

marketing costs (if not

per ton)

 

- 15.229

 

part of project cost)

 
 
 
 
 

Collection and internal

 
 
 
 

transfer (^Sd1.064

 
 
 
 

per ton)

 

- 1.064

 

Equals export parity

Export parity value

 
 
 

value at farm gate

at farm gate

 

^Sd84.625

 
 

Using conversion factors

 
 
 

C.i.f. at point of entry

C.i.f. Liverpool (taken as

 
 
 
 

estimate for all

 
 
 
 

European ports)

US$639.33

US$103.39

 

Deduct unloading at

 
 
 
 

point of import

Freight and

 
 
 

Deduct freight to

 

- 39.63

- 24.73

 

point of import

insurance

 
 
 

Deduct insurance

 
 
 
 

Equals f.o.b. at point

 
 
 
 

of export

F.o.b. Port Sudan

US$599.70

US$78.66

 

Convert foreign currency

Converted at official

 
 
 

to domestic currency

exchange rate of

 
 
 

at official exchange

^Sd1.000 = US$2.8726

^Sc1208.809

^Sd27.389

 

rate

 
 
 
 

(Table continues on the following pages.)

Table 7-2 (continued)

Steps in thecalculation

Relevant steps inthe Sudanese example

ValueLint

per tonSeed

Scartoa

Convert nontraded goods

Converted using

 
 
 

to equivalent domestic

standard conversion

 
 
 

value using conversion

factor of 0.9096

 
 
 

factors

 
 
 
 

Deduct local port

Port handling cost

 
 
 

charges

Lint: ^Sd5.564 per ton

5.058

 
 
 

Seed: ^Sd1.510 per ton

 

1.373

 

Deduct local transport

Freight to Port Sudan

 
 
 

and marketing

at ^Sd6.782 per ton

6.165

- 6.165

 

costs from project to

 
 
 
 

point of export (if not

 
 
 
 

part of project cost)

 
 
 
 

Equals export parity

Export parity value

 
 
 

value at project

at gin at project

 
 
 

boundary

site

^Sd197.586

^Sd19.851

-

Conversion allowance

Convert to seed cotton

79.034

11.712

1.002

if necessary

(^Sd197.586 x 0.4

I

 

I

 

+ ^Sd19.851 x 0.59

 
 
 
 

+ ^Sd110.200 x 0.909

 

^Sd91.748

 

Deduct local storage,

Ginning, baling, and

 

transport, and

storage (^Sd15.229

 

marketing costs (if not

per ton)

- 13.843

part of project cost)

 
 
 

Collection and internal

 
 

transfer (^Sd1.064

 
 

per ton)

- 0.967

Equals export parity

Export parity value

 

value at farm gate

at farm gate

^Sd76.938

LSd = Sudanese pounds.

Source: Adapted from World Bank, "Appraisal of the Rahad Irrigation Project," PA-139b (Washington, D.C., 1973; restricted circulation), annex 16, table 6. The format of the table is adapted from William A. Ward, "Calculating Import and Export Parity Prices," training material of the Economic Development Institute, CN3 (Washington, D.C.: World Bank, 1977), p. 9.

a. Scarto is a by-product of very short, soiled fibers not suitable for export and is sold locally at a price of ^Sd110.200 per ton.

b. For purposes of illustration, there is assumed to be a foreign exchange premium of 10 percent. Thus, the dollar value of the Sudanese pound at the official exchange rate of ^Sd1.000 = US$2.872 has been divided by 1.1 to give an assumed shadow exchange rate of ^Sd1.000 = US$2.611 (2.872 - 1.1 = 2.611), whereas the standard conversion factor is divided by 1 plus the foreign exchange premium, or 0.909 (1 - 1.1 = 0.909). In the appraisal report that is the source of this table, no foreign exchange premium was assumed.

c. Seed cotton is converted to lint, seed, and scarto assuming that a ton of seed cotton yields 400 kilograms lint, 590 kilograms seed, and 10 kilograms scarto.

and this forces use of a more costly domestic alternative and thus greatly reduces the contribution of the project to national income. The second is when an import quota imposed on products that compete with the project's output makes the contribution of the project investment to national income high, even though the cost of production per unit of output from the project is higher than the cost of competing imports.

When the domestic cost of an important project input is higher than the world market price because of a quota or prohibitive tariff, the potential contribution of the proposed investment to national income

Table 7-3. Economic Import Parity Value of Early Crop Maize, Central Agricultural Development Projects, Nigeria

(1985 forecast prices in 1976 constant terms)

Steps in the calculation

Relevant steps in Value

the Nigerian example perton

F.o.b. at point of export

Add local transport and marketing costs to relevant market Equals value at market Conversion allowance if necessary

Deduct transport

and marketing costs to relevant market

F.o.b. at point of export

Add freight to point of import Add unloading at point

of import Add insurance Equals c.i.f. at point of import Convert foreign currency

to domestic currency at shadow exchange rate Add local port charges

Using shadow exchange rate F.o.b. U.S. Gulf ports No. 2 U.S. yellow corn in bulk' US$116

Freight and insurance 31

(Included in freight estimate)

C.i.f. Lagos or Apapa US$147 Converted at an assumed shadow exchange rate of 4,q1 =

US$1.476 X100 Landing and port charges

(including cost of bags) 22 Transport (based on a

350-kilometer average)` 10

Wholesale value x#132

(Not necessary)

Primary marketing (includes assembly, cost of bags,

and intermediary margins)` - 12 Transport (based on a

350-kilometer average)` - 10 Storage loss (10 percent of

harvested weight) - 9

Deduct local storage, transport, and marketing costs

(if not part of project cost) Equals import parity value at farm gate

Add freight to point of import Add unloading at point

of import Add insurance Equals c.i.f. at point of import Convert foreign currency

to domestic currency at official exchange rate

Import parity value

at farm gate x#101 Using conversion factors

F.o.b. U.S. Gulf ports

No. 2 U.S. yellow corn in bulk' US$116

Freight and insurance 31

(Included in freight estimate)

C.i.f. Lagos or Apapa US$147 Converted at official

exchange rate of

:~il = US$1.626 X91

Table 7-3 (continued)Steps in the calculation

Relevant steps in Valuethe Nigerian example perton

Convert nontraded goods

Converted using standard

to equivalent domestic

conversion factor of 0.909b

value using conversion

 

factors

 

Add local port charges

Landing and port charges

 

(including cost of bags, X22) 20

Add local transport

Transport (based on a

and marketing costs

350-kilometer average, X10)` 9

to relevant market

 

Equals value at market

Wholesale value X120

Conversion allowance

 

if necessary

(Not necessary)

Deduct transport

Primary marketing (includes

and marketing costs

assembly, cost of bags,

to relevant market

and intermediary margins, *12)` - 11

 

Transport (based on a

 

350-kilometer average, X10)` - 9

Deduct local storage, transport,

Storage loss (10 percent of

and marketing costs

harvested weight) - 8

(if not part of project cost)

 

Equals import parity value

Import parity value

at farm gate

at farm gate x#92

Nigerian naira.

Source: Adapted from World Bank, "Supplementary Annexes to Central Agricultural Development Projects," 1370-UNI (Washington, D.C., 1976; restricted circulation), supplement 11, appendix 2, table 4. The format of the table is adapted from Ward, "Calculating Import and Export Parity Prices," p. 10.

a. Forecast from Price Prospects for Major Primary Commodities (1976, annex 1, p. 12; see World Bank 1982a).

b. For purposes of illustration, there is assumed to be a foreign exchange premium of 10 percent. Thus, the dollar value of the naira at the official exchange rate of *1 = US$1.62 has been divided by 1.1 to give an assumed shadow exchange rate of =P-~ 1 = US$1.47 (1.62 - 1.1 = 1.47), whereas the standard conversion factor is 1 divided by I plus the foreign exchange premium, or 0.909 (1 - 1.1 = 0.909). In the appraisal report that is the source of this table, no foreign exchange premium was assumed.

c. Shadow prices were assumed for transport and for primary marketing because in the financial analysis the market wage overvalued the opportunity cost of unskilled labor. The value given is the opportunity cost in naira (before applying the standard conversion factor).

will be reduced by the tariff or quota. Given the policy prevailing, the project analysis will be an accurate indicator of the project's worth. Take fertilizer, for instance. If it is expensive to produce domestically, this is an indication that fertilizer production uses a large amount of scarce domestic resources relative to the resources necessary to produce some other product that could be exported to earn the foreign exchange needed to import the fertilizer from a foreign supplier. But if the domestic fertilizer must, in fact, be used for the project to move forward, then it will take a lot of domestic resources to produce the project's agricultural output, and the project will not, accordingly, make as much of a contribution to the national income as it could were imported fertilizer available. If the quota or prohibitive tariff against the input were removed, then the project investment would look quite different. A change in trade policy, however, will have implications ranging far beyond the boundary of the project itself, implications for both efficiencies in the economy and for noneconomic objectives. A change in trade policy may bring a wide range of changes in other prices in the economy as well as in the price of fertilizer used on nonproject farms, and to be valid an investment analysis would have to be run with the new price relations and include nonproject farms. Predicting these changes could be very difficult if the change in trade policy were significant. At best, the project analyst could run his analysis again using a c.i.f. price for fertilizer and making a broad guess about what the changes might be in the rest of the economy both within and outside agriculture., He could then turn to those responsible for trade policy and say that his project analysis signaled a need to consider with care removing the quota against fertilizer. But note that the project analysis is only a signal, not a criterion for decision; much, much more must go into a reevaluation of trade policy than the analysis of one project.

The other important case in which a change in a quota proves very difficult for the project analyst is that of a quota against imports that would compete with the output of a proposed project. If the imports are prohibited, the output of the project will sell for more in the protected market, and what otherwise might not be a very attractive project may now make sufficient contribution to national income to be justified. Again, i f policies are not going to be changed, this is an accurate indicator of the contribution to the national income. But if the domestic cost per unit of project output-say, apples-valued at shadow prices is greater than the c.i.f. cost of imported apples, then this is an indication that it would be more efficient from the standpoint of the economy as a whole for the project to produce something else, export it to earn foreign exchange, and then use the foreign exchange to import apples. Under the circumstances, the project analyst may want to run his analysis again using an import parity price and perhaps also adjusting some of the other price relations in the direction he thinks might prevail under a change in trade policy. He may find that domestic production would not make enough of a contribution to national income at these prices to justify the investment required. He might also want to determine the domestic resource cost of the import substitute along the lines discussed in the section of chapter 10 devoted to that topic; this will show that it costs more to save a unit of foreign exchange by producing apples domestically than the shadow exchange rate indicates the foreign exchange to be worth. His analysis has now signaled that trade policies should perhaps be reviewed. Again, it is only a signal; the analysis of this one project does not itself provide a complete decision criterion. The trade policy change will have many other effects that will be felt far beyond the boundary of the project itself.

Valuing Intangible Costs and Benefits

The methodology outlined for converting financial prices to economic values is one that is most appropriate for tangible costs and benefits. When intangible costs or benefits enter into investment considerations, they raise difficult issues of valuation.

Intangible factors have come up frequently in earlier discussions of identifying costs and benefits and of valuing them. They comprise a whole range of considerations-economic considerations such as income distribution, number of jobs created, or regional development; national considerations such as national integration or national security; and environmental considerations that can be both ecological and aesthetic, such as the preservation of productive ecosystems, recreation benefits, or famous spots of scenic beauty. [Lee (1982) discusses ecological considerations to be kept in mind when designing agricultural projects for tropical regions.]

The question of how to treat intangible factors most often arises when we are considering the benefits of a project. Many development projects are undertaken primarily to secure intangible benefits--education projects, domestic water supply projects, and health projects are a few common ones. Intangible benefits are usually not the major concern in agricultural projects, although many agricultural and rural development projects include components such as education or rural water supply from which intangible benefits are expected. Whether in agricultural projects or in other kinds of projects, intangible benefits, even though universally agreed to be valuable, are nevertheless virtually impossible to value satisfactorily in monetary terms. Yet costs for these projects are in general tangible enough, and the considerations of financial and economic valuation we have discussed earlier apply unambiguously.

Intangible costs are not uncommon, however, and prove just as difficult to bring within a valuation system as benefits. Often costs are merely the inverse of the benefits that are sought: illiteracy, disease, unemployment, or the loss of a productive environment or treasured scenic beauty.

Some costs in agricultural projects, while tangible, are very difficult to quantify and to value. Siltation, waterlogging, salinization, and soil loss are examples. These costs should not be ignored, and if they are likely to be substantial they should be treated in the project analysis in a manner analogous to intangible costs.

When considering projects in which intangible benefits or costs are important, the least the project analyst can do is to identify them: lives that will be saved, jobs created, kind of education provided, region to be developed, location of a park, ecosystem or kind of scenery to be preserved.

Very often, the analyst can also quantify intangibles: number of lives saved, number of jobs created, number of students to be enrolled, number of people expected to use a park. Even such simple quantification is often a substantial help in making an investment decision.

Economists have tried repeatedly to find means to value intangibles and thus bring them within the compass of their valuation system. The benefits of education have been valued by comparing the earnings of an educated man with those of one who is uneducated. Health and sanitation benefits have been valued in the number of hours of lost work avoided by decreasing the incidence of disease. Nutrition benefits have been valued in terms of increased productivity. Population projects have been valued by attaching a value to the births avoided. Although work in these areas continues-especially with regard to environmental impact-few applied project analyses in developing countries currently attempt to use such approaches to valuing intangible costs and benefits. For one thing, such efforts generally greatly underestimate the value of the intangibles. The value of an education is much more than just the increase in income-ask any mullah, monk, or priest. Good health is a blessing far in excess of merely being able to work more hours. Good nutrition is desirable for more reasons than just increased productivity. Moreover, the methodological approaches used to value intangibles turn out to be unreliable and open to serious question. Finally, there may be moral issues involved-many who support population programs do so out of considerations that extend far beyond any benefit-cost calculation.

In contemporary practice of project analysis in developing countries, the only method used to any extent to deal with intangible benefits is to determine on a present worth basis the least expensive alternative combination of tangible costs that will realize essentially the same intangible benefit. This is often referred to as "least-cost combination" or "cost effectiveness" (for an application of the method to sanitation projects, see Kalbermatten, Julius, and Gunnerson 1982, chapter 3). If the same education benefits can be provided by centralized schools that realize economies of scale but require buses or by more expensive smaller schools to which students can walk, which schools are cheaper? Can the same health benefits be provided at less cost by constructing fewer large hospitals but more clinics manned by paramedical personnel? By constructing a waterborne sewerage system or by installing low-cost household sanitation facilities that do not require sewers? Can the same number of lives be saved more cheaply by buying up all the property rights in a flood plain and moving people out than by constructing dykes and levees? Given two park sites that would give similar recreation benefits-perhaps one that would require buying warehouse sites and another that would require extensive filling and flood control along a river-which would be cheaper? Once it is determined that the least expensive alternative has been identified and its costs valued, then the subjective question can be more readily addressed: is it worth it?

Interestingly enough, electricity projects are customarily analyzed using least-cost combination. The marginal value product of electricity is in general considered greatly understated by the administered price charged; in any event, much electricity is used for home lighting that is very difficult to value. In practice, most power projects simply compare alternative means of producing the same amount of power: steam generating stations versus a hydroelectric dam; a large generator with transmission costs and several years of idle capacity versus a series of smaller stations close to the demand centers.

A variation of the least-cost combination method can be used to deal with intangibles in multipurpose projects. From the total cost of the project are deducted all those costs that can be directly attributed to tangible benefits-flood damage avoided, irrigation, navigation, and the like. These costs are compared with their associated benefits to determine if the purpose is worthwhile at all. Is the flood damage avoided worth the direct costs incurred? Finally, the residual costs for the project are compared with the residual, intangible benefits. Is the number of lives saved by the project worth the residual cost that must be incurred? (A method of allocating residuals was outlined in the section on joint cost allocation in chapter 6.)

The problems with valuing intangibles are more common and more difficult to deal with in sectors other than agriculture. In agricultural projects, most of the benefits usually are tangible and can be valued. The costs and benefits can be compared directly to choose the highest-yielding alternative. There are, however, several aspects of intangible benefits that are frequently encountered in agricultural projects. Agricultural extension services, for example, are sometimes considered to give an intangible benefit in greater farmer education. For the most part, it is best to treat such costs that may give rise partly to intangible benefits-or, at least, the incremental amount of such costs-as necessary within a project if the total, tangible benefits are to be realized. If a dairy production project requires helping farmers to learn better sanitation procedures, then the extension agents who teach the procedures are essential to the success of the project, and the benefit of their effort is the tangible one of more and better milk.

In rural development projects, there are often components that are hardly essential to the main production objectives and that produce generally intangible benefits. This is the case when village schools, rural water supplies, rural clinics, or even agricultural research costs are included in a project. If these components are relatively small in comparison with total project costs, as they often are, then the problem of valuing the benefits may be ignored. But if such components form a significant part of total project cost, they probably should be separated out and treated on a least-cost combination basis. This procedure was followed in the analysis for the Korea Rural Infrastructure Project. The project included irrigation, feeder roads, community fuelwood plots, rural domestic water supply, and rural electrification. The irrigation, feeder roads, and community fuelwood components were analyzed by

comparing their tangible costs with their tangible benefits, but the components for rural domestic water supply and rural electrification were each dealt with separately on a least-cost combination basis.

Finally, if the proposed project is one in which the output is wholly intangible, a least-cost combination approach is appropriate. This would probably be the case for agricultural projects in which the major investment is in extension, agricultural education, rural water supply, rural health improvement, or research.

Decision Tree for Determining Economic Values

A "decision tree" for determining economic values is given in figure 7-1, parts A-D. Most issues of economic valuation in agricultural projects are covered by this diagram. The decision tree is used by taking an item to be valued in an agricultural project and tracing through the tree, following each alternative as it applies to the item until the end of the tree is reached, where a suggestion about how to value the item will be found.

To illustrate, we may trace through a few common elements in agricultural projects. Take fertilizer to be used in an irrigation project that will produce cotton. The fertilizer is tangible, involves real resource use, is traded, is a project input, and would be imported without the project. Therefore, it is valued at the import parity price. Or take agricultural labor to be used to apply the fertilizer. It is tangible, involves real resource use, is nontraded, is a project input, is nonproduced, is labor, and would be underemployed without the project. Therefore, it is valued by taking the marginal value product of the labor in its without-project employment. (Note that labor is defined as a tangible item, a possible source of confusion in using the decision tree.) Or take a tax on the fertilizer. It is tangible, is a direct transfer payment, is a payment to or from government, and is a tax. Therefore, it is omitted from the project economic account. Or, finally, take the cotton to be produced in the project. It is tangible, involves real resource use, is traded, is a project output, and will be an export. Therefore, it is valued at the export parity price.


TOCPREVNEXTINDEX