Estimating Social Prices for Tradables

The social price for an agricultural commodity is a border price-the price at which foreign suppliers would deliver the commodity to the domestic market or the price that foreign consumers would pay domestic suppliers to deliver the commodity to their markets. But if the world price is to represent an expected social opportunity cost for the domestically produced commodity, it may require adjustment. Observed world prices may not reflect the impacts of the domestic country's market power. In the absence of actual imports or exports of the domestically produced commodity, world price equivalents must be estimated. These prices will usually be derived from some observed world price. But to be truly equivalent, they must reflect the effects of international transport costs and differences in quality. Moreover, observed world prices may be different from those expected in the future.

Finding World Prices

One source of data on world prices is the country's international trade statistics. Implicit world prices (average per unit values) can be found by division of total values by quantities traded (of imports or exports). But biases in trade data can arise if the amounts traded are small, if the data are distorted because firms have improperly reported figures to avoid taxes or to repatriate earnings (overinvoicing or underinvoicing), or if the quality of the foreign produced commodity differs substantially from that of the domestically produced commodity. Interviews with officials from the statistics agency or from trading companies allow assessment of these possibilities. If own-country trade data are missing or biased, comparable world prices often can be estimated by examination of trade data of a nearby country. These prices are adjusted for international transport and insurance costs to simulate a cif import price or fob export price for the country of interest. A third possible source of direct world prices is price information from industry, government agencies, or international organizations. These groups regularly publish cif or fob prices at various locations, which can then be adjusted to allow for any international transport and insurance cost differences between the listed port and the relevant country port.

When direct world price information cannot be found, an alternative procedure is to estimate world prices indirectly, by removing the effects of distorting domestic policies. To find world prices indirectly, one

starts with the private prices of tradables and then estimates the quantitative impact of policies affecting the commodity. This process results in an implicit world price that reflects the removal of the distorting effects of policy. In terms of the PAM, I or J is measured in order to find E or F as residuals from A or B, respectively. The procedure works well when all policy transfers are easily measurable and known. If, for example, the sole distorting policy on an importable fertilizer input is a subsidy that is known to be 20 percent of the cif import price of fertilizer, the observed domestic price of fertilizer is 80 percent of the world price (calculated as the domestic price divided by 0.8).

This procedure is more difficult when quantitative trade restrictions are present or when transfers involve a combination of policies, including quotas. For instance, rice imports might face a 30 percent tariff and a quota of 500,000 metric tons. In adjusting an observed domestic price of 450 rupiah per metric ton, the researcher cannot simply divide 450 by 1.3 and claim the implied world price (cif local ports) is 346 rupiah. The quota on rice imports, if binding, would create upward pressure on domestic prices in addition to that already caused by the tariff. The implied result, 346 rupiah, sets an upper limit to the world price, but the actual world price could be much lower. In this circumstance, cif or fob rice prices in nearby countries must be obtained.

Location

Recognition of positive costs for transportation and handling is one potential complication in the identification of world prices. The price received by domestic producers for exportation to foreign markets (the fob price) is no longer the same as the price paid by domestic consumers for importation from foreign markets (the cif price). The foreign demand schedule becomes distinct from the foreign supply schedule, as shown in Figure 11.1. If the country is a net importer of the commodity, domestic demand and supply relationships are like those illustrated by the supply curve, S, and the demand curve, D. The domestic market price under free trade will be Pw; this term represents the social value of the commodity. If the country is a net exporter of the commodity, the domestic market supply curve will be represented by S", and the social value will become a lesser amount, P`;". If the supply curve intersects the demand curve at a price between Pw and P`;', as does S', no trade occurs; the social price of the nontradable commodity, P_D, is determined entirely by domestic market conditions.

When international transportation and handling costs are small, the cif-fob price band will be small as well. Whether fob or cif prices are used to represent social values is a minor concern. But when the price band is large, the choice of appropriate world prices becomes an issue for the calculation of social profitability. The problem for construction of PAMs arises because the location of the supply and demand curves under social price conditions is not known. The observed supply curve reflects the impacts of divergences in the markets of related commodities and markets for domestic factors and commodity inputs. A commodity that is imported or nontraded might be an exportable under social price conditions. Hence, judgments about quantities of domestic supply and demand under social price conditions become necessary elements in the choice of social output prices.

The possible wide differences between social profitability calculated at cif versus fob prices means that location can be a key criterion in the identification of commodity systems. Specification of a system as producing for the export market allows direct evaluation of competitiveness at the fob price. Social prices for internal domestic market locations are then estimated by subtraction of the social costs of transport and handling from the fob price. This procedure generates a schedule of social prices for the exportable. If internal prices were to fall below this fob-determined price, producers would increase exports, causing internal prices to rise; prices above the fob-determined price would cause a

diversion from exports to internal markets, forcing internal prices to fall. If, instead, production is substituted for imports, cif prices are the relevant border prices. Domestic transport and handling charges are then added to the cif price to estimate social prices for internal market locations. Arbitrage (between importing and consumption of domestic production) again assures that this set of social prices will be sustained.

When the commodity is nontradable, the social supply and demand curves intersect within the fob-cif band. In these circumstances, only upper and lower limits can be defined for the social prices. The actual price is determined by local supply and demand. Because the social curves are not known, the social output price bears some degree of indeterminacy. These indeterminacies are likely to be most important at isolated markets, because differences between the cif-determined and the fob-determined price schedules become increasingly large as internal markets are more distant or more isolated from the border.

Time Frame of Analysis

Explicit recognition of the time frame of analysis is the second aspect of identification of social commodity prices. Observed domestic market prices and world prices can be used to assess the profitability of the commodity system in any given year. From the policy-maker's perspective, however, long-run profitability is often more important. To capture longer-run interactions of policy and profitability, the analyst needs to use expected prices as the measures for calculation of input costs and output revenues. Generally, observed prices will be close or equal to expected prices, and the distinction between expected and actual prices will become a nonissue. Because most commodity world markets are large, long-run demand and supply shifts in world markets generate slow changes in commodity prices. But if the analysis is conducted in an unusual year, when extraordinary demand or supply shifts create abnormally high or low prices, the use of observed prices can give misleading measures of profitability and policy transfers.

Estimates of expected future prices can be obtained from commodity models of world markets. Price projections are made for a number of commodities by international organizations such as the World Bank. In principle, a price projection represents the current price adjusted for expected future changes in the demand and supply balance. If demand is expected to increase more rapidly than supply, expected future prices will be higher than current prices; in the opposite circumstance, expected future prices will decline. The difficulty with such exercises is the possibility that they will be self-defeating prophesies. If market participants believe the projection and thus expect future prices to rise, incentives are created to find new sources of supply or substitutes in consumption. If prices are expected to decline, producers will be encouraged to switch to other activities and new sources of consumer demand will be encouraged. In part because of such adjustments, eventual market prices usually differ substantially from their projected values.

Because of such complications, the current price is usually the best indicator of expected prices in the near future, unless it has been influenced by some large and unusual shock in the world demand-supply balance. The rationale for reliance on current prices is particularly strong when the commodity can be stored, because expected future prices will be linked to current prices. If expected future prices exceed current prices by more than the cost of storing the commodity, an incentive will exist for merchants to buy the commodity at the current time and store it for future sale. This action will increase current prices. If expected future prices differ from current prices by less than the cost of storage, incentives will exist to sell some of the stored commodity in the current period, thus driving down current prices. In either instance, intertemporal arbitrage should equate expected future prices to current prices plus storage costs.

In many agricultural commodity markets, world prices are influenced by subsidy and trade policies in foreign countries. Even if all countries in the world market were price-takers, the aggregate impact on world price could still be significant because so many countries use these policies. From the efficiency perspective, however, the magnitude of this price effect is not an important issue. For an individual country, optimal income is generated by using world prices, however imperfectly they might be established. The world price thus represents the opportunity cost of the commodity for the domestic economy. Domestic policy-makers might feel that foreign country policies are unfair to domestic producers; in this case, a decision to set a private market price greater than the social (world) price could use foreign policy considerations as a nonefficiency rationale. The particular price effects of foreign subsidy and trade policies thus become important only if future policy changes are expected to alter world prices. For example, reductions in subsidies in exporting countries or reductions in tariffs in importing countries would cause world prices to rise. Because prediction of the timing, magnitude, and price effects of such policy changes is usually impossible, most estimates of expected prices implicitly assume constancy of agricultural policies in other countries.

A final complication for expected price calculations arises because of world price variability and consequent risk premia. In some commodity markets, world prices fluctuate widely from year to year. Such conditions most often characterize residual markets, where world exports or imports (or both) are dominated by the policy actions of one or a few large countries. Relatively small changes in demand-supply balances in these countries can then translate into large changes in world market supply and demand. In response to such conditions, countries may try to pursue stable internal prices by varying positive subsidies to consumers of importables or to producers of exportables. When domestic production fluctuates, net demands from the world market or supplies to the world market are changed. If many countries pursue such stabilization policies simultaneously, and if net demand changes are correlated across countries (for example, when countries are subjected to similar weather), world price instability is further aggravated.

In the presence of substantial price variability, small countries may believe that long-run averages of world prices are an insufficient measure of the incentives afforded by world markets. If consumers are willing to pay a premium for price stability, the social value of output may be increased by a tax, the proceeds of which finance a domestic price stabilizing scheme. Such schemes can be pursued with a variety of methods (buffer stocks, financial buffer funds) and provide varying degrees of food security, depending on the size of the buffer. The choice of risk premium is usually based on the net cost of operating an arbitrarily specified buffer program.

Quality

Quality differences between the product in the domestic market and the competitor available from the world market constitute a third complication in the calculation of social commodity prices. Premia and discounts can usually be associated with particular characteristics, such as appearance, impurities, nutrients, and country of origin. If the product is an export, the export price data already reflect the quality of the domestic product. To find implicit world prices for importables, however, judgments are needed to establish comparable quality so that observed world market prices can be adjusted to match the quality of domestic output. The relative domestic prices of the imported commodity and its local counterpart reflect the marginal valuations of domestic consumers. Because preferences differ among countries, the world market may place a different relative price on the domestic commodity. But if many commodities of different quality are traded on the world market, the empirical task involves specification of the physical characteristics of the domestic commodity and the selection of a commodity of comparable quality from those available on the world market.

Large Country Effects

When a large country can influence world prices, the selection of appropriate social prices depends on domestic supply behavior and the characteristics of world market demand. Figure 11.2 considers the case of a large country exporter. Figure 11.2a describes the domestic market and Figure 11.2b the foreign market. Because foreign market demand is downward sloping, the foreign marginal revenue curve must be downward sloping as well. Domestic supply to the foreign market is derived by estimation of the difference between domestic supply and demand at different prices. Under competition, these supply prices are the same as marginal costs of production. Maximum profits are achieved when foreign marginal revenue equals domestic marginal costs; this point occurs when exports are Q*. To hold exports to this level, the government imposes an export tax equal to P* - MR*. Export tax revenues amount to (P* - MR*)Q*, or ABCE on the right-hand side of the figure. Under export taxation, the price received by producers is MR*, not P*. If the export tax is not used, competition among domestic producers will lead to an output of Q and a price of P'. In this situation,

income to the economy is less than it might be, because marginal production costs (equal to P') are greater than marginal revenue, MR'. But unless producers receive the export tax revenue, they will be better off with unregulated trade.

In principle, tax-inclusive prices should be used for social valuation, because their use creates the maximum national income. In the presence of an optimal export tax, private and social prices will be equal, and both will be less than the world price; for an optimal import tax, private and social prices will be in excess of world prices. In practice, however, the detailed empirical models of commodity market behavior necessary to permit the calculation of world and domestic prices under optimal taxes are often lacking. In part, the lack of models reflects a dearth of data. But for many agricultural commodities, judgments about market power are also confounded by the prominence of policy interventions. Particular countries can have a large share of world trade. If their share of world production is small, however, attempts to impose optimal taxes may result only in policy changes within competing nations that cause the tax-imposing country to lose market share rather than influence world prices. In most empirical situations, therefore, optimal tax arguments can be raised as a justification for an observed divergence between private prices and world prices. But such measures are not likely to indicate the optimal size of the divergence.

Using the Exchange Rate

At some point, the analyst needs to convert world prices into domestic currency; this conversion requires an exchange rate. Entries in the calculation of private profitability, the top row of the matrix, present no difficulties. The researcher normally does not even come across an exchange rate in private budgeting. If some items are denominated in foreign currency, the actual exchange rate used by the farmer or marketing agent-official or otherwise-is the correct candidate for conversion. Private profits measure observed incentives and results. When policy is not enforceable-for example, when parallel market rather than official exchange rates are used by participants in the system-there is little point in assuming that private actions are affected by nonbinding policy. Private profits are measures from actual markets, whether legal or illegal. Interest then centers on what exchange rate to use to convert world prices into domestic currency for social valuation (in the second row of PAM). Adjusting the exchange rate for the impacts of output price distortions and macroeconomic policy effects is a complex task. Fortunately, such corrections are not essential for the construction of the PAM. Exchange-rate changes cause changes in output prices that will be transmitted eventually to domestic factor prices. Social factor prices reflect marginal value products-the social price of output times the marginal physical products-and these prices will change in equal proportion because changes in exchange rates alter tradable-output prices in equal proportion. All tradable-commodity systems are similarly affected by the exchange rate, once factor prices have had time to adjust. The nonuniform effects of exchange-rate changes occur in systems for nontraded goods, because the nontraded-output prices are not directly affected, whereas all input costs are changed. Producers of nontraded commodities will face pressures to alter input combinations and reduce costs; otherwise, they will be forced to raise output prices and suffer the consequences of reduced demand.

The inputs and outputs of tradable-commodity systems are not identically affected, however, when the government fails to offset inflation with exchange-rate changes. In this circumstance, factor prices increase faster than the exchange rate depreciates and thus faster than prices of tradable outputs and inputs rise. PAM analysts then need to adjust private exchange rates (and the social prices of outputs and tradable inputs). This inflation adjustment simply corrects for past or projected movements in the country's real exchange rate (RER). The RER is found by comparison of the country's ratio of exchange-rate changes to wholesale (or consumer) price changes with the same ratios for its major trading partners.

An example serves to demonstrate the process of RER adjustments. If a country experiences an inflation rate of 50 percent while average inflation in the country's major trading partners is only 5 percent, differential inflation is about 43 percent: (150 / 105 - 1) x 100 percent. The country's exchange rate is pegged to the U.S. dollar and is initially (before the differential inflation) in equilibrium at 100 domestic currency units (DCUs) per U.S. dollar, or DCU 100 / $1. If the government maintains the fixed exchange rate, the real exchange rate will appreciate to DCU 70 / $1 (the end-of-year official exchange rate, DCU 100 / $1, divided by a ratio of the indices of the domestic inflation rate to the weighted average inflation rate in the principal trading partner countries, 150 / 105). If this year is used as the base year for PAM analysis, an exchange-rate correction of 43 percent will be required to estimate the equilibrium real exchange rate prevailing at the end of the year. The inflation adjustment factor will be 1.43, and the adjusted end-of-year rate will thus be DCU 143 / $1. If differential inflation is assumed to occur uniformly throughout the year, the average degree of overvaluation will be half of 43 percent and the average adjusted exchange rate will be about DCU 122 / $1. The social domestic currency prices of tradable outputs and inputs will then be 22 percent higher than private prices (122 / 100), reflecting the inflation tax that distorting exchangerate policy imposes on tradables.

A second circumstance in which the social prices of tradable commodities might be adjusted for exchange-rate effects occurs in the presence of multiple exchange-rate regimes. These regimes involve a set of official rates or, more commonly, an official rate and a parallel market rate. The exchange rate selected to convert tradable-commodity prices should match the rate that is determining the observed level of domestic factor prices. When the official rate is controlled and parallel markets for foreign exchange are small, the official rate provides the relevant standard. When parallel markets predominate, the official rate should be ignored in favor of the parallel market rate. Finally, when parallel and official markets are of comparable magnitudes, a blend of the two rates is appropriate. In all cases, the analyst must discover the particular rate used to generate the domestic prices for the tradables of the commodity system. For example, if trade data used for social values are based on the official exchange rate but parallel markets dominate in the economy, the social values of tradables (items E and F of the PAM) are multiplied by the ratio of the parallel rate to the official rate.

Finally, exchange-rate adjustments might also be desired to reflect certain short-run effects of exchange-rate changes. Prices of tradable commodities are usually thought to adjust more rapidly than domestic factor prices, thus altering profitability. In the PAM, short-run effects of the exchange rate involve adjustments to E and F; because G does not change, H changes to a greater extent than the exchange rate. These calculations can indicate incentives for existing systems to change and for new systems (different technologies or different regions) to initiate production.

An enormous amount of empirical detail is required to calculate the social exchange rate-the magnitude of the sustainable government budget deficit, the extent to which the existing deficit is financed by foreign capital inflows, the slopes of the supply and demand curves for foreign exchange, and the magnitude of protection afforded various commodities. As a result, estimations usually entail much guesswork. Other experts on the domestic economy may provide useful reviews of assumptions and calculations. Domestic government organizations,

such as the ministry of finance or the central bank, may have economic advisers who assess the economy at an aggregate level. These individuals should have information about the inflation rates and the structure of official and parallel market exchange rates. International organizations such as the World Bank and the International Monetary Fund may also have expert advisers doing such calculations. Box 11.1 provides a stylized example of the procedures needed to determine the social price of a tradable good.